AERRVol29 1

AGRICULTURAL ECONOMICS RESEARCH REVIEW
EDITORIAL BOARD
Chairman
Dr. Mruthyunjaya, Former National Director, NAIP (ICAR), G-502, NCC Urban,
Meadows-2, Doddaballapur Road, Bangalore - 560 064, Karnataka
Chief Editor
Dr. Pratap S. Birthal, Director, Institute of Development Studies, Jaipur - 302 004,
Rajasthan
Managing Editor
Dr. Naveen P. Singh, Principal Scientist, ICAR-National Institute of Agricultural

Economics and Policy Research, New Delhi - 110 012
Members
Dr. Suresh C. Babu, Senior Research Fellow, International Food Policy Research
Institute, Washington, DC, USA
Dr. P. Indira Devi, Director, Centre of Excellence in Environmental Economics,
Professor (Agricultural Economics), College of Horticulture, Kerala Agricultural
University, Thrissur - 680 656, Kerala
Dr. Madhur Gautam, Lead Economist, The World Bank, Washinton, DC 20433,
USA
Dr. Girish K. Jha, Principal Scientist, Division of Agricultural Economics, ICARIndian Agricultural Research Institute, New Delhi 110 012
Dr. P.K. Joshi, DirectorSouth Asia, International Food Policy Research Institute,
NASC Complex, Dev Prakash Shastri Marg, New Delhi - 110 012
Dr. Ranjit Kumar, Principal Scientist- Economics (VDSA), Market, Institutions &
Policy Program, International Crops Research Institute for Semi-Arid Tropics,
Patancheru, Hyderabad - 502 324, Telangana
Dr. M.A. Sattar Mandal, Senior Advisor to Food and Agriculture Organization of
United Nations, Dhaka, Bangladesh
Dr. Niti Mehta, Professor & Acting Director, Sardar Patel Institute of Economic &
Social Research, Thaltej Road, Ahmedabad - 380 054, Gujarat
Dr. Surabhi Mittal, Senior Scientist, CIMMYT-India, NASC Complex, DPS Marg,
New Delhi - 110 012
Dr. Gopinath Munisamy, Director, Market and Trade Economics Division, U.S.
Department of Agriculture, Economic Research Service, 1400 Independence Ave.,
SW, Mail Stop 1800, Washington, DC 20250-0002, USA
Dr. N. Nagaraj, Former Principal Economist, International Crops Research Institute
for the Semi-Arid Tropics, Patancheru - 502 324, Hyderabad, Telangana
Dr. Karl M. Rich, Department of International Economics, Norwegian Institute of
International Affairs (NUPI), Norway
Dr. Devesh Roy, Research Fellow, Markets, Trade and Institutions Division,
International Food Policy Research Institute, Washington, DC, USA
Dr. Smita Sirohi, Principal Scientist (Dairy Economics), ICAR-National Dairy
Research Institute, Karnal - 132 001, Haryana
ISSN 0971-3441
Online ISSN 0974-0279
Agricultural Economics Research Review
Agricultural Economics Research Association (India)

National Agricultural Science Centre Complex
Dev Prakash Shastri Marg, Pusa
New Delhi - 110 012

Vol. 29 (No.1) January-June 2016
Agricultural Economics Research Association (India) 2016
Financial support from:

Indian Council of Agricultural Research (ICAR), New Delhi
Published by:
Dr Suresh Pal, Secretary, AERA on behalf of Agricultural Economics Research Association (India)
Printed at:
Cambridge Printing Works, B-85, Phase II, Naraina Industrial Area, New Delhi 110 028
ISSN 0971-3441

[Journal of Agricultural Economics Research Association (India)]
Volume 29
Number 1
January-June 2016
CONTENTS
Research Articles
Patterns and Drivers of Dairy Development in India: Insights from Analysis of Household
and District-level Data
Avinash Kishore, Pratap S. Birthal, P.K. Joshi, Tushaar Shah and Abhishek Saini
Adoption Status and Influencing Factors of Mobile Telephony in Dairy Sector: A Study in
Four States of India
Prakashkumar Rathod, Mahesh Chander and D. Bardhan
15
TFP Growth of Wheat and Paddy in Post-Green Revolution Era in India: Parametric and
Non-Parametric Analysis
Surya Bhushan
27
Total Factor Productivity Growth and Returns from Research Investment on Soybean in India
Purushottam Sharma and B.U. Dupare
41
Issues Limiting the Progress in Negotiable Warehouse Receipt (NWR) Financing in India
Shalendra, M.S. Jairath, Enamul Haque and Anu Peter V.
53
Institutional Synergies in Processing and Value Addition: Role of a Producers Organisation

in Transforming Farm Economy in Rural Punjab
Udeshna Talukdar and Kamal Vatta
61
Farm Business Income across Land-size Classes and Land Tenure Status: A Field Study in
Assam Plains
Binoy Goswami
69
Impact of Nonfarm Activity on Rural Income Inequality in South 24 Parganas District of

West Bengal
S.M. Rahaman, S. Haldar, Sowmyashree B.V., S. Nandi, L. Malangmeih and B.K. Bera
83
Changing Cropping Pattern, Agricultural Diversification and Productivity in Odisha

A District-wise Study
Dinesh Kumar Nayak
93
Crisis of Sustainability or Perils of Ill-managed Open Access Fisheries? Analysis of

Long-term Catch Trends in Marine Fisheries of Maharashtra and India
M. Suresha Adiga, P.S. Ananthan, H.V. Divya Kumari and V. Ramasubramanian
105
Contd....
Contents contd....
Is Sustainable Sugarcane Initiative (SSI) Technology More Profitable than Conventional
Method for Sugarcane Production? An Economic Analysis
Arthi K., V. Saravanakumar and R. Balasubramanian
117
Effect of Change in Indian Rice Price on Nepalese Rice Market: A Partial Equilibrium
Model
Bhawani Mishra, Krishna L. Poudel and Dilli Raj Mishra
127
Export of Rice from India: Performance and Determinants

Anup Adhikari, M.K. Sekhon and Manjeet Kaur
135
Research Notes
Crop Diversification in Gadag District of Karnataka
N.D. Basavaraj, T.M. Gajanana and M. Satishkumar
151
Impact of Integrated Farming of Water Chestnut and Cat Fish on Livelihood of Farmers in
Seasonal Waterlogged Areas of Odisha
Souvik Ghosh, S. Roy Chowdhury, R.K. Mohanty and P.S. Brahmanand
159
Book Review
167
Guidelines for Submission of Papers/Abstracts
169
Author Index
Adhikari, Anup
135
Goswami, Binoy
69
Rahaman, S.M.
Ananthan, P.S.
105
Haldar, S.
83
Ramasubramanian, V.
105
Arthi K.
117
Haque, Enamul
53
Rathod, Prakashkumar
15
Balasubramanian, R.
117
Jairath, M.S.
53
Roy Chowdhury, S.
Bardhan, D.
Basavaraj, N.D.
15
151
Joshi, P.K.
Kaur, Manjeet
1
135
Saini, Abhishek
83
159
1
Saravanakumar, V.
117
Satishkumar, M.
151
Bera, B.K.
83
Kishore, Avinash
Bhushan, Surya
27
Malangmeih, L.
83
Sekhon, M.K.
135
Birthal, Pratap S.
Mishra, Bhawani
127
Shah, Tushaar
Brahmanand, P.S.
159
Mohanty, R.K.
159
Shalendra
53
Chander, Mahesh
15
Divya Kumari, H.V.

Dupare, B.U.
105
41
Nandi, S.
83
Sharma, Purushottam
41
Nayak, Dinesh Kumar
93
Sowmyashree B.V.
83
Peter, V. Anu
53
Suresha Adiga, M.
105
Gajanana, T.M.
151
Poudel, Krishna L.
127
Talukdar, Udeshna
61
Ghosh, Souvik
159
Mishra, Dilli Raj
127
Vatta, Kamal
61

Vol. 29 (No.1) January-June 2016 pp 1-14
DOI: 10.5958/0974-0279.2016.00014.8
Patterns and Drivers of Dairy Development in India: Insights

from Analysis of Household and District-level Data
Avinash Kishore*a, Pratap S. Birthalb, P.K. Joshia, Tushaar Shahc and Abhishek Sainid
International Food Policy Research Institute, South Asia Office, NASC Complex,
Pusa, New Delhi-110 012
b
National Institute of Agricultural Economics and Policy Research, Pusa, New Delhi-110 012
c
International Water Management Institute, Jal Tarang, Near Smruti Apartment, Anand-388 001, Gujarat
d
Independent Consultant, New Delhi
a
Abstract
Traditionally, Indian farmers kept bovines, especially cattle, for draught purposes in agriculture and
transportation with milk as an adjunct. However, with increasing farm mechanization and rising demand
for milk, the bovine functions have shifted more towards dairying. While bovine population has been
increasing, the chronic scarcity of feed and fodder reinforces the need for optimization of bovine population
for sustainable growth of dairying. In this paper, using district-level data from 1997 to 2007, we show
that this transformation from draught to dairying is underway in some parts of the country, and further
using household-level data, we find that smallholders have contributed disproportionately more to this
transformation. This transformation or intensification of dairying is demand-driven with urbanization
having a strong positive influence on dairy development. On the supply-side, factors like farm
mechanization, improved access to groundwater irrigation and crop diversification away from cereals,
are associated with a shift in the bovine economy from draught to dairying.
Key words: Dairy development, smallholders, urbanization, mechanization, crop diversification
JEL Classification: Q12, O18, P25
Introduction
India holds more than a quarter of worlds bovine
population, and with a production of more than 133
million tons in 2012-13 it is the largest producer of
milk in the world (GoI, 2014). Between 1981-82 and
2011-121, milk production increased more than fourfold, making it the largest agricultural commodity in
quantity as well as value terms (Birthal and Negi,
2012). Milk now accounts for over a quarter (26.4%
in 2011-12) of the total value of agricultural output2
and two-thirds of the total value of livestock
production. Every second rural household in India
owns at least one dairy animal, either cattle or buffalo,
* Author for correspondence
Email: avinash.kishore@gmail.com
and the ownership rate exceeds 70 per cent among the

households owning more than 0.5 hectare of land (GoI,
2006).
In spite of sustained growth in milk production,
the demand for milk is outpacing its supply. Gandhi
and Zhou (2010) have projected the demand for milk
to grow faster than its annual production. The
increasing demand-supply gap may lead to sharp rise
in the prices of milk. Mishra and Roy (2011) have
shown that rising price of milk has been the most
important contributor to food price inflation in India
since 1998. It may threaten the nutritional status of
people as milk is the main source of animal protein for
Indian households. Note that, milk accounts for nearly
three-fourths of the household expenditure on livestock
Vol. 29 (No.1)
January-June 2016
products (Mishra and Roy, 2011). Milk and milk

products have the highest income elasticity of demand
amongst food commodity groups, and therefore, the
demand is likely to increase rapidly as the economy
grows and incomes rise (Joshi and Kumar, 2012).
different types of bovines owned by households

allowing us to explore equity in bovine ownership.
Throughout the paper, we restrict our analysis to rural
areas that account for more than 95 per cent of the
total bovine population in India.
India has the worlds largest bovine population,

but the chronic scarcity of feed and foddertheir prices
have doubled since 2004-053and shrinking area and
deteriorating quality of grazing lands entail that the
milk output increases without further increase in the
number of cows and buffaloes. Instead, the total bovine
population should decline while the number and yield
of in-milk animals should increase. It means that Indias
dairy production system needs to become more
efficient.
Ideally, one would like to have data on yield and

production of milk by different types of dairy animals
to analyze the changes in dairy economy. However,
this type of data are not available at district-level. We,
therefore, rely mainly on demographic variables
available in the Livestock Censuses and survey reports
of the National Sample Survey Organization for our
analysis. Indias bovine economy is likely to become
increasingly focused on milk production as it develops
and intensifies into a specialized activity while the
draught function will become less important (Birthal
and Parthasarathy Rao, 2004). We believe that this
specialization will be achieved, among other things,
by increasing the share of in-milk bovines in the bovine
herd4. If this is true, then we can use the ratio of inmilk bovines to the total bovine populationcalled
herd efficiency ratio (HER) in this paperas an
indicator of dairy development. Other things being
equal, a region (or a farmer) with higher percentage of
in-milk bovines in its herd (i.e. higher HER), we claim,
has a more intensive and more efficient dairying than
the one with lower HER. We assume that a more
efficient dairying system will have higher proportion
of crossbred cows and buffaloes in it. It will have a
higher ratio of females and more in-milk females in
the female stock. Starting from these assumptions about
relationship between bovine population structure and
levels of dairy development (or dairy production
efficiency), we use HER to characterize the levels of
dairy development.
In this paper, we study trends and patterns in dairy

development in India and try to identify key drivers of
these patterns using district-level data from three
quinquennial rounds of Livestock Census,
supplemented by an analysis of the household-level
data from a large-scale survey on Land and Livestock
Holdings conducted by the National Sample Survey
Organization (GoI, 2006). The next section describes
data and assumptions made to examine and interpret
the emerging trends in Indias bovine economy. The
third section focuses on the changes in equity in
ownership of bovines across landholding classes. The
fourth section brings out the regional differences in
dairy development in India. Then, we discuss the
estimation strategy to identify drivers of dairy
development. Results are presented in section 6 and
concluding remarks are made in the last section.
Data and Key Assumptions

The Livestock Censuses of India conducted in
1997, 2002-03 and 2007 and the survey on Livestock
Ownership across Operational Land Holding Classes
in India, conducted by NSSO in 2003, are the two
main sources of data used in this paper. The Livestock
Census provides district-level data on livestock
demography. For dairy animals, it includes data on total
number of non-descript cows (NDC), crossbred cows
(CB) and buffaloes by sex (male-female), age-group
(< 1 year, 1-2.5 years), usage (work, breeding, work
and breeding, neither used for breeding nor work) and
lactation status (in-milk, dry, not-calved, others). The
NSSO survey has data on ownership of number of
Dairying in India: A Smallholders Enterprise

From 1981 to 2011, milk production in India has
grown at more than 4 per cent compounded annually,
surpassing growth rates in the global dairy output and
Indias own food grain production (Birthal and Negi,
2012). Even more striking, smallholder farmers are
driving this growth. From 1981 to 2003, farm
households cultivating less than or equal to one hectare
of land (termed as marginal farm households) in total
farm households increased from 41 per cent to 48 per
cent and their share in in-milk animals soared from 31
per cent to 52 per cent (Table 1). There was a time
Kishore et al. : Patterns and Drivers of Dairy Development in India
Table 1. Ownership of in-milk bovines by operational landholding size class in 2002-03

Category
of farmers
Marginal9 (< 1 ha)

Small (1-2 ha)
Medium (2-4 ha)
Large (> 4 ha)
Near-landless(< 0.04 ha)
Per cent
operating
households
Per cent
cattle
Per cent
buffalo
Per cent
in-milk
crossbreds
Per cent
in-milk
buffaloes
Per cent
in-milk nondescript
cows
Herd
Efficiency
Ratio
(HER)
69.60
16.30
9.10
5.10
15.40
52.90
21.30
14.75
11.05
8.58
50.01
20.88
15.37
13.78
11.96
59.68
18.03
12.44
10.37
17.10
55.09
18.54
14.00
12.68
10.60
49.63
20.56
15.22
14.45
12.90
0.203
0.159
0.167
0.163
0.229
Source: Estimated by authors from GoI (2006).
when the rural poor tended dry bovines of the rich

farmers on common and fallow lands, until they calved
again; this meant that, at any point in time, the poor
held a large share of unproductive bovines, while the
rich had bulk of the milking cows and buffaloes
(Binswanger and Rosenzweig, 1986). This practice is
much less common now. Presently, not only do the
marginal farmers own a disproportionate share of
milking bovines, they also own 60 per cent of Indias
crossbred cows (Tables 2 and 3).
We explore the relationship between herd
efficiency ratio and land ownership using unit-level
data from Land and Livestock Holding Survey
conducted in 2003 with data from a representative
sample of more than 26,000 rural households who
owned bovines (Table 4). In the regression model we
also control for the time-invariant district-level
characteristics (district dummies are not shown in the
table). We find that households who do not own any
agricultural land, have higher HER (by 0.054 points)
than the households who do. Among the land-owning
Table 2. Growth in marginal holdings and their share

of in-milk bovine stock, 1971-72 to 2002-03
Year
1971-72
1981-82
1991-92
2002-03
Per cent share

of marginal
holdings in
total in-milk
bovines
Households with
marginal holdings
as per cent of
total rural
households
20.0
31.0
44.0
52.0
32.9
41.1
48.3
47.9
Source: Estimated by authors
households, HER is higher for households with larger

land sizes. Similarly, households with relatively elder
and more educated heads and more irrigated land have
higher HER. Among different social (or caste) groups,
tribal households have the lowest HER, followed by
scheduled castes, other backward castes and nonbackward caste households in that order.
Table 3. Herd efficiency ratio (HER) across landholding classes, 1971-72 to 2002-03
Year
1971-72
1981-82
1991-92
2002-03
2007
Marginal farmers
(<1 ha)
Small farmers
(1-2 ha)
Medium farmers
(2-4 ha)
Large farmers
(>4.0 ha)
All farmers
0.190
0.164
0.231
0.237
0.176
0.134
0.213
0.214
0.187
0.159
0.222
0.225
0.215
0.160
0.256
0.260
0.21
0.16
0.22
0.23
0.25*
Source: Estimated by authors using data from NSSO

*Livestock Census, 2007.
Vol. 29 (No.1)
January-June 2016
Table 4. Regression showing household level

determinants of herd efficiency ratio (HER) in
India
Variables
If owns land (0=No; 1=Yes)
Landholding size (ha)
Groundwater irrigated area (%)
Surface irrigated area (%)
Education (years of schooling)
Figure 1. Inequity in ownership of land and bovines in
India, 2002-03
Bovine stock is more equitably distributed than

land (Figure 1). Marginal farmers, operating less than
one-fifth of Indias cultivable land5, own more than
half of in-milk bovines. Small farmers with land size
1-2 ha, own about one-fifth of the stock. Thus, with
nearly three-quarters of in-milk stock owned by small
and marginal farmers, dairying in India is
predominantly and increasingly a smallholders
enterprise and its rapid growth may help in more
equitable distribution of farm income and reduction in
poverty (Birthal and Negi, 2012).
Inequity in bovine holdings has also declined over
time. The Gini ratioa measure of the degree of
inequality taking values from 0 to1 of bovine holding
in India declined from 0.43 in 1961 to 0.37 in 1971
and further to 0.28 in 1991 (Sharma et al., 2003).
However, there has been a slight increase in inequity
of in-milk bovine ownership between 1992 and 200203: a reversal of the trend observed over the previous
three decades, and strengthening of the positive
association between in-milk bovine stock ownership
and size of household operational holdings during 1992
to 2002-03 (GoI, 2006). Our own estimates show that
the Gini ratio of bovine holdings had gone up to 0.36
in 2002-03almost the same level as in 1992. We do
not know the exact reasons for this trend reversal. Some
studies (e.g., Birthal and Taneja, 2006) suggest that
decline in the area and the quality of common grazing
lands and fallows is making animal husbandry more
difficult or even unviable for the near-landless and
marginal farmers. Increasing opportunities for such
Age (years)
Scheduled castes
Other backward castes
Others
Constant
Observations (No.)
R-squared
HER
-0.0535***
(0.00870)
0.00309***
(0.00109)
0.000139
(0.000222)
0.000201***
(6.75e-05)
0.00875***
(0.00122)
0.000638***
(0.000204)
0.0816***
(0.00953)
0.116***
(0.00766)
0.147***
(0.00821)
0.150***
(0.0189)
25,959
0.048
Notes: *** p<0.01, ** p<0.05, * p<0.1

Robust standard errors within parentheses
District dummy variables not shown in the table; If owns
land is a dummy variable =1 if a households owns
agricultural land; holding size measures total land owned
by the household; Groundwater irrigated area (%) and
Surface irrigated area (%) indicate percentage of land
irrigated by groundwater and surface water, respectively;
Scheduled tribes were dropped in the regression.
land-constrained households in the non-farm sectors

could be another possible reason for their exit from
bovine husbandry.
The landless and near-landless households, defined
here as those who own less than 0.02 hectare of land,
are most likely to own bovines in northern states of
Jammu & Kashmir, Punjab, Uttarakhand, Haryana and
Rajasthan, while they are least likely to own one in the
southern states of Andhra Pradesh, Karnataka, Kerala
and Tamil Nadu (Appendix 1).
A probit analysis (Table 5) of more than 16,000

near-landless or landless rural households shows that
within a state (we control for state dummies in the
regression), a Sikh or Hindu household that owns some
land, has a larger family size, belongs to other backward
castes (OBC) or non-backward castes, has cultivation
or agricultural labour as its main occupation, and is
less educated is more likely to own a bovine. Most
state dummies are statistically significant, suggesting
that the rate of bovine ownership among near-landless
and landless rural households is significantly different
in most states from that in Jammu & Kashmirthe
omitted state in the regression.
Recent Developments in Bovine Demography of
India
In the traditional Indian farming system, cattle have
been reared primarily for farm power needs. Crop and
bovine production systems in India are closely
intertwined with bovine playing a subsidiary role of
providing draught power and manure for crop
production using crop residues as feed. It is largely a
self-contained system with limited market integration.
This age-old system is changing, giving way to dairying
as a specialized activity (Birthal and Parthasarathy Rao,
2004)still maintained off crops, but not for it. In
many parts of India, dairying accounts for a significant
and growing part of farm income and farm
employment, especially for marginal farmers (Shukla
and Brahmankar, 1999; Siddhu and Bhullar, 2004;
Singh and Verghese, 2004; Birthal and Taneja, 2006).
The rise of milk-orientation of Indian bovine economy
is evident from the changing herd composition. The
number of bullocks, including bovines used for
draught, decreased by more than 10 millions between
1992 and 2007 and the female: male ratio increased
from 1:1 in 1992 to almost 2:1 in 2007. However,
different regions of India are in different stages of this
transition from traditional, self-contained mixed
farming system toward dairying as a specialized
economic activity. We report some of the key trends in
bovine herd structure in India and the inter-regional
variations therein in this section.
Slow Decline in Excess Cattle
India had 0.6 billion bovines in 2007more than
any other country in the world. Being a country scarce
in land and fodder but facing fast growth in demand
Table 5. Characteristics associated with higher

likelihood of a landless or near landless rural
household owning at least one bovine
Explanatory variables
Land (ha)
Household-size
Sex (0=male; 1=female)
Age (years)
Age squared
If married (0=No; 1=Yes)
If secondary education or above
(0=No; 1=Yes)
If Scheduled caste
If Other backward caste
If Others
If agricultural labourer
If non-agricultural labourer
If a cultivator
If Muslim
If Sikh
If Jain
Constant
Observations (No.)
State dummies
Clustered standard errors
Coefficient
44.84***
(9.916)
0.178***
(0.0161)
-0.415***
(0.134)
0.101***
(0.0175)
-0.000965***
(0.000178)
-0.142
(0.0932)
-0.525***
(0.133)
0.194
(0.125)
0.292**
(0.130)
0.261*
(0.155)
0.737***
(0.108)
-0.0745
(0.0787)
2.014***
(0.158)
-0.730***
(0.123)
0.417***
(0.0996)
-1.417**
(0.687)
-3.769***
(0.454)
16,366
YES
Yes
Note: *** p<0.01, ** p<0.05, * p<0.1

Robust standard errors within parentheses
If* variables are all dummy variables that are assigned value
of 1 if the head of the household belongs to the category.
Land measures the total land area owned by the household;
SC, OBC and Others are caste groups in India with STs
(scheduled tribes) as the omitted category. Similarly,
Muslim, Sikh and Jain are the religious groups with Hindus
as the omitted religious group.
Vol. 29 (No.1)
January-June 2016
Table 6. Bovine and land ownership, herd size and herd efficiency ratio (HER) values across households of different
social categories in India in 2003
Social group
Scheduled tribes
Scheduled castes
Other backward castes
Others
Per cent rural

households
that own at
least one
bovine
Mean land
owned per
household
(ha)
52.57
38.57
49.15
50.55
0.79
0.31
0.76
1.01
Average number of bovines owned per household
HER
value
All households
Bovine-owning households
1.87
0.96
1.67
1.74
3.56
2.49
3.40
3.44
for milk and milk products, India needs to reduce its

population of unproductive bovines and increase the
productivity for efficient use of existing resources.
However, Indias bovine population continued to grow
steadily till 1992, it stabilized in 1990s and registered
a small reduction between 1997 and 2003for the first
time since 1951only to record an increase again
between 2003 and 2007.
Slow Improvement in HER
The in-milk females constituted only 22 per cent
of the total bovine stock of India in 1997. Their share
increased to 23 per cent in 2002-03 and to 25 per cent
in 2007. The HER of 0.25 is still quite low and indicates
the persistent problems of low productivity and low
efficiency. The HER is relatively higher for buffaloes
(0.33) and crossbred or exotic cows (0.32) and lower
for non-descript cows (0.18). This is expected as the
non-descript cattle are still valued more for draught
power as compared to crossbred cattle and buffaloes
that are reared mainly for milk production.
Across social classes, the HER seems to follow
the existing social status gradient found in rural India
with lowest values for scheduled tribes (STs) followed
by scheduled castes (SCs), other backward castes
(OBC) and other castes (Table 6). The tribal households
are most likely to own a bovine and their average herd
size is also the largest, but they have the lowest herd
efficiency (HER) ratio, probably because of their
greater dependence on cattle for farm power needs and
cultural taboos on milk consumption. Geographically,
eastern India and tribal districts of the country continue
to have the most under-developed dairy economy with
HER well below the national average of 0.25, while
0.126
0.217
0.249
0.290
districts in northern states of Punjab and Haryana have

a higher HER (Table 7).
More Buffaloes than Cows
In India, buffaloes are increasingly becoming the
preferred dairy animals. Rapid farm mechanization
especially, increasing use of tractors, power tillers and
threshers, has been reducing farmers dependence on
animal power. Thriving rental markets for these
Table 7. Herd efficiency ratio (HER) by state and region
of India in 2007
State
HER value
Eastern region
Assam
Chhattisgarh
Odisha
West Bengal
0.18
0.12
0.12
0.18
Western region
Gujarat
Rajasthan
Maharashtra
Madhya Pradesh
India
0.28
0.30
0.24
0.22
0.25
Northern region
Punjab
Haryana
Uttar Pradesh
0.39
0.32
0.30
Southern region
Andhra Pradesh
Karnataka
Kerala
Tamil Nadu
0.28
0.27
0.3
0.27

Table 8. Ratio of female and in-milk buffaloes to cows
across selected states of India
State
Ratio of number of buffaloes to cows

All females
In-milk females
Haryana
Punjab
Andhra Pradesh
Gujrat
Uttar Pradesh
Rajasthan
4.99
3.48
2.01
1.77
1.67
1.18
4.92
3.28
2.08
1.77
1.83
1.28
machines allow even smallholders to use these for

agricultural operations. In 2007, India had more than
0.2 billion buffaloes accounting for more than one-third
(34.5%) of total bovine population6. If we look at
female bovines only, buffaloes constituted 44.0 per cent
of the adult female bovine population and 46.3 per cent
of total in-milk bovines. If this trend continues, soon,
India will have more in-milk buffaloes than in-milk
cows. The states of Haryana, Punjab, Andhra Pradesh,
Gujarat, Uttar Pradesh and Rajasthan already have
more female buffaloes than female cattle by 2007
(Table 8).
The data in Table 9 show that indigenous or nondescript cows account for more than half of the total
bovine population and more than 40 per cent of all
female (45%) and in-milk (43%) bovines. Their
population share, however, has been declining, giving
way to buffaloes and crossbred cows as the bovine
economy moves from draught to dairying. The share
of in-milk cows in all females is close to 40 per cent
for all three types of bovines, but the share of females
in total itself is much lower for indigenous cows. About
44 per cent of all indigenous cattle are males while
they account for only 24 per cent of crossbred cattle

and buffaloes. Male animals account for a high share
of indigenous population because a large number of
farmers still depend on animal power for agricultural
work and transportation. Nearly two-thirds (63%) of
the indigenous male cattle are kept mainly for draught
purposes. As machines substitute animals as motive
power for agriculture, the need for animal power will
decline and the share of indigenous cattle and the
fraction of male animals therein will go down, resulting
in a rise in the HER.
To summarize, the rapid dairy development in India
seems to be confined to arid and semi-arid northwestern and southern regions of the country while the
eastern and tribal regions continue to trail behind.
Districts with dynamic dairy economies, found in states
like Punjab, Haryana, Kerala, Tamil Nadu, Karnataka,
Andhra Pradesh, are marked by decline in bullock
population, increase in the population of buffaloes and
crossbred cows, relatively high ratio of milking to
female stock and decline in non-descript cows. In the
next section, we will try to identify the key drivers of
increasing orientation of bovine demography towards
dairying using district-level data.
Several studies have been carried out to understand
the changes in demographic structure of livestock at
the national and state levels (Sharma, 2004; Birthal
and Negi, 2012), but to the best of our knowledge none
of these has used district-level data. State-level data
could mask considerable inter-district variations in
bovine demography and its drivers that may exist
within a state. For example, the Kolhapur district in
Maharashtra has the distinction of having highest HER
in the country (0.46) while the Chandrapur district in
the same state has the lowest HER of 0.13. Hence, an
analysis of the bovine demography using district level
Table 9. Population shares of different types of bovines in India: 1997 to 2007
Year
1997
2003
2007
Indigenous cows
Share in
Ratio of
Ratio of
total
female to
in-milk
bovine
total
to total
population animals
female
(%)
animals
57.00
53.00
52.00
0.52
0.54
0.56
0.35
0.35
0.37
Cross-breds
Share in
Ratio of
Ratio of
total
female to
in-milk
bovine
total
to total
population
female
population
(%)
animals
animals
9.00
10.45
13.02
0.72
0.75
0.76
0.39
0.40
0.41
Share in
total
bovine
animals
(%)
33.74
36.54
35.00
Buffaloes
Ratio of
female
to total
female
Ratio of
in-milk
to total
animals
0.75
0.77
0.76
0.38
0.40
0.40
data is important to understand the causes of such

uneven development and to target efforts and
investment accordingly.
Empirical Strategy
We use the district-level data on bovine
demography from three rounds of Livestock Censuses
pertaining to the years 1997, 2003 and 2007 (GoI, 2003;
2005; 2010), and combine these with variables on
human population, land uses and agricultural variables
such as farm size, cropping pattern, fertilizer-use
intensity, area under surface water and groundwater
irrigation, density of tractors and power tillers, etc.7
for the corresponding years to understand the emerging
pattern in bovine demography and its drivers.
First, we estimate a pooled OLS or POLS model
where we pool data from three rounds of Livestock
Censuses assuming that observations across years are
not correlated over time for the same district, i.e.
Yit = + Xit + it
(1)
where, Yit is the dependent variable, HER of district i

in year t; is the constant term; Xit is a set of time
varying variables influencing Y in district i and it is
the error-term.
The assumption of errors being uncorrelated across
years for the same district is quite a strong. We relax
this in the subsequent models where we use the panel
data to estimate fixed effects model (Equation 2):
Yist = s + + st + Xist + ist
(2)
where, Yist is the herd efficiency ratio in district i in

state s and year t; s is the state fixed effect; controls
for national trend in the dependent variable; st is a
state-year interaction effect and ist is a district-year
specific error-term. Xist is a set of time varying control
variables for the district. The state fixed effects control
for time-invariant characteristics of the state that may
affect livestock demography and are correlated with
regressors, while the state-year interactions control for
annual shocks across districts in a state. A brief
description of the explanatory variables is given below.
Urbanization Other things held constant, per capita
milk consumption is generally higher in urban areas
and among the high income populations. Therefore, it
is likely that more urbanized districts will show
demand-led development of dairying economy8.
Vol. 29 (No.1)
January-June 2016
Population Density The literature predicts that

higher population density leads to a greater
intensification of agriculture (Boserup, 1965). In the
case of livestock, it is argued that interaction between
crop and livestock production systems is weak at low
population densities; it increases with population
density and finally declines giving way to specialized
crop and livestock activities (McIntire et al., 1992 cited
in Birthal and Parthasarathy Rao, 2004). Accordingly,
HER is expected to be higher when dairying is practised
as a commercial activity. Vaidyanathan et al. (1982)
argue that bovine male-female ratio is density
dependent in India. They posit that as human density
increases, draught animal density increases only up to
a threshold; thereafter, draught animals are discarded
and cows are retained. Thus, both Vaidyanathan et al.
(1982) and McIntire et al. (1992) suggest that milkorientation or HER of a district should increase with
increase in population density, once it crosses a
threshold level. Since most districts in India have
relatively high population densities, we would expect
a positive correlation between HER and population
density and negative correlation between bovine and
work animal density and density of the (human)
population.
Literacy Literate households are more likely to
adopt new technology that can enhance dairy
production. We expect districts with higher rural
literacy rates to have higher herd efficiency ratios and
lower density of working bovines.
Mechanization Historically, cows in India have
been valued to produce draught animals and milk as
adjunct (Vaidyanathan et al., 1982). If animals are the
main source of draught power, then there should be
more male cows in the herd. During the past four
decades, India has adopted crossbreeding technology
to enhance milk production, but crossbred males are
not considered suitable for draught purposes. The use
of male buffaloes for work is also limited. Farm
mechanization (mostly tractors and power tillers)
reduces the compulsion of rearing male cattle (mainly
non-descript). Therefore, we expect districts with
higher degrees of mechanization to have higher herd
efficiency ratios.
Irrigation The districts with better irrigation facility
tend to be more mechanized with higher cropping
intensity and crop yields, and therefore greater
availability of green and dry fodder and less need for
draught animals. The per capita income is also higher

in more irrigated areas (Bhattarai and
Narayanamoorthy, undated). Thus, irrigation creates
both supply and demand side conditions for more
advanced dairying. Groundwater irrigation allows
higher cropping intensity and year-round cultivation
of green fodder crops and is associated with better
yields compared to surface water irrigation. We expect
positive coefficients for both surface and ground water
irrigated areas, but the latter is likely to have a larger
coefficient.
Cropping Pattern Cereals are the main sources of
dry fodder. Other things held constant, districts with a
lower fraction of area cropped with cereals will have
lower availability of dry fodder. It is our surmise that
the pressure to get rid of low-yielding excess cattle
and to intensify dairying is higher when roughage is
relatively scarce. Therefore, we expect districts with
lower (and declining) area under cereal crops to have
more intensive dairying.
Fertilizer Use Higher use of chemical fertilizers is
a sign of more intensive agriculture and probably a
more milk-oriented dairying, as it contributes towards
improving availability of roughages.
Barren and Permanent Pastures Common grazing
lands (pastures) comprise an important source of fodder
for the smallholders and landless farmers who have
limited availability of harvested crop residues to feed
cattle and buffalo (Dikshit and Birthal, 2010). It is likely
that fodder is more easily available in the districts with
larger fraction of their geographical area as barren lands
and permanent pastures. If so, then we can expect less
intensive dairying in the districts with larger area under
pastures.
Road Density Better road connectivity makes milk
marketing easier, facilitating dairy development.
Therefore, we expect road density in a district to be
positively correlated with HER.
Results and Discussion

First, we discuss results from the pooled OLS
regression (column 1 of Table 10). The model explains
nearly half of the inter-district variation in herd
efficiency ratio (HER). All the included variables have
expected signs and except surface irrigation, road
density and pasture lands, are statistically significant
at 99 per cent confidence level. However, the OLS

estimates could be biased and inconsistent if there are
unobserved variables that may be correlated with
independent variables included in the model.
We try to mitigate the problem of unobserved
heterogeneity employing panel regression with state
fixed effects. We also include national and state-specific
time trends (see columns 2 and 3 in Table 10). Column
3 with both national and state-specific time trends is
our preferred model.
We find that more urbanized districts have a greater
orientation toward dairying. A one- percentage point
increase in the share of urban population in a district is
associated with 0.085 point higher HER. The districts
with higher rural population densities also have higher
HER, but the HER starts declining once the population
density exceeds a threshold level, as indicated by
Vaidyanathan et al. (1982) and McIntire et al. (1992).
The coefficient on squared of population density is
statistically significant too. Districts with higher rural
literacy rate also have higher herd efficiency. Irrigation
sources, both groundwater and surface water have
positive coefficients, but the coefficient on surface
irrigation variable is not significant. A one-percentage
point increase in area under groundwater irrigation is
associated with 0.05 point higher HER. Mechanization
(measured here by the density of tractors and power
tillers) is strongly and significantly correlated with
HER. However, the effect of increase in availability of
tractors and power tillers on HER starts shrinking after
a threshold is reached. The vibrant rental markets
ensure that the agricultural demand for machines is
saturated at relatively low machine densities than it
would be in the absence of markets. The districts where
cropping pattern is dominated by cereals tend to have
a lower HER, as is indicated by the negative and
statistically significant coefficient on area under cereals
crops. It seems that diversification toward non-cereal
crops and dairying happens in tandem. It is also possible
that fodder constraint in the less cereal-centric districts
forces the farmers to be more efficient in their herd
management as in peri-urban districts.
The coefficients on fertilizer use, landholding size
and under barren and pasture lands that are statistically
significant in the OLS model, lose their significance
in the models with state fixed effects and state-specific
time trends. We do not find statistically significant
10
Vol. 29 (No.1)
January-June 2016
Table 10. Determinants of variation in herd efficiency ratio (HER) across districts of India
Variable
Per cent urban population
Rural population density (population in 00/sq
km of geographical area)
Squared rural population density
Per cent sown area under cereals
Groundwater irrigated area (% of NSA)
Surface irrigated area (% of NSA)
Road density (km/sq km area)
NPK (kg/ha)
Average land size (ha)
Tractor density (No./ha of NSA)
Tractor_density squared
Literacy rate (percent)
Per cent area under barren land and pastures
(Pooled
Regression)
HER
(Panel with State

Fixed Effects)
HER
(Panel with State and

Time fixed effects)
HER
0.11169***
(0.015)
0.02144***
(0.003)
-0.00126***
(0.000)
-0.07684***
(0.008)
0.06953***
(0.012)
-0.00924
(0.013)
0.00018
(0.000)
0.00016***
(0.000)
0.00979***
(0.002)
1.11495***
(0.151)
-3.22057***
(0.511)
0.11888***
(0.016)
0.06051*
(0.029)
0.07583***
(0.016)
0.02213***
(0.004)
-0.00086**
(0.000)
-0.04390***
(0.011)
0.04740***
(0.013)
-0.00350
(0.014)
-0.00016
(0.000)
0.00002
(0.000)
0.00494*
(0.002)
0.33969***
(0.090)
-0.70158*
(0.277)
0.12389***
(0.019)
0.02444
(0.035)
-0.00077
(0.000)
1.69336*
(0.852)
0.08504***
(0.016)
0.01562***
(0.004)
-0.00072*
(0.000)
-0.04595***
(0.010)
0.04551***
(0.012)
0.01465
(0.014)
0.00009
(0.000)
0.00002
(0.000)
0.00212
(0.002)
0.20612*
(0.088)
-0.60596*
(0.259)
0.08303***
(0.024)
0.00243
(0.036)
-0.00044
(0.001)
1.06425
(1.748)
807
807
Yes
No
Yes
Yes
Year
Constant
Observations (No.)
R-squared
State fixed effects
State-specific time trend
0.08323***
(0.015)
807
0.51721
Note: *** p<0.001, ** p<0.01, * p<0.05

Standard errors within the parentheses
association between HER and road density, land-use

or landholding size, probably because of limited
variation in the shorter time-series.
To summarize, both supply and demand side
factors seem to affect milk-orientation of a districts
bovine economy. Bovine economies are more milkoriented in more urban districts with greater population
density, higher rural literacy, better access to
groundwater irrigation and higher levels of farm
mechanization. On the contrary, districts with cerealcentric cropping pattern tend to have fewer in-milk
11
animals in their bovine herd. Diversification to noncereal crops goes with diversification or greater dairy
orientation of bovine economy too.
Table 11. Inter-quartile range of district level herd

efficiency ratio (HER) in 2007
Conclusions and Policy Implications
Zero (Minimum)
0.085
Indias dairy economy has grown rapidly over the

past four decades. We highlight two positive aspects
of this recent growth. First, this growth has been
achieved with an increase in the overall productivity
of the bovine herd (measured here by HER) and
reduction in the number of male bovines. Farmers are
becoming more efficient in herd management. Animals
are being reared more for dairying than for the draught
power. In a land and fodder scarce country like India,
this improvement in efficiency is much needed for
sustainable development of dairying.
First
0.207
Second (Median)
0.265
Third
0.315
Fourth (Maximum)
0.463
Second, an increasing share of this growing

economy has been captured by the marginal and submarginal farmers. Though there has been a reversal in
this trend in recent years, the bovine ownership
continues to be more equitable than land ownership
and the marginal and near-landless or landless farmers
have been at forefront in the transformation of bovine
economy from draught to dairying. They have the
highest share of in-milk animals in their herds. If these
trends persist, rapid dairy development could lead to
more equitable distribution of farm income. It will also
help improve the nutritional status of the poor
households since ownership of diary animals has a large
positive impact on consumption of milk. This happens
because households tend to consume more of what they
produce themselves in areas with missing or poorly
developed markets like in rural India (Hoddinott et al.,
2014).
On the flip side, there are large inter-district
variations in the levels of dairy development (as
measured by HER (ranging from 0.085 to 0.43, see
Table 11) and the pace of change in the dairy economy.
The districts of eastern India and tribal India continue
to have a very small share of in-milk animals in their
bovine herd and they have lagged further behind the
rest of India over the past fifteen years. These are also
the districts with higher level of poverty where rapid
dairy development could make a notable difference.
Our analysis shows that proximity to larger demand
centres has a strong influence on dairy development.
Quartile
HER value in 2007
Source: Computed by authors using data from Livestock

Census 2007
Milk is a bulky and perishable commodity. Its

marketing is difficult and production is unlikely to take
off unless there is a ready access to the market (Birthal,
2008).
On the supply side, groundwater irrigation helps
farmers to intensify land use and ensures year-round
access to green fodder which is required to maintain
an efficient herd with high proportion of in-milk
animals at any point of time. Farm mechanization is
another big contributor to increase in herd efficiency
and development of dairy economy. It reduces the need
for or dependence on draught animals and allows
farmers to diversify towards dairying. Crop
diversification away from cereals is also associated
with increase in herd efficiency.
In India, and in many other countries of the world,
bovines are a major source of household income and
nutrition. The poorer households tend to have even a
higher dependence on bovines for their livelihoods.
As diets diversify and intake of milk and milk products
increases, the production needs to keep pace with the
rising demand. This can be achieved sustainably only
if the productivity of dairy animals increases. Farm
mechanization, crop diversification, improved access
to year-round irrigation and improvement in market
linkages help smallholders become more efficient milk
producers. A more efficient and sustainably growing
dairy economy would not only improve nutritional
status of households, but also help bring a greater share
of farm income to the farmers who own very small
landholdings. The policy should, therefore, focus on
increasing crop diversification and smallholders access
to irrigation, farm machines and markets.
12
End Notes
1
Based on AHS (2006) and Table 1 in http://

www.nddb.org/English/Statistics/Pages/MilkProduction.aspx: Estimates of Production and Per
Capita Availability of Milk 1950-51 to 2006-07, All
India
Excluding contributions from forestry and fisheries.
http://www.nddb.org/English/Statistics/Pages/Indexnumber-of-Wholesale-prices.aspx
Improvement in dairy production efficiency will also

involve having bovines with higher milk yield. As
mentioned earlier, we do not have data on milk yield
at the disaggregate level.
Directorate of Economics and Statistics, 2001
Indigenous and crossbred cows accounted for 55.2

per cent and 10.3 per cent of total bovines in India in
2007, respectively.
See ICRISAT Meso-data at http://vdsa.icrisat.ac.in/

vdsa-database.htm for district level time-series data
on a number of variables related to agriculture.
There might be districts that are not highly urbanized

but are close to big urban centres like Bharatpur and
Alwar in Rajasthan (close to Delhi), Mehsana in
Gujarat (Close to Ahmedabad) and Bangalore-rural
(close to Bangalore) and many others. Such districts
will have high levels of dairy development too.
We do not have district level data on per capita
income. Also, there is a possible problem of
endogeneity in including per capita income as an
explanatory variable in our regressions because
higher income means higher milk demand, but more
developed dairying could also lead to higher incomes.
This category includes the 15.4 per cent near-landless

households also.
Acknowledgement
The author thank the referee for helpful suggestions
on the earlier draft of the paper.
References
Bhattarai, Madhusudan and Narayanamoorthy, A. (Undated)
Impact of irrigation on agricultural growth and poverty
alleviation: Macro level analyses in India. Water Policy
Vol. 29 (No.1)
January-June 2016
Research Highlight 12. IWMI-Tata Water Policy

Program, India
Binswanger, Hans P. and Rosenzweig, Mark R.
(1986) Behavioural and material determinants of
production relations in agriculture. The Journal of
Development Studies, 22(3): 503-539.
Birthal, P.S. (2008) Linking smallholder livestock producers
to markets: issues and approaches. Indian Journal of
Agricultural Economics, 63(1):19-37.
Birthal, P.S. and Parthasarathy Rao, P. (2004) Intensification
of livestock production in India: patterns, trends, and
determinants. Indian Journal Agricultural Economics,
59(3): 555-566.
Birthal, P.S. and Taneja V.K (2006) Livestock sector in India:
Opportunities and challenges for smallholders. In:
Smallholder Livestock Production in India:
Opportunities and Challenges, Eds: P.S. Birthal, V.K.
Taneja and W. Thorpe. National Centre for Agricultural
Economics and Policy Research, New Delhi, India and
International Livestock Research Institute, Nairobi,
Kenya.
Birthal, P.S. and Negi, D.S. (2012) Livestock for higher,
sustainable and inclusive agricultural growth. Economic
and Political Weekly, 47 (26/27): 89-99.
Birthal P.S., Taneja, V.K. and Thorpe, W. (Eds) (2006)
Smallholder Livestock Production in India:
Opportunities and challenges. National Centre for
Agricultural Economics and Policy Research, New
Delhi, India and International Livestock Research
Institute, Nairobi, Kenya.
Boserup, Ester (1965) The Conditions of Agricultural
Growth: The Economics of Agrarian Change under
Population Pressure. G. Allen & Unwin, London, UK.
Dikshit, A.K. and Birthal, P.S. (2010) Indias livestock feed
demand: estimates and projections. Agricultural
Economics Research Review, 23(1):15-28.
Gandhi, V.P. and Zhou, Z. (2010) Rising demand for
livestock products in India: nature, patterns and
implications. Australasian Agribusiness Review, 18(1):
103-35.
GoI (Government of India) (undated) All-India Area,
Production and Yield of Foodgrains from 1950-51 to
2006-07. Directorate of Economics and Statistics,
Ministry of Agriculture, New Delhi. Available at: http:/
/dacnet.nic.in/eands/4.5(a)Allper cent20Indiaper
cent20Area,per cent20Productionper cent20andper
cent20Yieldper cent20ofper cent20Foodgrains.xls

GoI (Government of India) (2003) 16th Indian Livestock
Census 1997. Department of Animal Husbandry,
Dairying and Fisheries, Ministry of Agriculture, New
Delhi.
Delhi.
GoI (Government of India) (2006) Livestock Ownership
across Operational Land Holding Classes in India
2002-03. Report No. 493, Ministry of Statistics and
Program Implementation, New Delhi.
Delhi.
GoI (Government of India) (2014) Annual Report 2012-13.
Department of Animal Husbandry, Dairying and
Fisheries, Ministry of Agriculture, New Delhi. Available
at: http://www.dahd.nic.in/dahd/WriteReadData/
Annualper cent20Reportper cent202012-13per
cent20EFG.pdf (accessed on 17th June 2014).
Hoddinott, J.F., Headey, D.D. and Dereje, M. (2014) Cows,
Missing Milk Markets and Nutrition in Rural Ethiopia.
International Food Policy Research Institute,
Washington, D.C. Available at: http://www.ifpri.org/
sites/default/files/publications/esspwp63.pdf
Joshi, P.K. and Kumar, P. (2012) Food Demand and Supply
Projections for India, (Mimeo). International Food
Policy Research Institute New Delhi Office.
Mishra, Prachi, and Roy, Devesh (2011) Explaining Inflation
in India: The Role of Food Prices, (Mimeo).
International Monetary Fund, Washington, D.C.
13
Sharma, V.P. (2004) Livestock economy of India: current

status, emerging issues and long term prospects. Indian
Journal of Agricultural Economics, 59(3): 512-554.
Sharma V.P., Delgado C.L., Staal, S. and Singh, R.V. (2003)
Structural Changes in the Dairy Sector and Equity
Issues in Livestock Industrialization Project: Phase II
- Policy, Technical, and Environmental Determinants
and Implications of the Scaling-Up of Milk Production
in India. Foof and Agriculture Organization of United
Nations, Rome. Available at: http://www.fao.org/
wairdocs/lead/x6170e/x6170e30.htm#TopOfPage
Shukla, R.K. and Brahmankar, S.D. (1999) Impact
Evaluation of Operation Flood on Rural Dairy Sector.
National Council of Applied Economic Research, New
Delhi.
Singh R.P. and Verghese K.A. (2004) Diversification of
Economic Activities and Farm Household Income in
Rajasthan. Maharana Pratap University of Agriculture
and Technology, Udaipur, Rajasthan.
Sidhu, R.S. and Bhullar, A.S. (2004) Changing structure of
the farm economy in Punjab: impact of livestock on
income and employment. Indian Journal of Agricultural
Economics, 59(3): 578-587.
Vaidyanathan, A., Nair, K.N., Harris, Marvin, Abruzzi,
William S., Adams, Richard N., Batra, S.M., Chibnik,
Michael, Crotty, R., Doherty, Victor S., Freed, Stanley
A., Freed, Ruth S., Fruehling, Royal T., MacLachlan,
Morgan D., Nonini, Donald M., Odendhal, Stewart,
Prasada, Rao D.L. and Robkin, Eugene (1982) Bovine
sex and species ratios in India (and comments and
reply). Current Anthropology, 23(4): 365-383.
Received: July, 2015; Accepted: November, 2015
14
Vol. 29 (No.1)
January-June 2016
Appendix 1. Percentage of landless or near landless (< 0.02 ha) households across states in rural India who owned at least
one bovine in 2002-03
State
Andhra Pradesh
Assam
Bihar
Chhattisgarh
Gujarat
Haryana
Himachal Pradesh
Jammu & Kashmir
Jharkhand
Karnataka
Kerala
Madhya Pradesh
Maharashtra
Odisha
Punjab
Rajasthan
Tamil Nadu
Uttar Pradesh
Uttarakhand
West Bengal
India
Fraction of landless and near landless (<0.02 ha)

households that owned a bovine
Fraction of all rural households

that owned a bovine
0.046
0.111
0.095
0.049
0.115
0.23
0.037
0.148
0.041
0.098
0.024
0.09
0.043
0.053
0.319
0.142
0.051
0.14
0.121
0.065
0.186
0.275
0.593
0.47
0.618
0.498
0.627
0.657
0.771
0.568
0.467
0.166
0.606
0.396
0.489
0.656
0.643
0.225
0.635
0.734
0.432
0.476
Source: Estimated by authors using LLSS data from NSSO (2006)

DOI: 10.5958/0974-0279.2016.00015.X
Adoption Status and Influencing Factors of Mobile Telephony

in Dairy Sector: A Study in Four States of India
Prakashkumar Rathoda*, Mahesh Chanderb and D. Bardhanc
Department of Veterinary and Animal Husbandary Extension Education, Veterinary College, Bidar, Karnataka
b
Division of Extension Education, cDivision of Livestock Economics, Statistics and Information Technology,
ICAR-Indian Veterinary Research Institute, Izatnagar-243 122, Uttar Pradesh
Abstract
The study has analysed the adoption status of mobile telephones in dairying, and enlisted the constraints
as perceived by the dairy farmers. It has also identified the factors affecting adoption of mobile telephones
in India through the primary data collected from 360 dairy farmers of four states of north India, viz.
Haryana, Punjab, Uttar Pradesh and Uttarakhand. The adoption status has revealed that the majority of
respondents have been using mobile phones partially in dairying since 3-6 years, followed by adoption
since 0-3 years. A significant difference (p < 0.001) has been observed among the respondents across the
states with regard to adoption of mobile phones in dairying. The application of multinomial logit model
has revealed that the model was highly significant and fit for explanation. The variables distance to
veterinary institution or animal healthcare centre, landholding size and scientific-orientation were
significantly associated with the probability that the respondent will be a full adopter of mobile phone in
dairying. The study has recommended that appropriate measures need be adopted for effective use of
mobile phones in dissemination of livestock-related information in general and dairying in particular to
the farming community.
Key words: Dairy sector, livestock technology, multinomial logit model, mobile phones, innovation,
adoption status
JEL Classification: Q16, Q18
Introduction
Information and communication technology (ICT)
has significant potential to help farmers in acquiring
and accessing information which can be utilized to
enhance agricultural and livestock production. At
present, a wide range of ICT platforms is available for
accessing and sharing agriculture-related information
and knowledge in the forms of web pages, audio, video
Email: prakashkumarkr@gmail.com
The paper is based on Ph.D. thesis, Livestock Innovation
System: A Multi-stakeholder Analysis in Dairying submitted by the first author in 2015 to ICAR-Indian Veterinary
Research Institute, Izatnagar.
and text messaging. Among all the ICT tools, mobile

phone has emerged as one of the widely accepted and
adopted instruments in most parts of the world to ease
the information communication process among the
farming community (Hayrol et al., 2009). The
increasing penetration of mobile phones and delivery
of mobile-enabled information services to the farming
community can reduce information asymmetry as well
as complement the role of agricultural extension
services by saving time, offering instant out-reach and
ensuring continuity in information availability (Mittal
et al., 2010). On similar lines, Mishra (2010) has also
found that the most popular mode of availing
information related to agriculture and animal husbandry
is the mobile phone, followed closely by radio/
16
Vol. 29 (No.1)
January-June 2016
Table 1. Locale of the study

University selected for study
(for scientists and extensionists)
District under study

(for dairy farmers)
State
Indian Veterinary Research Institute (IVRI), Izatnagar (http://ivri.nic.in/)

G.B. Pant University of Agriculture & Technology (GBPUA&T), Pantnagar
(http://www.gbpuat.ac.in/)
National Dairy Research Institute (NDRI), Karnal
(http://www.ndri.res.in/ndri/Design/Index.html)
Guru Angad Dev Veterinary and Animal Sciences University (GADVASU),
Ludhiana (http://www.gadvasu.in/)
Bareilly
Udham Singh Nagar
Uttar Pradesh
Uttarakhand
Karnal
Haryana
Ludhiana
Punjab
television, while Sife et al. (2010) have emphasized

on the role of mobile phones in accessing market
information for agriculture and livestock produce.
However, despite huge potential inherent in mobile
telephony, there are some critical issues that need to
be addressed before scaling up these services.
Although many public and private sector
organizations are providing mobile-enabled services
in India (GoI, 2015), the role of mobile phones in the
dairy sector is still low. Lack of customization of these
services to serve specific needs of farmers and lack of
operational knowledge are some of the impediments
in effectively leveraging the potential of mobile phones
(Gakuru et al., 2009; Babu and Asokhan, 2011; Katke
and Padamlatha, 2012; Inigo et al., 2014).
In spite of dairy being an important component of
agriculture, the adoption of mobile phones by the dairy
farmers for information access, is an area which has
not received its due attention. Although some authors
like Rathod and Chander (2014), have pointed out the
importance of mobile telephony in the dairy sector,
several initiatives are still to be strengthened by the
public and private extension and research organizations
for effective delivery of information through mobile
phones. In the above context, the present study was
carried out with the specific objectives of assessing
the adoption status of mobile phones, determining the
major constraints perceived by the farmers; and
identifying the factors that affect adoption of mobile
phone use in Indian dairy sector. Some suggestions of
dairy farmers are also given to improve the diffusion
and adoption of mobile phone use in Indian dairy sector.
Materials and Methods

A combination of purposive and multi-stage
random sampling was adopted in the study to select
the respondents. Four agricultural / veterinary

universities and institutes, which are at the forefront
of research in the livestock sector, were selected. All
the selected universities / institutes have carried out
researches in development of ICT-based information
dissemination modules, including mobile telephony.
The districts in which these veterinary universities /
institutes are situated were thus selected to ascertain
the extent of adoption of mobile telephony as a means
of accessing information related to dairying. Table 1
presents the names of selected universities / institutes
and the districts in which the study was conducted. A
total of 6 villages from each district were selected
randomly at the rate of three villages from each block.
Fifteen farmers having at least two dairy animals were
then selected from each village using random and snow
ball method. Thus, the ultimate sample size was of 360
farmers from a total of 24 villages in four states of
north India.
The primary data were collected by interviewing
the selected respondents personally using a pre-tested
semi-structured interview schedule and also an open
schedule. The reference period for which information
was collected was from November, 2013 to June, 2014.
The data were also collected through observation
method and focussed group discussions held at the
villages.
Analytical Framework
To ascertain the extent of use of mobile phone for
getting information related to dairying, respondents
were asked to elicit their adoption level on a four-point
continuum, viz. full adoption (3), partial adoption
(2), discontinuation (1) and non-adoption (0). The
responses of the farmers regarding number of years of
mobile phone use were categorized into respective
Rathod et al.: Adoption Status and Influencing Factors of Mobile Telephony in Dairy Sector
categories, viz. adoption since past 0-3 years,

adoption since past 3-6 years, adoption since past
6-9 years and adoption since more than 9 years. The
data collected from sample respondents were coded,
tabulated, analyzed and presented in the form of tables.
The statistical tools, viz. frequency, percentage, mean,
standard deviation and Chi-square test were used for
analysis using SPSS version 20.0 package. The sociopersonal, economic and psychological characters of
dairy farmers were classified into low, medium and
high categories based on the mean and the standard
deviation.
Multinomial Logit Model
To identify the factors that influence the
respondents degree of adoption of mobile phone, the
multinomial logit model (as used by Pundo and Fraser,
2006) was fitted. The multinomial logit model focused
not only on the most important decision (use of mobile
phone or not), but also on the degree of adoption of
mobile phone in dairying. In the fitted model, the
dependent variable assumed three discrete values, viz.
0 (when the respondent did not adopt mobile phone in
dairying), 1 (when the respondent partially adopted
mobile phone-use in dairying) and 2 (when the
respondent fully adopted mobile phone-use in
dairying). Given the alternatives before a respondent,
the probability that an individual i choose the
alternative j can be expressed by Equation (1):
(1)
where,
Pr[Yi=j] = Probability that an individual i belongs to
either No adoption,Partial adoption
or Full adoption category
j = 1, 2, 3
i = 1, 2, 3, .. , 360
Xi = Vector of the predictor variables, and
j = Vector of the estimated parameters.
The multinomial logit model determines the effect
of independent variable on the probability that a farmer
will belong to one of the three categories, viz. nonadopter, partial adopter and full adopter. This model
was estimated by keeping the dependent variable 0 (i.e.
17
non-adopter) as the reference category. The e was

calculated, which gave the odds ratio (OR) associated
with change in the independent variables. The odds
mean the ratio of probability of happening to
probability of not happening of an event. The odds are
expressed as single number to the ratio to 1. The odds
of 2 associated with partial adoption, for example,
means that the likelihood of partially adopting an
innovation is twice that of not adopting. Zero-order
correlation matrix was obtained to ensure that multicollinearity did not pose any problem in estimating
parameters of the mathematical model. The variables
having higher multi-collinearity were dropped in the
final model to improve the values of the variables. Table
2 depicts the variables used in the model with their
expected signs.

Socio-personal, Economic and Psychological
Characteristics of Dairy Farmers
Table 3 reveals that the majority of dairy farmers
in the study area of all the four states, viz. Uttar Pradesh,
Uttarakhand, Haryana and Punjab, belonged to the
medium age group, ranging from 35 to 61 years and
the average age of dairy farmers was 48 years. The
majority of dairy farmers in pooled data have been
found illiterate, followed by those having education
up to high school. The majority (73%) of respondents
reported that they pursued both agriculture and animal
husbandry as their main source of livelihood. The next
major occupation was of agricultural labourers (12%).
Dairying as the main source of income was pursued
only by 5 per cent of the respondents. These findings
are in conformation with earlier studies reporting that
animal husbandry is mainly a smallholder phenomenon
and is followed predominantly as an adjunct to
agriculture under mixed farming system (Sone et al.,
2015; Singh et al., 2015; Bhoj et al., 2015).
The pooled data (Table 3) also indicated that the
majority of respondents had medium levels of familysize, landholding-size, herd-size, experience in dairying
and annual family income. The study also revealed that
the majority of farmers did not have any social
participation and had medium distance to a veterinary
institution or veterinary healthcare centre. It was
observed that the majority of the dairy farmers were in
medium category in terms of information-seeking
18
Vol. 29 (No.1)
January-June 2016
Table 2. Variables used in multinomial logit model and their expected signs
Variable
Definition and measurement
Expected
sign
Age
Age of respondent (in years)
Family size
Total number of members in a household
Land
Landholding size of household (in acres)
Livestock
Number of livestock owned by household in units
Distance
Average distance from veterinary institution or animal healthcare centre (in km)
Information sources
Information seeking behaviour of respondent from various sources on three point

continuum as frequent, less frequent and never
Decision
Decision-making ability of respondent related to dairy farming on three point

continuum as low, medium and high
Scientific
Scientific-orientation of respondent on three point continuum as low, medium

and high
Economic
Economic-orientation of respondent on three point continuum as low, medium

and high
Risk
Risk-orientation of respondent on three point continuum as low, medium and high
educnew=1
Illiteracy of respondent
educnew=2
Education up to 10 years of schooling
ocupnew=1
Agriculture as major occupation and animal husbandry as subsidiary occupation of

household
ocupnew=2
Animal husbandry as major occupation of household
partinew=1
No social participation of respondents
partinew=2
Respondent is member of at least one social organization
behaviour, decision-making ability, scientificorientation, economic-orientation and risk-orientation.

Also, there was a significant difference among the
respondents across the states with regards to all the
studied variables, except age and family size of the
respondents.
The one-way analysis of variance was used to
determine if there was significant difference between
states for socio-personal, economic and psychological
characters of dairy farmers. It was found that there was
a significant difference across states for all the studied
variables, except age. The higher values for familysize and distance from veterinary institution were
observed in Uttar Pradesh vis--vis other states. The
values for characteristics such as landholding-size,
livestock possession, decision-making ability,
scientific-orientation, economic-orientation and riskorientation were found higher in Punjab than in other
states (Table 4).
Adoption of Mobile Phone Use in Dairying

Across all the selected states, the majority (83.3%)
of respondents adopted mobile phones in dairying only
partially and only a small portion (13.0%) used mobile
phones to full extent (Table 5). Although all the farmers
possessed a mobile phone, they rarely used it for
accessing any information related to animal husbandry
or dairying. During the study period, it was also
observed that only a few respondents used mobile
phones for communicating with the universities /
institutes to gather scientific information. Further, very
negligible number of respondents received voice or text
messages from agricultural or veterinary institutions
through their mobile phones. The study also revealed
a significant difference (p <0.001) among the
respondents across the states with regards to use of
mobile phones in dairying which may be due to
variations in their socio-economic status, educational
level and information-seeking behaviour (Table 5).
19
Table 3. Socio-personal, economic and psychological characters of dairy farmers in four selected states
(N=360)
Variable
Category
Uttar Pradesh
Age (years)
Young
State
Uttarakhand Haryana
Pooled
Mean
S.D.
48.20
13.65
3.75
Punjab
22
(24.45)
48
(53.33)
20
(22.22)
68
(75.56)
21
(23.33)
52
(57.78)
17
(18.89)
61
(67.78)
20
(22.22)
57
(63.33)
13
(14.45)
70
(77.78)
19
(21.12)
49
(54.44)
22
(24.44)
64
(71.11)
82
(22.78)
206
(57.22)
72
(20.00)
263
(73.06)
03
(03.33)
05
(05.56)
01
(01.11)
13
(14.44)
0
(0)
06
(06.67)
71
(78.89)
13
(14.44)
09
(10.00)
26
(28.89)
52
(57.78)
03
(03.33)
02
(02..22)
01
(01.11)
05
(05.56)
03
(03.33)
19
(21.11)
01
(01.11)
09
(10.00)
73
(81.11)
08
(8.89)
17
(18.89)
16
(17.78)
50
(55.55)
07
(07.78)
01
(01.11)
06
(06.67)
05
(05.56)
01
(01.11)
08
(08.88)
0
(0)
15
(16.67)
64
(71.11)
11
(12.22)
02
(02.22)
18
(20.00)
62
(68.89)
08
(08.89)
03
(03.33)
09
(10.00)
08
(8.89)
06
(06.67)
03
(03.33)
0
(0)
05
(5.55)
77
(85.56)
08
(8.89)
03
(03.33)
08
(08. 89)
62
(68. 89)
17
(18.89)
01
(01.11)
19
(5.28)
23
(6.39)
11
(3.05)
43
(11.94)
01
(0.28)
35
(9.72)
285
(79.17)
40
(11.11)
31
(08.61)
68
(18.89)
226
(62.78)
35
(09.72)
07
(01.94)
78
( 86.67)
High
10
(11.11)
Social participation Nil
48
(53.34)
One organization
36
(40.00)
86
(95.56)
03
(0 3.33)
58
(64.44)
26
(28.89)
75
(83.34)
12
(13.33)
56
(62.23)
31
(34.44 )
64
(71.11)
25
(27.78)
16
(17.78 )
62
(68.89)
303
(84.17)
50
(13.89)
178
(49.44)
155
(43.06)
Middle
Old
Major occupation
Agriculture +
Animal
husbandry
Animal
husbandry
Business
Government
service
Labour
Any other
Family size (No.)
Small
Medium
Large
Landholding size
Landless
Small
Medium
High
Livestock
possession
(livestock units)
Low
Medium
31.8**
8.3
4.12
9.98
4.83
5.87
42.2**
4.59
3.55
25.0**
63.4**
Contd...
20
Vol. 29 (No.1)
January-June 2016
Table 3. Socio-personal, economic and psychological characters of dairy farmers in four selected states Contd
(N=360)
Variable
Category
Uttar Pradesh
Two or more
organization
Public leader
Distance from
veterinary
institution
Small
Medium
Long
Information
seeking behaviour
Low
Medium
High
Decision-making
ability
Low
Medium
High
Scientificorientation
Low
Medium
High
Economicorientation
Low
Medium
High
Risk-orientation
Low
Medium
High
State
Uttarakhand Haryana
Pooled
Punjab
03
(3.33)
03
(3.33)
0
(0)
06
(6.67)
0
(0)
0
(0)
03
(3.33)
0
(0)
30
( 33.33)
12
(13.33)
0
(0)
45
(50.0)
24
(06.67)
3
(0.83)
75
(20.83)
60
(66.67)
30
(33.33)
31
(34.45)
55
(61.11)
04
(4.44)
33
(36.67)
51
(56.67)
6
(6.66)
51
(56.67)
38
(42.22)
1
(01.11)
30
(33.33)
56
(62.22)
4
(4.45)
26
(28.89)
54
(60.0)
10
(11.11)
90
(100.0 )
0
(0)
24
(26.67)
58
(64.44)
08
(8.89)
20
(22.22)
59
(65.56)
11
(12.22)
16
(17.78)
55
(61.11)
19
(21.11)
29
(32.22)
56
(62.22)
5
(5.56)
31
(34.45)
56
(62.22)
03
(3.33)
60
(66.67)
0
(0)
12
(13.33)
65
(72.22)
13
(14.45)
18
(20.0)
56
(62.22)
16
(17.78)
06
(6.67)
55
(61.11)
29
(32.22)
18
(20.0)
46
(51.11)
26
(28.89)
22
(24.44)
44
(48.89)
24
(26.67)
45
(50.0)
0
(0)
02
(2.22)
59
(65.56)
29
(32.22)
12
(13.33)
62
(68.89)
16
(17.78)
0
(0)
49
(54.44)
41
(45.56)
0
(0 )
48
(53.33)
42
(46.67)
0
(0)
53
(58.89)
37
(41.11)
255
(70.84)
30
(8.33)
69
(19.16)
237
(65.84)
54
(15.0)
83
(23.06)
228
(63.33)
49
(13.61)
73
(20.28)
197
( 54.72)
90
(25.0)
77
(21.39)
206
(57.22)
77
(21.39)
79
(21.94)
207
(57.50)
74
(20.56)
Note: Figures within the parentheses indicate percentage in each variable
Mean
S.D.
4.27
5.05
187.7**
19.37
1.92
56.3**
21.59
2.55
18.0**
14.22 127.6**
1.73
12.94
1.72
83.7**
11.52
1.28
67.0**
21
Table 4. Analysis of variance for socio-personal, economic and psychological characteristics of dairy farmers in
selected four states
(N=360)
Characteristics
Uttar Pradesh
Mean SD
Uttarakhand
Mean SD
Haryana
Mean SD
Punjab
Mean SD
F-statistic
Age
Family-size
Landholding- size
Livestock possession
Distance from veterinary institution
Decision-making ability
Scientific-orientation
Economic- orientation
Risk-orientation
48.17 14.34
9.40 5.78a
3.46 4.07b
4.25 4.11b
9.00 7.48a
20.57 2.86c
12.54 1.35d
12.06 1.22c
11.13 1.08c
48.43 13.50
7.63 3.09b
3.83 4.75b
3.71 1.93b
4.33 2.15b
21.49 2.29b
14.11 1.60c
12.18 1.35c
10.89 0.88c
47.93 12.67
8.04 3.88b
4.61 4.65b
4.30 3.22b
2.59 2.36c
21.88 2.53ab
14.79 1.46b
13.20 1.92b
11.70 1.53b
48.26 14.27
8.13 2.96b
7.42 8.30a
6.14 4.07a
1.17 1.35d
22.41 2.19a
15.43 0.97a
14.34 1.24a
12.36 1.05a
0.02NS
3.14*
8.89**
8.50**
61.51**
8.86**
74.20**
47.91**
28.40**
Notes: The means bearing different superscripts (a, b, c and d) differ significantly in the same row.
* and ** depict significance at 5 per cent and 1 per cent levels, respectively
Table 5. Adoption status of mobile phones at field conditions (N=360)
Adoption category
Uttar Pradesh
Non-adoption
Partial adoption
Full adoption
07 (7.78)
79 (87.78)
04 (4.44 )
State
Uttarakhand
Haryana
05 (5.55)
77 (85.56)
08 (8.89 )
01 (1.11 )
75 (83.33)
14 (15.56 )
Pooled
13 (3.61)
300 (83.33)
47 (13.06)
25.1**
Punjab
0 (0)
69 (76.67)
21(23.33)
Notes: Figures within the parentheses indicate percentage of the state; No respondent in all the four states belonged to the
Discontinued adopter category.
Inigo et al. (2014) have reported that dairy farmers in

Tamil Nadu used mobile phones for seeking
information related to animal husbandry and dairying,
while Mittal and Tripathi (2009) have revealed that
mobile phones were being used primarily for social
purposes and occasionally for agricultural activities.
The study has also pointed out that information
regarding seeds was the most frequently accessed
information, followed by the market price of various
agro-commodities.
Adoption Period of Mobile Phone in Dairying
An attempt was also made to categorize the mobile
phone users according the number of years they have
been using this tool in dairying and the results are
presented in Table 6. The majority of respondents
adopted the innovation during 3-6 years (59.4%) prior
to the date of survey, followed by adoption during past
0-3 years (25.2%) and 6-9 years (10.2%). The mobile
phones were normally used to contact the veterinarians

for treatment of animals, while few farmers in Haryana
and Punjab received text or voice messages from
universities or institutions. This variation in the level
of mobile phone adoption might be due to variations
in their educational level and economic status and
information access in the study areas leading to a
significant difference (p <0.001) among respondents
across the states in the use of mobile phones in dairying
(Table 6). Inigo et al. (2014) have reported that dairy
farmers in Tamil Nadu used mobile phones for seeking
information related to animal husbandry and dairying,
but faced the constraints of language and network
coverage in effective utilization of mobile phoneenabled services.
Reasons for Adoption of Mobile Phone in Dairying
The reasons for adoption of mobile phone in
dairying in the study area were identified during group
22
Vol. 29 (No.1)
January-June 2016
Table 6. Adoption period of mobile phone in dairying (N=360)

Adoption period category
Uttar Pradesh
Non-adoption
0-3 years
3-6 years
6-9 years
More than 9 years
07
(7.78)
30
(33.33)
51
(56.66)
02
(2.22)
0
(0 )
State
Uttarakhand
Haryana
05
( 5.55)
21
( 23.34)
54
( 60.0)
08
( 8.89)
02
( 2.22)
01
(1.11)
22
(24.44)
56
(62.23)
10
(11.11)
01
(1.11)
Pooled
Punjab
0
( 0)
18
(20.0)
53
(58.89)
17
(18.89)
02
(2.22)
13
(3.61)
91
(25.27)
214
(59.44)
37
(10.28)
05
(1.39 )
32.4**
Note: Figures within the parentheses indicate percentage in a state
discussions and are enlisted below.
as perceived by respondents, are listed below.
Easy to operate and handle

Can receive credible information through text or
voice message
Competitive price for maintenance of a mobile
phone
Compatibility to receive text messages in the local
language
Experts can be contacted as per convenience
Veterinarian or expert can be approached in
emergencies
The retailers/input dealers can be contacted
Mobile phones use by default for dairy or animal
husbandry activities
Illiteracy/ low educational level

Poor and disrupted power supply
Operational difficulties
High maintenance cost
Lack of mobile phone compatibility to receive
SMS in local language
Lack of immediate interaction for better
understanding
Human element is missing
Lack of practical exposure
According to Jensen (2007), introduction of mobile

phones to Kerala fishermen decreased price dispersion
and wastages by facilitating the spread of information
which could make the markets more efficient by
reducing risk and uncertainty. Wankhade et al. (2009)
and Naik et al. (2012) have also reported that increased
penetration of mobile phones and mobile phoneenabled information services to the farming community
reduced the information asymmetry and could
complement the role of extension services as it saved
time, effort and cut steps of extension process.
With regard to the use of mobile phones in dairying,

Mittal and Tripathi (2009) have pointed out that
although mobile phone act as a catalyst to improve
farm productivity and rural incomes, the quality,
timeliness and trustworthiness of information were the
three important aspects that required special focus to
meet the needs and expectations of the farmers. Inigo
et al. (2014) have reported that dairy farmers in Tamil
Nadu faced the constraints of language and network
coverage in effective utilization of mobile phoneenabled services in animal husbandry and dairying.
Sharma and Arya (2005) have found that high cost,
domination of English language, and complexity of
information were the major constraints of mobile
phone-use in animal husbandry and dairying.
Reasons for Different Adoption Levels of Mobile

Phone in Dairy Sector
Identification of Factors Influencing Mobile Phone

Use in Dairying
The reasons for partial adoption/ discontinuation

and non-adoption of mobile phones in the dairy sector,
The factors that significantly influenced the

respondents likelihood of belonging to any one of
adopter categories of mobile phones in dairying (viz.

non-adopter, partial adopter and full adopter), were
identified using the multinomial logit model and the
results of this analysis are presented in Table 7. The
23
Chi-square value of 65.8 showed that likelihood ratio

statistic are highly significant (p < 0.001), suggesting
that model was a good fit for explanation. The variables
significantly associated with the probability that the
Table 7. Identification of factors influencing mobile phone-use in dairying using multinomial logit model
N = 360
Variable
(S.E)
Intercept
Age
Family size
Landholding- size
Herd-size
Distance
Information sources
Decision-making ability
Scientific-orientation
Economic-orientation
Risk-orientation
Educational level 1
Educational level 2
Main occupation-1
Main occupation-2
partinew=1
partinew=2
Wald chi2
Nagelkerke R2
log pseudo likelihood
25.189
(1692.45)
-0.003
(0.026)
-0.155
(0.072)
0.28
(0.197)
0.096
(0.131)
-0.019
(0.059)
0.065
(0.29)
-0.197
(0.16)
0.408
(0.272)
0.24
(0.314)
-0.168
(0.432)
-12.107
(664.114)
-10.55
(664.115)
1.291
(0.74)
0.201
(1.518)
-13.947
(1556.699)
-13.936
(1556.698)
Partial adoption
Wald
Odds ratio
0
0.015
0.997
4.613
0.856
2.007
1.323
0.535
1.101
0.106
0.981
0.051
1.067
1.513
0.821
2.26
1.504
0.584
1.271
0.152
0.845
5.52E-06
2.62E-05
3.042
3.638
0.018
1.223
8.77E07
8.87E07
(S.E)
7.012
(664.161)
-0.011
(0.029)
-0.091
(0.079)
0.277*
(0.199)
0.133
(0.138)
-0.123*
(0.078)
0.302
(0.305)
-0.217
(0.175)
0.395*
(0.304)
0.176
(0.34)
-0.178
(0.459)
-12.175
(664.115)
-9.485
(664.115)
0.903
(0.841)
0.552
(1.618)
-1.077
(0.798)
-1.123
(0)
Full adoption
Wald
Odds ratio
0
0.139
0.989
1.331
0.913
1.93
1.319
0.926
1.142
2.496
0.884
0.983
1.352
1.542
0.805
1.696
1.485
0.268
1.193
0.15
0.837
5.16E-06
7.60E-05
1.151
2.466
0.116
1.737
1.821
0.341
0.325
65.8
0.25
763.65
Notes: Figures within the parentheses indicate standard errors

*, ** and *** denote significance at 10 per cent, 5 per cent and 1 per cent levels, respectively
24
respondent will be a full adopter of mobile phone in

dairying were: distance to veterinary institution or
animal healthcare centre (P <0.10), landholding-size
(P < 0.10) and scientific-orientation (P < 0.10). The
sign of regression coefficient for the variable distance
to veterinary institution was negative while for other
two variables, viz. landholding-size and scientific
orientation, was positive. The landholding-size was
positively associated with full adoption of mobile
phone, implying that with increase in landholding-size,
the farmers moved towards increased adoption level
of mobile phones and continued to be in full adoption
category. The probability of full adoption increased
by 31.9 per cent with an increase of one unit in
landholding-size. This indicated that farmers would be
more interested to use mobile phones in dairying if
their landholding-size increases.
The negative sign associated with the variable
distance to veterinary institution in case of full
adoption indicated that as the distance to a veterinary
institution increased, the probability of full adoption
of mobile phone in dairying decreased and that of
partial and non-adoption increased. The longer distance
to veterinary institution implied lesser chance of contact
with livestock extension functionaries, which probably
explained the negative association of distance from
veterinary institution with higher degree of adoption
of this innovation.
The odds ratio associated with the variable
distance to veterinary institution suggested that with
one kilometre increase in distance to veterinary
institution, the likelihood of full adoption of mobile
phone decreased by 88.4 per cent. The scientificorientation of the farmers was positively associated
with full adoption, implying that with increase in
scientific-orientation, the farmers moved towards
increased adoption level and continued to be in full
adoption category. The probability of full adoption
increased by 48.5 per cent with increase of one unit in
scientific-orientation. The earlier studies have reported
that the number of contacts with extension officers, as
a proxy measure for access to agricultural information,
positively contributed to awareness generation and
subsequent adoption of new technologies (Adesina et
al., 2000; Yirga, 2007). This factor indicated the fact
that poor information sources and less contact with
extension sources led to poor adoption of innovations.
Vol. 29 (No.1)
January-June 2016
Farmers Suggestions for Effective Use of Mobile

phones in Dairying
The dairy farmers had various problems with the
existing innovations and hence, they demanded support
from different stakeholders like research and extension
institutes, government departments, etc. Following are
the suggestions of dairy farmers for effective use of
mobile phones in the dairy sector.
Regular supply of power in villages
Educating the farmers for mobile phone operation
Low/nominal maintenance cost
Evolving a process for a direct contact with the

experts
Mobile phone compatibility to receive messages

in local language
Immediate interaction for better understanding

must be made more effective
Engage educated rural youths to provide practical

exposure and training on mobile phone use
Conclusions
The study has revealed that the majority of
respondents use mobile phones only partially in
dairying and a small portion has their full adoption in
the study area. The majority of respondents have
adopted mobile phones in dairying during the past 3-6
years, followed by since 0-3 years. The variables
distance to veterinary institution, animal healthcare
centre, landholding-size and scientific-orientation
have been found significantly associated with the
probability that the respondent will be a full adopter
of mobile phone in dairying. Since the farmers faced
various problems leading to lesser use of mobile phones
in dairying, there is a need to address them for
increasing the use of mobile phones in dairying. The
study has also recommended that effective use of
mobile phones may be popularised for dissemination
of livestock-related information in general and dairying
in particular in India to the farming community.
Acknowledgements
The authors sincerely thank the Director, ICARIVRI, Izatnagar, for providing the necessary facilities
for conducting this research work. They are grateful
to all the respondents for sharing their valuable views

during the study. They thank the anonymous referee
for helpful suggestions.
References
Adesina, A.A., Mabila, D., Nakamleu, G.B. and Endamana,
D. (2000) Econometric analysis of the determinants of
adoption of alley farming by farmers in the forest zone
of Southwest Cameroon. Agriculture, Ecosystems and
Environment, 80: 255-265.
Babu, V. and Ashokan (2011) Scenario of information and
communication technology among the dairy farmers
of Puducherry. In: Proceedings of International
Conference on Innovative Approaches for Agricultural
Knowledge Management: Global Extension
Experiences. New Delhi. pp. 352 353.
Bhoj, S., Bardhan, D. and Kumar, A. (2013) Determinants
and implications of rural womens participation in
microfinance programme: An analysis of dairy self-help
groups in Uttarakhand state of India. Livestock
Research for Rural Development, 25 (10).
Article #185. (http://www.lrrd.org/lrrd25/10/
bhoj25185.htm)
Gakuru, M., Winters, K. and Stepman, F. (2009) Innovative
farmer advisory services using ICT. In: W3C Workshop
on Africa Perspective on the Role of Mobile
Technologies in Fostering Social Development.
Maputo, Mozambique. 1-2 April.
GoI (Government of India) (2015) mKisan. A portal of
Government of India for farmer-centric mobile-based
services. http://mkisan.gov.in/dashboard.aspx.
Accessed on 21-04-2015.
Hayrol Azril, M.S., Md. Salleh, H. and Inon, B. (2009) Level
of mass media usage (television, radio and newspaper)
among Malaysian agro-based entrepreneurs. Journal
Communication Knowledge Communique, 1: 28-38
Inigo, G.A.V., Chander, M., Thakur, D. and Bardhan, D.
(2014) Potential of mobile telephony in utilization of
dairy related information: Empirical evidence from
Tamil Nadu. Indian Journal of Dairy Science, 67 (1):
91-96.
Jensen, R. (2007) Digital provide: Information technology,
market performance and welfare in the South Indian
fisheries sector. The Quarterly Journal of Economics,
122 (3): 879-924.
Katke, K. and Padmalatha, N.A. (2012) Information need
of Indian farmer and role of mobile enabled agricultural
services. In: Proceedings of 4 th Conference on
25
Excellence in Research and Education. Indian Institute

of Management, Indore.
Mishra, N. (2010) Livestock Information Dissemination
among Farmers through ICT: An Exploratory Study
in Tarai Area of Uttarakhand. M.V.Sc Thesis.
G.B. Pant Agricultural University, Pant Nagar,
Uttarakhand.
Mittal, S. and Tripathi, G. (2009) Role of mobile phone
technology in improving small farm productivity.
Agricultural Economic Research Review, 22
(Conference No.): 451-459.
Mittal, S. Gandhi, S. and Tripathi, G. (2010) Socio-economic
Impact of Mobile Phones on Indian Agriculture.
Working Paper No. 246. Indian Council for Research
on International Economic Relations, New Delhi.
Naik, V.R., Padaria, Chandrashekar, N.R. and Naresh Babu,
N. (2012) Mobile A catalyst in the transfer of
agriculture technology. In: Proceedings of M4D 2012,
Eds: Vikas Kumar and Jakob Svensson. pp-151.
Pundo, M.O. and Fraser, G.C.G. (2006) Multinomial logit
analysis of household cooking fuel choice in rural
Kenya: The case of Kisumu district. Agrekon, 45 (1):
24-37.
Rathod, P. and Chander, M. (2014) Identification of socioeconomically important dairy innovations in India: A
perspective of scientists. In: Proceedings of the First
International Conference of the Asia and Pacific Islands
Rural Advisory Services (APIRAS) and the Fifth
Congress of Extension and Education in Agriculture
and Natural Resources Management: Facilitating
Information and Innovations for Empowering Family
Farmers, Ed: Esmail Karamidehkordi. University of
Zanjan, Iran. 101 p.
Sharma, G.R.K. and Arya, H.P.S. (2005) Constraints in
utilization of internet communication among livestock
farmers. Indian Veterinary Journal, 82 (5): 561.
Sife, A.S., Kiondo, E. and Macha, J.G.L. (2010) Contribution
of mobile phones to rural livelihoods and poverty
reduction in Morogoro region, Tanzania. The Electronic
Journal on Information Systems in Developing
Countries, 42(3): 1-15.
Singh, P., Bardhan, D. and Tripathi, S.C. (2015) Constraints
faced by members and non-members of dairy
cooperative societies in using modern ICT tools: A
comparative analysis of plain and hill regions of Nainital
district of Uttarakhand. Journal of Rural Development,
34 (2): 149-166.
26
Sone, P., Bardhan, D. and Kumar, Avadhesh (2015) Role of

goats in livelihoods of rural poor in Uttarakhand hills
of India: An analysis with special reference to
multivariate typology of households based on farm and
socio-economic characteristics. Indian Journal of
Animal Sciences, 85 (8): 913-919.
Wankhade, A., Khare, N., Pannase, S. and Shrivastava, D.
(2009) ICT based Kisan Mobile Sandesh
An innovative approach. In: Proceedings of
Vol. 29 (No.1)
January-June 2016
National Seminar on ICT for Agriculture and Rural

Development. Pasighat, Arunachal Pradesh.
Yirga, C.T. (2007) The Dynamics of Soil Degradation and
Incentives for Optimal Management in Central
Highlands of Ethiopia. Ph.D. thesis. Department of
Agricultural Economics, Extension and Rural
Development, University of Pretoria, South Africa.
Received: July, 2015; Accepted: January, 2016

DOI: 10.5958/0974-0279.2016.00016.1
TFP Growth of Wheat and Paddy in Post-Green Revolution Era

in India: Parametric and Non-Parametric Analysis
Surya Bhushan
Development Management Institute, Patna-800 004, Bihar
Abstract
This paper has applied the parametric and non-parametric approaches to estimate the Total Factor
Productivity (TFP) growth for major wheat and paddy crop producing states in India for the period
19812010. Both the approaches have revealed that the shift due to technological adoption is a vital
source for overall productivity growth. Both the approaches used have produced almost similar results at
spatial and temporal directions, showing robustness in TFP estimation. Further, the high-yield states,
particularly Punjab, have depicted a decline in the TFP growth for wheat and paddy crops in recent times,
which raises an alarm on the long-term sustainability of paddy-wheat production system in this Green
Revolution star state. An obvious extension to this study would be the application of this approach to
incorporate more crops and states or at the district level. Another interesting work could be incorporating
higher order policy variables such as subsidies, government investment, variables representing resource
endowment, infrastructure, groundwater extraction, etc. in the efficiency equation of SFA.
Key words: TFP, Malmquist, DEA, SFA, India
JEL Classification: O13, O47, D24, Q10, Q18
Introduction
The agricultural productivity growth is most
significant among the key development challenges
before Indias economy, especially to such concerns
as food availability and rural poverty since the early1990s. Given the binding of land constraint, agricultural
growth in India depends on making land (for crops)
more productive. The TFP growth in agriculture
increases income for the rural communities, which
promotes their spending on the non-farm sector (Ellis,
2000; Himanshu et al., 2011). More specifically, it is
likely to lead the rural farm communities to support
Email: surya.bhushan@gmail.com
This paper is drawn from my doctoral dissertation entitled,
Agricultural Productivity and its Environmental Impacts in
India: A Parametric and Non-Parametric Analysis, awarded
by Jawaharlal Nehru University, New Delhi in 2015.
non-farm commodities and services such as consumer

goods and services, inputs and services to agricultural
production, and processing and marketing services.
This paper has looked into the major trends and
factors of change in the total factor productivity (TFP)
growth of the two dominant crops, viz. wheat and paddy
in the Indian agricultural landscape. Using the panel
data of Indian states for the post-Green Revolution
period 1981-2010, we have estimated parametric
stochastic frontier analysis (SFA) (Aigner et al., 1977)
and non-parametric data envelopment analysis (DEA)
(Charnes et al., 1978), and have compared the results
on the TFP measures obtained from both approaches.
The rationale for using two competing approaches was
to countercheck whether results obtained by one could
be confirmed by the other.
By estimating input or output distance functions,
non-parametric DEA can be used to decompose TFP
28
growth into movements of the frontier (technical gains

or innovation) and movements toward the frontier
(efficiency gains or catching up) using the Malmquist
index (see Coelli and Rao, 2005 for more details). The
merit of SFA, on the other hand, is that it considers
stochastic noise in data (e.g. labour or capital
performance variations) and also allows for the
statistical testing of hypotheses concerning production
structure and degree of inefficiency. Its main
limitations, however, are that it requires an explicit
imposition of a particular parametric functional form
representing the underlying technology and also an
explicit distributional assumption for the inefficiency
terms. Contextually, the parametric SFA is likely to be
more appealing than the non-parametric DEA in the
cases where data suffer from serious measurement
errors, random events and difficulty in identifying
inputs and outputs, and DEA may be better choice when
random disturbances are less of an issue, and price
information is not available (Fre et al. 1994).
Productivity Measurement in Indian

Agriculture
Indian agriculture has witnessed tremendous
growth during the past several decades with the
Vol. 29 (No.1)
January-June 2016
adoption of Green Revolution technology during late1960s. The sources and effects of growth have been of
considerable interest to researchers, policy makers, and
practitioners. With the availability of micro-level farm
data, particularly of CACP, in India, few crop-specific
TFP studies have been conducted since early-1990s
(Sindhu and Byerlee, 1992; Kumar and Mruthyunjaya,
1992; Kumar and Rosegrant, 1994; Kumar et al., 1998;
Kumar, 2001; Joshi et al., 2003). However, the Indias
agricultural productivity estimates lack unanimity
(Table 1). This may be due to differences in the
estimation methods and the data used, which has
resulted into generating debates on the trend of Indias
agricultural productivity.
As can be seen from the Table 1, depending on the
methodology and level of aggregation, the results vary
substantially across studies, so careful scrutiny is
required to reconcile and interpret the findings. The
majority of studies have used Trnqvist Index to
estimate total factor productivity (TFP) growth, with
exceptions who used DEA based Malmquist Index. It
is also interesting to note that the three Trnqvist
approaches using FAO data from Table 1 examine the
same time periods but generate widely different TFP
growth estimates. Coelli and Raos (2005) average
Table 1. A brief summary of Indian agricultural TFP studies

Study
Data
sources
Sectors
Method
Period
Rosegrant and Evenson (1992)

Evenson et al. (1999)
Mukherjee and Kuroda (2003)
Joshi et al. (2003)
Indian
Indian
Indian
Indian
Crops (15)
Crops (18)
All agriculture
Rice
Trnqvist
Trnqvist
Trnqvist
Trnqvist
Malmquist
Trnqvist
Malmquist
Trnqvist
1975-1985
1977-1987
1973-2003
1980-1990
1990-1999
1980-1990
1990-1999
1980-2000
1980-2000
1981-2000
1971-2000
Malmquist
Trnqvist
Malmquist
Trnqvist
Malmquist
1981-2005
1981-2000
1980-2000
1985-2006
1980-2000
Wheat
Coelli and Rao (2005)
FAO
All agriculture
Bhushan (2005)
Kumar and Mittal (2006)
Indian
Indian
Bhushan (2009)
Avila and Evenson (2010)
Nin-Pratt et al. (2010)
Chand et al. (2011)
Fuglie (2012)
Indian
FAO
FAO
Indian
FAO
Wheat
Paddy
Wheat
Wheat
All agriculture
All agriculture
Crops & livestock
All agriculture
TFP growth
(%)
0.98
1.05
2.19
3.50
2.08
2.44
2.14
1.40
0.90
1.08
0.68
1.10
2.41
0.69
0.53
1.39
Bhushan : TFP Growth of Wheat and Paddy in Post-Green Revolution Era in India
annual Trnqvist estimate of 0.90 per cent is less than

Fuglies (2012) estimate of 1.39 per cent per annum
TFP growth, which itself is significantly lower than
Avila and Evensons (2010) estimate of 2.41 per cent
per annum. Coelli and Rao (2005) and Nin-Pratt et al.
(2010) have estimated TFP using Malmquist indexes.
But, their results differ, for instance, Nin-Pratt et al.s
estimate indicate the growth to be half of what Coelli
and Rao estimate. However, as both Malmquist
estimates were extracted from broader, global analyses,
individual country estimates may be affected by the
dimensionality issue or the number of commodities and
countries (Lusigi and Thirtle, 1997). Moreover, Coelli
and Rao (2005) note that if shadow prices are indeed
correctly estimated, for many countries the estimates
may significantly differ from the sample average due
to country-specific factor abundance or scarcity. These
two issues may be why the TFP growth estimates from
Coelli and Raos differ by method, their Trnqvist index
indicating slower TFP growth than does their
Malmquist index.
The differing years and crops included make direct
comparisons of TFP growth difficult for the Indian data
based studies. Recently, Ray and Ghose (2010) have
used DEA approach to obtain Pareto-Koopmans
measures of technical efficiency of agricultural
production of states in India over the period 1970-2001.
The output technical efficiency at India level averaged
over all the years, was about 85.5 per cent of the Pareto
optimal level. Similarly, the input-oriented technical
efficiency was 0.86, implying that about 14 per cent
reduction in the average level of inputs would be
possible. However, the paper did not say anything about
the other key component of the TFP growth, the
technical progress.
A few studies have also attempted to measure TFP
growth on a select group of crops or livestock products,
e.g., by Fan et al. (1999) and Bhushan (2005; 2009).
These studies have evaluated TFP growth at the state
level, which is important for accounting for the entire
agricultural sector and allows for a broader
representation of growth. The present paper is an
extension to Bhushans (2005, 2009) studies in terms
of time period, coverage of crops, and methodological
expansion in estimating TFP growth of paddy and
wheat in India.
29
Data and Methodology

Data Sources
For the estimation of TFP growth at the crop and
state levels, time-series data were collected for the
period 1981-82 to 2011-12 from the Comprehensive
Scheme for Studying Cost of Cultivation of Principal
Crops brought out by the CACP, Directorate of
Economics and Statistics, Ministry of Agriculture,
Government of India. The detailed features of CACP
dataset are described in Bhushan (2015). The missing
year data on inputs-use and yields per hectare were
predicted using the interpolation based on the
polynomial trends available in the data. The quantity
data were given priority over price data, wherever both
were available to avoid the anomalies in price
information: Output (in kg), human labour (in hours),
animal labour (pair hours), chemical inputs including
fertiliser and manure (in kg) and machine labour per
hectare were computed in hours. All the data were
transformed by taking centered moving averages of
three periods to avoid any short-term weather
fluctuations, particularly rainfall effect.
Table 2 presents the data summary of the major
variables used in estimating the TFP growth of wheat
and paddy from 1980-81 to 2010-11. The selected five
states, viz. Haryana, Madhya Pradesh, Punjab,
Rajasthan and Uttar Pradesh, constituted around 80 per
cent of the area and 85 per cent of the output of wheat
produced in the country. Yield-wise, Punjab and
Haryana are high-productive states, while Rajasthan
and Uttar Pradesh fall under middle-productive states
and Madhya Pradesh is in the low-producing category.
One important and obvious attribute of high wheat crop
yield states is higher use of material inputs like,
chemical fertilisers, machine labour and intensity of
irrigation in comparison to traditional inputs like human
labour and animal labour. The paddy data summary
also tells a similar story, except for the case of Andhra
Pradesh and Karnataka, where the use of machine,
human, and animal labour is substantial. Punjab and
Haryana present a familiar picture of high machine
combined with low human and animal labour in paddy
production.
Methodological Skeleton
The sources of productivity growth of agriculture
can be split into two major components (Nishimizu
and Page, 1982):
30
Vol. 29 (No.1)
January-June 2016
Table 2. Summary of mean of inputs used per hectare in wheat and paddy crops in the selected states of India, 19812010s
State
Yield
(kg)
Chemical
fertilisers
(kg)
Punjab
Haryana
Rajasthan
Uttar Pradesh
Madhya Pradesh
3,879
3,663
2,955
2,908
1,768
205
170
85
133
71
Punjab
Andhra Pradesh
Karnataka
Haryana
West Bengal
Uttar Pradesh
Odisha
Bihar
Assam
Madhya Pradesh
5,640
4,571
4,059
3,911
3,214
3,042
2,608
2,149
2,191
1,700
189
179
178
175
82
100
62
62
6
49
Human
labour
(hours)
Wheat
309
342
557
524
361
Paddy
570
1,080
1,104
662.5
1,192
832
1,058
863
686
622
Machine
labour@
(hours)
Animal labour
(Pair hours)
Per cent
irrigated area$
516
485
314
378
189
9
25
57
65
72
96.6
97.9
93.8
92.3
62.1
442
286
208
338
71
170
34
85
22
51
10
79
123
13
160
55
227
127
256
126
99.2
95.5
63.2
99.4
37.5
58
37.6
43.0
21.8
18.8
Note: @For machine labour data, the cost of maintenance of farm machinery which included diesel, electricity, lubricants,
depreciation, repairs and other expenses, if any, was used. The machine labour was further indexed to machine labour input
price indices
$
The per cent irrigated area was calculated as the ratio of total irrigated wheat (paddy) area to total wheat (paddy) sown area
(a) The efficiency gains, i.e. the growth in the factor

of production, indicating the movement along the
best practice production frontier, and
(b) The technological gains, i.e. shifting of the
production frontier outward (inward) in case of
technological progress (regress).
Based upon Farrells (1957) original idea on
technical efficiency, later studies are extended to focus
on the methods of estimation of production functions
(Afriat, 1972; Aigner et al., 1977). The difference
between actual production level of a firm and the
frontier measures its technical inefficiency. The frontier
can be fixed or stochastic and the estimation
methodology can take a parametric or non-parametric
approach. Thus, both parametric and non-parametric
approaches differ in the assumptions they make
regarding the shape of the efficient frontier and the
existence of random error.
Non-Parametric DEA Model: Malmquist Index

Introduced by Caves et al. (1982) in its empirical
usage, the Malmquist Index is constructed by
measuring the radial distance of the observed output
and input vectors in periods t and t+1 relative to two
reference technologies: technology in the period t and
technology in the period t+1. In this paper, we have
measured TFP growth using the Malmquist index
method described in Fre et al. (1994) and Coelli et
al. (2005, Ch. 11). We used the following model
specified by Fre et al. (1994):
MOC (xt, yt, xt+1, yt+1) = E (xt+1, yt+1, xt, yt) T (xt+1,
yt+1, xt, yt)
(1)
where, E (.) refers to the relative efficiency change
under the constant returns to scale (CRS). This
measures the catching-up to the best practice frontier
for each observation between two time periods t and
t+1, and T(.) represents the technical change that
measures the shift in the frontier of technology (or

innovation) between two time periods evaluated at xt
and xt+1. We have used the software DEAP (version
2.1) developed by Tim Coelli (1996 b) to estimate
efficiency and productivity indices.
The SFA addresses technical efficiency and

recognizes the fact that random shocks (as labour or
capital performance) beyond the control of producers
may affect the production output. These models were
simultaneously introduced by Aigner et al., (1977) and
Meeusen and van den Broeck (1977). We have used
the software FRONTIER (4.1) developed by Tim Coelli
(1996 a).
In this paper, we have used translog stochastic
frontier production 1 model with time-varying
inefficiency in panel data as:
ln Yit = 0 + jjlnxijt + jjj(lnxijt)2 + jkjklnxijt
lnxikt + jjtlnxijtt + t + tt2 + vit uit
(2)
where, ln denotes the natural logarithm, Yit indexes the
wheat and paddy productivity per hectare for the ith
state at time t, xjk, denotes various input variables; and
0, j, jj, jk, jt, t, tt are the unknown parameters to
be estimated. The introduction of time trend, t,
interacted with input variables allows for non-neutral
technical change in the model.
To account for the unobserved, non-time varying
factors (or fixed effects), we included a set of input
variables used in production in the specification. Due
to the interest in linking the environmental
sustainability and technical efficiency in production,
we followed empirical development in the technical
efficiency analysis literature by Kumbhakar et al.
(1991) and Huang and Lui (1994). The technical
inefficiency function, uit that varies across the states
can be written as:
(3)
where, it is the random error-term, distributed as

N(0,2), xit, input variables (chemical fertilisers,
1
machine labour and irrigated area in the current model),

s are the parameters of input variables to be estimated.
The technical efficiency measures,
TEit=E[exp(-uit)|eit],
where (eit=vit-uit), can be used to calculate the efficiency
change component.
Parametric SFA Model: Malmquist Index
uit = 0 + ilnxit + it
31
Now with the help of above technical efficiency

measures, the efficiency change can be estimated using
following equations (Coelli et al., 2005: 301):
(4)
(5)
and
Malmquist TFP =
(Efficiency Change) (Technical Change) (6)

Wheat
Before examining the parameter estimates of the
production frontier and the factors that affect
inefficiency of the farmers, we investigated the validity
of the model used for analysis. Then we performed the
joint tests using the likelihood ratio (LR) tests. The
null hypotheses related to three tests of the production
specifications. The results are presented in Table 3.
The first null hypothesis tests whether the CobbDouglas production functions is adequate to explain
the underlying technology of wheat production in the
selected Indian states. The second hypothesis tests
whether there is no technical progress effect. The third
hypothesis tests whether technical change is neutral
and the fourth hypothesis tests whether technical
efficiency is neutral. The test results showed that all
three null hypotheses could be rejected, thus indicating
that a trans-log production technology was accepted
and was applicable here. The subsequent Table 4
presents the SFA estimates of translog production
function used.
The above specification is somewhat restrictive. It has, though, following attractive features:
The input elasticities vary over time to capture changes in the production structure.
The specification of technological change is general in the sense that it allows one to test whether technical change is biased
towards particular inputs.
Technical (in)efficiency is allowed to vary over time.
32
Vol. 29 (No.1)
January-June 2016
Table 3. Statistics for tests of hypothesis: SFA for wheat

Null hypothesis
Likelihood
function
LR test
statistics
Critical value
at 1% level#
Decision
Given pooled sample
307.027
Data can be explained by Cobb-Douglass

production specification.
H0: 6=...=20=22=0
251.637
110.781
20.001,16 = 38.566
Reject H0
There is no technical progress effect.

H0: 21=22=23= 24=25=26=27=0
269.346
75.362
20.001,7 = 23.551
Reject H0
Technical change is neutral.

H0: 23= 24=25=26=27=0
291.825
30.404
20.001,5 =19.696
Reject H0
Technical efficiency is neutral.

H0: 1= 2=3=0
291.004
32.047
20.001,3 = 15.357
Reject H0
Note: #The correct critical values for the hypothesis involving =0 were taken from Table 1 in Kodde and Palm (1986: 1246).
To test functional forms, a likelihood ratio (LR)

test was used. The translog is preferred as a more
flexible functional form that allows for interaction of
inputs, unlike the Cobb Douglas which does not allow
for input interactions and assumes elasticity of
substitution between inputs equals one (Table 3).
The variance parameters, 2, of the likelihood
function are estimated in terms of
and
is statistically significant and different from

zero (Table 4).
This implies that the one-sided random inefficiency
component completely dominates the measurement
error and other random disturbances. The technical
efficiency component, , indicates that the difference
between actual and potential outputs is primarily due
to technical and state productivity differentials in wheat
productivity across state. The rejection of the null
hypothesis, H0: =0, implies the existence of stochastic
production function. The coefficient of time is 0.0585,
which indicates mean technical progress of 5.8 per cent
per year. However, this is statistically insignificant. The
coefficient of time squared is negative and significant
(at 5% level), indicating that the rate of technical change
increases at a decreasing rate. Again, the efficiency
response suggests that subsidy-driven input, like
chemical fertiliser-use now negatively affects output
growth, indicating that ironically, instead of boosting
productivity, subsidies might now be contributing to
lower productivity, compromising sustainability and
future productivity growth.
Following the methodologies described above,

Equations (1) and (6) were used to estimate TFP by
DEA and SFA approaches, respectively (Table 5).
The two approaches have depicted TFP growth of
0.93 per cent (SFA) and 2.03 per cent (DEA) (Table
5). However, on excluding Punjab, the mean TFP
growth of 0.84 per cent (SFA) and 0.86 per cent (DEA)
were found similar for the entire modelling period,
indicating robustness of the methodologies used. The
dominant nature of technical change vis--vis
efficiency change was observed for both the
approaches. The technical efficiency growth had been
zero or negative throughout; it indicates that as the
production frontier continued to shift outwards,
production kept pace, but the yield gap was not
closed further. Similarly, in the case of input
contribution to wheat output growth, except for Punjab,
both DEA and SFA have shown almost similar growths,
suggesting input intensification in the wheat
production in the Indian states.
Paddy
The SFA specification for paddy was slightly
different from that of wheat, due to huge yield gap
between high- and low-productive states. Further, due
to different agro-climatic zones, the paddy cultivation
in low-yield states, particularly in Assam and Bihar,
has been regularly affected by floods. As can be seen
from Table 2, Punjab has average yield more than
double that of these states. Thus, dummy was used to
control for these states in the panel SFA. Then we
performed the joint tests using the likelihood ratio (LR)
33
Table 4. SFA estimation output for wheat

Variables
Parameters
Frontier function
Ordinary least squares
Coefficient
Standard error
t-ratio
Coefficient
Standard error
t-ratio
Constant
-42.6871
0.9850
-43.34
-42.9622
13.3357
-3.22
lnCH
1.5219
1.2074
1.26
3.7574
1.9735
1.90
lnHL
10.3612
0.7040
14.72
9.9131
2.2975
4.31
lnM
2.5596
0.9641
2.65
1.1003
1.4616
0.75
lnAL
0.9332
0.4133
2.26
1.3008
0.6238
2.09
lnIRR
1.6747
0.8600
1.95
1.8684
4.8977
0.38
lnCH^2
0.1968
0.2465
0.80
-0.0845
0.2452
-0.34
lnHL^2
-0.7100
0.2260
-3.14
-0.7445
0.4086
-1.82
lnM^2
2.5791
0.3804
6.78
2.1357
0.6982
3.06
lnAL^2
-0.0134
0.0413
-0.32
-0.0383
0.0354
-1.08
lnIRR
10
5.1192
0.6722
7.62
3.8022
1.5987
2.38
lnCH*LnL
11
0.1540
0.4511
0.34
-0.2621
0.5560
-0.47
lnCH*lnM
12
-1.3357
0.5790
-2.31
-0.9580
0.6866
-1.40
lnCH*lnAL
13
-0.2078
0.2729
-0.76
-0.4440
0.1940
-2.29
lnCH*lnIRR
14
0.6750
0.6090
1.11
0.6373
0.9573
0.67
lnHL*lnM
15
-0.3327
0.4604
-0.72
-0.1845
0.6225
-0.30
lnHL*ln AL
16
0.0585
0.1371
0.43
-0.0468
0.1734
-0.27
lnHL*lnIRR
17
-2.4483
0.6315
-3.88
-1.8415
1.7127
-1.08
lnM*lnAL
18
0.0268
0.2051
0.13
0.1161
0.1706
0.68
lnM*lnIRR
19
-5.9817
0.7438
-8.04
-4.8729
1.8156
-2.68
lnAL*lnIRR
20
-0.2981
0.2609
-1.14
-0.1232
0.4082
-0.30
21
0.0585
0.1371
0.43
-0.0468
0.1734
-0.27
T^2
22
-2.4483
0.6315
-3.88
-1.8415
1.7127
-1.08
lnCH*T
23
0.0268
0.2051
0.13
0.1161
0.1706
0.68
lnHL*T
24
-5.9817
0.7438
-8.04
-4.8729
1.8156
-2.68
lnM*T
25
-0.2981
0.2609
-1.14
-0.1232
0.4082
-0.30
lnAL*T
26
-0.1463
0.0676
-2.17
-0.1530
0.0846
-1.81
lnIR*T
27
-0.0007
0.0003
-2.47
-0.0009
0.0004
-2.14
Constant
-1.26721
0.349403
-3.63
lnCH
-0.06321
0.040333
-1.57
lnM
-0.09271
0.099788
-0.93
lnIRR
0.48227
0.206038
2.34
Sigma
0.002242
0.00026
8.63
Gamma
1.25E-06
801840.6
Inefficiency Model
Variance parameters
Note: CH=Chemical fertiliser, HL= Human labour, AL=Animal labour, M= Machine labour, IRR=Irrigation, T=Time,
ln= Natural logarithm
34
Vol. 29 (No.1)
January-June 2016
Table 5. Sources of TFP growth for wheat, 1981-2010

State
Haryana
Punjab
Uttar Pradesh
Rajasthan
Madhya Pradesh
India
Output
growth
(%)
1.77
1.26
1.43
2.01
2.75
1.84
DEA
SFA
Efficiency Technical TFP
Input
change
change change contribution change
change change
(%)$
(%)$
(%)$
(%)
(%)$
(%)$
(%)$
0.00
0.00
0.10
0.00
0.00
0.02
1.30
6.90
0.80
2.30
-1.10
2.01
1.30
6.90
0.90
2.30
-1.10
2.03
0.46
-5.28
0.53
-0.29
3.89
-0.18
-0.14
-0.09
-0.01
0.08
-0.23
-0.08
1.43
1.30
0.92
1.93
-0.52
1.01
1.29
1.21
0.91
2.01
-0.75
0.93
Input
contribution
(%)
0.48
0.04
0.51
-0.01
3.52
0.90
Notes: Growth rates are (1-index)%

$The average numbers were calculated by using geometric mean of corresponding indices estimated through DEA and
SFA.
Output trend growth of wheat yield growth was calculated by running log-linear regression on time.
Since productivity growth is defined as output growth divided by the input growth, the contribution of inputs to output
growth can be calculated by dividing the output growth index by the Malmquist productivity index. If it is less than one (or
in percentage terms negative), then total inputs actually decline.
Table 6. Statistics for tests of hypothesis: SFA for paddy

Null hypothesis
Given pooled sample
Data can be explained by Cobb-Douglass
production specification.
H0: 6=...=20=22=0
There is no technical progress effect.
H0: 21=22=23= 24=25=26=27=0
Technical change is neutral.
H0: 23= 24=25=26=27=0
Controlling for Punjab, Assam and Bihar.
H0: 28= 29=30=0
Technical efficiency is neutral.
H0: 1= 2=3=0
Likelihood
function
LR test
statistics
Critical value
at 1% level#
Decision
389.029
158.299
461.458
20.001,16 =38.566
Reject H0
303.385
171.287
20.001,7 =23.551
Reject H0
334.700
108.658
20.001,5 =19.696
Reject H0
494.871
211.684
20.001,3 =15.357
Reject H0
463.464
62.814
20.001,3 =15.357
Reject H0
Note: # The correct critical values for the hypothesis involving =0 were taken from Table 1 in Kodde and Palm (1986:
1246).
tests as done for the wheat. Using the similar battery

of LR tests as discussed for wheat, we found that
translog functional form was favoured over the CobbDouglas form (Table 6).
The inclusion of Assam and Bihar dummy was also
supported by the LR test result. The summary of
estimates of SFA for paddy is presented in Table 7.
The SFA results for paddy look very similar to

wheat crop for variance parameters and inefficiency
measure. The coefficient of time is negative and
insignificant, indicating that the technical change is
decreasing. The coefficients of time interacted with the
chemical fertiliser are positive and with human labour,
near negative, suggesting that technical change has
35
Table 7. SFA estimation output for paddy

Variables
Parameters
Constant
0
lnCH
1
lnHL
2
lnM
3
lnAL
4
lnIRR
5
lnCH^2
6
lnHL^2
7
lnM^2
8
lnAL^2
9
lnIRR
10
lnCH*lnL
11
lnCH*lnM
12
lnCH*lnAL
13
lnCH*lnIRR
14
lnHL*lnM
15
lnHL*ln AL
16
lnHL*lnIRR
17
lnM*lnAL
18
lnM*lnIRR
19
lnAL*lnIRR
20
T
21
T^2
22
lnCH*T
23
lnHL*T
24
lnM*T
25
lnAL*T
26
lnIR*T
27
Assam_Dum
28
Punjab_Dum
29
Bihar_Dum
30
Inefficiency Model
Constant
0
lnCH
0
lnM
1
lnIRR
3
Variance parameters
Sigma
2
Inefficiency
Frontier function
Ordinary least squares
Coefficient
Standard error
t-ratio
Coefficient
Standard error
t-ratio
-5.6640
2.1228
1.3081
-1.1653
0.0058
2.4724
0.1665
0.2758
0.0023
0.0030
0.6057
-0.4139
-0.1958
-0.2170
-0.1424
0.5620
-0.0681
-1.4715
-0.0702
-0.1245
0.4152
-0.0060
0.0000
0.0060
-0.0092
0.0046
0.0042
0.0015
0.5849
0.3839
-0.2086
1.1668
0.8018
0.9191
0.3255
0.4703
1.0416
0.0381
0.3079
0.0195
0.0393
0.1317
0.3046
0.0450
0.1039
0.2179
0.1505
0.2375
0.4219
0.0601
0.0829
0.1526
0.0365
0.0002
0.0026
0.0065
0.0014
0.0017
0.0032
0.0743
0.0280
0.0180
-4.85
2.65
1.42
-3.58
0.01
2.37
4.37
0.90
0.12
0.08
4.60
-1.36
-4.35
-2.09
-0.65
3.74
-0.29
-3.49
-1.17
-1.50
2.72
-0.17
-0.01
2.31
-1.42
3.38
2.52
0.47
7.88
13.72
-11.60
-4.6706
3.8986
0.3563
-1.8193
-0.0053
2.3731
0.1987
0.4740
-0.0018
-0.0115
0.5417
-0.8965
-0.2178
-0.2269
-0.2550
0.7432
-0.0528
-1.3786
-0.0908
-0.0433
0.4729
-0.0100
-0.0003
0.0024
-0.0062
0.0074
0.0024
0.0006
0.7083
0.3649
-0.1963
6.0049
0.9255
2.4046
0.3591
0.5667
0.8224
0.0447
0.5387
0.0211
0.0439
0.1529
0.3814
0.0505
0.1256
0.2439
0.1634
0.2902
0.3781
0.0601
0.0883
0.1709
0.0376
0.0002
0.0030
0.0071
0.0018
0.0020
0.0047
0.0913
0.0263
0.0163
-0.78
4.21
0.15
-5.07
-0.01
2.89
4.45
0.88
-0.08
-0.26
3.54
-2.35
-4.31
-1.81
-1.05
4.55
-0.18
-3.65
-1.51
-0.49
2.77
-0.26
-1.69
0.79
-0.88
4.01
1.23
0.13
7.76
13.88
-12.07
0.0543
0.0092
-0.0047
-0.0046
0.0195
0.0016
0.0013
0.0016
2.79
5.94
-3.60
-2.89
0.0044005
1
0.000496
7.61E-06
8.87
131492
Notes: CH=Chemical fertiliser, HL= Human labour, AL=Animal labour, M= Machine labour, IRR=Irrigation, T=Time,
ln= Natural logarithm
s are the estimated parameters of the selected variables
36
Vol. 29 (No.1)
January-June 2016
Table 8. Sources of TFP growth for paddy, 1981-2010

State
Andhra Pradesh
Assam
Bihar
Haryana
Karnataka
Madhya Pradesh
Odisha
Punjab
Uttar Pradesh
West Bengal
India
Output
growth
1.50
1.40
1.20
1.40
1.70
1.30
2.00
0.80
1.70
1.80
1.50
DEA
SFA
Input
change
change change contribution change
change change
(%)
(%)
(%)
(%)
(%)
(%)
(%)
0.10
0.00
0.10
0.00
-1.00
0.00
0.00
0.00
-0.10
0.40
-0.10
0.30
-9.60
-1.20
1.50
-0.80
0.10
-4.20
4.40
0.00
-0.70
-1.10
0.40
-9.60
-1.10
1.50
-1.80
0.10
-4.20
4.40
0.00
-0.30
-1.10
been fertiliser-using but labour-saving over this period.

This indicates an outward shift of the isoquant at a
faster rate over time in the fertiliser-intensive part of
the input space. The TFP growth of paddy estimated
using similar approach followed in wheat TFP
calculation, is given in Table 8.
Unlike for wheat, the two approaches for paddy
crop depicted a slightly different TFP growth 0.1
per cent (SFA) and -1.1 per cent (DEA) (Table 8) at
mean for the entire modelling period of 1981-2010.
The sources of TFP growth also differed from one
another. In the case of DEA, the technical regress
seemed to be the main culprit of negative TFP growth,
whereas for SFA, the positive technical progress was
offset by the almost no change in technical efficiency.
Further, directionally, the state level TFP growth,
except for Karnataka has shown an almost similar
pattern under both the methods. A substantial variation
of TFP growth was seen among the states.2
Again, a look at the input contribution revealed
that the inputs-use had actually increased at the overall
level for both DEA and SFA, showing input
intensification in paddy production output growth.
This implies moving along the production surface with
a given technology. This confirms the findings of
Kalirajan and Shand (1997) that the output growth in
2
1.10
12.10
2.30
-0.10
3.60
1.20
6.50
-3.40
1.70
2.10
2.60
0.5
0.1
-0.1
0.2
-0.6
0.2
-0.4
-0.4
0.1
0.0
0.0
1.1
-1.6
-0.1
0.8
0.9
-0.2
-0.3
1.0
0.4
-0.1
0.2
1.5
-1.5
-0.2
1.0
0.3
0.0
-0.7
0.6
0.5
-0.1
0.1
Input
contribution
(%)
0.0
2.9
1.4
0.4
1.4
1.2
1.5
1.2
1.1
1.3
1.3
agriculture is increasingly dependent on the input

growth in most of the states in India. The input based
growth is unsustainable in the long-run. Kumar et al.
(2004) have also raised the concern over the
indiscriminate use of natural resources in the
intensively cultivated areas of the Indo-Gangetic Plains.
The situation is worse in the low-yield rain-fed states,
where the use of modern inputs, like machine labour
and chemical inputs, is still way below the national
average. This has important policy implication for the
crucial need of public investment in irrigation and water
management in these states, as the majority of farmers
in these states are small and marginal and use input
resources at sub-optimal level.
A Comparison of DEA and SFA
The temporal and spatial decomposition of TFP
growth at the state level using both DEA and SFA has
suggested that our results are fairly robust to the choice
of methodology (Tables 9 and 10).
In comparison to previous decades, in the recent
decade of 2000s, Punjab, one of the Green Revolution
star states, witnessed a decline in the TFP growth for
both paddy and wheat. This raises an alarm over the
long run sustainability of paddy-wheat system, which
brings together conflicting and complementary
Barua and Das (1996) have also observed the persistence of regional inequality in India due to differences in agricultural
productivity and infrastructure.
37
Table 9. Mean temporal and spatial DEA and SFA TFP growth for wheat
State
Haryana
Madhya Pradesh
Punjab
Rajasthan
Uttar Pradesh
India
1980s
SFA (%)
1990s
2000s
2.03
-1.61
1.27
2.37
0.84
0.97
2.13
-0.24
1.93
1.64
1.67
1.42
-0.03
0.21
0.35
2.50
0.27
0.65
1980s
DEA
1990s
2000s
0.80
-7.60
7.31
3.40
1.09
0.88
2.28
0.80
10.91
0.43
1.46
3.10
0.63
3.25
2.75
3.23
0.16
2.00
Table 10. Mean temporal and spatial DEA and SFA TFP growth for paddy
State
Andhra Pradesh
Assam
Bihar
Haryana
Karnataka
Madhya Pradesh
Odisha
Punjab
Uttar Pradesh
West Bengal
India
1980s
SFA
1990s
2000s
1980s
DEA
1990s
2000s
2.18
-2.68
-0.60
1.47
0.32
-2.79
-1.09
0.15
2.81
0.05
-0.04
1.50
-1.39
0.02
0.34
0.13
0.33
-1.18
1.45
-0.67
0.31
0.08
1.16
-0.51
-0.12
1.07
0.47
1.90
-0.37
-0.27
-0.48
-0.82
0.20
-0.98
-19.68
-4.93
-1.65
-4.62
-5.23
-8.67
3.72
0.18
0.48
-4.35
2.92
-4.70
1.46
-0.60
-1.09
1.61
-2.50
12.59
-0.12
-0.86
0.78
-0.79
-4.52
-0.20
6.68
0.04
3.64
-1.73
-2.67
-0.08
-0.49
-0.06
practices the concern already raised by many

(Murgai, 2001). The issue of sustainability of paddy
productivity is fast emerging across other high-,
medium- and low-yield states like Andhra Pradesh,
Uttar Pradesh, West Bengal and Bihar. Much of the
system operates at low yield because of inadequate
nutrients and inappropriate water management
(Timsina and Connor, 2001). The decline in Bihar
seems to be surprising in terms of paddy production
which is yet to reach its potential.3 The decline might
be due to series of floods and droughts faced by the
state in the 2000s, which created a drag on the
productivity in this state.
In the case of wheat producing states, only
Rajasthan and Madhya Pradesh have shown a gain in
TFP in the recent decade. The other three high-yield
3
states face a decline in TFP trend in the 2000s from the

previous decade. Mueller et al. (2012) have emphasized
the crucial role of nutrients and water management in
pathways towards sustainable intensification in
agricultural production.
Further, the trend patterns of efficiency change,
technical change and TFP change have been found to
be similar for both the approaches, despite more yearto-year fluctuations in the DEA results.
The overall DEA TFP series for the Indian states
is the more volatile than that of SFA. This might be
due to the fact that the DEA method is more sensitive
to year-to-year changes, while the stochastic frontier
method appears to smoothen these effects to some
degree (Bayarsaihan and Coelli, 2003). Alternatively,
Banerjee (2008) has listed several possible reasons on the low level of intermediate input use, such as fertiliser, unwillingness
to take risks, unavailability of credit, lack of right internal or external incentives for long-range planning, distortions in the land
market or lack of understanding of the benets of fertiliser.
38
the stochastic frontier models, constrained by

parameterisation, are unable to respond quickly to a
sudden shift in the technology. The uneven and volatile
DEA estimation of TFP in the wheat-producing states
of India might be due to the sudden change in the
composition of input-output combinations. Further
research is required to ascertain the reasons for these
anomalies. Coelli et al. (2005) have suggested a
window method for attempting to obtain more stable
DEA frontiers by pooling the data from two or three
adjacent periods to construct the required frontiers.
Concluding Remarks and Way Forward

The analysis presented in this paper is the depiction
of agricultural TFP growth across Indian states for two
dominant crops, paddy and wheat, using nonparametric DEA and parametric SFA approach for the
period1981-2010. Some important conclusions that
have emerged from these findings are:
First, led by tremendous growth in the frontier
states, Haryana and Punjab, technological progress has
been the primary, consistent driver of productivity
growth over the past three decades at India level. For
wheat, the TFP growth by DEA method was 2.03 per
cent and technical progress 2.01 per cent, whereas for
paddy, it was -1.1 per cent for both. Part of technical
regress, especially in most of the paddy producing
states, could have been due to declining quality of
natural resources, such as soil fertility and water
mining, which warrants further investigation.
Second, technical efficiency has been plateaued
throughout, with occasional improvement or decline,
indicating that as the production frontier continued to
shift outwards, production kept pace, but the yield
gap was not closed further in other words, no
change in efficiency occurred. Thus, from our empirical
results, we conclude that the decline and subsequent
improvement in productivity in paddy and wheat
production in the Indian states is most likely a result
of combination of problems with management and
incentive structure, as well as the absence of
breakthrough in agricultural technology.
Third, the drop in TFP estimates for wheat and
paddy in the recent decade for Punjab, irrespective of
methodology used, reveals an alarming picture on the
sustainability of paddy-wheat production system in the
state. The environmental impacts causing saturation
Vol. 29 (No.1)
January-June 2016
in productivity needs to be explored further with more

micro-level investigations. Further, among the lowyield states, such as Rajasthan, Karnataka, and Madhya
Pradesh, which have shown an improvement in the
technical progress and so TFP in the later period of
2000, also suffer a technical regress in the later period.
The poor technical change is most likely the result of
lack of investment in technology, while low technical
efficiency is generally due to management and
incentive problems as well as poor information
dissemination (Brooks et al., 1991; Foster and
Rosenzweig, 1996; Acharya, 1997; Desai and
Namboodiri, 1997; Munshi, 2004).
Fourth, from the methodological perspective, the
technical change turned out to be a more dominant
source of TFP growth, independent of choice of
methodologies.
In the last, the paper does provide a quantitative
understanding of agricultural production system over
the recent past and in the longer-run is a useful first
step toward gaining a sense of what we can expect in
the years ahead.
An obvious extension to this study would be the
application of this approach to incorporate more crops
and states or at the district level. Another interesting
work could be incorporating higher order policy
variables such as subsidies, government investment,
variables representing resource endowment,
infrastructure, groundwater extraction, etc., in the
efficiency equation of SFA. Further, our understanding
of the consequences of agricultural TFP growth on the
economies and environments is limited and worthy of
further exploration at the micro and macro levels.
Acknowledgements
The author is grateful to Prof. Abhijit Sen, Prof.
Kirit Parikh, Prof. R S Deshpande, and the anonymous
referee of this journal for their helpful comments and
valuable insights and suggestions on the earlier draft.
The usual disclaimer applies.
References
Acharya, S.S. (1997) Agricultural price policy and
development: some facts and emerging issues. Indian
Afriat, S. (1972) Efficiency estimate of production functions.
International Economic Review, 13: 568-598.
Aigner, D.J., Lovell C.A.K. and Schmidt, P. (1977)
Formation and estimation of stochastic frontier
production function models. Journal of Econometrics,
6: 21-37.
Avila, A.F. and Evenson, R.E. (2010) Total factor
productivity growth in agriculture: The role of
technological capital, In: Handbook of Agricultural
Economics, Eds: R.E. Evenson and P. Pingali. Elsevier
Science, Amsterdam. pp. 37693822.
Banerjee, A.V. (2008) Big answers for big questions: The
presumption of growth policy. Brookings Conference
on What Works in Development? Thinking Big and
Thinking Small. Accessed from: http://www.
brookings.edu/~/media/Events/2008/5/29%20
global%20development/2008_banerjee.PDF
Barua, A. and Das, S.K. (1996) Regional disparity, economic
growth and liberalization: A study of the Indian
economy. Journal of Development Studies, 32(3): 364390.
Bayarsaihan, T. and Coelli, T.J. (2003) Productivity growth
in pre-1990 Mongolian agriculture: Spiralling disaster
or emerging success? Agricultural Economics, 28(2):
121137.
Bhushan, S. (2005) Total factor productivity growth of wheat
in India: A Malmquist approach. Indian Journal of
Agricultural Economics, 60(1): 32-48.
Bhushan, S. (2009) Non-parametric productivity growth of
wheat in India during post-green revolution. The Indian
Economic Journal, 56(4): 55-77.
Bhushan, S. (2015) Agricultural Productivity and
Environmental Impacts in India: A Parametric and
Non-parametric Analysis. PhD dissertation
(Unpublished). Jawaharlal Nehru University, New
Delhi.
Brooks, K., Guasch, J.L., Braverman, A. and Csaki, C.
(1991) Agriculture and the transition to the market.
Journal of Economic Perspectives, 5(4): 149-161.
CACP (Commission for Agricultural Costs and Prices)
(various issues) Comprehensive Scheme for the Study
of Cost of Cultivation of Principal Crops in India.
Department of Agriculture & Cooperation, Ministry of
Agriculture, Government of India, New Delhi.
Caves, D.W., Christensen, L.R. and Diewart, W.E. (1982)
The economic theory of index numbers and the
measurement of input, output, and productivity.
Econometrica, 50(6): 1393-1414.
Chand, R., Kumar, P. and Kumar, S. (2011) Total Factor
Productivity and Contribution of Research Investment
39
to Agricultural Growth in India. Policy Paper 25.

National Centre for Agricultural Economics and Policy
Research, New Delhi.
Charnes, A., Cooper, W.W. and Rhodes, E. (1978)
Measuring the efficiency of decision making units.
European Journal of Operational Research, 2: 429-444.
Coelli, T.J. (1996a) A Guide to FRONTIER version 4.1: A
Computer Program for Frontier Production Function
Estimation. CEPA Working Paper 96/07. Department
of Econometrics. University of New England. Armidale,
Australia. Accessible at http://www.uq.edu.au/
economics/cepa/software.php April 2014.
Coelli, T.J. (1996b) A Guide to DEAP version 2.1: A Data
Envelopment Analysis (Computer) Program. CEPA
Working Papers No.8/96. Accessible at http://
www.uq.edu.au/economics/cepa/software.php April
2014.
Coelli, T.J. and Rao, D.S.P. (2005) Total factor productivity
growth in agriculture: Malmquist index analysis for 93
countries, 1980-2000. Agricultural Economics, 32: 115134.
Coelli, T.J., Rao, D.S.P., ODonnell, C.J. and Battese, G.E.
(2005) An Introduction to Efficiency and Productivity
Analysis. Second Edition. Springer, Heidelberg,
Germany.
Desai, B.M. and Namboodiri, N.V. (1997) Determinants of
total factor productivity in Indian agriculture. Economic
and Political Weekly, December, 27: A165-A171.
Ellis, F. (2000) Rural Livelihoods Diversity in Developing
Countries. Oxford University Press, Oxford.
Fan, S., Hazell, P. and Thorat, S. (1999) Linkages between
Government Spending, Growth and Poverty in Rural
India. IFPRI Report 110. International Food Policy
Research Institute, Washington, DC.
Fre, R., Grosskopf, S. Norris, M. and Zhang, Z. (1994)
Productivity growth, technical progress, and efficiency
change in industrial countries. American Economic
Review, 84: 66-89.
Farrell, M. (1957) The measurement of productive
efficiency. Journal of Royal Statistical Society, Series
A. CXX(3): 253-290.
Foster, A.D. and Rosenzweig, M.R. (1996) Technical change
and human capital returns and investments: Evidence
from the green revolution. American Economic Review,
86(4): 931-953.
Fuglie, K.O. (2012) Productivity growth and technology
capital in the global agricultural economy. In:
Productivity Growth in Agriculture: An International
40

Perspective, Eds: K.O. Fuglie, E. Ball and S.L. Wang.
Chapter 16. CABI, Wallingford, UK.
Himanshu, Lanjouw P., Mukhopadhyay, A. and Murgai, R.

(2011) Non-farm Diversification and Rural Poverty
Decline: A Perspective from Indian Sample Survey and
Village Study. Working Paper 44. Asia Research Centre,
London School of Economics and Political Science,
London.
Huang, C. and Liu, J. (1994) Estimation of a non-neutral
stochastic frontier production function. Journal of
Productivity Analysis, 5: 171-180.
Joshi, P.K., Joshi, L., Singh, R.K., Thakur, J., Singh, K. and
Giri, A.K. (2003) Analysis of Productivity Changes and
Future Sources of Growth for Sustaining Rice-Wheat
Cropping System. National Agricultural Technology
Project ((PSR 15; 4.2). National Centre for Agricultural
Economics and Policy Research. New Delhi.
Kalirajan, K.P. and Shand, R.T. (1997) Sources of output
growth in Indian agriculture. Indian Journal of
Kodde, D.A. and Palm, F.C. (1986) Wald criteria for jointly
testing equality and inequality restrictions.
Econometrica, 54: 1243-48.
Kumar, P. (2001) Agricultural performance and productivity.
In: Indian Agricultural Policy at the Crosswords, Eds:
S.S. Acharya and D.P. Chaudhri. Rawat Publication,
New Delhi. pp. 353-476.
Kumar P. and Mruthyunjaya (1992) Measurement and
analysis of total factor productivity growth in wheat.
Indian Journal of Agricultural Economics, 47(7): 451458.
Kumar, P. and Rosegrant, M.W. (1994) Productivity and
sources of growth for rice in India. Economic and
Political Weekly, 29(52): A183-A188.
Kumar, P., Joshi, P.K., Johansen, C. and Asokan, M. (1998)
Sustainability of rice-wheat-based cropping system in
India. Economic and Political Weekly, XXXIII: A152A158.
Kumar, P., Kumar, A. and Mittal, S. (2004) Total factor
productivity of crop sector in the Indo-Gangetic Plain
of India: Sustainability issues revisited. Indian
Economic Review, 39: 169-201.
Kumar P. and Mittal, S. (2006) Agricultural productivity
trends in India: Sustainability issues. Agricultural
Economics Research Review, 19 (Conference No.): 7188.
Kumbhakar, S.C., Ghosh, S. and McGuin, J.T. (1991) A
generalized production frontier approach for estimating
Vol. 29 (No.1)
January-June 2016
determinants of inefficiency in U.S. dairy farms.

Journal of Business & Economic Statistics, 9(3): 279286.
Lusigi, A. and C. Thirtle (1997) Total factor productivity
and the effects of R&D in African agriculture. Journal
of International Development, 9(4): 529-38.
Meeusen, W. and Broeck, J. van den (1977) Efficiency
estimation from Cobb-Douglas production functions
with composed errors. International Economic Review,
18: 433-444.
Mueller, N.D., Gerber, J.S., Johnston, M., Ray, D.K.,
Ramankutty, N. and Foley, J.A. (2012) Closing yield
gaps through nutrient and water management. Nature,
490: 254-257.
Mukherjee, A.N. and Kuroda, Y. (2003) Productivity growth
in Indian agriculture: Is there evidence of convergence
among states. Agricultural Economics, 29(1): 45-53.
Munshi, K. (2004) Social learning in a heterogenous
population: Technology diffusion in Indian green
revolution. Journal of Development Economics, 73(1):
185-213.
Murgai, R. (2001) The green revolution and the productivity
paradox: Evidence from the Indian Punjab. Agricultural
Economics, 25(23): 199-209.
Nin-Pratt and Fan, Yu (2010) Comparisons of agricultural
productivity growth in China and India. Journal of
Productivity Analysis, 33: 209-223.
Nishimizu, M. and Page, J.M. (1982) Total factor
productivity growth, technical progress and technical
efficiency change: Dimensions of productivity change
in Yugoslavia, 1965-78. Economic Journal, 92: 920936.
Ray, S.C. and Ghose, A. (2010) Production Efficiency in
Indian Agriculture: An Assessment of the Post-Green
Evolution Years. Working Paper 2010-26. Department
of Economics Working Paper Series. University of
Connecticut. October.
Rosegrant, M.W. and Evenson, R.E. (1992) Agricultural
productivity & sources of growth in South Asia.
American Journal of Agricultural Economics, 73(3):
757-761.
Sindhu, D.S. and Byerlee, D. (1992) Technical Change and
Wheat Productivity in the Indian Punjab in Post-Green
Period. Working Paper 92-02. Economics. International
Maize and Wheat Improvement Center. Mexico.
Timsina, J. and Connor, D.J. (2001) Productivity and
management of rice-wheat cropping systems: Issues and
challenges. Field Crops Research, 69: 93-132.
Received: December, 2015; Accepted: March, 2016

DOI: 10.5958/0974-0279.2016.00017.3
Total Factor Productivity Growth and Returns from

Research Investment on Soybean in India
Purushottam Sharma* and B.U. Dupare
ICAR-Indian Institute of Soybean Research, Khandwa Road, Indore-452 001, Madhya Pradesh
Abstract
The study has estimated total factor productivity growth of soybean and returns to research investment
on the crop in India. For study, time-series data on cost of cultivation of soybean in major states were
collected from the reports of Commission on Agricultural Costs and Prices for the period 1980-81 to
2011-12. The Tornqvist-Theil index has been used to compute input, output and TFP indices and compound
annual growth rates on triennium average datasets were estimated. The regression analysis was carried
out to identify the sources of TFP growth. The study has indicated moderate growth of TFP in soybean
(1.2% per year) with 10.5 per cent share in output growth. The decade-wise analysis has revealed that
TFP growth and the share of TFP in growth of output is increasing in the recent decade. The research
investment and irrigation have turned out to be the significant variables affecting TFP positively. Although,
marginal value output of research has been found less than one, the internal rate of return for research
investment is increasing in recent decades, suggesting enhancement in research allocation and irrigation
infrastructure for productivity improvement and edible oil security in the country.
Key words: Total factor productivity, Tornqvist-Theil index, returns to research investment, soybean,
India
JEL Classification: Q16, Q13, D24
Introduction
Productivity growth in agriculture is of paramount
importance as higher yields are associated with
declining rural poverty, suggesting that impact of
growth in agricultural production on poverty remains
high (Himanshu et al., 2010). The agricultural
productivity continues to be an important driver of rural
poverty reduction, especially it helps rise agricultural
Email: purushottamji@gmail.com
The paper is drawn from the larger institutional research study
on Socio-Economic Analysis of Growth in Soybean Crop
Productivity and Impact of Research conducted by authors
at ICAR-Indian Institute of Soybean Research, Indore,
Madhya Pradesh. Authors would like to thank anonymous
referees of the journal for providing valuable comments on
the earlier version of the paper.
wages. The slow agricultural growth could be due to

reduced demand for food, slow technological change
in agriculture, lack of employment opportunities for
part time smallholders, limited technology adoption
by full-time farmers (Binswanger-Mkhize, 2012). In
India, the productivity of oilseed crops is low because
the crops are largely grown on marginal lands with
minimal use of inputs.
The oilseed crops play an important role in
agricultural development of India, sharing 14 per cent
of the countrys gross cropped area and accounting for
about 3 per cent of the gross domestic product and
nearly 6 per cent of the value of all agricultural products
(ICAR-IIOR, 2015). Soybean, introduced for
commercial cultivation in India in 1970-71, has
established itself as a leading oilseed crop in the rainfed
agro-ecosystem of central and peninsular India (Bhatia
42
et al., 2008). The crop has gained tremendous strides

in terms of growth in area and production in the country.
The area under soybean has increased at an annual
growth rate of 13.74 per cent and production by 15.1
per cent during 1970-71 to 2013-14 (based on fiveyear moving average data), whereas the growth in
productivity has been merely at 1.19 per cent only. The
government policies and research support from
National Agricultural Research System (NARS) paved
the way for more than trebling the average productivity
of soybean in the country, from 436 kg/ha in 1970-71
to 1353 kg/ha in 2012-13 (GoI, 2015b).
The soybean is the largest oilseed crop in the
country sharing 38 per cent of total oilseeds crop area
and 42 per cent of total oilseeds production during TE
2012-13, contributing about 28 per cent to the total
vegetable oils and two-thirds of the oil meals supplies
during this period. However, the country has been
meeting its edible oil demand by importing more than
50 per cent of its requirement. The per capita
consumption of vegetable oil is increasing very rapidly
in the country due to increasing population and
improving economic status of the consumers. The per
capita availability for edible oils has increased to about
16.8 kg/year compared to 3.2 kg/year in 1960-61 (GoI,
2015a). The consumption of edible oils reached 21.06
million tonnes in 2013-14 (Nov- Oct) and is likely to
increase further with changing dietary habits,
enhancing income levels and increasing population
(Sharma and Bhatia, 2015). The per capita consumption
of edible oils increased faster (5.65 %) than the
production of oilseeds (4.55 %) during the previous
decade, leading to a higher growth rate in imports of
edible oils (7 %) (Reddy and Bantalian, 2012). It is,
therefore, necessary to utilize the domestic resources
fully to maximize production and ensure edible oil
security for the country.
The change in soybean production in the country
was mainly contributed by the area expansion and the
yield expansion contributed less than one-fifth share.
The area expansion in oilseeds was facilitated by the
encouraging results achieved on introducing new
varieties/crops and reducing yield variability (Chandel
and Rao, 2003). But now the chances of area expansion
are limited, and hence, yield increase is the best
alternative to boost supply of edible oils and improve
farmers income. Under this backdrop, the present study
was undertaken with the following objectives: (i) to
Vol. 29 (No.1)
January-June 2016
analyse the total factor productivity growth of soybean

in India and its determining factors, and (ii) to estimate
the returns from research investment in soybean in the
country.

Time-series data on item-wise cost of cultivation,
use of inputs and their prices for soybean grown in
major states of India were collected for the period 198081 to 2011-12 from the reports of Comprehensive
Scheme for the Study of Cost of Cultivation of Principal
Crops, Commission on Agricultural Costs and Prices,
Ministry of Agriculture, Government of India. Timeseries data on public funding on soybean research were
collected from the ICAR-Budget Books, annual reports
of ICAR-Directorate of Soybean Research and All
India Coordinated Research Project on Soybean. Also,
time-series data on infrastructural variables (market
density), kharif-rainfall, price parity, and proportion
of irrigated area were collected from various
publications of central and state governments of
soybean-producing states.
The total factor productivity measures the growth
in total output which is not accounted for by the
increase in total inputs. The TFP index is computed as
the ratio of index of aggregate output (output index) to
the index of aggregate inputs (input index). Therefore,
growth of TFP index is the growth of output index less
the growth of input index.
To calculate input and output index, TornqvistTheil index was used for data on output and inputs of
soybean crop. The Tornqvist index of TFP is the
frequently used index to compute TFP growth. It does
not require the assumption of neutral technical change
and allows for variable elasticity of substitution
(Evanson et al., 1999). Another advantage of this index
is that it accounts for the change in quality of inputs
because current factor prices are used in constructing
the weights. The quality improvements in inputs are
incorporated to the extent that these are reflected in
higher wages and rental value (Capalbo and Vo, 1988).
The output and input indexes were calculated for the
soybean crop in major states and at all-India level using
TFPIP program developed by Coelli et al. (1998; 2005).
The Tornquist-Theil index is a superlative index,
which is exact for the linear homogeneous translog
production function (Diewart, 1976). The TFP growth
Sharma and Dupare : Total Factor Productivity Growth and Returns from Research Investment on Soybean
is measured from the Tornquist-Theil TFP index (Desai,

1994). The logarithmic form of the Tornquist-Theil TFP
index is given by Equation (1):
It = It-1 n(m) j=1 (Qj,t/ Qj,t-1)1/2(Sj,t + Sj,t-1)
(1)
where, Sj,t = Qj,t*Pj,t/ [n(m) j=1 Qj,t*Pj,t]

Here, It is the index value of output/input for the current
year t; It-1 is the index value of output/ input for the
previous year t-1; Qj,t is the quantity of jth output/
input for the year t, when j varies from 1 to n for
outputs and from 1 to m for inputs used in the
production of soybean; Qj,t-1 is the quantity of jth output/
input for the crop for the previous year t-1; Sj,t is the
share of jth output value in total value of production
(VOP) or jth input in total cost; and Pj,t is the current
price of jth output/ input.
The total factor productivity is calculated using
index for input-use and output. If QIt is the index for
output in year t and XIt is the index for inputs for the
same year, the total factor productivity (TFP) index is
equal to QIt/XIt. Thus, TFP index was computed as the
ratio of an index of aggregate outputs to an index of
aggregate inputs. Specifying the index equal to 100 in
a particular year (1980-81 in the present study) and
accumulating the measure based on Equation (1)
provides the TFP index for soybean in India. The annual
growth rates of triennium moving average data of
output index and input index were calculated fitting
the exponential trend equation. The growth rate of TFP
was worked out by deducting growth rate of input index
from the growth rate of output index.
Both grain and straw yields were included in output
index of soybean. The quantity of straw yield was
calculated from grain yield using harvest index as per
Chandel (2003). The shares in total revenue estimated
using farm harvest prices were applied as the weights
to aggregate the outputs. Inputs included in input index
were land, seed, fertilizers and manure, plant protection
chemicals, human labour, animal labour, machine
labour and irrigation. Inputs were aggregated using
their shares in total cost of cultivation as weights. The
total factor productivity of soybean for all India was
worked out from area weighted average input costs
and returns for different states including Madhya
Pradesh, Maharashtra, Rajasthan, Chhatisgarh and
Uttar Pradesh for available data.
43
Total Factor Productivity Decomposition

The total factor productivity is influenced by
research, extension services, human capital,
infrastructural development, price policy and weather
(Chand et al., 2011). As an input to public investment
decision, it is important to understand the relative
importance of these productivity-enhancing factors in
determining TFP growth. The sources of TFP growth
were estimated and, based on these estimates, the
relative TFP growth accounting and marginal rates of
return on investment of these productivity-enhancing
factors were computed.
To identify the determinants of total factor
productivity, the TFP index was estimated as a function
of productivity-enhancing factors representing
research, infrastructural development, price policy and
rainfall. The five parameters namely research stock
(Res_Stock), kharif-rainfall (Rf_kharif), density of
regulated markets in soybean-growing states (Market),
price parity with the competing crop (Price_Parity) and
percentage irrigated area under the crop (Area_Irrig)
were regressed against TFP index. All the parameters
were in log form except those present in percentages.
The other parameters tried in the model were rural
literacy rate, road density, cropping intensity, but were
dropped as they showed high multicollinearity with
other variables. To assess the determinants of TFP, the
regression Equation (2) was specified as:
TFPI = f(Res_Stock, Rf_kharif, Market, Price_
Parity, Area_Irrig)
(2)
The research stock was constructed using a lagged
scheme following Evenson (1991) and Chand et al.
(2011) by summing up research investment of five
years by assigning weights as 0.2 in the year t-2, 0.4 in
the year t-3, 0.6 in the year t-4, 0.8 in the year t-5 and
1.0 in the year t-6. These coefficients were used to
estimate TFP growth and the estimated values of
marginal product (EVMP) were ultimately used to
compute marginal internal rate of return to investment
on oilseeds research.
Returns to Research Investment
The value of marginal product for soybean research
was estimated as per Equation (3):
EVMP (Research_Stock) = bi (V/Res_Stock)
(3)
44
Vol. 29 (No.1)
January-June 2016
Table 1. Growth in yield of soybean in major states of India

Period
Andhra
Pradesh
Madhya
Pradesh +
Chhatisgarh
TE 1980
TE 1990
TE 2000
TE 2010
TE 2013
970.0*
953.8
1285.0
1719.4
777.3
760.2
1053.6
1125.8
1188.1
-2.7
2.3
2.5
0.01
1.9
5.3
1.6
1980s
1990s
2000s
1980-2013
Karnataka
Maharashtra
Rajasthan
Yield (kg/ha)
560.3
623.2
887.9
901.7
1280.6
1272.5
626.5
1021.0
1160.5
939.5
1446.3
1469.8
Compound annual growth rate (%)
8.0
7.0
5.2
2.6
-3.8
-3.4
5.1
2.2
3.1
2.5
Uttar Pradesh
+
Uttarakhand
Gujarat
India
467.5
902.6
717.6
1276.0
1439.6
696.4
813.2
647.1
855.5
766.0
764.4
1106.0
1096.4
1296.7
7.4
-8.5
8.1
1.5
7.3
-0.8
-2.7
2.0
0.6
2.6
3.0
1.8
*Data pertains to TE 1992-93
where, V is the value of soybean production associated

with total factor productivity (value of soybean output
multiplied by the share of TFP in total output),
Res_Stock is the research stock and bi is the TFP
elasticity of research stock estimated from Equation
(2).
The stream of benefits was generated under the
assumption that the investment made in research in
the period t-i will start generating benefits after a lag
of 5 years and the benefits will increase during the
next nine years, remain constant for next nine years
and will start decreasing thereafter. Following Evenson
(1991) and Chand et al. (2011), an investment of one
rupee in the year t-i will generate a benefit equal to 0.1
EVMP in the year t-i+6, 0.2 EVMP in the year t-i+7,
and so on till t-i+13, and it will be 0.9 EVMP in the
year t-i+14. After this, the benefits will be equal to
EVMP up to the year t-i+23. Thereafter, the benefits
will start declining and will be 0.9 in the year t-i+24
and so on. This stream of benefits can then be
discounted at the rate r at which the present value of
benefits is equal to one. Thus, r was considered as
the marginal internal rate of return (MIRR) to soybean
research investment.

In this paper, growth in total factor productivity of
soybean in India has been worked out along with its
determining factors. The returns from investment in
soybean research in India have also been computed.

To understand the change in cost structure of soybean
cultivation, the growth in different components of
soybean cultivation has been analysed.
Growth in Yield of Soybean in Major States of
India
During the initial phase of commercial cultivation
of soybean in the country, its productivity was 425 kg/
ha (during 1970-71), which increased to 766 kg/ha
during TE 1980, and 1297 kg/ha during TE 2013,
indicating almost tripling of yield from the initial years
(Table 1). The highest productivity of soybean was
found in Andhra Pradesh at 1719 kg/ha during TE 2013,
which had increased from 954 kg/ha during TE 2000.
Other states with higher productivity are Rajasthan,
Maharashtra and undivided Uttar Pradesh. The soybean
yield levels had shown a decline in the states like
Andhra Pradesh, undivided Uttar Pradesh and Gujarat
during TE 2000 compared to in TE 1990. Overall, there
has been an increasing trend in soybean productivity,
and the rate of growth in yield is picking-up with the
concerted efforts of research, extension and farmers
of the country.
Between 1980-81 and 2012-13, the yield of
soybean has increased at the annual rate of 1.8 per cent
per annum. It was lower during the initial decade and
has picked up in the recent decade. The growth in
soybean yield was observed highest in the state of
Maharashtra (3.1%), followed by Rajasthan (2.5%),

Andhra Pradesh (2.5%), Karnataka (2.2%) and Gujarat
(2%). During the recent decade, the soybean yield
growth rate was seen picking up in undivided Madhya
Pradesh, Rajasthan, undivided Uttar Pradesh, and
Andhra Pradesh.
There has been a tremendous growth in production
of soybean in India mainly due to expansion in area,
whereas contribution of yield is minimal with the slow
growth in productivity. Post-introduction of soybean,
the cropping system has shifted from rainy season
fallow followed by winter season wheat or chickpea
system to soybean followed by wheat or chickpea
(Bhatia et al., 2008). This has enhanced the cropping
intensity which has resulted in profitability per unit
area.
Changes in Cost of Cultivation of Soybean in India
The real cost of soybean production, cost incurred
on major inputs, and real returns (at 2004-05 constant
prices) along with the proportion of different inputs in
total cost of cultivation of soybean were worked out to
understand the changes in cost and profitability over
time. The traditional inputs make up about three-fourths
of the total cost of soybean cultivation in India. The
proportion of modern inputs have increased from mere
10 per cent during the decade of 1981-1991 to about
18 per cent in the recent decade (Table 2). A higher
proportion of labour inputs and a lower share of
productive and protective inputs was observed in the
cost of soybean cultivation. This signifies that there
was increase in operational inputs due to rise in wage
rate, hiring charges of tractor and hike in diesel price,
leading to a curtail in expenditure on productive or
protective inputs or higher dependency on farm-saved
inputs (especially seeds) due to poor resource base.
This phenomenon might also be adding to the slow
growth in soybean productivity.
The real average cost incurred on different
components of soybean cultivation and real gross
returns over the years are presented in Table 3. It is
evident from Table 3 that the real operational cost of
soybean cultivation has increased gradually. The
operational cost of soybean cultivation (at 2004-05
constant prices) has increased from ` 3637/ha in TE
1982-83 to ` 10882/ha in TE 2011-12, at annual
compound growth rate of 3.63 per cent.
45
Table 2. Trends in cost structure of soybean production

in India: 1980-81 to 2011-12
(% of total cost)
Components
1981-91
1991-01
2001-11
Traditional inputs
Land
Seed
Human labour
Animal labour
Manure
76.6
31.7
14.1
16.9
10.4
3.5
74.7
28.3
13.6
22.5
8.2
2.1
71.9
24.7
11.1
22.6
10.8
2.7
Modern inputs
Fertilisers
Pesticides
Irrigation
Machine labour
Othersa
9.6
5.5
0.2
0.5
3.4
13.8
14.5
6.5
0.9
0.3
6.8
10.8
17.8
4.2
2.3
0.4
10.9
10.3
Source: Calculated by authors from data collected from

reports of CACP.
a
denotes cost items such as interest on working and fixed
capital, depreciation, and other miscellaneous items.
The total real cost of soybean cultivation in India

has increased from ` 7231/ha in TE 1982-83 to ` 16431/
ha. The real gross returns have increased to ` 19903/
ha recently from ` 10727/ha during TE 1982-83. The
growth was found to be faster in real total cost (2.62%/
annum) as compared to in real gross returns (2%/
annum), resulting in lower profits for farmers from the
crop. The growth in real cost of inputs such as plant
protection chemicals and machine use has been found
to be faster. The real cost of human labour in cultivation
of soybean has increased fast (3.94 % per annum), even
though the use of human labour (hours/ha) has declined
significantly.
The growth has increased faster in real operational
cost (3.63 % per annum) than in fixed cost (1.12 %)
and gross returns (1.99 %), indicating that input-use
has increased faster than productivity. This implied that
soybean production is increasingly becoming inputintensive with low yield response. This may be due to
the factors such as following a mono-cropping system
(Soybean- wheat/ chickpea) for a longer duration in
the major producing areas leading to fast depletion of
soil health and increased incidence of insects and pests
(Dupare et al., 2010), inadequate and imbalanced soil
46
Vol. 29 (No.1)
January-June 2016
Table 3. Average real cost and returns in soybean production in India, TE 1982-83 to TE 2011-12
(`/ha)
Real cost and returns
Gross returns
Operational cost
- Human labour
- Animal labour
- Machine labour
- Seed
- Fertilizers & manure
- Plant protection chemicals
- Irrigation charges
Fixed cost
Total cost
Net returns
Returns over operational cost
TE 1982-83
TE 1991-92
TE 2001-02
TE 2011-12
CAGR (%)
10727
3637
1297
813
90
805
523
13
10
3595
7232
3495
7091
13988
6139
1828
784
584
1531
1144
37
63
4663
10802
3186
7849
12792
8166
3071
1326
1000
1382
1009
107
70
3937
12103
689
4627
19903
10882
3917
1264
2051
1743
997
598
32
5550
16432
3471
9021
1.99
3.63
3.94
2.53
10.78
1.94
1.93
23.64
7.63
1.12
2.62
Source: calculated by authors from data collected from reports of CACP.

Note: Growth is annual compound growth rates for the period 1980-81 to 2011-12. All India item-wise cost and returns
have been worked out as area weighted average input costs and returns data for different states including Madhya Pradesh,
Maharashtra, Rajasthan, Chhatisgarh and Uttar Pradesh for available data.
nutrition (Sharma et al., 1996; Tiwari, 2001; Hegde

and Sudhakara Babu, 2009), low production
technology adoption (Sharma et al., 2006; Dupare et
al., 2011; Kumar et al., 2012; Singh et al., 2013), water
stress at critical crop growth stages (Tiwari, 2001;
Bhatia et al., 2008), and so on. Oilseeds are energyrich crops while in India these are grown under energystarved conditions (Hegde and Sudhakara Babu, 2009)
and the low response of oilseeds to fertilizer application
may be due to hidden hunger for secondary and micronutrients.
Total Factor Productivity Growth of Soybean in
India
The increase in input use, to a certain extent, allows
the agricultural sector to move along the production
surface. The balanced use of modern inputs is expected
to induce an upward shift in the production function to
the extent that a technological change is embodied in
them. The TFP measures the extent of increase in the
total output, which is not accounted for by increases in
the total inputs.
The input, output and TFP index of soybean in
India was worked out and is depicted on Figure 1. It is
evident from Figure 1 that there was a robust growth

in output index of soybean from 1980-81 to 2011-12,
but it was mainly because of increased inputs-use which
brought into inefficiencies in their use. A higher growth
in output index coupled with almost parallel growth in
inputs-use index led to near stagnant total factor
productivity of soybean in India (Figure 1). However,
the output index in recent years was found to be higher
than the index of inputs-use.
Annual growth rates in output, input and TFP
indexes (compound annual growth rate based on threeyear moving averages) of soybean in India are
presented in Table 4. The total period was divided into
three different decades to understand the dynamics of
TFP over time. The analysis has revealed that the annual
growth in output as well as inputs-use index has slowed
down in the recent decade. However, the growth was
found to be higher in output index than the inputs-use
index, leading to a positive growth in TFP for soybean
in recent decades (Table 4). The higher output growth
triggered by technological change has resulted in
positive soybean TFP growth in the recent period.
Overall, TFP index of soybean has grown moderately
(1.2 % per annum) in India. Also, the contribution of
TFP to total output growth was negligible during 1980s,
47
Figure 1. Input, output and TFP index of soybean cultivation, 1980-81 to 2011-12
Table 4. Annual growth rates of input-use, output and TFP index of soybean in India: 1980-81 to 2011-12
Decade
Output
Input-use
TFP index
Share of TFP
in output growth
Madhya Pradesh (1981-2012)

Maharashtra (1995-2012)
Rajasthan (1997-2012)
All India
1982-1991
1992-2001
2002-2012
1982-2012
9.94
10.19
5.82
8.28
10.79
4.23
1.66
-0.60
1.59
20.11
-5.61
37.63
19.60
9.52
9.08
11.40
19.44
7.89
6.47
10.20
0.16
1.62
2.61
1.20
0.83
17.06
28.72
10.52
Note: TFP for all India have been worked out from area weighted average input costs and returns data for different states
including Madhya Pradesh, Maharashtra, Rajasthan, Chhatisgarh and Uttar Pradesh for available data.
which increased fast and reached about 29 per cent in

the recent decade. This indicates that the productivity
growth rather than the input growth has been the main
driver of soybean production in India during the recent
period. For the overall period under analysis, nearly
10.5 per cent of the total soybean output growth was
contributed by TFP in India. The TFP growth of
soybean in Madhya Pradesh and Rajasthan was found
positive, while in Maharashtra, it was negative on
higher input-use.
Determinants of TFP of Soybean in India
To analyse the determinants of total factor
productivity of soybean in India, the aggregate/ area
weighted average data of the districts, covering more
than 90 per cent of soybean area, were used for the

independent factors such as kharif rainfall and density
of regulated markets. For TFP growth, the various
factors are responsible are technology improvement,
infrastructural development, weather, price policy and
human capital. Thus, TFP index was decomposed
against five explanatory variables related with these
factors, viz., research (Res_Stock), infrastructure
(Market, Area_Irrig), price policy (Price_Parity) and
weather (Rf_Kharif). The dependent variable was taken
in natural log of TFP index. All explanatory variables
were specified to the base of natural logarithms, except
the variable defined in percentage/ ratio terms such as
proportion of soybean area irrigated and price parity.
Maize was taken as a competitive crop of the soybean
48
Table 5. Determinants of TFP of soybean in India, 198081 to 2011-12

Variable
Regression coefficient
Pr > |t|
3.114**
0.227*
0.152
0.081
-0.082
0.055**
0.44
2.04
0.0230
0.0090
0.4285
0.8806
0.2686
0.0267
Intercept
Res_Stock
Rf_kharif
Market
Price_Parity
Area_Irrig
R2
Durbin-Watson D
for price parity variable. The estimated regression

results are presented in Table 5. Although, some other
factors such as extension stock, road density, cropping
intensity, rural literacy, etc. were also included in the
model, but were not considered due to multicollinearity.
Also, the exclusion of these variables did not much
affect R2 value.
The regression results indicated that expenditure
on soybean research had a positive and significant
impact on total factor productivity of soybean, implying
thereby that public expenditure on soybean research
assumes a greater role in accelerating the productivity
of soybean in the country. The per cent soybean area
irrigated had a positive and significant effect on TFP.
Irrigation played a crucial role in kharif crops,
particularly under prolonged drought situations. Thus,
providing irrigation at the crucial stage can boost the
productivity of the crop.
Returns from Research Investment in Soybean in
India
The impact of soybean research was measured in
terms of economic returns to investment. Measuring
Vol. 29 (No.1)
January-June 2016
economic returns of agricultural research comprises

two steps; the first step includes decomposing the
growth of TFP to various factors including research,
and the second step includes estimating the marginal
product for investment on research by the product of
research stock elasticity and average product value to
research. The results of regression were used to
determine the relative contribution of research
investment in TFP growth of soybean. Since the
function fitted was semi-log linear, the technical
coefficients of the variables gave the production
elasticities. Using the elasticity of TFP with respect to
research stock variable, the value of marginal product
and marginal internal rate of returns from soybean were
calculated and are presented in Table 6.
The analysis has revealed that marginal returns to
investment on soybean research in India were very low.
For the overall period, 1980-81 to 2011-12, increment
of one rupee in research stock generated, on an average,
additional income of ` 0.26 only, indicating very low
rate of returns to public research investment. The
decade-wise analysis has indicated that the marginal
rate of returns from soybean research has improved in
the recent decade (Table 6). Less than one marginal
value product of research for groundnut, and rapeseed
& mustard was also reported by Chand et al. (2011).
Similarly, the internal rates of returns to soybean
research investment are moderate (10 %).
The ratio of soybean research investment to the
value of output from soybean at nominal prices has
indicated that not even 0.001 per cent of value of output
from crop was invested in research. The results suggest
that further higher investment on soybean research will
generate soybean supply at a faster rate. Hence, to meet
the fast growing demand for edible oils and in turn
oilseeds, the policy signals are clear; Invest more on
Table 6. Estimated value of marginal product of research stock and marginal internal rate of returns to research
investment in soybean in India: 1980-81 to 2011-12
Decade
EVMP
IRR
(%)
Average research investment

(in lakh `)
Research investment to
value of output ratio
1981-1991
1991-2001
2001-2011
1981-2011
0.008
0.39
0.71
0.26
Negative
13.1
18.2
10.0
72.57
227.77
634.39
314.19
0.0016
0.0005
0.0004
0.0005
Note: VMP less than 1 indicates that research in the commodity is not generating enough output to justify investment.
research and create irrigation infrastructure, may be

through localized water harvesting to boost domestic
production of the crop and the edible oil pool.
Low or stagnant growth in TFP of soybean crop
was reported earlier by Chandel (2007) and Chand et
al. (2011). Chandel (2007) has reported that the input
index increased marginally faster (18.29 %) than the
output index (18.23 %), leading to marginally negative
TFP growth. Similarly, Chand et al. (2011) have
revealed that the total factor productivity (TFP) growth
for soybean was 0.83 during post-TMO period and
declined to 0.62 during post-WTO period. The value
of marginal product of oilseed research stock was found
to be less than ` 1 during all the periods, and started
declining during post-TMO and continued to decline
during post-WTO periods. The internal rate of returns
through oilseeds research investment was around 18
per cent for groundnut during all the periods, while it
continuously declined for mustard crop from 27 per
cent during pre-TMO period, 17 per cent during postTMO period and 13 per cent during post-WTO period
(Chand et al., 2011).
The lower growth in total factor productivity for
oilseeds can be attributed to lower investment in
oilseeds research. Indian research system invests
merely 4.2 per cent of total agricultural research
investment on oilseeds research (Chandel and Rao,
2003), whereas oilseeds contribute about 10 per cent
of total value of output from agriculture crops. Even
within oilseed crops research, resource allocation was
found to be disproportionate. The share of research
investment for mustard and sesame had increased,
while share of all other oilseed crops had declined. The
research investments for crops like rapeseed and
mustard, groundnut and soybean were lower than their
contribution in the value of output (Chandel and Rao,
2003). Rao et al. (2000) have also reported R&D input
index of 0.36, reflecting more budget and manpower
for the soybean crop for which R&D output index was
moderately high (0.56). Some studies have reported
low adoption of soybean production technology
(Sharma et al., 2006; Dupare et al., 2010; 2011; Kumar
et al., 2012; Singh et al., 2013) leading to realization
of lower yield, calling for concerted efforts on
extension services along with policy support for timely
availability of quality inputs to the farmers.
Most of the change in total soybean production in
the country was mainly contributed by area expansion
49
with 18 per cent contribution of yield increase. The

chances of area expansion are limited now and hence,
efforts need to be enhanced towards yield increase and
resource conservation. The modern research tool such
as biotechnology, nanotechnology, etc., need to be
deployed to improve the genetic potential of crop and
input-saving technologies aiming at enhancing soybean
production per unit of factors of production. This
requires putting higher research investment in soybean
with well-defined and goal-oriented efforts. The
concerted efforts towards this direction will improve
not only farmers income, but will also help realize
higher production of soybean in the country and in
turn, higher capacity utilization of ailing oil processing
units, saving foreign exchange through reduced burden
on edible oil imports and even earning significant
amount of foreign exchange through exporting soy
products.
Conclusions
The study has analysed the growth in total factor
productivity of soybean in India using Tornqvist-Theil
index approach from the data collected from reports
of Commission on Agricultural Costs and Prices, New
Delhi. Efforts have also been made to workout returns
from research investment in soybean in the country.
The analysis has indicated that the annual growth in
output index is higher than the inputs-use index, leading
to positive moderate growth in TFP for soybean in India
(1.2 % per annum) and higher in the recent decade.
The contribution of TFP to total output growth was
nearly 10.5 per cent.
The analysis of determinants for TFP has indicated
that expenditure on soybean research and proportion
of irrigated area under soybean have a positive and
significant impact on total factor productivity, implying
thereby that public expenditure on soybean research
and irrigation assumes a greater role in accelerating
the productivity of soybean in the country. The internal
rates of returns to soybean research investment are
moderate (10 %). The returns to research have
increased over time with increase in soybean research
investment, implying that higher investment on
soybean research will facilitate soybean supply at a
faster rate. Hence, to meet the fast growing demand
for edible oils and in turn oilseeds, higher investments
on research and creating irrigation infrastructure are
needed.
50
References
Bhatia, V.S., Singh, P., Wani, S.P., Chauhan, G.S., Kesava
Rao, A.V.R., Mishra, A.K. and Srinivas, K. (2008)
Analysis of potential yields and yield gaps of rainfed
soybean in India using CROPGRO-Soybean model.
Agricultural and Forest Meteorology, 148: 1252-1265.
Binswanger-Mkhize, H.P. (2012) India 1960-2010:
Structural change, the rural nonfarm sector, and the
prospects for agriculture. In: Stanford Symposium Series
on Global Food Policy and Food Security in the 21st
Century. Accessed from https://woods.stanford.edu/
sites/default/files/files/India1960-2010.pdf.
Capalbo, S.M. and Vo, T.T. (1988) A review of the evidence
on agricultural productivity and aggregate technology.
In: Agricultural Productivity: Measurement and
Explanation, Resources for the Future, Eds: Susan M.
Capalbo and John M. Antle. Washington, DC, USA.
Chand, Ramesh, Kumar, P. and Kumar, S. (2011) Total
Factor Productivity and Contribution of Research
Investment to Agricultural Growth in India. Policy
Paper No. 25, ICAR-National Centre for Agricultural
Economics and Policy Research, New Delhi.
Chandel, B.S. (2003) Impact Assessment of Oilseeds
Research in India. Institute Project Report, National
Academy of Agricultural Research Management,
Hyderabad.
Chandel, B.S. (2007) How sustainable is the total factor
productivity of oilseeds in India?, Indian Journal of
Chandel, B.S. and Rama Rao, D. (2003) Investment in
oilseeds research in India. Economic and Political
Weekly, 38(43): 4618-4622.
Coelli, T., Rao, D.S.P. and Battese, G.E. (1998) Introduction
to Efficiency and Productivity Analysis. Kluwer
Academic Publishers, Boston.
Coelli, T.J., Rao, D.S.P., ODonnell, C.J. and Battese, G.E.
(2005) An Introduction to Productivity and Efficiency
Analysis. Second Edition, Springer, USA.
Desai, B.M. (1994) Contributions of institutional credit, selffinance and technological change to agricultural growth
in India. Indian Journal of Agricultural Economics,
49(3): 457-475.
Diewert, W.E. (1976) Exact and superlative index numbers.
Journal of Econometrics, 4(2): 115-145.
Dupare, B.U., Billore, S.D. and Joshi, O.P. (2010) Farmers
problems associated with cultivation of soybean in
Madhya Pradesh, India. Journal of Agricultural
Sciences and Technology, 4(6): 71-78.
Vol. 29 (No.1)
January-June 2016
Dupare, B.U., Billore, S.D., Joshi, O.P. and Verma, S.K.

(2011) Adoption of improved soybean production
technology in Madhya Pradesh: A critique. Journal of
Oilseeds Research, 28(2): 125-130.
Evenson, R.E. (1991) Agricultural technology, population
growth, infrastructure and real incomes in north India.
In: Research and Productivity in Asian Agriculture, Eds:
R.E. Evenson and C.E. Pray. Cornell University Press,
Ithaca, USA.
Evenson, R.E., Pray, C. and Rosegrant, M.W. (1999)
Agricultural Research and Productivity Growth in
India. Research Report No. 109. International Food
Policy Research Institute, Washington, D.C.
GoI (Government of India) (2015a) Economic Survey-201415. Ministry of Finance, New Delhi.
GoI (Government of India) (2015b) Agricultural Statistics
at a Glance 2014. Directorate of Economics and
Statistics, Department of Agriculture and Cooperation,
Ministry of Agriculture, New Delhi.
Hegde, D.M. and Sudhakara Babu, S.N. (2009) Declining
factor productivity and improving nutrient-use
efficiency in oilseeds. Indian Journal of Agronomy,
54(1): 1-8.
Himanshu, Lanjouw P., Mukhopadhyay A. and Murgai R.
(2010) Non-farm Diversification and Rural Poverty
Decline: A Perspective from Indian Sample Survey and
Village Study Data. Working Paper 44. Asia Research
Centre, London School of Economics and Political
Science, London, UK.
ICAR-IIOR (2015) Vision 2050. ICAR-Indian Institute of
Oilseeds Research, Hyderabad.
Kumar, A., Rathod, M.K. and Kalantri, L.B. (2012)
Behaviour of farmers in adoption of recommended
technology for soybean. Indian Research Journal of
Extension Education, Special issue (II): 223-227.
Rao, D.R., Shastry, K.R. and Kiresur, V.R. (2000)
Technological Forecasting of Future Oilseeds Scenario
in India. Project Report, National Academy of
Agricultural Research Management, Hyderabad.
Reddy, A.A. and Bantilan, M.C.S. (2012) Competitiveness
and technical efficiency: Determinants in the groundnut
oil sector of India. Food Policy, 37(3): 255-263.
Sharma, H.O., Nahatkar, S.B. and Patel, M.M. (1996)
Constraints of soybean production in Madhya Pradesh:
An analysis. Bhartiya Krishi Anusandhan Patrika,
11(2): 79-84.
Sharma, H.O., Patidar, M. and Nahatkar, S.B. (2006)
Constraints of soybean production technology
in Vindhyan Plateau agro-climatic region of
Madhya Pradesh. Research on Crops, 7(1): 100110.
Sharma, P. and Bhatia V. S. (2015) Oilseed and edible oil
economy in India: Enhancing supplies through
augmenting productivity. In: Souvenir 53rd All India
Convention on Oilseeds, Oils Trade and Industry,
organized by The Central Organization for Oil Industry
and Trade (COOIT) during 24-25 October, at
Nagpur.
51
Singh, M., Dwivedi, A.P., Mishra, A., Singh, R.P., Singh,

D., Singh, S.R.K. and Chand, P. (2013) Adoption level
and constraints of soybean production technology in
Sagar district of Madhya Pradesh. Journal of
Community Mobilisation and Sustainable
Development, 8(1): 94-99.
Tiwari, S.P. (2001) Shattering the production constraints in
soybean-based cropping systems. JNKVV Research
Journal, 35(182): 1-10.
Received: January, 2016; Accepted: April, 2016

DOI: 10.5958/0974-0279.2016.00018.5
Issues Limiting the Progress in Negotiable Warehouse

Receipt (NWR) Financing in India
Shalendra, M.S. Jairath, Enamul Haque and Anu Peter V.
CCS-National Institute of Agricultural Marketing, Jaipur 302 033 Rajasthan
Abstract
Although the warehouse receipt system (WRS) has the potential to improve the access to institutional
credit of farming community, specially the smallholders lacking acceptable collateral for availing loan
from formal channels, the progress of loan against NWR is observed to be slow. The paper has analysed
the issues responsible for limited popularity of pledge finance amongst farmers. The agencies having
integration with the market like the collateral management service providers, APMCs, Agricultural
Marketing Board, etc., have been found to be a better window for disbursement of pledge finance. The
physical availability of warehouses, complicated procedure and poor awareness level are some of the
factors that limit the use of loan against a negotiable warehouse receipt (NWR). Awareness about
advantages of NWR loans needs to be generated by involving banks, Krishi Vigyan Kendras, State
Agricultural Marketing Boards, Regulated Market Committees, NGOs, co-operatives, FPOs and even
APMCs. In order to inculcate storage habit and promote storage among the producers, warehousing
should be encouraged under public-private partnership mode. There is a strong need to push NWR beyond
a negotiable instrument as it offers immense potential to trade, short-term loan to farmers, balanced
supply of agri-commodities and enhances farmers returns.
Key words: Negotiable warehouse receipt, agricultural finance, agricultural credit, pledge finance
Introduction
in agricultural economy have gained importance over

the years (Shraddha, 2013). The initiatives like
nationalization of commercial bank, setting up of
Regional Rural Banks and many other policy
interventions like priority sector lending have resulted
in increased flow of institutional credit to agriculture
over the years. The structure of flow of formal credit
to agriculture has also changed over the years, from
the cooperatives during 1950s and 1960s to scheduled
commercial banks and RRBs as the major sources of
direct and indirect institutional credit to agriculture in
recent years (Thejeswini et al., 2014).
Email: shalendra_cpsingh@rediffmail.com
The paper is based on the primary survey conducted for the
in-house research study on analysing the factors responsible
for slow growth of Negotiable Warehouse Receipt (NWR)
Finance in India
The development of efficient, sustainable and

widely accessible rural financial systems remains a
major challenge for factors like high cost of delivering
financial services to small and widely disbursed
Credit is important for the present time technologyintensive agriculture. The availability of credit can help
in enhancing the use of inputs (Mishra, 1994), and
adoption of modern technologies (Rajeev and Dev,
1998), cultivation of remunerative crops, and
increasing cropping intensity (Rajendra et al., 1995),
improving net returns per unit area and generating more
capital stock at farms (Baba et al., 2014a;b).
Innovations in agricultural credit for its significance
54
customers and lack of suitable collateral (Onumah,

2003). Only 21 per cent rural households had access
to formal credit and the majority of bank loans were
collateralized (RBI, 2005). The warehouse receipt
system (WRS) has the potential not only to improve
the access of farming community to institutional credit
but also to work as trade facilitator and thus helping in
improving the per capita household income through
efficient marketing of the agricultural commodity.
Negotiable Warehouse Receipt (NWR) will help in
developing an effective tool to take care of short-term
post-production credit needs of the farming community.
Availability of short-term credit to farmers will provide
a real boost to the increase in flow of formal credit.
The importance of short-term credit has been reflected
by the refinance figures of NABARD for the year 201415 which reveal that nearly 75 per cent of the total
refinance is for short-term component (NABARD,
2014-15).
In spite of the requirement of credit by producerfarmers and government efforts to promote the use of
negotiable warehouse receipt by creating storage
infrastructure in the rural area, providing appropriate
legal and regulatory environment and initiating various
schemes to popularize pledge finance among farmers,
the progress in taking loans against NWR has been
very slow and confined to a limited geographical area.
Vol. 29 (No.1)
January-June 2016
specifically for the study. The secondary information

was accessed from various other publications of the
state and central governments. Simple descriptive
statistical techniques like averages, percentages and
graphical analysis were used to analyse the information
collected and draw logical inferences.

Need and Importance of NWRs
Making credit available to the farmers mainly in
the agricultural ecosystem dominated by smallholders,
has always been a challenge. The lack of appropriate
assets to offer as security has further aggravated the
situation. The warehouse receipt system has the
potential to offer a viable channel to deliver credit to
smallholder farmers as it allows the use of stocks as
collateral against loan. The same has been the intension
of the government behind creation of storage capacity
in the country.
Methodology
Warehouse receipt is the instrument that allows the

farmers to extend the sales period of modestly
perishable products well beyond the harvesting season.
The warehouse receipt can be used as collateral for
short-term borrowing to obtain working capital.
Collateralizing agricultural inventories will lead to an
increase in the availability of credit, reduce its cost,
and mobilize external financial resources for the sector.
The process of warehousing sector and associated
benefits are limited in many developing countries
because of institutional and structural shortcomings like
lack of incentives for the development of private
storage industry owing to government intervention in
agricultural sector and lack of proper legal, regulatory,
and institutional environment to support a system of
Warehouse Receipts (RBI, 2005).
The study is based mainly on the primary

information collected from farmers and bankers
considered under the study. The information was
collected during the last quarter of 2014 from a sample
of 60 farmers spread across 11 districts of Rajasthan
and Madhya Pradesh. The farmers were selected
randomly from a list provided by the warehouses of
beneficiaries availing storage facility. The selection of
bank was also made on the same principle. The
information was collected through personal interview
method using a well-structured questionnaire designed
The banks are the major players in fulfilling the

objectives of the Government of India as they have to
provide pledge loan to farmers against the agricultural
produce stored in the warehouses. Despite continuous
efforts by the government, the pledge loan against
warehouse receipt has not become popular in India as
the bankers are not confident with the management
and maintenance of the warehouses in the country (RBI,
2005). With the aim to regulate and promote the growth
of warehousing sector in India and also to introduce
the negotiability of warehouse receipts, the
With this background, this paper has analysed the

issues responsible for the limited popularity of pledge
finance amongst farmers and has evaluated the reasons
for limited access to benefits available through NWRs
like pledge finance. The study has also analysed the
perception of key stakeholders like farmers and bankers
on the subject.
Shalendra et al. : Issues Limiting the Progress in Negotiable Warehouse Receipt Financing in India
Warehousing (Development and Regulation) Act 2007

was enacted in October, 2010 by the Government of
India.
Table 1. Status of pledge finance in India, 2013-14

State
Growth and Spread of Pledge Finance

Pledge finance in general is bailment of goods as
a security for payment of a debt. In the Indian
agricultural scenario, pledge finance has immense
importance as most of the small and marginal farmers
are resource-poor and lack collaterals to offer as
security to the financial institutions for availing loan.
The NWRs can be used as collateral. In many Indian
states, pledge finance has been introduced in recent
past (by pledging stored farm produce).
Among the 28 Indian states, only 12 have availed
the benefits of pledge finance scheme for farmers and
other stakeholders dealing in agricultural commodities
(Table 1). In these states, the scheme has been
popularized by agencies such as Central Warehousing
Corporation (CWC), State Warehousing Corporation
(SWC), State Agricultural Marketing Departments/
Boards, collateral management service provider, i.e.
National Collateral Management Service Limited and
National Bulk Handling Corporation (NBHC). Some
of the states like Punjab, Andhra Pradesh, Rajasthan
and Haryana, have achieved reasonable progress.
Figure 1, depicting the share of different agencies
in pledge finance and the number of beneficiaries,
reveals that the major share of pledge finance is being
managed by the agencies providing collateral
55
Punjab
Andhra Pradesh
Rajasthan
Haryana
Karnataka
Madhya Pradesh
Tamil Nadu
Odisha
West Bengal
Kerala
Maharashtra
Gujarat
Total
Pledge
Per cent
finance
share
(in crore ` )
2915
1314
1169
974
450
312
241
109
76
57
39
21
7675
37.97
17.12
15.22
12.69
5.87
4.06
3.14
1.42
0.99
0.74
0.51
0.27
100.00
No. of
beneficiaries
358
19813
612
265
13107
895
5765
206
81
1516
390
62
43070
Source: GoI (2014)
management services like National Collateral

Management Service Limited (NCML) and National
Bulk Handling Corporation (NBHC). However, state
level agencies like State Warehousing Corporation
(SWC), Central Warehousing Corporation (CWC) and
State Agricultural Marketing Board/Agricultural
Produce Market Committees (APMCs) are found to
be better equipped for extending benefit to a larger
number of beneficiaries. The figure of average loan
per beneficiary also suggests that the loan disbursed
APMC = Agricultural Produce Market Committee; CWC = Central Warehousing Corporation;

NBHC = National Bulk Handling Corporation; NCML = National Collateral Management
Services Limited
Figure 1. Share of different agencies in total pledge finance disbursed and number of beneficiaries, 2013-14
56
Vol. 29 (No.1)
January-June 2016
APMC = Agricultural Produce Market Committee; CWC = Central Warehousing Corporation; NBHC
= National Bulk Handling Corporation; NCML = National Collateral Management Services Limited
Figure 2. Average loan disbursed per beneficiary by different agencies
through collateral agencies is being availed mainly by

the traders.
The loan disbursed through State and Central
Warehousing Corporations might have gone to the
farming community, but mainly to large or medium
farmers as suggested by the size of average loan
disbursed per farmer. The loan disbursed through
APMCs has been found to be most suitable for small
and marginal farmers, as suggested by the size of loan
per beneficiary (Figure 2).
The agencies integrated with the market have
proved to be a better window for disbursement of
pledge finance like the collateral management service
providers. The collateral management service agencies
for their business structure seem to be giving preference
to traders over farmers. Marketing facilitating public
agencies like APMC seem to be more focused on
distributing loans to smallholder farmers.
Perception of Farmers about Different Sources of
Credit
The perception of farmers about different means
of credit available to them was analysed in terms of
understanding of procedure, convenience in availing

and ease of documentation (Table 2). The findings
suggest that the convenience of farmers for KCC and
cooperatives is due to their understanding about the
necessary procedure and documentation required to
avail the loan from these sources. The pledge finance
has not been found popular among the farming
community due to poor awareness about the
documentation required to avail the loan. The farmers
also found Kissan Credit Card (KCC) better in terms
of interest rate charged. Based on this perception, the
majority of farmers have shown preference to the KCC
with the remaining showing their interest in
cooperatives.
Rationale behind Farmers Decision to Utilize
Warehousing Facilities
In order to understand the limited success of credit
against NWR, it is important to identify the factors
that encourage the farmers to store the commodity. The
analysis suggests that the majority of farmers stored
an agri-commodity with the expectation of reaping
benefits of favourable price movement. However, this
depends on the availability of warehouse near the
Table 2. Perception of farmers about different means of credit

Credit source
KCC
Cooperatives
Pledge
Moneylenders
Commission agents
Perception of farmers (%) on

Understanding
Convenience
Awareness on
of procedure
documentation
96
87
20
57
13
96
76
17
24
2
41
30
22
Interest rate
(%)
Preference
4
8
8
93
7
0
Table 3. Reasons behind farmers decision on storing
agricultural commodity
Reasons
Perception
of farmers
(%)
For storing commodity

To reap benefits of price movement
Availability of warehouse nearby
Economic capacity to retain commodity
For immediate disposal of commodity
Urgent cash need
Absence of warehousing facilities in the
vicinity
Good price during harvest time
66
21
14
23
75
2
farmers field and his capacity to retain the commodity

(Table 3). The farmers opined that to encourage storage
and pledge finance, such facilities should be made
available within a radius of 5-10 km of production
centre.
The absence of warehousing facilities near the
production centre was reported as the major factor
behind immediate disposal of agricultural produce by
around three-fourth farmers. Immediate need for
57
money was another reason for not retaining the harvest

by them. Only 2 per cent of the farmers found prices at
harvest time favourable in comparison to prices
expected after storage.
Analysis of Gains Realized by Farmers
The storage decision of farmers varied from crop
to crop. For storage, the most popular crop was rice
(60%), followed by bajra (27.27%), soybean (25%),
mustard (14.29%) and mung (14.29%).
The percentage of quantity stored in relation to
total production of a crop was also analysed and is
presented in Table 4. The percentage of total production
of a commodity stored was recorded maximum for rice
(43%), followed by rabi-mustard (41%), chana (40%),
bajra (37%), mung (31%) and soybean (25%). The
commodities which were stored in small quantities
were kharif-mustard (6%), guar (3%) and wheat (3%).
It was interesting to find that guar growers stored it
in either house or field. However, the period of
storage in the case of guar was high. Guar being a
storage-friendly crop, many farmers store the seed for
more than a year. In the case of cotton, barley, coriander
and cumin, farmers are yet to come forward for
storage.
Table 4. Benefits accrued on utilization of warehousing facilities

Commodity
Commodity
sold
immediately
after harvest
(%)
Commodity
stored
(%)
Bajra
Guar
Mung
Mustard
Rice
Soybean
63
97
69
94
57
75
37
3
31
6
43
25
Chana
Mustard
Wheat
60
59
97
40
41
3
Average price
realized
immediately
after harvest
(` /q)
Average price
realized after
storage
(` /q)
Kharif crops
1108
1500
4272
5500
6433
5000
2843
3200
1300
1917
3250
4000
Rabi crops
2733
2600
2900
3500
1465
1983
Average
Accrued
Percentage
storage
benefit out
of farmers
cost*
of storage over availing
(` /q)
harvest period
storage
price
facility
(%)
(%)
72
163
178
106
76
115
28.9
25.0
-25.0
8.8
41.6
19.5
27.27
5.56
11.11
14.29
60.00
25.00
104
108
76
-8.7
16.9
30.1
14.29
14.29
23.08
Note: Here the cost included transportation, storage and interest opportunity. Interest opportunity was calculated by assuming
that a farmer sales immediately after harvest and puts the amount for a term-deposit at interest rate of 8 per cent per annum.
58
The benefit accrued on storage was found

maximum in rice (41.62%), followed by wheat
(30.12%), bajra (28.87), guar (25.00%), soybean
(19.54%) and rabi-mustard (16.94%). The
commodities for which farmers could not reap the
benefit of storage or suffered price loss were mung (25.05%) and chana (-8.7%).
Perceived Reasons for Slow Progress in NWR
Finance
Various reasons for limited popularity of NWR,
as perceived by farmers, are presented in Table 5. It
reveals that farmers in general are not aware about
NWR facility and its various benefits. The use of
storage facilities among farmers may be enhanced by
educating them on the importance and benefits of
warehousing in terms of price enhancement and
liquidity available under NWR. The lack of scientific
warehousing was another reason for poor response to
the initiative. High transportation cost to distant
warehouses and complex documentation also limited
the use of NWR. Immediate cash requirement was also
observed as an important reason for the limited success
of NWR facility.
Table 5. Reasons for not availing NWR facilities
(Responses of non-beneficiaries of NWR)
Reason
Lack of awareness
Lack of storage facilities
Transportation cost
Documentation involved
Immediate need of money
Response (%)
92
100
100
100
100
Perception of Bankers
It was observed that most of the leading private
banks offered loan against NWR. This business though
was skewed towards traders, as only around 8 per cent
applications were received from the farming
community. The loan against NWR was available to
producer-growers which varied in the range of 50-70
per cent. Such a variation was observed more in the
case of private banks where it was in the range of 4050 per cent based on the trade acceptance of the
commodity and extent of price fluctuations witnessed
during the past three years. In addition, high degree of
Vol. 29 (No.1)
January-June 2016
variation was observed for factors like poor availability

of relevant information on marketing components like
prices, arrivals and national and international trade
scenario of a particular commodity. Also, the decision
was primarily based on the local trade conditions
considering the impact on transportation expenditures
and other logistic arrangements required for marketing
the produce in case of default. The study reveals that
before extending loan against NWR, the banks
considered factors like ownership of warehouse
(public/private sector), business relationship,
accreditation by WDRA and insurance of the
commodity in that order.
Conclusions and the Way Forward

The finance against NWR is related not only to
warehousing and banking but also to the market.
Likewise, the agencies having integration with the
market have been found better equipped for
disbursement of pledge finance like the collateral
management service providers, APMCs and
warehousing corporation. The collateral management
service agencies, for their business structure, seem to
give preference to traders over farmers, while
marketing facilitating public agencies like APMC seem
to be more focused on smallholder farmers. The
farmers understanding about the necessary procedure
and documentation required to avail the loan, has made
means like KCC and cooperatives more popular across
the farming community.
The expected favourable price movement over time
is the factor encouraging farmers to store the produce,
but is contained by unavailability of warehouses near
the production area. The foodgrain and oilseed crops
have been found to be more popular from storage point
of view and maximum benefits on storage have also
been observed in these crops. The availability of
warehouses, complicated procedure and poor
awareness are some of the reasons limiting the use of
loan against NWR. There is a strong need to establish
warehousing infrastructure within the production area
to popularise the use of warehousing and enhance
availing of loan against NWR. Awareness about
advantages of NWR loans needs to be generated by
involving banks, Krishi Vigyan Kendras, State
Agricultural Marketing Boards, Regulated Market
Committees, NGOs, co-operatives, FPOs and even
APMCs.
The facility of loan against the stored produce is

being utilized mainly by the traders as reflected by the
higher number of applications received by the Banks
from traders. It may be an outcome of limited use of
warehouse facility by the farming community. In order
to inculcate storage habit and promote storage among
the producers, warehousing should be encouraged
under public-private partnership mode. All the leading
banks do offer loans against NWR though it is skewed
towards traders with wide variations in the percentage
of loan disbursed against the market value of the
commodity stored. The reasons for degree of variation
have been cited as poor availability of relevant
information on marketing components like prices,
arrivals and trade scenario of a particular commodity
and decision being based primarily on the local trade
conditions. There is a strong need to push NWR beyond
a negotiable instrument as it offers immense potential
to trade, short-term loan to farmers, balanced supply
of agri-commodities and enhances farmers returns.
Acknowledgements
The authors acknowledge the support and guidance
provided by the Director General, CCS-National
Institute of Agricultural Marketing, Jaipur, during the
research project. They are also thankful to the
anonymous referee for his valuable comments.
References
Baba, S.H., Wani, M.H, Wani, S.A., Zargar B.A. and N.A.
Qammer (2014a) Institutional credit to mountain
agriculture: Issues of structural changes and impact in
Jammu & Kashmir. Agricultural Economics Research
Review, 27 (Conference No.): 111-122.
Baba, S.H., Wani, M.H., Zargar, B.A., Malik, H.A. and
Showket, A. Mir (2014b) Institutional credit and capital
formation in mountain agriculture: Evidences from
Jammu & Kashmir. Indian Research Journal of
Extension Education, 14(02): 71-77.
GoI (Government of India) (2014) Pledge Finance. Ministry
of Agriculture and Farmers Welfare, New Delhi ( http:/
/agricoop.nic.in/Admin_Agricoop/Uploaded_File/
all_sta.pdf)
Jairath, M.S., Peter, V. Anu and Haque, Enamul (2014) An
imperative institutional innovation for addressing issues
of agricultural marketing A case study of
Warehousing Development Regulatory Authority.
Agricultural Economics Research Review,
27(Conference No.): 35-44.
59
Mishra, P. K. (1994) Crop insurance and crop credit: Impact

of the comprehensive crop insurance scheme on
cooperative credit in Gujarat. Journal of International
Development, 6(5): 529-568.
NABARD (National Bank for Agriculture and Rural
Development) (2014-15) Annual Report. Mumbai.
Onumah, Gideon E. (2003) Improving access to rural
finance through regulated warehouse receipt systems
in Africa. (Accessed from http://afraca.org/download/
general_rural_finance_publications/RegualtedWarehouse-Receipts-in-Africa-Improving-Access-toFinance-Best-Practices-by-Gideon-Onumah.pdf)
Rajeev, M. and Dev, S. (1998) Institutional and noninstitutional credit in agriculture: Case study of Hugli
district of West Bengal. Economic and Political Weekly,
33(47): 21 November.
Rajendra, S., Podder, H.S., Vijaya Kumar, H.G., Shankara
Murthy and Venkjatesh Murthy, T.N. (1995) Impact of
credit on cropping pattern and cropping intensity on
farms in un-irrigated and irrigated villages in Bijapur
district of Karnataka An application of linear
programming technique. Agricultural Economics
Research Review, 8(1) 29-42.
RBI (Reserve Bank of India) (2005) Report of the Working
Group on Warehouse Receipts & Commodity Futures.
Department of Banking Operations and Development,
Mumbai.
Shraddha, S. (2013) Financial innovation in Indian
agricultural credit market: Progress and performance
of Kisan Credit Card. International Journal of Ongoing
Research in Management and Information Technology,
13: 1-3.
Thejeswini, R., Kiresura, V.R., Preeti, N.S., Teggi, M.Y. and
Trilokanatha, G.A. (2014) Agricultural credit in India
Innovations in design and delivery of products and
services. Agricultural Economics Research Review, 27
(Conference No.): 75-83.
Online search
http://agricoop.nic.in
http://agmarknet.nic.in/
http://agricoop.nic.in/NCF/NCF%20Report%20-%203.pdf
http://ageconsearch.umn.edu/bitstream/183449/2/
201400194.pdf
http://afraca.org
Received: February, 2016; Accepted: May, 2016

DOI: 10.5958/0974-0279.2016.00019.7
Institutional Synergies in Processing and Value Addition:

Role of a Producers Organisation in Transforming
Farm Economy in Rural Punjab
Udeshna Talukdara* and Kamal Vattab
Department of Agricultural Economics, Assam Agricultural University, Jorhat-13, Assam
b
Centres for International Project Trust (CIPT), New Delhi-110 016
Abstract
The study has highlighted the role of a producers organisation, namely Farm Produce Promotion Society
(FAPRO), in improving farm economy in rural Punjab. The study conducted in the Hoshiarpur district of
Punjab during 2010-11, is based on the primary data collected from 140 farmers and has examined the
change in cropping pattern and increase in income and employment due to value addition at farm level,
as against cereals cultivation in the state. It has observed that turmeric as a high-value perennial crop
occupies 26.4 per cent area of GCA and the gross return of turmeric was almost double of the return from
rice and wheat. The establishment of FAPRO for value addition of turmeric has raised the farm income
and employment compared to rice-wheat cropping system. The FAPRO adds value through powder making
of turmeric and selling rhizomes for seed purpose. The Society has also been found to capitalize through
other value-added activities. Field study has indicated that FAPRO needs a strong policy intervention
with linkages between farm and the factory through skill development, increase of capacity utilization,
mechanization of processing process, development of storage facilities, increasing of operational efficiency
and reduction of overhead cost with proper post- harvest management. It will encourage farmers to
continue turmeric cultivation in the cropping system of Punjab. Such farmers organisations need to be
replicated in other parts of the country in the co-operative sector.
Key words: Producers organization, value addition, farm economy, turmeric crop, Punjab
JEL Classification: Q13, Q18, Q19
Introduction
The passive role of agri-institutions like processing
and marketing co-operatives adds value to agricultural
commodities, increases marketability of products and
strengthens farm economy. It has been identified as an
essential stage in the value chain of non-durable
Email: udeshna_talukdar07@yahoo.com
The paper is a part of the unpublished MSc thesis entitled,
An Economic Analysis of Farm Produce Promotion Society (FAPRO) A Case Study, submitted by the first author to the Department of Economics and Sociology , Punjab
Agricultural University, Ludhiana in the year 2013.
perishable products which get obscure with high

wastage, affecting the normal profit of the crop. The
processing and post-harvest management of
agricultural commodities has gained impetus due to
high wastages before they reach the market (Dodamani,
2007 and Anonymous, 2011). The economy of the
country will also get accelerated with increased export
if there is proper policy push for agro-processing
(Singh, 2004). Value addition through processing of
agricultural produce will capitalise the economy of
rural India and minimize migration from village to
towns (Thorat et al., 2003). Processing is an important
marketing function for agricultural commodities
62
creating form utility and its importance has grown

manifold in the recent times (Acharya et al., 2011 and
Brahm, 2000). Out of the total food marketed, the
value-added products in India shared about 35 per cent
(Chandrasekharam, 2001) and it was the largest in terms
of production, consumption, export and growth
prospects. With annual production of 205 million
tonnes of fruits and vegetables, India is the second
largest country in farm production in the world. In
contrast, countries like USA (65%), China (23%) and
Philippines (78%) are far ahead of India in reducing
wastages and enhancing value-addition and shelf-life
of farm products.
In India, processing of agricultural produces is
highly decentralized in small households having small
capacities (Sidhu, 2005). The co-operative marketing
and processing societies, established to overcome the
problems arising out of marketing system of
agricultural produce, are making good progress in
scientific and efficient marketing. Moreover, the
progress of regulated markets is not uniform in all the
areas of country. The need for strengthening cooperative organization has, therefore, been recognized
for marketing of agri-produce, in addition to cooperative marketing. The value of agricultural produce
marketed through co-operative marketing societies, has
increased many fold more than 75-times during the
past forty years (Baviskar, 2009). The produce
marketed through these societies accounts for 8-10 per
cent of the marketed surplus. The co-operatives are
one of the core segments of Punjab economy. An
important aspect is the processing of farm produce at
farmers level so that they can get a better price for
their produce and earn higher profit. Therefore, there
is an urgent need to introduce processing units at a
level where farmers are directly involved and can get
direct benefits from the processing units. Since it cannot
be done at individual level, the promotion of farmers
societies for marketing is imperative.
Farms Produce Promotion Society
The establishment of Farms Produce Promotion
Society (FAPRO) in the Ghugial village in Hoshiarpur
district of Punjab in collaboration with the State
Department of Agriculture and more than 300 farmers
in the year 2001 and registered under Societies
Registration Act XXI of 1860, has paved the way for
value-addition, processing and marketing of farm
Vol. 29 (No.1)
January-June 2016
produce of the district. With the objectives of selfmarketing and processing of farm produce, the society
established processing plants for honey and turmeric
in 2005 under the Rastriya Shram Vikas Yojana
(RSVY) of Government of India. It encourages
integration by combining production, processing,
value-addition and marketing of finished products for
increasing profitability in farming, generating
employment and increasing market power. The FAPRO
is generating direct or indirect employment for more
than 500 farm families in the area. The organization
has a special focus on helping the marginal and small
farmers, who are the most vulnerable section of the
farming community.
Apart from turmeric and honey, FAPRO is also
involved in processing and marketing of agricultural
products such as turmeric powder, honey, gram floor,
pulses, pakora, soybean products, jaggery and other
seasonal products such as groundnut and kinnow. The
organization intends to expand the quantity of marketed
products by onsite sales, outlets sales and participation
in farmers fairs and exhibitions in the nearby areas.
The organization aims to expand its operations by
increasing the coverage of their current activities and
starting some new activities such as packaging of
processed products, use of mobile units for marketing
of produce, production of cattle feed, etc. As the success
of such producers organizations can pave the way for
a significant rise in farm incomes and crop
diversification in the state, the present study was carried
out to find the economic viability of the FAPRO and
its overall impact on the farmers income in Punjab.

Both primary and secondary data were collected
for the study. The primary cross-section data were
collected pertaining to the year 2010-11. Ten villages
were selected randomly out of the villages belonging
to the member- farmers of FAPRO. Out of these 10
villages, 140 turmeric cultivators were identified and
40 of them were selected for the study proportionately
based on farm size. Thus, 1 marginal, 11 small, 17
medium and 11 large farmers were selected. Primary
data were collected by personal interview method using
a pretested structured questionnaire. The secondary
data were collected from the published audited account
sources of the FAPRO society.
Talukdar and Vatta : Institutional Synergies in Processing and Value Addition
The data collected through the interview schedule

were converted to master tables which facilitated
tabulation of data in the desired form. The collected
data were then grouped into tables and were analyzed
using various statistical tools like percentage and
simple average. The analysis was mainly confined to
value addition of turmeric by FAPRO in Punjab.
63
The cropping pattern being followed by the

members of FARPO seemed to be more diversified than
the overall cropping pattern of Punjab state, as revealed
from Table 1. In gross cropped area (GCA), the share
of turmeric was higher (26.4%) than under each of rice
and wheat (20.8%) or kharif and rabi vegetables (11.6%
each). The high value of Simpson Index of Diversity
(SID: 0.813) showed that the cropping pattern of the
FAPRO members was more diversified. These farmers
were growing turmeric and vegetables and also were
not confined to the monoculture of rice and wheat.
turmeric from rice and wheat crops (Talukdar et al.,

2015). The cost of cultivation of turmeric, wheat and
rice in Punjab is presented in Table 2. The seed
appeared to be the most expensive component of
cultivation of turmeric, as compared to rice and wheat
crops. The expenses on labour-use were also
higher for turmeric. In all, the variable costs were
` 71919/ ha for turmeric, ` 19204/ha for wheat and
` 23966/ha for rice, which implies that the costs were
almost 67 per cent higher in turmeric cultivation than
in cultivation of rice and wheat. The gross returns from
turmeric were almost double the combined gross returns
from rice and wheat, these being ` 2,93,930 /ha,
` 86,697/ha and ` 62,985/ha for turmeric, wheat and
rice, respectively. The returns over variable costs were
` 222011/ha, ` 67492/ha and ` 39018/ha for turmeric,
wheat and rice, respectively. It clearly demonstrates
that turmeric cultivation is much more remunerative
as compared to rice-wheat rotation. The cultivation of
turmeric by the FAPRO members not only helped in
raising the farm income but also the extent of
employment, as turmeric cultivation is more labourintensive than rice or wheat rotation.
Economics of Turmeric Cultivation at Farm Level
Capital Investment in Value-addition of Turmeric
The establishment of FAPRO has encouraged the

farmers to diversify area under high-value crops like
The extent of investments made by the FAPRO

for value addition in turmeric and cost and returns
associated with such value-addition was also examined.
It was found that a total investment of ` 78.5 lakh was
made by the FAPRO for value-addition in turmeric.
The major components of investment were dryer (` 60
lakh), de-humidifier (` 8 lakh), automatic-cumvibrator- sevier (` 0.60 lakh), diesel-fired oven (` 1.4
lakh) and automatic packing line (` 8.5 lakh).
Investment on the drier was the most important
component with a proportion of 76.4 per cent, followed
by automatic packing line (10.8%) and de-humidifier
(10.2%) of the total capital investment.
Results and Discussions

Cropping Pattern, Extent of Diversification and
Returns
Table 1. Cropping pattern of member-farmers of

FAPRO, 2010-11
(ha)
Name of crop
Turmeric
growers
Kharif crops
Rice
5.46
Turmeric
6.94
Sugarcane
1.36
Vegetables
3.04
Fodder
0.49
Sub-total
17.29
Rabi crops
Wheat
5.46
Vegetables
3.04
Fodder
0.49
Sub-total
9
Gross cropped area
26.29
Simpson Index (SID)
-
Per cent share

of GCA
20.8
26.4
5.20
11.6
1.9
65.8
20.8
11.6
1.9
34.2
100.0
0.813
Quantity of Input and Output during Valueaddition

The value-addition in turmeric is primarily done
under five different steps, viz. washing, boiling, drying,
polishing and grinding, in addition to branding, packing
and labelling and standardization. Table 3 depicts the
quantity of input and output during different steps of
value addition in turmeric. It was found that at every
step there was a loss in weight, except in boiling where
64
Vol. 29 (No.1)
January-June 2016
Table 2. Costs on cultivation of turmeric, rice and wheat of sample farmers, 2010-11
(`/ha)
Costs structure
Turmeric
Land preparation
Sowing
Fertilizer and weed control
Irrigation
Harvesting and threshing
454
49400
5063
1057
Cost on human labour
Land preparation
1143
Sowing
988
Fertilizer application and weed control
1817
Irrigation
988
Harvesting and threshing
7377
Transportation
988
Interest on variable costs @ 9 % for half crop period
3094
Total variable costs
71919
Gross returns
293930
Returns over variable costs
222011
(67.58)
Wheat
Rice
3396
1180
4853
741
2470
2852
686
5399
741
-
988
918
988
2470
247
123
827
19204
86697
67492
(20.54)
876
741
1111
4940
4720
864
1032
23966
62985
39018
(11.88)
Note: Figures within the parentheses indicate percentage share of total return over variable cost for all crops
Table 3. Quantity of input and output during valueaddition in turmeric at FAPRO
was raised by 174.28 per cent after processing as

powder (Appendix I).
Steps
Income from Value-addition in Turmeric
Raw material
Washing
Boiling
Drying
Polishing
Grinding
Quantity of starting
material (kg)
Quantity of end
material (kg)
100
100
98
100
20
18
100
98
100
20
18
16
it gained. Table 3 reveals that one quintal of raw

turmeric provides 16 kg of turmeric powder.
Value-addition, Returns and Input-Output Ratio
It was found that the total cost of processing,
including marketing and distribution, was ` 9.16
thousand/q. The market price of turmeric powder being
` 12 thousand/q, the net profit realization was ` 2.84
thousand/q. Thus, the cost -return ratio was1:1.31. This
was also similar to the benefit-cost ratio of processing
of turmeric in Karnataka (Lokesh et al., 2004). It was
observed that the value of raw turmeric for ` 700.00
Income from Turmeric Powder
The FAPRO processed raw turmeric and produced

224 quintals of turmeric powder from 1400 quintals of
raw turmeric in a year. The annual gross income was
` 26.88 lakh with a net income of ` 6.36 lakh. The
production of turmeric powder was found costly and
97.78 per cent of total cost was shared by operational
cost. This high production cost was due to underutilization of plant capacity (only about 5%) and,
therefore, FAPRO should take measures to enhance
plant capacity utilization (Appendix I).
Income from Turmeric Seeds
FAPRO also sells rhizomes of turmeric for seed
purpose. Table 4 shows that gross income earned from
selling turmeric seeds per year was ` 57.60 lakh with a
net return of ` 46.14 lakh. The organization processed
1200 quintals of turmeric for seed purpose and
recovered 46 per cent of turmeric seeds which were
sold at ` 4800 per quintal.

Table 4. Economics of value addition from turmeric seeds
Cost head
1.
Fixed costs
Depreciation on plant
2.
Land rent
3.
Salary of permanent workers
4.
Others
Total fixed costs
5.
6.
7.
Operational costs
Costs of raw turmeric (1200 q @ ` 7/kg)
Direct cost of labour (1200 human days
@ ` 200/ human day)
Packing, labelling and marketing
Total operational costs

Total costs
Seed recovery (%)
Gross return from seed (1200 q @ ` 4800/q)
Net profit over variable cost
B-C ratio
Value (`)
6582
(0.57)
2961
(0.26)
20886
(1.82)
8642
(0.75)
39071
(3.40)
840000
(73.27)
240000
(21.00)
26670
(2.33)
1106670
(96.60)
1145741
(100)
46
5760000
4614259
1:5.02
Notes
1. Figures within the parentheses indicate percentage to
total
2. Depreciation The plant was used for different activities
and the depreciation per quintal of turmeric powder
produced during the year was estimated proportionately
based on the total depreciation estimated by the
management (CA) of the Society.
3. Land rent Prevailing land rent at Ghugial village was
considered per unit of the produce
4. Salary of permanent workers Permanent workers were
engaged in the Society as workers and managers on fixed
pay basis. The total estimated cost for pay and wages by
the management was proportioned and was estimated
per quintal basis of turmeric.
5. Others These included expenses for electricity,
advertisement and market development, telephone,
travelling, news papers, repairs and maintenance and
miscellaneous expenses per unit of the produce.
65
Table 5. Generation of employment from processing of

turmeric at FAPRO, 2010-11
(Human days/tonne)
Particulars
Labour
use
Washing and boiling
0.4
(8.3)
2.0
(41.7)
0.8
(16.6)
1.6
(33.4)
4.8
672
1200
Drying
Polishing and grinding
Packaging and labelling
Total labour-use
Labour-use at FAPRO for turmeric powder
Labour-use at FAPRO for turmeric seed
production
Total employment generation at FAPRO
1872
Note: Figures within the parentheses are the percentages of

total labour use.
Contribution of FAPRO to Employment

Generation
The establishment of producers organizations such
as FAPRO enhances the profitability of existing
activities through value addition and introduction/
promotion of some new activities. These organizations
also help in generating employment opportunities
firstly directly at the premise of the organization due
to labour needs for value-addition activities and second
indirectly by encouraging farmers to diversify labourintensive crops and enterprises which help in
augmenting household incomes. In the present case,
the employment generation through FAPRO is
perceived to have occurred at the FAPRO premises
where labour is required for turmeric processing. In
addition, employment is also generated during the
marketing of various value-added products.
Employment Generation at FAPRO Premises
The employment generation for processing of
turmeric is highlighted in Table 5. It shows the
requirement of labour to carry out various operations
of turmeric processing. A significant amount of labour
is also involved in separating turmeric to be used as
seed from the raw turmeric.
66
Value-addition of other Agro-produces at

FAPRO
Apart from processing of turmeric and honey,
FAPRO also undertakes value-addition of some other
agro-produces, though it was restricted largely to
cleaning and packaging. It helped in utilizing the space
and human labour and also the marketing network of
FAPRO which required various commodities in which
value-addition is undertaken by FAPRO to the extent
of 2 to 46 per cent (Appendix II.).
Conclusions
The paper attempts to highlight the role of a
producers organization, namely Farm Produce
Promotion Society (FAPRO), in improving farm
economy in a village of Hoshiarpur district in Punjab.
The paper has studied value-addition and employment
generation through processing of raw turmeric in the
area. The value- addition is being done through
processing of raw turmeric to powder form and
marketing of seed rhizomes. The study has revealed
that the value of turmeric was raised to 174 per cent
through processing and other post-harvest operations.
The FAPRO could also generate a total of 1872
human days of labour in a year. Some other produces
for which value-addition was being undertaken by
FAPRO included mirch, dalia, papad, pokora and
warian. The field survey and economic analysis of
value addition at FAPRO indicated that the organisation
needs a strong policy intervention with linkages
between farm and the factory through skill
development, enhancing of capacity utilization,
installation of mechanical drier and scientific storage
facilities and reduction of overhead cost. Adoption of
high performance multipurpose machine and product
development strategies like better packing, labelling,
branding with HACCP certification and imposition of
national and international quality standards will create
additional marketability of the products with long-term
viability of FAPRO. This will encourage farmers to
continue turmeric as a high-value crop in the cropping
system of Punjab. Such farmers organisations need to
be replicated in other parts of the country in the
cooperative sector.
Vol. 29 (No.1)
January-June 2016
Acknowledgements
The authors are indebted to the unknown referee
for his critical evaluation and constructive suggestions
for improvement in the paper.
References
Acharya, S.S. and Agarwal, N.L. (2011) Agricultural Marketing in India. Oxford and IBH Publications, New Delhi.
Anonymous (2011) Ministry of Food Processing Industries.
Statistical data base:http:mofpi.nic.in.
Baviskar, S.B. (2009) Cooperatives in Maharashtra:
Challenges ahead. Economic and Political Weekly, 21:
4217-21.
Brahm P. (2000) Growth of fruit and vegetables processing
industry in India: A major technological change in
agricultural marketing. Indian Journal of Agricultural
Marketing 14: 72-79.
Chandrasekharam, D. (2001) Food processing industry and
geothermal: Indian scenario. Geo Heat Center Bulletin,
23: 8-12.
Dodamani, M.T. (2007) Production and Value-addition in
Naturally Coloured Cotton under Contract Farming
An Economic Analysis. MSc Thesis. University of
Agricultural Sciences, Dharwad.
Lokesh, G.B. and Chandrakanth, M.G. (2004) Economics
of production, marketing and processing of turmeric in
Karnataka. Indian Journal of Agricultural Marketing,
18: 32-44.
Sidhu, M.S. (2005) Fruits and vegetable processing industry
in India An appraisal of the post-reform period.
Economic and Political Weekly, 28: 3056-61.
Singh, K. (2004) Emerging trends in agro processing sector.
Indian Journal of Agricultural Marketing, 59: 655-68.
Talukdar, U. and Vatta, K. (2015) Producers organisation
and economics of cultivation of turmeric as a high-value
crop against rice-wheat cropping system for increasing
farm income: A case study in Hoshiarpur district of
Punjab. Economic Affairs, 60: 29-32.
Thorat, V.A, Tilekar, S.N. and Bhosale, S.S. (2003) Potential
of Kokum processing for employment and income
generation A case study. Indian Journal of
Agricultural Marketing, 58: 602.
Received: April, 2015; Accepted: December, 2015
67
Appendix I. Economics of production of turmeric powder at FAPRO, 2010-11

(Value in000 rupees)
Cost head
(i) Depreciation
(ii) Land rent
(iii) Salary of permanent workers
(iv) Others
A) Total fixed cost
Operational costs
(i) Raw material consumed (1400 q @ ` 700/q)
(ii) Direct human labour costs
(a) Washing and boiling
(b) Sun drying and storage
(c) Polishing
(d) Grinding
(e) Packing
(iii) Cost of electricity and fuel
(a) Washing
(b) Boiling
(c) Polishing
(d) Grinding
(e)Packing
(iv) Cost of bagging
(v) Cost of marketing
(vi) Repair and maintenance
(vii) Taxes and fees
(viii) Stationeries
(ix) Interest on working capital
B) Total operational cost
C) Total cost (A+B)
D) Total production of powder (q)
E) Total cost of processing per quintal (`/q)
F) Price of turmeric powder (`/q)
G) Net profit (`/q)
H) B-C ratio
Value
Per cent in total
8
4
24
10
46
0.38
0.17
1.18
0.49
2.22
980
47.76
11
56
11
11
45
0.55
2.73
0.55
0.55
2.18
2
403
25
45
22
90
38
22
5
17
223
2006
2052
224
9
12
3
1:1.31
0.10
19.64
1.22
2.18
1.09
4.37
1.84
1.06
0.24
0.85
10.87
97.78
100
68
Vol. 29 (No.1)
January-June 2016
Appendix II: Value-addition of other commodities being undertaken at FAPRO, 2010-11

Activities
Packed organic basmati

Besan
Cattle Feed
Chana
Chick pea
Crushed wheat grain
Lentil
Green gram
Dry pea
Jaggery
Jaggery special
Spice powder
Mash
Lentil (seed)
Chilli
Moth
Green gram (seed)
Green gram organic
Namkeen
Pakora
Pakorian
Papad
Kidney bean
Red bean
Rongi
Shakkar
Soya nuts
Other spices
Quantity
marketed
(kg)
Purchase
price
(`/kg)
Sale price
(`/kg)
Value-added over
purchase price
(`/kg)
79
19221
48 bags
4404
7074
216
5445
4691
349
1780
49
294
3690
2278
280
2866
1494
47
108
2129
824
75
1805
1952
173
200
450
146
115
42
675
40
42
30
45
61
35
37
45
297
51
40
102
45
59
80
80
70
73
112
56
57
63
43
110
110
130
50
688
47
48
43
53
69
39
41
56
370
60
48
161
55
68
94
93
101
106
154
62
64
69
49
142
145
16
8
13
7
6
13
8
8
4
4
11
72
9
9
59
10
8
14
13
32
33
42
6
7
6
6
32
35
Per cent value

addition over
purchase price
13.54
19.29
1.97
16.97
13.76
44.43
18.31
12.60
10.14
11.39
25.35
24.29
18.26
21.38
58.27
23.40
14.28
17.45
16.75
45.87
45.61
37.31
10.25
11.93
9.49
13.11
29.08
31.86

DOI: 10.5958/0974-0279.2016.00020.3
Farm Business Income across Land-size Classes and Land

Tenure Status: A Field Study in Assam Plains
Binoy Goswami
Faculty of Economics, South Asian University, New Delhi-110 021
Abstract
Using farm level data from the plains of Assam, the paper has estimated farm business income across
different land-size classes and land tenure status. The analysis has been carried out at the aggregate level
as well as at disaggregate level for three specific crops, viz. winter paddy, summer paddy and winter
vegetables. It has been found that sharecropping and fixed rent tenancy contracts have a negative and
significant impact on farm business income. The lower level of farm business income on leased-in land,
especially under sharecropping, can be attributed to payment of a significant amount as rent, which is
higher than even that stipulated in the tenancy law. Accordingly, certain reforms in the existing tenancy
law have been suggested. The study has suggested a shift in the cropping pattern from the presently
predominant winter paddy to more remunerative crops such as vegetables, which is also desired for a
healthy transition of Assam agriculture from subsistence cultivation to a profitable venture.
Key words: Farm business income, land-size, land tenure, sharecropping, fixed rent tenancy, paddy,
Assam
Introduction
The analysis of income generation through crops
cultivation has important implications. The knowledge
about differential levels of income from cultivation of
different crops may help the farmers to utilize their
often limited resources efficiently. This in turn may
contribute to increase in income level and uplifting of
living standard of the farmers. Moreover, understanding
of the factors that contribute to variations in income
levels across farm households may help the policy
makers to design policies on improving the economic
conditions of farmers in an effective manner. Against
this backdrop, the present paper, based on data
generated through a primary survey in the plains of
Assam, has estimated farm business income generated
through farming and from different crops individually.
Email: binoygoswami@sau.ac.in
The paper has identified the factors that cause variations

in income levels across farm households. Among the
factors that may potentially affect income generation,
two factors, viz. land-size and land tenancy, have been
given more attention in this paper.
The study has been conducted in the state of Assam,
the largest state in North-East India, where agriculture
still contributes a substantial proportion (24.44% in
2009-10) to the gross state domestic product (GSDP)
and more than 50 per cent of the workforce is engaged
in agriculture (GoA, 2012). Considering the importance
of the sector, the issue of what impacts the income
generation in crops cultivation and thereby the lives
of more than half of the workforce, becomes very
pertinent. In this context, the role of various forms of
land tenancy contracts and land-size is worth examining
for the following reasons. Assams agriculture is
dominated by paddy cultivation and tenancy is
widespread in the paddy production in the state
70
(Bezbaruah, 1994; Lal Bahadur Shastri National

Academy of Administration, 1994; Gautam, 1995;
Kuri, 2003). The present study has found that about
one-third of the sample area is under tenancy with
sharecropping as the dominant form of tenancy
contract. In terms of land-size, almost all the sample
farmers are either marginal, small or at best medium
farmers with the average size of operational holding
of 1.34 ha. There is not a single farmer in the sample
operating on more than 6 ha of land. In view of the
widespread incidence of tenancy and predominance of
smallholders farming, the study has analysed farm
business income across the available land-size classes
under different forms of tenancy contracts in Assam
agriculture.

In Assam, about 81 per cent of the total
geographical area comprising Brahmaputra Valley and
Barak Valley is plain and the remaining 19 per cent
constitutes hills. The hills and the plains are
distinctively different in terms of both agricultural
practices and institutions. While shifting cultivation is
still predominant in the hills, settled cultivation is
practised in the plains. On the other hand, land is owned
privately in the plains, but the transition to individual
ownership of land from community ownership is not
yet complete in the hills. Hence, the present study was
limited to the plains of Assam.
The study is based on the primary data collected
using multi-stage sampling technique during 2011-12.
At the first stage, four non-contiguous districts, three
from Brahmaputra Valley and one from Barak Valley
and then one development block from each of these
districts, were selected purposively. These selected
districts fall in four distinct agro-climatic zones, viz.
Dibrugarh in Upper Brahmaputra Valley Zone,
Morigaon in Central Brahmaputra Valley Zone, Nalbari
in Lower Brahmaputra Valley Zone and Cachar in
Barak Valley Zone. The development blocks selected
from these districts were Borboruah, Mayang, Chamata
and Narshimpur. Then three villages were selected from
each block and finally 7-10 per cent of farm households
from each village were selected randomly for the
survey. In this way, a total of 221 farm households
were selected for interview.
Vol. 29 (No.1)
January-June 2016
Line of Analysis
The paper has estimated two variants of farm
business income, viz.
(i) Farm business income1 (FBI-1) which denotes
return over variable costs per hectare of land
without considering cost of family labour, and
(ii) Farm business income2 (FBI-2) which is return
over variable costs plus the imputed costs of family
labour per hectare of land.
The farm business income from the overall
cultivation as well as from three individual crops, viz.
winter paddy, summer paddy and winter vegetables,
was analysed at two levels. At the first stage, FBI-1
and FBI-2 were analysed in terms of land-size classes
and land tenure status through tabular analysis. Then,
a multiple regression analysis was carried out to
examine the effect of land tenure status and land size
on FBI-1 and FBI-2 more rigorously by controlling
for interferences of other factors. Both FBI-1 and FBI2 were regressed on land-size and tenure status besides
certain control variables. Since farm business income
may depend on factors other than land-size and land
tenure status of farmers, it is important that the effects
of those variables are controlled for to isolate the effects
of land-size and tenure status. The theoretical
justifications for inclusion of these variables along with
their definitions are given below.
The independent and control variables considered
in the regression analysis were broadly divided into
five categories, viz., farmers characteristics, tenure
status, input intensity, enabling factors and locational
dummy. The farmers characteristics included his age
(AGE), education level1 (EDU) and land size in hectare
(FS). Since the age of a farmer reflects his experience,
it was expected to contribute positively to farm business
income. Likewise, education should also have a
positive impact. The land size could have either a
positive or negative effect2 (Goswami, 2012). The
tenure status included two variables, viz., area under
sharecropping as a percentage of operational holding
(ASC) and area under fixed rent as a percentage of
operational holding (AFR)3. Sharecropping is expected
to have a negative effect on income generation
(Marshall, 1920). However, the effect of fixed rent
contract on farm business income might be either
positive or negative positive as it might contribute
Goswami : Farm Business Income across Land-size Classes and Land Tenure Status
FBI-1 i = 0 + 1AGE i + 2EDU i + 3FS i +

4ASCi + 5AFRi + 6LABi + 7TILLi
+ 8 IRRI i + 9 HYV i + 10 NPK i
+11EXTi +12CREi + 13CIi + 14D1i +
15D2i + 16D3i + U i
(1)
to productivity yield4 and negative if the tenants needed

to pay a higher rent and had higher input intensity.
The input intensity included the following five
variables: labour cost per hectare of operational holding
(LAB)5, tilling cost per hectare of operational holding
(TILL), area under irrigation as a percentage of
operational holding (IRRI), area under HYVs as
percentage of operational holding (HYV) and fertiliser
consumption measured in terms of NPK per hectare of
operational holding (NPK).
It was assumed that the farmer was rational and
given his experience in farming, he would apply inputs
till the level where the contribution of the inputs to
income is positive. The enabling factors included access
to extension service (EXT) and access to credit
(CREDIT). While access to credit should enable the
farmers to apply productivity-enhancing inputs, the
access to extension services should enable the farmers
to apply these inputs in a scientific way. Besides,
farmers having access to extension service might also
have access to market information. Thus, these two
factors were expected to influence farm business
income positively. Access to extension has been used
as a dummy variable, where D = 1 if the i-th farmer
had received any direct benefits from the governments
extension service network; D = 0, otherwise6. Similarly,
access to credit is also a dummy variable, where, D =
1, if the i-th farmer had access to institutional credit
and otherwise D = 0. Finally, since the data used in the
regression analysis came from a sample of households
covering four different agro-climatic zones, three
locational dummies have been introduced to control
for the impact of agro-climatic variations and
differences in cropping pattern and soil quality on farm
business income. Thus, taking Dibrugarh as the
reference category, the three dummies used were: D1,
D2 and D3, where D1 = 1 for Morigaon, 0 otherwise;
D2 = 1 for Nalbari, 0 otherwise; and D3 = 1 for Cachar,
0 otherwise. A-priori, it was not possible to predict as
to what signs the coefficients of these locational
dummies would take. In addition, the regression of farm
business income had another variable which was
cropping intensity (CI). The higher level of cropping
intensity would mean higher farm business income,
implying that the variable would have a positive impact
on farm business income.
Thus, after incorporating the variables mentioned
above, the following two linear multiple regression
equations were arrived at for estimation.
71
FBI-2 i = 0 + 1AGE i + 2EDU i + 3FS i +

4ASCi + 5AFRi + 6LABi + 7TILLi
+ 8IRRI i + 9HYV i + 10 NPK i +
11EXTi + 12CREi + 13CIi + 14D1i +
15D2i + 16D3i + U i
(2)
where, Ui is the random disturbance-term which is
assumed to be normally distributed with zero mean.
The estimates of the parameters were obtained
using software STATA 10.0.
Since the data used in this exercise came from a
cross-section of farmers, before estimating the model,
the Breusch-Pagan/Cook-Weisberg test was applied to
check for the presence of heteroskedasticity in the data
set. The result of the test showed the presence of
heteroskedasticity and consequently, the robust
standard error was estimated.
Profile of Sample Farmers in Assam
Table 1 shows the distribution of the sample farm
households and areas under different size classes of
operational holding. Table 1 reveals that in terms of
both number and area, most of the sample farmers were
marginal (<1 ha) or small (1-2 ha) farmers.
There are four categories of households in the
sample in terms of their tenure status. Of all, the ownerTable 1. Percentage distribution of farm households and
areas under different size classes of operational
holdings
Operational
holdings
(in ha)
0-1
1-2
2-3
3-4
4-5
5-6
Sample
households
(%)
Sample
areas
(%)
38.01
38.46
16.30
3.62
2.71
0.90
16.59
36.27
26.78
8.64
8.48
3.24
Source: Authors calculations based on field survey data
72
Vol. 29 (No.1)
January-June 2016
Table 2. Tenure status-wise cropping pattern in Assam

(in per cent)
Tenure status
Owner Operator
Sharecropping
Fixed rent
Overall
Winterpaddy
Summerpaddy
Winter
vegetables
Rapeseed &
mustard
Potato
Jute
58.35
88.37
22.12
58.18
21.67
9.40
39.36
22.09
9.15
21.54
9.37
7.72
1.31
16.20
7.78
2.21
0.48
0.39
1.74
0.90
0.44
0.39
0.84
Note: (i) The figures represent the percentage shares of the crops in the total cropped area;
(ii) - means no farmer was available in that size class.
Source: Authors computation
operator-cum tenant is the predominant category (37.10

%), followed by owner operator (33.48 %), pure tenant
(16.30 %) and owner operator-cum-lessor (13.12 %).
It may be mentioned that for analysis of income
generation in terms of tenure status, we have considered
mainly three categories of operational holdings.
Operational holdings of the owner operators and the
owner operator-cum-lessor and the owned land portion
of the owner operator-cum-tenant have been considered
as owned land. On the other hand, the operational
holdings with pure tenants and the leased-in portion of
the owner operator-cum- tenant have been considered
as leased-in land. Again, within leased-in operational
holding, we have taken following two categories: land
under sharecropping and land under fixed rent.
Table 2 presents the cropping pattern of three
categories of farmers. The major crops that the owner
operators grew were: winter paddy (58.35 %), followed
by summer paddy (21.67 %) and winter vegetables
(9.15 %). The sharecroppers, however, predominantly
grew winter paddy (88.37)7. On the other hand, the
fixed rent tenants mainly grew the following crops:

summer paddy (39.36 %), winter paddy (22.12 %),
winter vegetables (21.54 %) and rape and mustard
(16.20 %).
Farm Business Income of Sample Households in
Assam
FBI-1 and FBI-2 from Overall Cultivation
Table 3 presents farm business income from all

crops grown during one year by farm households across
different size-classes of operational holdings under
different tenure status. The figures in Table 3 can be
been compared in the following two ways: (i) FBI-1
and FBI-2 in a land-size class across tenure status, and
(ii) FBI-1 and FBI-2 by a particular group of farmers
across size-classes of operational holding.
While farm business income (both FBI-1 and FBI2) under fixed rent in the land-size classes of 0-1 ha
and 1-2 ha was lower as compared to that on owned
land, reverse was the case in the subsequent land-size
Table 3. FBI-1 and FBI-2 under different tenure statuses across different size-classes of operational holdings
(`/ha)
Income type
FBI-1
FBI-2
Tenure status
Owner
Sharecropper
Fixed rent
Owner
Sharecropper
Fixed rent
Source: Authors Computation
0-1
1-2
28414
4377
27041
25530
575
23311
20281
3382
10081
17884
104
7002
Land-size classes (in ha)

2-3
3-4
19712
21570
17323
20338
19244
-1568
41019
16948
-2447
40033
4-5
5-6
13880
11975
-
33002
32278
-
classes. On the other hand, farm business income,

especially the FBI-2 of the sharecroppers was
negligible in comparison to other two groups of farmers
across all land-size classes. In fact, the sharecroppers
in land-size class of 3-4 ha incurred losses. The reasons
behind such a meagre income of sharecroppers as
compared to other two groups of farmers are: (i) they
cultivate only winter paddy (Table 2) and hence their
cropping pattern is less diversified, and (ii) they have
to pay half of the produce as rent.
The study has also revealed that yield and input
intensities on sharecropped land in all size-classes are
the lowest as compared to the owner operators and the
fixed rent tenants (Appendix Table 1). This could be
one of the reasons for the small farm business income
under sharecropping. However, this reasoning does not
hold good in the case of farm business income on
owned land and land under fixed rent tenancy. While
farm business income under fixed rent in the land-size
classes of 0-1 ha and 1-2 ha is lower vis-a-vis owned
land and opposite is the case in the subsequent landsize classes, yield and input intensities on the land under
fixed rent tenancy in all size-classes are higher than
those on owned land (Appendix Table 1).
One reason as to why the farm business income in
the lower land-size classes is less under fixed rent
tenancy relative to owned land, in spite of higher input
intensity and yield, may be payment of a substantial
amount as rent which the owner operators do not incur.
Nevertheless, some fixed rent tenants (especially in
higher land-size classes with better resources) may
generate more income if yield gains from intensive
application of better quality inputs outweigh the higher
overall cost of production as compared to the owner
operators. The information provided in Appendix Table
1 shows that though yield is higher on land under fixed
rent than owned land in all size-classes, the difference
is much larger in the higher land-size classes.
The study has found that farm business income of
owner operators falls as land-size increases, except for
the largest land-size of 5-6 ha. On the other hand, the
cost of cultivation first falls from ` 20443/ha in landsize class of 0-1 ha to ` 16929/ha in land-size class of
1-2 ha (Appendix Table 1) and then increases in all the
subsequent land-size classes. Thus, the fall in farm
business income in the higher land-size classes may,
at least partly, be attributed to the increase in costs of
cultivation8. The jump in farm business income of
73
owner operators in the largest land-size class (5-6 ha)

may be due to higher yield realized by these farmers
due to economies of scale. It can be observed from
Appendix Table 1 that while the increase in per hectare
cost of cultivation between land-size classes of 4-5 ha
and 5-6 ha is ` 4664, the corresponding increase in
yield is ` 24,967. Here, the higher yield might have
outweighed the increase in cultivation cost sufficiently
to result in higher income in that land-size class.
In the case of sharecroppers, an inverse relationship
has been observed between land-size and income; both
FBI-1 and FBI-2 decreased as land-size increased. In
fact, the sharecroppers in the land-size class of 3-4 ha
(the largest for sharecroppers) incurred a loss. Although
cultivation cost in this land-size class is lower, the fall
in yield is much bigger vis--vis in the smaller landsize classes, causing the sharecroppers in this land-size
class to incur loss. On the other hand, the relationship
between the land size and farm business income for
the fixed rent tenants is not clear. While the pattern of
farm business income on various land-size classes does
not match with that of cost of cultivation, it is similar
with that of yield (Appendix Table 1). This suggests
that farm business income for the fixed rent tenants
may be influenced more by yield than by cost of
cultivation.
The above analysis being at the aggregate level,
may not reveal certain crop-specific information.
Hence, we conducted crop-specific analysis of farm
business income by selecting three important crops,
viz. winter paddy, summer paddy and winter
vegetables.
Farm Business Income from Selected Crops
Analysis of FBI-1 and FBI-2 from Winter Paddy
Table 4 presents land-size-wise farm business

income from cultivation of winter paddy by owner
operators and sharecroppers. Though winter paddy is
grown on 22.12 per cent of the cropped area under
fixed rent, but the number of farmers growing the crop
was small (only 15), and hence this group of farmers
was not been included in the analysis. Except for the
land-size class of 0-1 ha, the FBI-1 was more on the
owned land than the sharecropped land. In fact, when
family labour was accounted for in FBI-2, the figures
in all land-size classes turned negative for the
sharecropped land. The reason for this could be
74
Vol. 29 (No.1)
January-June 2016
Table 4. Tenure status-wise FBI-1 and FBI-2 from cultivation of winter paddy in different land-size classes
(`/ha)
Income type
Tenure status
0-1
FBI-1
FBI-2
Owner
Sharecropper
Owner
Sharecropper
2227
3088
377
-195

1-2
2-3
3-4
4243
2719
2641
-271
1322
-172
-
10385
-1334
8282
-2081
4-5
-5196
-5860
-
payment of 50 per cent of produce as rent. From the

study, the relation between land-size and farm business
income on the owned land is not clear. However, a
comparison of farm business income and yield and
costs of cultivation (Appendix Table 2) on owned land
suggests that variations in farm business income across
land-size classes are probably influenced more by the
variations in yield than in cost of cultivation. For
example, the owner operators in the land-size class of
4-5 ha in which they incurred loss, experienced a fall
in yield of ` 10,985/ha and an increase in cultivation
cost of only ` 3,157/ha, as compared to those in the
land-size class of 3-4 ha. On the other hand, FBI-1 is
more in the lower land-size classes under
sharecropping, the reason for which is the use of more
family labour in the small plots (see Appendix Table
2).
The losses incurred in terms of FBI-2 on the
sharecropped land (in fact, FBI-1 also became negative
in the land-size class of 3-4 ha) increased as land-size
increased which could be due to the problems
associated with the management of a bigger farm. In
the sharecropped land, no clear association of farm
business income across land-size classes with costs of
cultivation and yield was observed. However, it may
be pointed out that the fall in yield was substantial as
compared to the fall in costs of cultivation in the landsize class of 3-4 ha (where the loss for sharecroppers
was highest) vis--vis 1-2 ha (Appendix Table 2).
One point that needs to be mentioned at this
juncture is that not only the farm business income from
winter paddy cultivation is negative on the
sharecropped land, the farm business income on owned
land is also negligible. Yet, winter paddy is the
predominant crop grown on the sharecropped and
owned lands9. One would wonder as to why a rational
farmer should grow a crop when he incurs a loss or

cannot have enough income from it. One plausible
explanation is that it is grown basically for selfconsumption.
Analysis of FBI-1 and FBI-2 from Summer Paddy
Table 5 presents farm business income from
cultivation of summer paddy on owned land and land
under fixed rent. Both FBI-1 and FBI-2 were found
considerably higher on owned land relative to the land
under fixed rent. Both FBI-1 and FBI-2 on land under
fixed rent in the land-size class of 1-2 ha were negative.
It may be due to the higher costs of production on land
under fixed rent than on owned land (see Appendix
Table 3) besides the rent that the tenants had to pay
additionally. On the other hand, no clear association
between land-size and farm business income could be
found for the owned land and for the land under fixed
rent. However, a comparison of farm business income
with cultivation cost and yield on owned land suggests
that variations in farm business income across landsize classes may be influenced more by variations in
yield than in costs of cultivation. In the case of land
under fixed rent, variations in both cost of cultivation
and yield might have contributed to the variations in
farm business income across land-size classes as it has
been observed that an increase (decrease) in farm
business income is associated with increase (decrease)
in yield and a fall (increase) in cost of cultivation.
Analysis of FBI-1 and FBI-2 from Winter
Vegetables
Table 6 reveals that farm business income from
cultivation of winter vegetables on land-size class of
0-1 ha is more on owned land as compared to on land
under fixed rent; the opposite is the case in the land-
75
Table 5. Tenure status-wise FBI-1 and FBI-2 from cultivation of summer paddy in different land-size classes
(`/ha)
Income type
FBI-1
FBI-2
Tenure status
Owner
Fixed rent
Owner
Fixed rent
0-1
1-2

2-3
32013
18562
29606
15958
20961
-7237
19661
-8681
57359
21570
54112
20338
3-4
4-5
16740
16740
-
39759
39147
-

Table 6. Tenure status wise FBI-1 and FBI-2 from cultivation of winter vegetables in different land-size classes
(`/ha)
Income type
Tenure status
0-1
FBI-1
FBI-2
Owner
Fixed rent
Owner
Fixed rent
75894
43475
67145
39871

1-2
2-3
41297
54562
37797
44702
18413
18413
3-4
78815
76843
size class of 1-2 ha. The higher farm business income

on owned land in the size class of 0-1 ha could be due
to use of more family labour by the owner operators in
this land-size class as compared to the fixed rent tenants
(see Appendix Table 4). In the case of winter vegetables
also, the relation of land-size with farm business income
is not clear. However, comparing incomes, yields and
costs of cultivation, it can be observed that the fall in
income from ` 75894/ha in 0-1 ha category to ` 41297/
ha in 1-2 ha category on owned land could be due to
the fall in yield and increase in costs of cultivation. On
the other hand, the variations in farm business income
across land-size classes on the lands under fixed rent
seem to have more to do with variations in yield than
in cost of cultivation (see Appendix Table 4)
Thus, the analysis at the aggregate level shows that
while farm business income on land under fixed rent
in the smaller land-size classes is lower as compared
to that on owned land, the reverse is the case in the
subsequent land-size classes. The farm business income
on the sharecropped land, however, is the lowest
relative to owned land and land under fixed rent
tenancy. In terms of land-size, while in case of the fixed
rent tenants, the relationship between the land-size and

farm business income is not clear, the performances of
the owner operators and the sharecroppers reveal a
unambiguous inverse relationship if the land-size class
of 5-6 ha in case of the owner operators is ignored.
On the other hand, the dis-aggregate level analysis
carried out for three selected crops reveals as follows.
In the case of winter paddy, except for the land-size
class of 0-1 ha, the FBI-1 is higher on the owned land
relative to the sharecropped land. In fact, after
accounting for the cost of family labour, farm business
income accruing to households becomes negative in
all land-size classes for the sharecropped land.
In the summer paddy, both FBI-1 and FBI-2 are
higher on owned land relative to the land under fixed
rent. Lastly, it has been found in case of winter
vegetables that farm business income is more on owned
land in the land-size class of 0-1 ha as compared to the
land under fixed rent; it is the opposite is the case of
land- size class of 1-2 ha. No clear association between
land-size and farm business income, except for the
sharecroppers in case of winter paddy, could be
observed at the dis-aggregate level. Thus, the facts that
76
emerge from this analysis are: (i) At the aggregate level

sharecroppers have the lowest income and there is an
inverse relationship between land- size and farm
business income on sharecropped land, and (ii) FBI-2
on sharecropped land is lower than on owned land in
the case of winter paddy and both FBI-1 and FBI-2 are
higher on owned land than on land under fixed rent
tenancy in the case of summer paddy.
Econometric Analysis
To get a deeper insight into the relation of landsize and tenure status with farm business income,
regression analysis was carried out. Though the
regression analysis has been done at the aggregate level,
the possibility of crop-specific information distorting
the results was very minimum for the following two
reasons: (1) the technology used for cultivating a
specific crop by a farmer, irrespective of his land tenure
status, is more or less same, and (ii) incorporation of
certain variables either as independent or control
variables would indirectly represent the impact of crops
grown. As for example, incorporation of area under
sharecropping as a proportion of the total operational
holding of a household, area under HYVs as a
proportion of total operational holding and per hectare
fertilizer consumption by a household as either
independent or control variables should be sufficient
to account for the impact of the cultivation of winter
paddy (On one hand, winter paddy is the predominant
crop grown on sharecropped land, use of HYVs and
fertilizer in the cultivation of this crop is very minimum
irrespective of the tenure status of the farmers on the
other hand) by a household on the productivity of
operational holding and income generation of that
household. The results of regression analysis are
summarized in Table 7.
It has been found that the coefficients of area under
sharecropping and area under fixed rent are negative
and significant in the regression equations of both FBI1 and FBI-2. While the coefficient of area under
sharecropping is significant at 1 per cent level of
significance, the coefficient of area under fixed rent is
significant at 10 per cent level of significance. Thus,
these results imply that if the leased-in area constitutes
a larger proportion of the total operational holding of
a household, especially if leased-in under
sharecropping, the household would have relatively
lower farm business income. The reason for lower
Vol. 29 (No.1)
January-June 2016
Table 7. Results of regression analysis for FBI-1 and

FBI-2
Test of
heteroskedasticity
Dependent variables
Independent
variables /constant
AGE
EDU
FS
ASC
AFR
IRRI
LAB
TILL
HYV
NPK
EXT
CREDIT
CI
D1
D2
D3
CONSTANT
R2
F
BP/CW test
Chi2 [1] = 64.28
Prob. = 0.0000
BP/CW test
Chi2 [1] = 69.76
Prob. =0.0000
FBI-1
FBI-2
Estimates of coefficients/values
76.51
(102.41)
1729.62
(1304.39)
-2067.69
(1581.06)
-150.94***
(28.17)
-97.66*
(58.01)
85.29
(57.09)
0.04
(0.52)
-0.03
(0.93)
159.12***
(53.02)
2.06
(22.81)
10450.39
(13472.5)
1285.80
(4072.06)
159.52**
(64.11)
-13171.28**
(5790.90)
-16738.16***
(4991.28)
-4547.38
(3832.89)
-3969.44
(7274.05)
0.35
6.73***
[16, 197]
84.22
(99.81)
1895.05
(1247.87)
-1624.27
(1561.45)
-159.52***
(27.96)
-105.83*
(55.09)
90.38
(56.21)
-0.11
(0.48)
-0.20
(0.90)
161.55***
(52.23)
3.49
(22.36)
11541.93
(13804.16)
1813.08
(4003.23)
152.57**
(63.68)
-10464.19*
(5736.15)
-14997.76***
(4955.59)
-3270.65
(3752.752)
-7230.30
(7211.68)
0.37
7.18***
[16, 197]
Notes: The figures within ( ) and [ ] are heteroskedasticity

consistent robust standard error and degrees of freedom,
respectively.
***, ** and * indicate significance at 1 per cent, 5 per cent
and 10 per cent levels, respectively.
income on the land under sharecropping and fixed rent

is the payment of a substantial amount as rent to the
lessors. It was interesting to note that the coefficient
of land size did not appear to be significant in any of
the two regression equations.10
Two other variables that contribute positively to
farm business income are area under HYVs and
cropping intensity. The area under HYVs has a
coefficient which is significant at 1 per cent level of
significance in the regression equations for both FBI1 and FBI-2, whereas the coefficient of cropping
intensity is significant at 5 per cent level of significance
in both the equations. Among the locational dummies,
the coefficients of D 1 and D 2 are negative and
significant in the regression equations of FBI-1 and
FBI-2.

The study has found that sharecropping and fixed
rent tenancy contracts have adverse impact on farm
business income in Assam. The lower level of farm
business income on leased-in land, especially on
sharecropped land, can be attributed to high amount
(50% of produce in case of sharecropping) of rent being
paid by the tenants. The implementation of tenancy
law prevailing in the state with provisions for regulation
of land rents has never been much effective. In fact,
informal or concealed tenancy has made the law
redundant. All the tenancy contracts in our study were
informal. The emergence of concealed tenancy may
be attributed to a restrictive provision in the tenancy
law. The existing law has the provision of a tenant
becoming an occupancy tenant and ultimately the
owner of the land if he holds the land continuously for
three years. Owing to this stringent condition, the
lessors do not want the tenancy contracts to be recorded
making it impossible for the tenants to realise the
benefits of the law. Hence, if the tenants are to be
protected from the burden of exorbitant rent, the first
step would be to reform this stringent provision in the
tenancy law and record the tenancy contracts.
It has been observed that cropping intensity has a
positive impact on farm business income. In the present
study, however, the cropping intensity has been found
to be very low. The cropping intensity of owner
operators, sharecroppers and fixed rent tenants has been
found as 131 per cent, 112 per cent and 120 per cent,
77
respectively. Even at the state level, the cropping

intensity is much lower (148% in 2010-11), implying
that there is a potential for improvement. In fact, the
Ministry of Agriculture of Government of Assam has
targeted an increase of cropping intensity for the state
of Assam from 148 in 2010-11 to 174 by 2016-17.
However, given the dominance of winter paddy in the
cropping pattern, it remains doubtful whether the target
would be achieved since winter paddy can be cultivated
during one particular season only.
The people of Assam being predominantly rice
consumers; rice is the major crop with more than 70
per cent of the gross cropped area under it. Again within
rice, winter paddy is the main crop since this is largely
grown for self-consumption. In fact, as the present
study has shown, in spite of not generating sufficient
income (the sharecroppers actually incur losses), the
farmers continue to grow winter paddy for selfconsumption and in recent times, the area under winter
paddy has continued to increase. On the other hand,
increment in the area under summer paddy, vegetables
and other horticultural crops has been very minimal,
though incomes from these crops are relatively more.
Hence, a shift in the cropping pattern from winter paddy
to summer paddy and vegetables is required11. These
should help in increasing the cropping intensity also
as mostly short-duration HYVs are used while
cultivating these crops.
If food security of the farmers could be ensured
through proper execution of the public distribution
system or through direct cash transfer, one may expect
that there would be a shift in cropping pattern and
increase in cropping intensity. Given the access to
subsidised food grains, the rational choice of farmers
would be not to produce those crops if their cultivation
is not profitable. In the context of Assam, given the
fact that winter paddy is a subsistence crop, we can
therefore expect a shift from winter paddy to summer
paddy and other horticultural crops. Such a change will
however get materialized only when it is complemented
with the supporting infrastructure such as irrigation
facilities, credit, cold storage, transportation and
marketing facilities. Ignoring these needs will unfold
a serious crisis in Assams agrarian sector. This
comprehensive change is also inevitable for ensuring
that the Government of Assam is able to attain its target
for achieving the targeted increase in the cropping
intensity. The change in cropping pattern and associated
78
increase in cropping intensity will in turn ensure that

the agriculture sector in Assam makes a healthy
transition from its current subsistence level to a
profitable venture.
Vol. 29 (No.1)
5
The cost of labour has been expressed in value

(monetary) terms. Although not substantial, there
are variations in the wage rates prevailing across
field study locations. We have taken the average
of the existing wage rates in the four field study
locations and then multiplied the number of man
days both hired and family by the average wage
to obtain the total cost of labour.
Six questions relating to the extension service

while interviewing the farmers were asked.
Farmers responses to these questions indicated
whether they had received any direct benefits from
the extension service. On the basis of the farmers
responses to these questions, the variable D was
assigned the value 0 or 1.
Goswami and Bezbaruah (2013) have explained

as to why the sharecroppers predominantly grow
winter paddy in the following way:
Sharecropping is usually the preferred form of
contract when the crop grown is the conventional
winter paddy. Winter paddy is grown during the
rainy season and harvested during winter. As a
result, it is subjected to greater risk and uncertainty
caused by weather conditions than crops grown
in the other seasons. Since, under sharecropping,
the risk associated with the crop is also shared
along with the output, the tenants prefer
sharecropping when they grow winter paddy.
It can be observed from Appendix Table 1 that

labour cost is the largest component of total costs
of cultivations.
For the state of Assam, while paddy is the main

crop, within paddy winter paddy is the one which
has the highest portion of rice acreage under it. In
fact, area under winter paddy has increased over
the time. Share of rice in total cropped area
increased from 73.64 per cent in 2007-08 to 75.34
per cent in 2009-10 (Goswami, 2012). On the other
hand, the share of winter paddy in total rice acreage
increased from 67.16 per cent in 2000-01 to 72.94
per cent in 2012-13 (GoA, 2013). The area under
summer paddy has also increased during the same
period but only marginally (from 3.2 lakh ha in
2000-01 to 3.99 lakh ha in 2010-11).
10
For a possible explanation as to why land size does

not have any impact on productivity and income
generation, one may need to look at two significant
Acknowledgement
The author is thankful to the anonymous referee
for his suggestions.
End-Notes
1
In terms of education level of a farmer, there are

five categories, viz., illiterate, below primary,
primary to high school, matriculates and
undergraduates and graduates and above. In the
regression analysis these five categories have been
assigned values from 0-4, respectively.
While, on one hand, the small farms may have

the advantage of abundant labour which helps in
raising many crops and consequently income, they
may face the constraints on financial resource on
the other hand. The big farms usually do not have
the financial constraint but may confront
managerial problems. Hence the sign of the
coefficient of this variable cannot be anticipated
a-priori.
Operational holdings were divided into three parts

and not necessarily all parts occurred in each
observation. These were the parts under
sharecropping, fixed rent tenancy and owner
operated. In the regression analysis it is neither
possible nor necessary to include these three parts
as separate independent variables. If all the three
parts are included, they add up to 100 per cent in
each observation, resulting in perfect multicollinearity situation. In view of the focus of the
paper, the two independent variables included
were percentages of operational holdings under
sharecropping and under fixed rent tenancy.
Indeed given other control variables, the effect on
the dependent variable of owner operated part of
the operational holding was captured by the
constant term in the respective regression
equations.
Unlike the sharecroppers, the fixed rent tenants

dont suffer from the incentive problem and supply
adequate efforts so as to maximize economic
surplus.
January-June 2016
changes which the organization of agricultural

production has undergone. One of those changes
is the emergence of various markets for the
services of factors of productions. Among all such
markets, the most crucial ones are the markets for
the services of tractor/power tiller and that of the
instruments for irrigation. The emergence of
markets for the services of inputs has made
cultivation much easier than before. The farmer
does not have to possess the factors of productions;
he can rent-in the services of the inputs from
markets. The other change that has occurred is in
the process of crop cultivation. A decade ago or
so, the process of cultivation was such that the
farmer would stand behind the plough by himself
and perform each part of the cultivation process.
He would, at best, hire a labourer to work
alongside him during the time of transplantation
and harvesting. Now-a-days, however, it has been
observed that most of the farmers prefer to contract
out almost all the parts of the cultivation process.
Contracting out is preferred by the farmers as it
minimizes the necessity to monitor to a large extent
which otherwise would have been a costly affair
had the labour been hired in a wage contract. Thus,
the farmers job has become less labourintensive;
in fact the farmers role has got reduced to that of
a manager only. Thus, on one hand, emergence of
markets has allowed the farmers to hire the
services of inputs and thereby it has made the
resource constraint that the small farms usually
face less binding. On the other hand, the possibility
of contracting out parts of cultivation process has
nullified the disadvantage of not having sufficient
family labour which the big farms may face
(Goswami, 2016).
11
Another reason as to why a shift towards summer

paddy and winter vegetables may help the farmers
is that these crops involve little weather risk. This
would induce the farmers to apply more of better
quality and costly inputs like HYVs, fertilizers,
pesticides, irrigation and so on. Application of
these inputs though increases the costs of
production; it minimizes production risk and
increases the production, productivity and
79
consequently, income. It is evident from our

analysis (Tables 4, 5 and 6) that notwithstanding
the loss incurred by the fixed rent tenants in the
size class of 1-2 ha in the case of summer paddy,
the farm business income in summer paddy is
considerably higher, and more so in the case of
winter vegetables as compared to winter paddy.
References
Bezbaruah, M.P. (1994) Technological Transformation of
Agriculture A Study in Assam. Mittal Publications,
New Delhi.
Gautam, H.C. (1995) Agrarian relations: A study on some
aspects of land tenancy system in Assam. Indian Journal
of Agricultural Economics, 50 (4): 682-687.
Goswami, B. (2012) Economic Implications of Tenancy: A
Study in Assams Agrarian Set-Up. PhD Thesis
(unpublished), submitted to Gauhati University,
Guwahati, Assam.
Goswami, B. (2016) Overcoming land size induced
constraints through endogenous institutional
innovations: Findings from a field study in Assam
plains, India. Economics Bulletin, 36 (1): 411-428.
Goswami, B. and Bezbaruah, M.P. (2013) Incidence, forms
and determinants of tenancy in the agrarian set-up of
the Assam plains. Economic and Political Weekly,
XLVIII (42): 60-68.
GoA (Government of Assam) (2012) Economic Survey of
Assam. Directorate of Economics and Statistics.
GoA (Government of Assam) (2013) Profile of Agri-Horti
Sector of Assam. Department of Agriculture.
downloaded from www.agriassam.in on 27.06.2014.
Kuri, P.K. (2003) Factor market imperfections and
explanation of tenancy: Testing an econometric model
using evidence from Assam of North East India. Indian
Journal of Agricultural Economics, 58 (2): 234-245.
Lal Bahadur Shastri National Academy of Administration
(1994) Land Reforms in Assam: An Empirical Study
(1988-1991). Land Reforms Unit, Mussoorie.
Marshall, A. (1920) Principles of Economics. Macmillan,
London, UK.
Revised received: October, 2015; Accepted: May, 2016
80
Vol. 29 (No.1)
January-June 2016
Appendix Table 1: Land tenure status-wise expenditures on key inputs in overall cultivation in different land-size
classes in Assam
(`/ha)
Expenditure type
Yield per hectare
Labour
% of family labour in total labour
Services of capital goods*
Other purchased inputs
All inputs
Land tenure
status
0-1 ha
1-2 ha
Land-size class
2-3 ha
3-4 ha
4-5 ha
5-6 ha
Owner
45974
34812
34501
35056
36719
61686
Sharecropper
31078
31746
21077
Fixed rent
65858
50219
59759
83020
Owner
9303
7778
7679
7808
9058
12891
Sharecropper
7925
8266
6499
Fixed rent
12164
13626
15388
12158
Owner
34
37
31
40
26
Sharecropper
49
42
14
Fixed rent
37
26
12
7043
5414
5621
6765
6924
9822
Owner
Sharecropper
5318
5440
5673
Fixed rent
13175
11139
8601
8523
Owner
4097
3737
3877
3535
8763
6696
Sharecropper
1720
2063
813
Fixed rent
7333
8208
3481
12969
Owner
20443
16929
17177
18108
24745
29409
Sharecropper
14963
15769
12985
Fixed rent
32672
32973
27470
33650
Note: *Expenditures on the services of capital goods include costs of hiring the services of power tiller, tractor and pumpset.
81
Appendix Table 2: Land tenure status-wise expenditures on key inputs in winter paddy in different land-size classes
in Assam
(`/ha)
Expenditure type
Yield per hectare
Labour
Services of capital goods
All inputs
Land tenure
status
0-1 ha
1-2 ha
Owner
Sharecropper
Owner
sharecropper
Owner
sharecropper
Owner
sharecropper
Owner
sharecropper
Owner
sharecropper
11748
24123
5022
7022
30
50
4181
3906
2167
1328
11370
12256
12988
25591
4428
6858
28
44
4229
4735
1690
1479
10347
13072
Land-size class
2-3 ha
3-4 ha
11209
4924
24
4698
1760
11382
20347
17928
4272
5528
34
14
6241
4826
1552
692
12065
11046
4-5 ha
9362
4084
16
5423
5715
15222
Appendix Table 3: Land tenure status-wise expenditures on key inputs in summer paddy in different land-size
classes in Assam
(`/ha)
Expenditure type
Yield per hectare

Labour
All inputs
Land tenure
status
0-1 ha
1-2 ha
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
55776
56455
10715
11366
27
29
12344
6908
3110
11169
26169
29443
48202
45816
13341
17488
12
8
12158
7166
3042
14405
28541
39059
Land-size class
2-3 ha
3-4 ha
83663
59759
14477
15388
22
8
11247
4400
3827
7682
29551
27470
44820
12602
0
12378
3100
28080
-
4-5 ha
74699
17008
4
13606
4938
35552
-
82
Vol. 29 (No.1)
January-June 2016
Appendix Table 4: Land tenure status-wise expenditures on key inputs in winter vegetables in different land-size
classes in Assam
(`/ha)
Expenditure type
Yield per hectare

Labour
All inputs
Land tenure
status
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
Owner
Fixed rent
0-1 ha
Land-size class
1-2 ha
2-3 ha
3-4 ha
109402
90884
15204
11772
53
31
7892
6096
19162
25395
42258
43263
87253
112049
16859
21065
24
47
6815
6235
25781
35564
49455
62864
112049
7948
25
6588
16934
31470
77613
22410
0
4773
10540
37723

DOI: 10.5958/0974-0279.2016.00021.5
Impact of Nonfarm Activity on Rural Income Inequality

in South 24 Parganas District of West Bengal
S.M. Rahamana*, S. Haldarb, Sowmyashree B.V.c, S. Nandid,
L. Malangmeihe and B.K. Beraf
Department of Agricultural Economics (PG), Bihar Agricultural University, Bhagalpur-813 210, Bihar
College of Horticulture, Orissa University of Agriculture and Technology, Chiplima-768 025, Odisha
c
Karnataka University, Dharwad-580 003, Karnataka
d
Basantapur S.S. Sikshaniketan, Malamba, Bardhaman-713 426, West Bengal
e
Division of Agricultural Economics, Indian Agricultural Research Institute, New Delhi-110 012
f
Department of Agricultural Economics, Bidhan Chandra Krishi Viswavidyalaya, Nadia-741252, West Bengal
a
Abstract
The impact of non-farm activity on rural income inequality in the South 24 Parganas district of West
Bengal has been studied. The non-agricultural activities play a key role in shoring up income and
employment opportunities for the rural work force in this district. The average annual income of nonfarm
families (` 108662) is almost double to that of farm families (` 56081). The results of Gini coefficient
decomposition have indicated that nonfarm income is the largest component (44.05%) of total income,
followed by farm income (39.41%). The non-farm income contributes largest to the overall inequality
(58.27%) in which non-farm self-employment activity increases the total income inequality, while nonfarm wage employment decreases it. The analysis of factors influencing participation of households in
different income generating activity showed that human capital endowed with formal and informal
education are engaged more in nonfarm activities. The study has suggested development of both farm
and non-farm sectors simultaneously to reduce rural income inequality.
Keywords: Gini-coefficient, income inequality, nonfarm activity, rural income, probit model, West Bengal
JEL Classification: C35, E24, J21, J31, J43, O14, O15, Q12
Introduction
In India, the rural economy is predominantly based
on agriculture and allied activities, but various nonagricultural activities also play a significant role in
providing employment and income to the rural sector,
particularly to labour force that is largely semi-skilled
Email: rkmvur@gmail.com
The paper is part of Ph.D. thesis entitled Economics of Farming Systems in Coastal Regions of West Bengal of the first
author, submitted to Department of Agricultural Economics,
Bidhan Chandra Krishi Viswavidyalaya, Nadia-741 252 in
2014 and degree awarded in 2015.
and belongs to both farm and non-farm households

(Malangmei et al., 2015). The rural transformation in
India has been rapid, especially after the 1960s, when
the rural nonfarm sector emerged as one of the major
sources of income that spurred economy-wide
employment growth. Unfortunately, the rapid economic
growth in the rural areas was associated with stunted
structural transformation, due to which the rural
nonfarm sector has the sagging employment growth.
In 2009-10, the National Sample Survey Organization
(NSSO) estimated that around 42.5 per cent of rural
households primarily earned their livelihood from nonfarm sources, whereas the share of non-farm
84
employment in rural workforce was 32.1 per cent. It is

quite interesting to note that there has been 40-50 per
cent increase in rural workforce participation in
nonfarm activities and the same proportion of rural
population has diversified its household income to
nonfarm sources (GoI, 2001).
For the past some years, the marginal and small
farmers have been feeling the necessity of non-farm
income as their meagre land was not sufficient for
generating income adequate to maintain the family
(Rajshekhar, 1995; Pandey and Singh, 2013). Further,
the agricultural sector has been unable to absorb the
growing rural work force due to falling output
elasticities of employment within the sector (Singh et
al., 2003). Participation in nonfarm activities enhances
the capacity of farm households to overcome negative
shocks and reinvest in farm activity. It mitigates income
fluctuations and enables adoption of more profitable
but risky agricultural technologies that encourages
transformation of traditional agriculture to modern
agriculture. The smallholders and landless households
depend on rural nonfarm activities to secure livelihood
during the slack agricultural season (Bhakar et al.,
2007).
It is also observed that educated youths take more
interest in engaging themselves in nonfarm activities
vis-a-vis their uneducated counterparts. Again,
households in distant villages from the peri-urban and
urban centres are more dependent on traditional
agricultural activities for income and livelihood.
Binswanger Mkhize (2013) envisioned that farm sizes
in India will continue to decline and as a consequence,
households will strive for income growth via
technological and allocational changes towards
diversified high-value agricultural commodities and the
non-farm sector.
Nonfarm Sector and its Contribution to Indian
Economy
The NCAER statistics on proportion of nonfarm
income to rural households indicate an increase of 20
to 27 per cent during 1971-72 to 1981-82. Further, 70
per cent of the increase in real household income is
attributed to the rise in non-agricultural incomes during
this period. Papola (1992), while analysing trends in
rural nonfarm employment using the NSSO data, found
that workers who recorded rural nonfarm employment
as their principal employment grew at 5 per cent per
Vol. 29 (No.1)
January-June 2016
annum between 1977-78 and 1987-88, while the share

of nonfarm jobs rose from 17.9 to 23.4 per cent in the
rural areas. The nonfarm activities not only
significantly increased total household income but also
functioned as a safety net through diversifying income
sources. Interestingly, large farmers repeated
investments on nonfarm sectors spurred into further
employment generation in this sector (Bhakar et al.,
2007). Therefore, the synergy of farm and nonfarm
activities in the rural areas actually helped people to
absorb themselves in employment-generating activities
in villages only. Hence, a positive attitude for
agriculture and rural development can be envisioned
with suitable policies and institutional reforms. This
change will ultimately improve the working
performance of many rural developmental programs
at the village level (Zhu and Luo, 2006).
The Problem Statement
It has been found that repeated non-performance
of farm activities for marginal farmers induce them to
participate in nonfarm activities and interestingly
enough that they earn higher income compared to those
with better resource endowment. Households are
motivated to undertake rural nonfarm activity by either
pull or push factors. If the nonfarm sector has high
returns, the pull factors will be strong and if farm
activity cannot provide enough income for households
or households need to diversify their income sources,
the push factors may kick in. Many marginal or landless
farm families are less capable of mitigating negative
shocks, and are more risk averse. In order to have
additional income as well as to diversify and undertake
activities with returns that may have a low or negative
correlation with those of farming, poor households may
have stronger incentives to participate in nonfarm
activity only.
The economy of coastal areas of West Bengal
mainly depends on agriculture (including crop culture,
fisheries, forestry, etc.), which influences the
livelihoods of millions of rural households in the
region. Nearly 20 per cent of the net state domestic
product (NSDP) is contributed by this primary sector,
in which crop culture contributes 16 per cent in East
Medinipur, 12 per cent in North 24 Parganas and 15
per cent in South 24 Parganas (Mandal et al., 2011).
The contribution of agriculture to districts NSDP has
shown a declining trend over the period 2002-03 and
Rahaman et al.: Impact of Nonfarm Activity on Rural Income Inequality in West Bengal
2006-07. The employment pattern in the rural coastal

areas (unskilled labourers) of West Bengal shows the
scarcity of labour during the peak planting (JulyAugust) and harvesting (December- January) periods
of kharif rice, which is the major crop of the region.
During rabi and summer seasons, the lessremunerative farm activities seldom fulfil the
households requirements of securing livelihoods and
the surplus labour force depends on non-agricultural
livelihood options. This push factor instigates migration
to nearby cities and towns during the lean periods of
agriculture. The household members are engaged in
several nonfarm wage earning activities (MGNREGS
work, construction work, porter, factory worker, shop
attendant, house maid, security guard, etc.) and selfemployment activities (shopkeeper, electrician,
mechanic, cable operator, own business, private tuition,
driving own vehicle for carrying goods and passengers,
hawker, vegetable and fruit vending, etc.). Therefore,
this study has attempted to answer the underlying
question whether nonfarm activity influences the
overall household income inequality in rural areas with
special reference to farming households in the South
24 Parganas district of West Bengal.

In West Bengal, the total 5.58 lakh ha of cultivable
saline area, highest (42.53%) comes under the
jurisdiction of South 24 Parganas district, followed by
Midnapore (32.58 %), North 24 Parganas (17.03 %)
and Howrah (7.79%) districts (Bandyopadhyay et al.,
2003). Among these districts, South 24 Parganas has
the lowest cropping intensity of 134 per cent. The
highest coastal salinity-affected area with lowest
cropping intensity, predominance of marginal and small
farmers, varying soil pH from 4.5 to 7.5, etc. were
the underlying reasons for the selection of this district
for study. In this study, primary data were collected
through a household level survey in six randomly
selected villages, namely Mandalpara, Kayaler Chak
and Pichhakhali of Kultali community development
(CD) block and Kheria, Paschim Bhangankhali and
Uttar Bhangankhali, for the agricultural year 2012-13.
From each village, 30 households were selected
randomly that ultimately rounded to total 180 rural
households who were personally interviewed adopting
a pretested structured schedule.
85
Analytical Tools
To study the impact of non-farm income on rural
inequality, the non-farm income was considered as an
exogenous transfer. We decomposed the total
household income and studied the distribution of each
income source and its contribution to total income
inequality.
(i) Decomposition of Gini Index
Following Pyatt et al. (1980) and Stark et al.

(1986), the Gini index of total income, Gt, can be
decomposed as:
(1)
where, Sk is the share of income component k in total
income, Gk is the Gini coefficient of income component
k, and Rk is the Gini correlation of income component
k with total income defined as:
(2)
Using Gini decomposition technique, the extent
of contribution to the overall income inequality, as
influenced by a particular income source, the relative
concentration coefficient of income component k can
be determined by formula (3).
(3)
where, gk is the relative concentration coefficient of
income component k in the overall inequality. The
income component k worsens the overall income
inequality if gk > 1 and has an equalizing effect if
gk < 1.
The Gini decomposition technique explores the
contribution of an income component in increasing or
decreasing the overall income inequality. The most
contributory factor to the total can be understood by
the regression-based approach to inequality
decomposition which quantifies the relative
contribution of various income determinants to the
inequality in a given income component (Adams,
2001).
86
(ii) Factors Affecting Participation in Different Income

Generating Activities
The participation equation is estimated with a

dummy variable equal to 1, if the household participates
in the activity, and 0 otherwise, is regressed on the
independent variables, viz. net cropped area (in acres),
number of family members, age of household-head,
years of education of household-head, number of
household members trained, distance from the nearest
railway station, number of working males and females,
and membership of any self-help group (SHG) in the
village. The rural families are more inclined to various
nonfarm activities for higher income. Thus, a probit is
fitted separately for each case to reveal the intensity of
participation in nonfarm income generating activities.
If Z i is a vector of independent variables of the
participation equation, then we have:
(4)
where, P*i is the non-observed continuous latent
variable and Pi is the observed binary variable, with a
value of 1 if the household participates in the nonfarm
activity and 0 otherwise.

Categorization of Farmers
To categorize sample farmers, the classification
given by Reddy et al. (2004) was followed. In West
Bengal, due to land reforms and socio-cultural status
of farmers, there was almost no existence of so-called
medium or large farmers having more than 10 acres of
landholding. The respondents were post-stratified into
marginal, small and semi-medium farmers based on
the size of operational holdings. The number of
Vol. 29 (No.1)
January-June 2016
respondents under different categories, presented in

Table 1, showed that the majority of sample farmers
belonged to the marginal group with operational
holding-size of less than or equal to 1.25 acres
(43.89%), followed by small farmers (40.56%) and
semi-medium farmers (15.55%). The marginal and
small farmers were predominantly mono-cropped with
cropping intensity of 119 per cent (Rahaman and
Haldar, 2014). This signifies the fact that the household
members, not engaged in round-the-year farming
activity, functioned as a pull factor for involvement in
nonfarm activities.
Basic Features of Sample Households
The income of rural household was from farm,
nonfarm and other sources (salaried job, pension,
remittances, land rent, etc.). The farm sector included
agriculture, livestock, forestry, fishing and hunting.
Formal or informal wage-employment and selfemployment were the two major sources of non-farm
income. Table 2 presents some key indicators about
sample respondents. Across the total farm households
in the sample, 29 per cent had only farm income and
71 per cent had both farm and non-farm income. The
average annual income of the households that
participated in non-farm activity (` 108662) was higher
than those households who participated in the farm
activity only (` 56081). However, agriculture provided
employment to a larger number of farm family
members, but on an average, non-agricultural activity
contributed to the major part of total income of farm
households.
Table 2 summarizes the characteristics of sample
households and indicates the significance of their
participation in non-farm activity for income and
livelihood. The average number of workers was three
Table 1. Categorization of sample farmers based on landholding-size

Category of farmers
Number
of farmers
Marginal farmers
Small farmers
Semi-medium farmers
Total
79 (43.89)
73 (40.56)
28 (15.55)
180
Size of holding (wet land in acres)

Ownership holding
Operational holding
0.58
1.40
3.06
1.32
Note: Figures within the parentheses indicate per cent to the total.
0.91
1.72
3.58
1.68
Cropping
intensity (%)
123.54
120.61
113.98
119.02
87
Table 2. Description of sample statistics, 2012-13

Particulars
Household income (`)

Total income
Farm income
Nonfarm income
Other income
All
households
Households that do
not participate in
nonfarm activity
Households that
participate in
nonfarm activity
94237 (78906)
37135 (34158)
41510 (55109)
15592 (33912)
56081 (39321)
33965 (19784)
22116 (43215)
108662 (85256)
38334 (38250)
57203 (57364)
13125 (29598)
3 (1)
1 (1)
1(1)
1 (1)
4.94 (4.19)
1 (1)
16.61 (7.91)
2.10 (0.57)
1.19 (0.85)
1.58 (1.06)
0.39 (1.15)
0.03 (0.09)
1.73 (1.04)
27
4 (2)
2 (1)
1 (1)
2 (2)
6.18 (4.07)
1 (1)
15.95 (7.99)
1.91 (0.54)
1.34 (1.35)
1.69 (1.22)
0.51 (0.75)
0.17 (0.79)
2.04 (1.80)
153
Characteristics of households
Average No. of working population
3 (2)
Males
2 (1)
Females
2 (1)
Number of dependents
2 (2)
Average number of schooling years of household-head
5.84 (4.12)
No. of members that have received some technical training
1 (1)
Distance from village centre to the nearest railway station (km) 16.13 (7.93)
Distance from village centre to the nearest bus stop (km)
1.96 (0.55)
Average ownership holding-size (acre)
1.32 (1.29)
Average net cropped area (acre)
1.68 (1.19)
Average leased-in land (acre)
0.49 (0.82)
Average leased-out land (acre)
0.15 (0.73)
Average gross cropped area (acre)
2.00 (1.71)
Number of observations
180
Note: Figures within the parentheses indicate standard deviation
per household, comprising two males and one female

member with two dependent members per family. The
average years of education of household-head was 5.84,
that shows completion of primary level schooling. The
average ownership holding per household was 1.32
acres, while the gross-cropped area was 2.0 acres. The
pressure on land surface due to its increasing demand
for non-agricultural activities coupled with decreasing
interests of rural youth in farming are the glaring
reasons for decreasing share of farm income in total
family income (Akpan et al., 2015). In comparison with
farm activity, the households that participated in nonfarm activities had a larger number of workers, who
were relatively educated and residing in peri-urban
areas.
Decomposition of Gini Coefficient
households. The first column shows that non-farm

income is the largest component of inequality in the
total income (44.05%), followed by farm income
(39.41%); and as expected, non-farm income also
contributed largest to the overall inequality (58.27%).
The Gini coefficient of 0.39 for total income reveals a
value that lies within the range seen in many other
developing countries. A recent computation by the
World Bank has indicated that Gini coefficients for
SAARC countries ranged from 0.30 for Pakistan to
0.39 for Bhutan (World Bank, 2014). The Gini
coefficients for various income components like
fishery, agricultural labour, self-employment, salary
income, pension & remittance and land rental income
were much higher than that of total income because
not all households derived income from each of the
mentioned sources.
The results of Gini coefficient decomposition

presented in Table 3 by income source for all
The aggregate non-farm income, however, may

hide the effects of its respective components on the
88
Vol. 29 (No.1)
January-June 2016
Table 3. Inequality decomposition by income source for all rural households

Income sources
Total income
Farm income
Agriculture
Livestock
Fishery
Agricultural labour
Nonfarm income
Self-employment
Non-agricultural labour income
Other incomes
Salary income
Pension, remittances, etc.
Land rent
Income share
(Sk)
Ginis
concentration
ratio (Gk)
Gini
correlation
(Rk)
Contribution
to
inequality
Relative
concentration
coefficient
100.00
39.41
19.47
7.29
6.69
5.95
44.05
30.94
13.11
16.55
11.51
3.76
1.28
0.39
0.39
0.37
0.18
0.69
0.77
0.61
0.76
0.34
0.31
0.90
0.93
0.92
0.89
0.73
0.88
0.45
0.94
0.84
0.79
0.91
0.47
0.23
0.14
0.40
33.07
13.70
2.90
5.33
11.13
58.27
47.87
10.39
8.67
6.15
1.31
1.21
0.90
0.70
0.40
0.80
1.87
1.32
1.55
0.79
0.38
0.53
0.35
0.94
overall income inequality. The nonfarm selfemployment income contributed to the increase in
income inequality, while non-farm wage employment
income decreased it, highlighting the importance of
distinguishing between self-employment and wageemployment income while assessing the effect of nonfarm income on income inequality. The results,
however, suggested the existence of entry barriers into
non-farm self-employment. The results also
corroborated the findings of Morduch and Sicular
(2002) on self-employment income, a dis-equalizing
effect on total income, whereas cultivation practices,
livestock-rearing and fishery activities decreased the
total income inequality.
Thus, it is clear that non-farm income has
potentially substituted farm income in many
households. Amongst the two major types of nonfarm
activities, participation in wage-paying activity is often
conditioned by the spatial mobility, as the members
concerned may leave their household and work outside;
while the self-employment activity is more likely to
be local family work (Mehta, 2002). In general,
participation in wage-paying activity is an individual
decision, while the participation in self-employment
activity is the choice of entire household.
Analysis of Factors Affecting Participation in

Different Income Generating Activities
To analyse the probability of participation in
various nonfarm activities by the households subject
to different factors that influenced the participation in
a given activity, the probit regression technique was
employed using statistical package R 3.1.1. The results
(Table 4) depict that the households with a higher
number of working male and female members were
more likely to participate in the non-farm activity. The
total cultivable land available to a family being
insufficient to earn a living for the households (Table
2), both farm and nonfarm families do lease-in land.
Other things being equal, a significant proportion of
households having lower opportunity cost of staying
in village only, do migrate to the nearby cities for
various non-agricultural works. On the other hand, the
rural households endowed with better human and
physical capital, had greater capability to establish
themselves as rural entrepreneurs.
Form Table 3, the variable average number of
years of schooling of household-head measured the
human capital endowment that positively influenced
participation in non-farm activity, especially in many
self-employment generating activities like village tutor,
89
Table 4. Regression results of participation in non-farm activities (Probit model)

Particulars
Net cropped area (acre)

No. of dependents in family
Age of the household-head
Average number of years of education of household-head
Number of household members trained
Distance from nearest railway station
No. of working males
No. of working females
Membership in any self-help group
Constant
McFadden R-squared
Akaike criterion
Hannan-Quinn
Log-likelihood
Schwarz criterion
Regression 1:
Nonfarm activity
Regression 2:
Self-employment
activity
Regression 3:
Wage paying
activity
0.381**
(0.181)
0.259*
(0.133)
-0.03*
(0.015)
0.111**
(0.046)
0.802**
(0.407)
-0.02
(0.026)
0.554***
(0.209)
0.613**
(0.277)
1.18*
(0.69)
-1.026
(0.985)
0.221
123.447
134.514
-51.723
150.721
0.107
(0.12)
0.032
(0.074)
-0.01
(0.012)
0.08**
(0.036)
0.076
(0.295)
-0.027
(0.021)
0.076
(0.132)
-0.139
(0.17)
0.405
(0.51)
-0.213
(0.698)
0.072
161.666
172.734
-70.833
188.940
0.06
(0.148)
0.403***
(0.131)
-0.026*
(0.015)
-0.026
(0.042)
-0.19
(0.393)
0.024
(0.024)
0.634***
(0.187)
0.708***
(0.239)
-0.046
(0.637)
-1.935**
(0.898)
0.265
134.268
145.336
-57.134
161.542
Note: Figures within the parentheses indicate standard error.

***, ** and * denote the significance at 1 per cent, 5 per cent and 10 per cent levels, respectively.
various entrepreneurship (electrical and electronics

goods sale and repairing) and many more. The
education of household-head was considered as a
crucial factor because all the decisions of household
members were either directly dictated or greatly
influenced by him/her. Hence, education provides the
motivation of participation in more remunerative nonfarm activities to alleviate poverty and avail
opportunity to compensate the asset disadvantage
(Rahaman et al., 2015). Thus, an educated/skilled
person fits easily in a non-farm job, whereas an
uneducated/unskilled is left with much choice other
than to opt for non-farm wage employment. The
number of households who had received some technical
training or had membership in any SHG, influenced
the quality of labour force positively in participating

in non-farm activities. Further, any sort of training with
knowledge of technology and management improved
the competence of the household members and
facilitated their participation in nonfarm works. In
short, human capital endowed with formal or informal
education, was engaged more in the non-farm activities.
The number of dependents of a family positively
influenced the participation in nonfarm wage earning
activities like housemaid, embroidery works, etc. for
enhancing livelihood security. The distance from
nearest railway station and bus stop was found statically
non-significant, indicating a good connectivity of
villages with the mega city Kolkata and its suburban
areas and it encouraged local migration for seeking
90
non-farm employment. The distribution of non-farm

income was unequal among the rural households
because not all the sample households participated in
the non-farm activities. Thus, the opportunity cost of
rural labour was very weak locally for participation in
the farm and non-farm activities.
Conclusions
The study has revealed that participation in nonfarm activities can provide opportunity to the
households with low farm productivity to supplement
their income. The households endowed with greater
human/ technical capital can opt for some skilled wage
earning activities including self-employment. Also,
with rising population, declining land-man ratio,
degraded soil fertility and dwindling land and labour
productivity, agriculture alone would not able to
provide adequate income and employment to the rural
households. Under such circumstances, to ensure a
regular flow of income for a decent living and to
achieve food and nutritional security, the household
members have to depend more on non-farm income
generating activities to supplement the farm income.
Thus, rather than raising inequality, the non-farm sector
can neutralise or atleast reduce income inequality in
the rural areas.
The study has observed the contribution of farm
sector to average family income as around 40 per cent
that can cater to the basic requirement of the family,
but a greater portion of household income is being
generated through non-farm sector. Therefore, it will
be appropriate to follow an integrated approach for the
development of both farm and non-farm sectors by
developing appropriate infrastructures and other
income-generating facilities in the non-farm sector.
Vol. 29 (No.1)
January-June 2016
Bandyopadhyay, B.K., Maji, B., Sen, H.S. and Tyagi, N.K.

(2003) Coastal Soils of West Bengal: Their Nature,
Distribution and Characteristics. Bulletin No. 1/2003,
Central Soil Salinity Research Institute, Canning Town,
South 24 Parganas, West Bengal. 62p.
Bhakar, R., Banafar, K.N.S., Singh, N.P. and Gauraha, A.K.
(2007) Income and employment pattern in rural area of
Chhattisgarh: A micro view. Agricultural Economics
Research Review, 20 (2): 395-406.
Binswanger Mkhize, H.P. (2013) The stunted structural
transformation of the Indian economy: Agriculture,
manufacturing and the rural nonfarm sector. Economic
& Political Weekly, 48: 5-13.
GoI (Government of India (2001) Concepts and Definitions
Used in NSS. National Sample Survey Organization,
Ministry of Statistics & Programme Implementation,
New Delhi. Available at: http://mospi.nic.in/
Mospi_New/upload/concepts_golden.pdf (accessed on
12.02.2015).
Malangmei, L., Rahaman, S.M., Haldar, S. and Bera, B.K.
(2015) Understanding sustainable rural livelihood
security in Indian context: A review. International
Journal of Current Research, 7(10): 21025-33.
Available at: http://www.journalcra.com/article/
understanding-sustainable-rural-livelihood-securityindian-context-review.
Mandal, S. Bandyopadhyay, B.K., Burman, D., Sarangi, S.K.
and Mahanta, K.K. (2011) Baseline report of the NAIP
Project: Strategies for Sustainable Management of
Degraded Coastal Land and Water for Enhancing
Livelihood Security of Farming Communities. Central
Soil Salinity Research Institute, South 24 Parganas,
West Bengal. 47p.
Mehta, G.S. (2002) Nonfarm Economy and Rural
Development. Giri Institute of Development Studies,
Lucknow.
References
Morduch, J. and Sicular, T. (2002) Rethinking inequality

decomposition with evidence from rural China. The
Economic Journal, 112: 93-106.
Adams, R.H. (2001) Nonfarm Income, Inequality and

Poverty in Rural Egypt and Jordan. Policy Research
Working Paper Series No. 2572. The World Bank,
Washington, DC.
Pandey, R.K. and Singh, B.K. (2003) Impact of urbanization

on rural income, employment and expenditure: A
sectoral analysis. Agricultural Economics Research
Review, 16 (Conf. No.) 70-82.
Akpan, S.B., Patrick, I.V., James, S.U. and Agom, D.I.

(2015) Determinants of decision and participation of
rural youth in agricultural production: A case study of
youth in southern region of Nigeria. Russian Journal
of Agricultural and Socio-Economic Sciences, 7(43):
35-48.
Papola, T.S. (1992) Rural nonfarm employment: An

assessment of recent trends. The Indian Journal of
Labour Economics, 353: 238-45.
Pyatt, G., Chen, C.N. and Fei, J. (1980) The distribution of
income by factor components. Quarterly Journal of
Economics, 95(3): 451-474.
91
Rahaman, S.M. and Haldar, S. (2014) Sewage fed farming

systems and sustainable livelihood in Peri-urban
Kolkata. Economic Affairs, 59(2): 147-161. Available
at: http://ndpublisher.in/admin/issues/EAV59I2e.pdf.
Singh, A., Vasisht, A.K. and Jain, P.K. (2003) Inter-state

variations in non-agricultural employment in rural
India: An exploratory analysis. Agricultural Economics
Research Review, 16 (Conf. No.): 60-69.
Rahaman, S.M., Haldar, S., Pal, S. and Ghosh, A. (2015)

Role of rural off-farm employment in earning income
and livelihood in the coastal region of West Bengal,
India. Available at: http://ageconsearch.umn.edu/
handle/212697
Stark, O., Taylor, E. and Yitzhaki, S. (1986) Remittances

and inequality. Economic Journal, 96: 722-40.
Rajshekhar, D. (1995) Economic Mobility of Rural

Households. Ph.D. thesis (unpublished), Jawaharlal
Nehru University, New Delhi.
Reddy, S.S., Ram, P.R., Sastry, T.V.N. and Devi, I.V. (2004)
Agricultural Economics. Oxford & IBH Publishing Co.
Pvt. Ltd, New Delhi.
World Bank (2014) World Development Indicators.

Available at: http://wdi.worldbank.org/table/2.9
(accessed on 12.01. 2015).
Zhu, N. and Luo, X. (2006) Nonfarm Activity and Rural
Income Inequality: A Case Study of Two Provinces in
China. World Bank Policy Research Working Paper
3811.
Received: July, 2015; Accepted: February, 2016

DOI: 10.5958/0974-0279.2016.00022.7
Changing Cropping Pattern, Agricultural Diversification and

Productivity in Odisha A District-wise Study
Dinesh Kumar Nayak
Centre for the Study of Regional Development, School of Social Sciences,
Jawaharlal Nehru University, New Delhi 110 067
Abstract
The paper examines the structure and nature of cropping pattern, crop diversification, crop concentration,
productivity level and inter-districts disparity in the state of Odisha based on the secondary data collected
for the period 1980 2005 from different published sources. The study has used Herfindahl index,
location quotient, Gini coefficient and panel data regression for analysis. The study has revealed that
most of the districts in Odisha are experiencing a lateral movement towards crop specialization and crop
diversification is seen only in tribal-dominated / technologically less-developed districts. The study has
observed a reduction in inequality during the studied period and has concluded that districts in Odisha are
converging as far as agricultural productivity is concerned. The study has identified the major determinants
of agricultural productivity in Odisha and has suggested some policy measures for increasing agricultural
productivity in the state.
Key words: Cropping pattern, crop diversification, crop concentration, agricultural productivity, Odisha
JEL Classification: Q1, Q 15, Q19, Q24
Introduction
Odisha is primarily an agrarian economy and
agriculture holds the key to overall development of
the state, which remains backward, unstable, rain-fed,
traditional and prone to natural calamities like droughts,
floods and cyclones. The Odishas agricultural scenario
has become stagnant over a long period of time
(Pattnaik and Shah, 2010; Swain, 1999). Tripathy and
Sarap (1994, p.969) state that an important determinant
of growth in agricultural income is the rise in
agricultural output, and change in crop pattern. These
in turn depend on institutional as well as technological
factors.
Agriculture continues to be the mainstay of Odisha
economy with contribution of about 25.6 per cent to
Email: dineshcsrd@gmail.com
the state gross domestic product during 2000-2005. The

performance of agriculture in the state of Odisha since
1980-81 shows that the overall pattern of structural
changes is associated with the process of economic
development (Table 1). During the past three decades,
the contribution of agriculture and allied sector to the
net state domestic product (NSDP) declined
consistently and continuously from 56.6 per cent during
1980-1985 to 29.7 per cent during 2000-2005 and
further to 19.4 per cent in 2009-10. On the other hand,
the contribution of industry and services sectors
increased respectively from around 15.4 per cent and
28.0 per cent in the 1980-1985 to around 20.7 per cent
and 49.6 per cent in 2000-2005 and reached 26.3 per
cent and 54.4 per cent, respectively in 2009-10.
Despite decline in its share of NSDP of Odisha,
agriculture continues to be a major source of livelihood
to a significant segment of population (around 60%).
Therefore, agricultural development holds the key to
94
Vol. 29 (No.1)
January-June 2016
Table 1. Sectoral composition of net state domestic product (NSDP) in Odisha: 1980 to 2010
(per cent)
Sector
Agriculture and allied
Agriculture
Forestry and logging
Fishing
Industry
Services
Per capita NSDP at 1999-2000
prices (in `)
1980-1985
1985-1990
1990-1995
1995-2000
2000-2005
2009-2010
56.58
48.89
6.97
0.73
15.41
28.01
7960
51.65
45.93
4.82
0.90
16.59
31.75
9084
41.76
36.93
3.43
1.41
20.83
37.41
8921
37.20
32.67
2.78
1.75
19.77
43.03
9855
29.70
25.57
2.65
1.48
20.67
49.63
11323
19.35
16.75
1.67
0.93
26.25
54.40
19456
Source: Calculated by author, taking NSDP at factor cost data from http://www.mospi.gov.in/cso_rept_pubn.htm, Accessed:
7 May, 2011
the overall development of the state by way of creating

employment, generating income, providing raw
materials to the industrial sector and ensuring food
security for the poor. Agriculture in Odisha is
characterized by low productivity due to dominance
of traditional agricultural practices, inadequate capital
formation, low investment, inadequate irrigation
facilities and uneconomic size of holdings. The recent
years have experienced not only low productivity but
also declining productivity.
The paper examines the structure and nature of
changing cropping pattern and crop diversification,
crop concentration, productivity level and inter-districts
disparity in the state of Odisha during the past three
decades (1980-2005). These have been studied through
measuring across districts and state as a whole,
temporal percentage changes in area share of
agricultural crops, crop diversification, crop
concentration index, productivity index and disparity
in agricultural productivity. The methods employed for
the study included Herfindahl Index for measuring crop
diversification, locational quotient measure for crop
concentration (Ghosh, 2011), and Sapre and Deshpande
index (Sapre and Deshpande, 1964) for measuring
ranks of the districts as per their agricultural
productivity. To compare inter-districts disparity,
various inequality measures were calculated; these
included relative mean deviation, coefficient of
variation, standard deviation of logs, Gini coefficient
and Theil index. The correlation and panel data
regression was used for the comparative analysis.

This study has conducted analysis at disaggregated
level by districts and has analyzed 30 crops grown in
Odisha covering 96-98 per cent of gross cropped area.
The crops selected for the study were (i) Cereals: rice,
wheat, maize, jawar, ragi, bajra, small millets; all kharif
and rabi pulses: biri, mung, kulthi, arhar, gram; (ii)
Oilseeds: til, mustard, groundnut, castor, linseed, niger,
sun flower, safflower; (iii) Vegetables: onion, potato,
sweet potato; sugarcane; mesta; fibre crops: sunhemp,
jute and cotton; and (iv) Cash crops: chilies, corriender,
garlic, ginger, turmeric. The study period, 1980 to 2005
was divided into five sub-periods and quinquennial
averages were worked out. The data were obtained from
published sources of Directorate of Agriculture and
Food Production, Government of Odisha,
Bhubaneswar.
Crop Diversification
To study crop diversification, Herfindahl index was
used. It is defined as:
proportion of area under ith crop,
, where, pi is the
, in
which Ai is the area under ith crop and

denotes
the total cropped area. The value of Herfindahl index
varies between 0 and 1; the value 1 depicts perfect
specialization and 0, shows perfect diversification.
Crop Concentration
Crop concentration means the variation in the
density of crops in an area or region at a given point/
Nayak : Changing Cropping Pattern, Agricultural Diversification and Productivity in Odisha
period of time (Ghosh, 2011). The concentration of a

crop in an area largely depends on its terrain,
temperature, moisture, price and income, social factors,
government policy, type of soils and many others. The
most commonly method to study crop concentration
is the Location Quotient (LQ) method.
where, Aij is the gross cropped area under the ith crop
in jth district, Aj is the gross cropped area in the jth
district, and jAj is the gross cropped area in the state
which is the summation of GCA of each district. Using
this Location Quotient method, Crop Concentration
Index (CCI) was calculated. If CCI value is higher than
unity, the component areal unit accounts for a share
greater than it would have had if the distribution was
uniform in the entire region and therefore, the areal
unit has a concentration of great agricultural
significance.
Agricultural Productivity
To assess agricultural productivity, Sapre and
Deshpande index was used. This method is widely used
because, along with crop yield levels rank, it takes
into account the proportion of area under crop. Instead
of using simple average ranks, weighted average of
ranks is used. The lower is the value of index, the higher
is the productivity level. Because, if a district has yield
level in rice which is highest in the state, its weightage
is 1; whereas if a districts yield level in rice is ranked
10th in the state, it will get weightage 10. The formula
for calculating this index can be expressed as:
where, r1,.,rn represent the rank of crops as per their

yield level in the district i in comparison to other
district and p1,.,pn represent the proportion of area
devoted to these crops in the district i.

Cropping Patterns in Odisha
The cropping pattern has shown a changing trend
in Odisha in recent years (Table 2). The area under
95
rice has increased from 48.1 per cent of the gross

cropped area of the state in 1980-1985 to 55.6 per cent
during 2000-2005. There is a marginal increase in the
area under cash crops.
The temporal changes in cropping pattern in
Odisha reveal crop diversification in the state. During
the period 1980-2005, the area under some cereals
(wheat, ragi, jowar, bajra and small millets); rabi pulses;
oilseeds (castor, linseed, and safflower); and cash crops
(jute, mesta, tobacco) declined by varying degrees, and
increased under selected cereals (rice and maize);
oilseed crop sunflower; cash crop cotton; and
condiment and spice crop ginger. The vegetable crops
like potato, sweet potato and onion have not revealed
any clear pattern of changes during the period studied.
Based on a similar periodical trend change, the
cropping pattern across the selected 30 crops revealed
three major groups. One, where area under crops
increased from 50 per cent to around 60 per cent; Two,
crops in which area decreased from around 25 per cent
to 16 per cent, and Three, crops group sharing between
12 and 15 per cent area, did not reveal any clear trend
of increasing or decreasing area.
It can be observed from Table 2 that even if there
is diversification as reflected through the periodical
changes in the cropping pattern of around ten crops
(wheat, ragi, jowar, bajra, small millets, castor, linseed,
safflower, jute, mesta and tobacco), it is outweighed
by the specialized and concentrated crops. Although
the number of these specialized crops is small (viz.,
rice, maize, sunflower, cotton and ginger), the total
share of their area is high. Among individual crops,
rice has the highest share in area, and this has increased
over time in the two and half decades under
consideration. An important implication of this
observation is that contrary to the association of crop
diversification with agricultural development,
specialization is associated with higher
commercialization and agricultural development in
Odisha.
Crop Diversification in Odisha
The computed values of Herfindahl index are
presented in Table 3 district-wise for the state of Odisha.
A perusal of Table 3 reveals that during the first period
(1980-1985), except Cuttack, Dhenkanal, Koraput,
Phulbani, all other districts were relatively specialized
96
Vol. 29 (No.1)
January-June 2016
Table 2. Changes in cropping pattern in Odisha: 1980 to 2005

Crop
groups
Crops
Cereals
Rice
Wheat
Ragi
Maize
Jowar
Bajra
Small millets
Total cereals
48.12
0.83
3.55
1.81
0.39
0.11
2.42
57.25
48.06
0.50
2.91
1.83
0.36
0.09
1.24
54.98
47.31
0.27
2.51
2.02
0.22
0.07
0.72
53.13
53.97
0.21
2.41
1.91
0.19
0.05
0.58
59.32
55.58
0.21
2.39
2.21
0.14
0.04
0.41
61.00
Pulses
Kharif pulses
Rabi pulses
Total pulses
Total food grains
5.92
15.19
21.11
78.35
6.96
15.13
22.09
77.08
7.24
14.91
22.15
75.27
6.64
13.49
20.13
79.45
6.89
12.41
19.30
80.30
Oilseeds
Groundnut
Til
Castor
Mustard
Linseed
Niger
Sunflower
Safflower
Total oilseeds
2.79
2.74
0.51
1.45
0.39
1.97
0.01
0.05
9.92
4.02
3.29
0.39
1.54
0.36
2.05
0.04
0.04
11.72
3.59
4.35
0.29
1.72
0.34
2.08
0.05
0.03
12.44
3.34
3.42
0.29
1.49
0.35
1.83
0.07
0.03
10.82
2.73
3.09
0.21
1.32
0.25
1.70
0.07
0.02
9.40
Vegetables
Potato
Sweet potato
Onion
Total vegetables
0.10
0.63
0.46
8.09
0.11
0.63
0.49
7.72
0.11
0.54
0.48
9.01
0.11
0.60
0.59
6.16
0.10
0.54
0.29
7.08
Cash crops
Sugarcane
Jute
Mesta
Sunhemp
Cotton
0.57
0.54
0.48
0.13
0.04
0.51
0.39
0.43
0.15
0.05
0.44
0.35
0.34
0.14
0.06
0.53
0.22
0.36
0.14
0.30
0.29
0.15
0.32
0.13
0.53
Condiments and
spices
Chilies
Coriander
Garlic
Ginger
Turmeric
Total condiments and spices
Tobacco
0.95
0.19
0.19
0.07
0.28
1.67
0.21
0.98
0.21
0.23
0.08
0.27
1.78
0.17
1.01
0.23
0.20
0.11
0.26
1.81
0.13
1.01
0.26
0.19
0.15
0.30
1.91
0.10
0.92
0.21
0.13
0.18
0.31
1.74
0.06
1980-1985
Share in total cropped area (%)

1985-1990 1990-1995 1995-2000
2000-2005
Source: Calculated from area data obtained from published sources of Directorate of Agriculture and Food Production,
Government of Odisha
Note: List of crops taken here cover 96-98 per cent of GCA in the districts of Odisha.
97
Table 3. District-wise crop diversification index in Odisha, 1980-2005

District
Balasore
Bolangir
Cuttack
Dhenkanal
Ganjam
Kalahandi
Keonjhar
Koraput
Mayurbhanj
Phulbani
Puri
Sambalpur
Sundergarh
Odisha
1980-1985
1985-1990
HI index value
1990-1995
0.69
0.68
0.61
0.61
0.69
0.65
0.65
0.58
0.68
0.48
0.68
0.69
0.69
0.65
0.66
0.61
0.60
0.54
0.64
0.63
0.62
0.57
0.65
0.50
0.70
0.64
0.64
0.61
0.62
0.60
0.58
0.50
0.65
0.67
0.58
0.57
0.62
0.48
0.72
0.60
0.61
0.60
1995-2000
2000-2005
0.74
0.66
0.64
0.57
0.70
0.64
0.63
0.58
0.69
0.54
0.76
0.64
0.69
0.65
0.73
0.69
0.67
0.55
0.67
0.68
0.59
0.60
0.72
0.53
0.75
0.69
0.74
0.66
Source: Authors calculations based on area data for crops taken from Directorate of Agriculture & Food Production,
Odisha, Bhubaneswar
districts with HI-index value higher than the states

average value. In the second period (1985-1990),
except four districts, viz. Cuttack, Dhenkanal, Koraput,
and Phulbani, all other districts have depicted HI-index
value higher than 0.6 (i.e. relatively more specialized);
the district Phulbani has been found to be the least
specialized district. In the third period (1990-1995),
Puri, Ganjam, Balasore, Kalahandi, Mayurbhanj and
Sundergarh have been found highly specialized
districts, while Phulbani, Koraput, Keonjhar,
Dhenkanal and Cuttack were the least specialized
districts. In the fourth period (1995-2000), there is only
one district, viz. Kalahandi where HI value has
declined. In the fifth period (2000-2005), the HI values
increased to the state average value in seven districts,
while, HI values declined vis-a-vis previous period in
six districts. In this period, all the thirteen districts have
depicted HI-index value greater than 0.5. A comparison
of HI values for the periods 1980-1985 and 2000-2005
reveal that except Phulbani, all other districts have
experienced a significant degree of crop specialization.
The districts Balasore and Puri, known as rice bowl
of Odisha, have depicted higher values of HI, close to
0.7 in the first period. Five other districts have also
experienced a similar kind of trend and this signifies
crop specialization in the area. From first to second

period, there is a clear trend of decline in HI values for
all the districts, except Phulbani and Puri. In the next
phase of second to third period, except Ganjam,
Kalahandi, and Puri, the rest 10 districts continued the
declining trend in their HI values. From third to fourth
period, the HI values for all districts have shown a
significant change towards crop specialization, with
the exception in Kalahandi district. The last phase has
shown a mix of decline and increase in the number of
districts in terms of their HI values. The average value
of HI index for all the districts was computed as 0.65
in the first period, and it went down to 0.61 in second
period and further to 0.60 in the third period, but in the
later two periods, it again scored 0.65 and 0.66 values.
Hence, we may conclude that in Odisha most of
the districts are experiencing a lateral movement toward
crop specialization and the crop diversification is not
happening. This finding needs to be seen in the context
of the fact that most tribal-dominated/less-developed
districts (such as Phulbani) have depicted a higher level
of diversification. The agricultural development has
mostly taken the form of shift toward rice production
in most districts (such as Sambalpur and Bargarh) of
the state.
98
HI value
Herfindahl index
Cropping intensity
Crop yield
Net irrigated area
Number of landholdings
Area operated
Chemical fertilizer consumption
Share of total foodgrain in GCA (%)
Share of total oilseeds in GCA (%)
Share of total vegetables in GCA (%)
Share of total condiments and spices in
GCA (%)
Share of other total crops in GCA (%)
1.000
0.421**
-0.011
0.305*
0.292*
0.123
0.339**
0.972**
-0.847**
-0.154
-0.624**
January-June 2016
Table 4 shows the correlation between HI values

and factors affecting crop diversification in Odisha. It
shows that correlation coefficient for share of total
oilseeds is highest (0.972), followed by the share of
condiments and spices, vegetables, other total crops
and crop yield and these are correlated positively for
making diversification in crop cultivation. On the other
hand, crop specialization observed in Odisha, is
supported positively by the share of total foodgrains,
cropping intensity, chemical fertilizer consumption, net
irrigated area, number of landholdings, and area
operated.
Table 4. Correlates of crop diversification in Odisha

Correlate
Vol. 29 (No.1)
-0.061
Source: Computed data.

Note: * and ** show significance at 0.05 per cent and 0.01
per cent levels, respectively.
Determinants of Crop Diversification

Among the host of factors, crop yield and cropping
intensity are the most important determinants of crop
choice. If yield level of crop increases, then these crops
attract a major chunk of area in a cropping season. But,
if cropping intensity increases, then crop diversification
takes place. Hence, we hypothesized that crop yield
will be negatively and cropping intensity will be
positively correlated with crop specialization index.
The regression estimates for crop diversification

are given in Table 5, where the dependent variable HI
indicates that higher is the value less will be the crop
diversification and vice versa. The independent
variables are cropping intensity, irrigation, fertilizer,
share of area under fertilizer and area share under
condiments and spices, and foodgrains. The problem
of multicollinearity has been taken care of and hence,
the mean value of variance inflationary factor (VIF)
has been estimated as 2.18, showing no
multicollinearity.
The overall significance of F value is quite
satisfactory and R-squared value showing 95 per cent
variation has been determined by these independent
variables. Fertilizer consumption has not emerged
significant for crop diversification. Although marginal,
irrigation and area under condiments and spices crops
Table 5. Regression estimates for crop diversification in Odisha

Dependent variable: Herfindahl index
Independent variable
Cropping intensity
Irrigation
Fertilizer
Share of area under foodgrains
Share of area under condiments
& spices
Constant
Coefficients
P>|t|
Variance inflationary
factor
0.000 (0.000)
-0.000 (0.000)
0.000 (0.000)
0.012 (0.001)
-0.003 (0.001)
0.009**
0.018**
0.648
0.000***
0.081**
2.06
3.17
2.23
1.86
1.59
-0.371 (0.042)
0.000
Prob>F
R-squared
value
0.000***
0.95
Source: Authors calculations.

Note: *** and ** Indicate significance at the 1 per cent and 5 per cent levels, respectively; Standard errors are reported in
parentheses of coefficients;
Mean VIF = 2.18
Number of observations = 65.
99
Table 6. Percentage share of GCA under total foodgrains in Odisha across districts against other crops
(per cent)
District
1980-1985
1985-1990
1990-1995
1995-2000
82.15
81.32
77.25
77.47
82.18
78.25
76.81
74.08
81.91
68.37
79.32
81.89
82.20
78.35
80.05
76.70
76.00
70.71
78.45
77.94
76.96
73.79
80.21
66.76
83.28
78.76
79.26
77.08
77.19
75.85
74.04
66.81
81.06
73.61
74.34
74.23
77.67
64.97
84.26
75.59
76.30
75.27
85.36
80.10
78.78
71.89
83.02
78.76
76.70
75.10
82.61
70.96
87.28
78.71
82.10
79.45
Balasore
Bolangir
Cuttack
Dhenkanal
Ganjam
Kalahandi
Keonjhar
Koraput
Mayurbhanj
Phulbani
Puri
Sambalpur
Sundergarh
Odisha
2000-2005
84.66
82.54
80.86
71.85
80.80
81.70
74.71
75.77
84.46
70.66
85.75
82.32
85.29
80.30
Source: Calculated from area data obtained from published sources of Directorate of Agriculture and Food Production,
Government of Odisha
influence crop diversification (at 5 % level). Area share

under foodgrains (at 1 % significance level) and
cropping intensity (at 5 % significance level) exercise
crop specialization.
The extent of crop diversification varies across
districts, depending upon agro-climatic conditions and
also equally important factors are crop yield and
cropping intensity. In this context, Table 6 shows that
there has been a significant share of total foodgrains
in gross cropped area, which shows the occurrence of
reverse of diversification, i.e. crop specialization. It is
seen that not a single district has less than 65 per cent
gross cropped area under total foodgrain crops. It
implies that this is the sole reason of crop specialization
in the state. It was therefore decided to see which crops
are getting concentrated across the districts in
comparison to the state as a whole.
Crop Concentration
The crop concentration index (Table 7) shows that

out of 30 crops grown in Odisha, the focus is only
on following 7 crops: Rice, Maize, Ragi, Niger,
Mustard, Til, Groundnut and Turmeric.
In all the 13 districts, the value of concentration
index for rice is higher than for rest of the crops
grown in Odisha in all the five sub-periods from
1980 to 2005.
The CCI value for maize has been found greater

than unity in only two districts Phulbani and
Keonjhar for all the five sub-periods.
As per CCI value, ragi has depicted concentration

only once in the first period (1980-1985) and only
in two districts, Phulbani (1.53) and Ganjam
(1.05).
Niger crop has shown concentration in only two

districts, Phulbani and Keonjhar, although it has
shown presence in all the sub-periods.
The concentration of mustard crop has been

observed only in one district, Phulbani, in all the
five sub-periods, while groundnut has depicted
concentration in only one period (1985-1990) and
one district, Dhenkanal.
Til and turmeric crops have shown concentration

in Dhenkanal and Phulbani districts, respectively.
Productivity Level and Inter-district Disparities

Table 8 presents the ranking of the districts in
Odisha as per Sapre and Deshpande index. It shows a
declining trend in agricultural productivity in Odisha
for the period 1980 to 1995 and increasing trend for
the period 1995 to 2005. The results remain somewhat
inconclusive as the overall productivity has decreased
between 1980 and 2005 and the value of index has
Sundergarh
Keonjhar
Mayurbhanj
Balasore
Bolangir
Phulbani
Dhenkanal
Sambalpur
Puri
Ganjam
Kalahandi
Cuttack
Koraput
Phulbani
Keonjhar
Phulbani
Ganjam
Phulbani
Phulbani
Dhenkanal
Phulbani
Rice
Mustard
Til
Groundnut
Turmeric
Dhenkanal
Phulbani
1.29
1.13
12.74
11.89
10.59
9.01
8.02
7.20
6.39
6.38
5.55
4.44
4.08
3.64
3.17
1.56
1.20
2.33
1.12
1.50
1.05
CCI
Sundergarh
Keonjhar
Mayurbhanj
Phulbani
Balasore
Puri
Bolangir
Sambalpur
Dhenkanal
Kalahandi
Ganjam
Cuttack
Koraput
Phulbani
Keonjhar
Phulbani
Keonjhar
Phulbani
Kalahandi
Dhenkanal
Phulbani
Districts
12.56
12.27
10.42
9.06
8.35
7.40
7.23
5.91
5.03
4.22
4.07
3.85
2.99
1.68
1.22
2.76
1.51
2.00
1.54
1.49
1.13
CCI
Phulbani
Sundergarh
Phulbani
Keonjhar
Mayurbhanj
Balasore
Bolangir
Puri
Sambalpur
Dhenkanal
Ganjam
Kalahandi
Cuttack
Koraput
Phulbani
Keonjhar
Phulbani
Keonjhar
Phulbani
Dhenkanal
Districts
1995-2000
1.49
13.20
12.90
12.33
11.75
11.29
8.01
7.51
6.38
5.08
5.00
4.91
4.58
3.30
1.81
1.23
2.17
1.48
1.80
1.36
CCI
Phulbani
Sundergarh
Phulbani
Mayurbhanj
Balasore
Keonjhar
Bolangir
Puri
Dhenkanal
Sambalpur
Kalahandi
Ganjam
Cuttack
Koraput
Phulbani
Keonjhar
Phulbani
Keonjhar
Phulbani
Dhenkanal
Districts
2000-2005
1.14
14.75
13.14
12.56
11.67
10.36
8.07
7.69
6.56
6.37
5.83
4.82
4.55
3.49
1.69
1.13
1.20
1.13
1.42
1.42
CCI
Source: Authors calculations.

Note: Though here calculation of crop concentration index is made for 30 crops, in the above table only the major crops concentration index has been shown.
1.31
Sundergarh
Keonjhar
Mayurbhanj
Balasore
Phulbani
Bolangir
Sambalpur
Puri
Dhenkanal
Kalahandi
Ganjam
Cuttack
Koraput
Phulbani
Keonjhar
Phulbani
Keonjhar
Phulbani
Dhenkanal
Districts
1990-1995
Vol. 29 (No.1)
Niger
Ragi
14.52
12.90
11.08
9.83
8.11
7.36
7.10
7.00
5.62
4.48
4.38
3.63
2.92
1.49
1.13
1.53
1.05
2.06
1.52
1.11
CCI
1985-1990
Maize
Districts
Crop
1980-1985
Table 7. Crop concentration index (CCI) of major crops in Odisha 1980-2005
100
January-June 2016
101
Table 8. Ranking of districts as per productivity index, 1980-2005

Ranking
1980-1985
1985-1990
1990-1995
1995-2000
2000-2005
Cuttack
1.44
Cuttack
1.85
Cuttack
2.04
Cuttack
1.82
Balasore
3.50
Sambalpur
3.83
Puri
4.83
Sambalpur
4.22
Sambalpur
3.64
Cuttack
3.69
Koraput
5.09
Sambalpur
5.19
Ganjam
5.72
Ganjam
5.43
Sambalpur
4.56
Ganjam
5.60
Koraput
5.78
Dhenkanal
6.09
Koraput
6.04
Koraput
5.65
Puri
6.16
Dhenkanal
6.25
Koraput
6.15
Balasore
6.86
Mayurbhanj 6.10
Mayurbhanj
6.69
Balasore
6.57
Puri
6.51
Puri
7.00
Phulbani
7.01
Dhenkanal
6.73
Mayurbhanj
6.92
Balasore
7.14
Bolangir
7.58
Ganjam
7.03
Phulbani
7.07
Phulbani
7.59
Bolangir
7.68
Phulbani
7.88
Kalahandi
8.16
Bolangir
7.49
Ganjam
7.84
Mayurbhanj
7.74
Sundergarh
8.28
Keonjhar
8.20
10
Balasore
7.57
Bolangir
7.89
Keonjhar
8.71
Mayurbhanj 8.44
Bolangir
8.44
11
Kalahandi
9.63
Keonjhar
8.83
Phulbani
9.09
Keonjhar
8.66
Puri
8.61
12
Keonjhar
10.53
Kalahandi
9.50
Kalahandi
9.29
Dhenkanal
9.06
Dhenkanal
8.84
13
Sundergarh
10.69
Sundergarh
10.00
Sundergarh
9.54
Kalahandi
9.20
Sundergarh 9.72
Odisha
6.81
Odisha
6.85
Odisha
6.92
Odisha
6.91
Odisha
6.89
Source: Calculated form the data Directorate of Agriculture and Food Production, Odisha, Bhubaneswar.
Note: Lower the index value higher is the productivity level. Hence, author has presumed here that taking the productivity
index values of Odisha as bench-mark the districts having productivity index value between 0 and 6 are considered more
productive, while those having productivity index value higher than 6 are regarded low productive. .
Figure 1. Disparity in agricultural productivity among districts of Odisha, 1980-2005

Source: Based on Table 8.
Note: Lower is the value of index, higher is the agricultural productivity level and vice-versa.
102
Vol. 29 (No.1)
January-June 2016
Table 9. Spatio-temporal disparity in agricultural productivity in Odisha

Inequality measures
Relative mean deviation
Coefficient of variation
log standard deviation
Gini coefficient
Theil index
1980-1985
1985-1990
1990-1995
1995-2000
2000-2005
0.38
0.44
0.57
0.25
0.13
0.30
0.36
0.46
0.20
0.09
0.28
0.35
0.44
0.19
0.08
0.34
0.41
0.53
0.23
0.11
0.35
0.42
0.49
0.24
0.11
Sources: Based on Table 8 data.
increased marginally. Nevertheless, the Sapre and

Deshpande index values for the districts capture
important characteristics of inter-district variations in
Odishas agricultural productivity.
It is visible that the districts of Sundergarh,
Bolangir, Dhenkanal, Kalahandi, Keonjhar,
Mayurbhanj and Phulbani have less agricultural
productivity. On the other hand, the districts like
Cuttack, Sambalpur and Koraput have higher
agricultural productivity, followed by Puri, Ganjam and
Balasore districts. More observations can be made on
the graph over temporal behaviour of Odisha by
districts. The disparity in productivity is distinctly
visible in the radar diagram (Figure 1). The points
closer to the centre in the diagram are more productive.
The distance from centre to the radius is 12 unit.
Table 9 depicts spatio-temporal disparity in
agricultural productivity in Odisha based on various
inequality measures. A perusal of Table 9 reveals that
the values of Theil index, coefficient of variation,
relative mean deviation, Gini coefficient and log
standard deviation depict a similar kind of slightly
declining trend during the first three periods, and
increase in these values in the fourth period, but still
lower than in the first period. As a whole, all the
inequality measures show a reduction in inequality
during the study period 1980 to 2005. Hence, we may
say that the districts in Odisha are converging as far as
agricultural productivity is concerned.
The agrarian structure and productivity interlinkage can be described better based on Table 10,
which shows correlation among the yield rate of
foodgrains, fertilizer consumption, irrigation
(percentage NIA to NSA), proportion of share-cropping
in total operated area, proportion of leased-in land in
total operated area, and proportion of marginal, small
and large size landholdings in total operated area. The

agrarian structure is based on technological and
institutional factors; here the variables fertilizer and
irrigation represent technological factors and the other
variables, viz. share-cropping, leased-in land, marginal,
small and large size landholdings represent institutional
factors.
A perusal of Table 10 reveals that the technological
factors, fertilizer and irrigation, are correlated
positively with the yield rate and show significance at
0.01 level of probability. But, the marginal landholding
size is negatively correlated with the yield rate and is
significant at 0.05 level of probability. The leased-in
land and yield rate are positively (24%) correlated and
are significant at 0.10 level of probability.
The correlation of yield rate with other variables
of institutional factors have not been found significant.
Fertilizer and irrigation, the technological factors are
positively (55%) correlated and are significant at 0.01
per cent level of probability. The leasing-in of land
has been adopted largely in the technologicallydeveloped regions as the correlation coefficients of
fertilizer and irrigation are positive and significant at
0.01 level of probability. The correlation coefficients
of marginal size of landholding with fertilizer and
irrigation are significant at 0.01 level. It shows that
fragmentation of landholding is less in the
technologically-developed area and tenancy is
associated with the low technology adoption. The
correlation coefficient of large landholding size with
fertilizer is significant at 0.10 level and with irrigation
is significant at 0.01 level of probability. It supports
that agricultural technology is being adopted by large
farmers. The large farmers have been found involved
in share-cropping. The share-cropping has been found
quite significant in technologically-developed regions.
103
Table 10. Correlates of productivity, technological and institutional factors, 1980-2005

Correlate
Yield
rate
Fertilizer
Irrigation
Yield rate
Fertilizer
Irrigation
Share-cropping
Leased-in land
Marginal landholding size
Small landholding size
Large landholding size
1.00
0.44*** 1.00
0.32*** 0.55*** 1.00
-0.06
0.08
0.35***
0.24*
0.63*** 0.42***
-0.29** -0.33*** -0.56***
-0.05
-0.09
-0.18
0.08
0.21*
0.35***
Share- Leasedcropping in land
1.00
0.34***
-0.61***
-0.12
0.37***
Marginal
Small
Large
landholding landholding landholding
size
size
size
1.00
-0.34***
-0.22*
0.29**
1.00
0.22*
-0.27**
1.00
-0.20
1.00
Source: Calculated from Agricultural Statistics, Statistical Abstract and Agricultural Censuses.
Note: ***, ** and * denote significance at 0.01 per cent, 0.05 per cent and 0.10 per cent levels of probability, respectively.
The next logical step was to probe into the causes

for inter-district variations in agricultural performance
in Odisha. To identify such factors, we estimated a
panel data regression equation by taking yield rate of
foodgrains for the five sub-periods in 13 districts of
Odisha as a dependent variable and fertilizer use,
cropping intensity, irrigation, percentage of leased-in
land in total operational area and four dummies, d 1, d
2, d 3 and d 4, for the years 2000-01, 1995-96, 199091 and 1985-86, respectively as the explanatory
variables. The cropping intensity and irrigation were

taken as proxy for infrastructure, percentage of leasedin land for institutional factor, fertilizer-use for
technological factor and dummies for the previous
years for capturing favourable climatic conditions in
Odisha.
In the panel data regression (Table 11), fertilizer
consumption (positively), d 1 dummy for the year
2000-01 (negatively) and d 2 dummy (positively) for
Table 11. Determinants of agricultural productivity in Odisha

Dependent variable : Yield rate
Independent variables
d1
d2
d3
d4
Fertilizer
Cropping intensity
Irrigation
Leased-in land
Constant
Coefficients
P>|t|
Variance inflationary
factor
-166.39 (77.01)
151.15 (69.16)
59.14 (66.77)
110.58 (64.03)
5.13 (1.44)
-2.15 (2.30)
2.07 (2.34)
3.12 (6.54)
987.87 (242.58)
0.035**
0.033**
0.380
0.090
0.001***
0.354
0.381
0.635
0.000***
2.40
1.94
1.81
1.66
2.56
2.92
2.81
2.33
Prob > F
value
R-squared
0.000***
0.468
Source: Authors calculation.

Note: *** and ** indicate significance at the 1 per cent and 5 per cent levels, respectively.
Standard errors are reported in parentheses of coefficients;
Mean VIF = 2.30
Number of observations = 65.
104
the year 1995-96 were the only variables that exercised

maximum influence on the yield of foodgrains by
affecting it significantly at 1 per cent and 5 per cent
levels, respectively. Other factors have not been found
statistically significant. Thus, the results indicate
fertilizer consumption to be the most important
determining factor of agricultural productivity, which
again proves that yield in Odisha agriculture is
technology-driven and not because of institutional
agrarian structure. The negative coefficient for the
dummy variable d 1 shows that 2000-01 was a bad
year for disaster-prone, flood, and drought affected
Odisha. The positive value of coefficient for dummy d
2 shows the year 1995-96 to be good from agricultural
point of view. The R-squared value 0.468 indicates that
around 47 per cent of inter-district variations in yield
could be explained by this regression model.
Concluding Remarks
The study has revealed that Odisha is basically a
mono-crop (rice) state. The reasons behind increasing
preference for production of rice are irrigation facilities
and provision of minimum support price (MSP) by the
government for this crop. An assured and increasing
price of rice vis--vis other crops has resulted in a
higher preference for rice cultivation. Thus, there has
been a move towards specialization rather than
diversification as a result of agricultural development
in the state.
Agricultural productivity level across the districts
of Odisha is converging as per the Sapre and Deshpande
index and the values for various inequality measures
have been found out in the analysis. The ranks of
districts as per productivity level urge for micro level
research across low-productivity districts by adopting
the model of high-productive districts (Balasore,
Cuttack and Sambalpur).
The inter-district variations in yield rate are mostly
related to difference in fertilizer consumption and
climatic conditions. Therefore, agricultural
Vol. 29 (No.1)
January-June 2016
productivity can be enhanced by increasing

technological factors and promoting cropdiversification. To combat exogenous factors such as
disasters like floods and droughts, the Government of
Odisha should accord due attention to plantation
programmes and related measures. The study has
suggested that the government should take proper steps
on management of the existing institutional drawbacks.
Acknowledgements
The author is highly grateful to his MPhil
supervisor Dr Deepak K. Mishra, Professor at CSRD,
SSS, Jawaharlal Nehru University, New Delhi, for his
support in carrying out this study. The suggestions
given by the Editor, AERR, are gratefully
acknowledged. The usual disclaimer apply.
References
Ghosh, B.K. (2011) Essence of crop diversification: A study
of West Bengal agriculture. Asian Journal of
Agricultural Research, 5(1): 28-44.
Nayak, Dinesh Kumar (2011) Agrarian Structure and
Agricultural Productivity in Orissa: An Analysis of
Inter-district Variations. MPhil Dissertation, submitted
to Jawaharlal Nehru University.
Pattnaik, I. and Shah, A. (2010) Is there a glimpse of
dynamism in Orissas agriculture? Economic and
Political Weekly, 45(26 & 27): 756-59.
Sapre, S.C. and Deshpande, C.D. (1964) Inter-district
variation in agricultural efficiency in Maharashtra state.
Indian Journal of Agricultural Economics, 19(1): 24252.
Swain, M. (1999) Tenancy structure in Orissa: Implications
for agricultural growth. Artha Vijnana, 41(3): 245-61.
Tripathy, P.K. and Sarap, K. (1994) Development profile of
Orissas agriculture. In: Compressive History and
Culture of Orissa, Vol II (Part 2), Ed: P.K. Mishra.
Kaveri Books, New Delhi. pp. 969-77.
Received: February, 2015; Accepted: December, 2015

DOI: 10.5958/0974-0279.2016.00023.9
Crisis of Sustainability or Perils of Ill-managed Open Access

Fisheries? Analysis of Long-term Catch Trends in Marine
Fisheries of Maharashtra and India
M. Suresha Adigaa*, P.S. Ananthana, H.V. Divya Kumarib and V. Ramasubramaniana
Central Institute of Fisheries Education, Versova, Andheri (W), Mumbai-400 061, Maharashtra
b
Department of Aquaculture, College of Fisheries, Mangalore-575 002, Karnataka
Abstract
The dynamics in the landings of marine fisheries resources in Maharashtra over the past five decades
(1961-2010) have been examined after classifying them into 24 different resource groups. The decadewise compound growth rates (CGR) of different resource groups and coefficient of variation have been
calculated. The present status of 25 resource groups has been examined separately based on a simple
criterion. The study has revealed that marine fishery in Maharashtra state is facing crisis since late-1990s.
Most of the fish stocks that were classified as abundant and less abundant at the country level, have come
under declining category in the case of Maharashtra. It indicates that state-wise scenario is different from
the national scenario and state-wise understanding of marine fishery resources is very important for
formulating regulatory and management measures.
Key words: Compound growth rate, decadal fish landings, fisheries, stock status, Maharashtra, India
JEL Classification: Q22
Introduction
The marine fish landings in India have fallen by 4
per cent to 3.78 million tonnes (MT) in 2013 from the
all-time high of 3.94 (Mt) in 2012 (CMFRI, 2014).
Between 1996 and 2009, the marine fish landings have
been fluctuating or became almost stagnant. The review
of studies on macro level growth trends and stock
assessments has revealed that most of the fisheries
resources are under stress or over-exploited to
maximum sustainable yield (MSY) level; some are
fully exploited and only a few remain under-exploited
(Srinath, 2003; Srinath and Balan, 2003; Dehadrai and
Yadava, 2004; Pillai, 2006). The demersal fisheries,
that constitute about 45 per cent of total landings, were
over-fished mainly by the mechanized sector
(Vivekanandan and Jayasankar, 2008). The economic
Email: suresha1947@gmail.com
loss of juvenile fishing in India was estimated at

` 85,558 crore (US$190 million) which is an indication
of the extent of over-exploitation (Najmudeen and
Sathiadhas, 2008) and perhaps the impending crisis in
the fisheries sector. Vivekanandan et al. (2009) have
reviewed the status of sustainability of 98 species by
constructing a Sustainability Index (SI). On a scale of
1-6, most fishes (68%) had SI values ranging between
3 and 4, indicating a medium level of sustainability. It
was found that 12 species had SI values below 3, which
may be considered as the vulnerable species.
Based on the rate of exploitation, James (2010)
has cautioned that several single species fisheries in
India may tilt the marine fish production in any year.
The dynamics of marine fish landings in India during
the past 60 years were analyzed by Sathianandan et al.
(2011), who are largely optimistic about the fish landing
scenario at the macro level. Of the 26 resource groups
106
studied following the method suggested by Mohamed

et al. (2010), it was found that 18 resource groups fell
under the abundant class 5 groups fell under lessabundant class, and 1 group each was under
declining, depleted and collapsed classes, with
important resource groups being either under
abundant or less-abundant class. However, the
assessment of commercially important marine fish
landing trends at the state or provincial level is crucial,
at least for two reasons: one, fishing in territorial waters
is governed by the respective state governments, and
two, availability of different fishery resources and their
rate of exploitation are different across states. No
detailed long-term catch trend analysis seems to have
been carried out in the state of Maharashtra, though
changes on a yearly basis are captured in annual reports
of CMFRI (CMFRI, 2012). Understanding the pattern
and crucial milestones in fish landings over a period
of time is sine quo non would provide insights into the
level of resource exploitation with reference to the
potential as well as help prepare appropriate
management measures.
Vol. 29 (No.1)
January-June 2016
was estimated to see the trend in contribution of major

group to state production over a period of time. The
decade-wise trend in marine fish landings and major
developments in fishery in the state were then
summarized.
In this paper, the dynamics of marine fish landings

in Maharashtra have been analyzed over the past five
decades. The status of stocks and its sustainability
depend on biological, ecological and environmental
factors. Various methodologies are available to study
the stock status and sustainability of fisheries, which
are very complex and need intensive data. Time series
landings data on major resource groups provide a broad
understanding on the status of stock.
The present sustainability status of different

resources for the state was estimated based on the
criteria proposed by Mohamed et al. (2010) by
classifying resources into five groups, viz., abundant,
less-abundant, declining, depleted and collapsed.
According to them, since stock abundance data were
not available for all the species, it was assumed that
catch is proportional to abundance and the historical
maximum catch of species (e.g. for past 50 years) was
taken as the baseline catch. This assumption reiterates
that abundance would be close to the figure for
maximum catch. For a comparison, the recent average
catch of that species (recent 3 years) was compared to
that of the baseline catch in percentage. Deciding a
cut-off percentage, though arbitrary, is necessary to
classify the stocks. The decision on the cut-off
percentage was based on the range of percentages seen
and a-priory knowledge of the stock catch and
abundance. Based on this, the stocks were classified
as the stocks as abundant (recent average catches
>70% of the historical maximum); less-abundant (5069%); declining (11-49%); depleted (6-10%) and
collapsed (< 5%). In the present study, percentage
contributions of these five groups to average total state
catch for the past ten years (2001-2010) and for past
three years (2008-2010) were identified.
Materials and Methodology
The data on marine fish production in Maharashtra

for the past fifty years (1961-2010) were collected from
the publications of Central Marine Fisheries Research
Institute, Cochin. The marine species were classified
into 24 groups. For Maharashtra, compound growth
rates (CGR) were estimated decade-wise for both
overall catch and total landings of each species group
during 1961 to 2010. The CGR was also estimated for
the total and decade-wise landings of India from 1961
to 2010 mainly to compare with the state growth trend.
The coefficient of variation (CV) was worked out for
each decade and the decadal mean catch was identified
to see the trends over the period 1961-2010. The
decade-wise percentage contribution of each group to
the total state landings during the respective decades
Total Marine Fish Landings in Maharashtra

Over the past five decades, Indias marine fish
catch has witnessed a tremendous growth with average
landings increasing from 8,32,426 tonnes during 19611970 to 27,38,943 tonnes during 2001-2010. During
the period 1961-1970 to 1990-2000, the marine fish
production in Maharashtra also showed an increasing
trend. Maharashtras contribution to Indias average
landings decreased from 16.63 per cent during 19611970 to 12.59 per cent during 2001-2010. The decadal
mean catch, coefficient of variation, growth rates in
India and Maharashtra marine landings are summarized
in Tables 1 and 2, respectively. Over the decades, the
state average landings increased significantly whereas
the compound growth rate decreased. The analysis of
Suresha Adiga et al. : Crisis of Sustainability or Perils of Ill-managed Open Access Fisheries?
107
Table 1. Decadal mean catch and coefficient of variation (CV) for marine fisheries in Indian and Maharashtra
Period
India
1951-1960
1961-1970
1971-1980
1981-1990
1991-2000
2001-2010
1961-2010
Mean catch (tonnes)

Maharashtra
664527
832426
1259624
1692597
2408687
2738943
1786455
138398
247097
306175
349536
344888
277219
Table 2. Decade-wise compound growth rates (CGR) of

marine fish catches for India and Maharashtra,
1961-2010
Period
1951-1960
1961-1970
1971-1980
1981-1990
1991-2000
2001-2010
1961-2010
CGR
India
Maharashtra
4.90
5.15
2.26
5.00
2.04
3.55
3.11
4.03
3.09
3.16
0.41
-4.69
2.15
growth trend in Maharashtra marine fish landings for

different decades for the period 1961-2010 showed an
increase in the decade of 1961-1970, as indicated by
the compound growth rate of 4.03 per cent. During the
decade 2001-2010, the catches declined with a negative
growth trend of -4.69 per cent and the maximum
percentage of variation in catch was observed in the
same decade. The historical maximum annual landing
in the state was recorded in the year 2002, and after
that landings decreased.
Beginning of Mechanization in Fisheries: 19611970
Prior to 1960, the majority of marine resources was
landed mainly by the traditional non-mechanized
fishing crafts and artisanal gears like dolnet, rampani
and gill net. The fishing activities in that period were
carried out in the inshore and near shore waters within
20-25 m depth. During the decade of 1961-1970,
mechanization of the traditionally used bag net, gill
India
CV
Maharashtra
Percentage of
Maharashtra to India
19.15
15.62
9.91
15.84
7.50
11.85
41.39
16.12
14.21
11.04
10.29
17.29
31.67
16.63
19.62
18.09
14.51
12.59
15.52
net and long line fishing crafts was vigorous. The

annual landings in the state of Maharashtra increased
from 1.16 lakh tonnes in 1961 to 1.92 lakh tonnes in
1970 with a decadal mean catch of 1.38 lakh tonnes.
Around 4.03 CGR was observed during the decade and
the inter-year growth rate during the decade showed a
significant difference (Table 3).
The major resources contributed to the fishery
during 1961-1970 were: crustaceans (31.7%),
bombayduck (19.41%), clupeids (8.37%), croakers
(6.64%), pomfrets (5.54%), mackerel (3.49%),
ribbonfish (3.45%), catfish (3.32%), eels (3.05%) and
elasmobranchs (3.02%). The majority of resources
showed a positive trend in growth as indicated by the
CGR mainly because of mechanization of nearly 50
per cent of traditional crafts and introduction of
trawling with extension of fishing beyond 30-40 m
depth during this period. The inter-year growth rate
during this decade for the overall landings showed a
significant difference. Four resources, viz. eel, unicorn
cod, perches and goatfish, showed negative trends in
the growth. The decade showed a wider variation in
landings of major groups, indicating relative instability
in the catch during this period.
Synthetification of Gears: 1971-1980
During 1971-1980, the annual marine landings
increased from 2.15 lakh tonnes in 1971 to 2.93 lakh
tonnes in 1979 with a decadal average of 2.47 lakh
tonnes. The mechanization of fishing crafts continued
vigorously during this period and also traditionally used
cotton nets were completely replaced by the synthetic
nets, which led to increased operations of dolnet by
2-3 times and increase in the landings of the majority
108
Vol. 29 (No.1)
January-June 2016
Table 3. Mean fish catch, coefficient of variation (CV), compound growth rate (CGR) and per cent contribution of
various major fish groups in Maharashtra state during 1961-1970
Fish group
Overall
Crustaceans
Bombayduck
Clupeids
Croakers
Pomfrets
Mackerel
Ribbonfish
Catfish
Eel
Elasmobranchs
Unicorn cod
Perches
Carangids
Threadfins
Seerfish
Goatfish
Silverbellies
Lizardfish
Tunas
Flatfish
Whitefish
Mullets
HB&FB
Molluscs
Mean catch
(tonnes)
CV
CGR
% of total state
landings
138398
43891
26858
11577
9191
7667
4832
4768
4600
4224
4195
3211
2147
1739
1297
1254
927
660
289
265
248
210
176
131
76
15.62
18.59
13.04
11.58
19.46
47.89
149.35
42.55
54.96
68.43
28.87
41.19
57.29
104.31
78.68
51.58
124.77
132.62
146.52
82.29
168.68
85.68
73.86
92.67
122.28
4.03
4.02
1.52
3.02
0.61
8.17
12.64
10.04
15.76
-10.41
6.50
-9.99
-11.81
23.44
1.84
14.61
-1.79
13.66
24.10
2.48
41.39
22.18
14.76
26.05
15.80
31.71
19.41
8.37
6.64
5.54
3.49
3.45
3.32
3.05
3.03
2.32
1.55
1.26
0.94
0.91
0.67
0.48
0.21
0.19
0.18
0.15
0.13
0.09
0.06
of fish species groups (Table 4). The increase in total

fish landings was also due to multi-day trips by the
shrimp trawlers (2-3 days/trip) that ventured beyond
territorial waters in the depth range of 25-40 metres.
The landings of demersal resources like perches,
croakers and mollusks, increased as trawlers began to
target these demersal resources in addition to shrimps.
Meanwhile, the mean catch of resources like mackerel,
eel, unicorn cod, goatfish, silverbellies, halfbeak and
fullbeak (HB and FB) and mullets decreased in this
decade as compared to in the previous decade. The
resources in the order of abundance were: crustaceans
(36.41%), bombay duck (18.5%), clupeids (9.66%),
croakers (6.73%), pomfrets (4.42%), catfish (4.19%)
and ribbonfish (3.15%). Though the mean catch
increased in crustaceans and catfish, the overall growth
rate showed a negative trend, as indicated by the CGRs

(-0.71% and -3.91%), mainly because of higher
fluctuations in landings between the years.
Introduction of Purse Seiners and Multi-day
Trawling: 1981-1990
The decadal mean catch increased from 2.47 lakh
tonnes in 1971-1980 to 3.06 lakh tonnes in 1981-1990
with a growth rate of 3.16 per cent and the inter-year
growth rate showed a significant difference during the
decade. Out of 24 groups of fisheries, 20 groups showed
an increasing trend in their landings during the 1980s
(Table 5). The demersal resources like croakers,
pomfrets, molluscs, perches, flatfish, lizardfish and
eels, showed a drastic increase in this decade compared
to the previous decade. The increased landings in
109
Fish group
Overall
Crustaceans
Bombayduck
Clupeids
Croakers
Pomfrets
Catfish
Ribbonfish
Elasmobranchs
Eel
Perches
Seerfish
Mackerel
Carangids
Threadfins
Unicorn cod
Molluscs
Flatfish
Lizardfish
Tunas
Goatfish
Silverbellies
Whitefish
HB & FB
Mullets
Mean catch
(tonnes)
CV
CGR
% of total state
landings
247097
89976
45717
23865
16638
10923
10358
9681
7791
3044
2834
2415
2232
2173
1939
1776
1472
1376
1004
805
577
517
484
72
65
14.21
14.48
31.35
19.35
19.09
39.55
28.81
21.08
26.99
45.11
57.54
34.29
71.19
24.27
44.03
112.32
104.17
53.82
64.8
82.56
55.88
43.76
38.99
55.51
74.39
3.09
-0.71
10.84
2.82
3.50
10.44
-3.91
2.71
7.50
8.83
15.35
10.50
-22.97
1.49
1.11
-42.22
30.42
11.26
22.87
22.79
-6.36
0.95
3.23
-7.95
-3.59
36.41
18.50
9.66
6.73
4.42
4.19
3.92
3.15
1.23
1.15
0.98
0.90
0.88
0.78
0.72
0.60
0.56
0.41
0.33
0.23
0.21
0.20
0.03
0.03
demersal groups were due to the multiday activity (24 days/trip) of trawlers becoming a standard practice
in the state, while fishing was extended further to 70
m depth. The landings of mackerel recovered in this
decade with a growth rate of 61.54 per cent due to the
introduction of purse seines in the mid-1980s.
Increased fishing effort, introduction of purse
seines and multi-day trawling were the major factors
that increased marine landings during the decade. The
decadal mean landing of bombayduck, elasmobranch
and whitefish increased, but the overall trend in growth
was negative (-8.9%, -1.94% and -5.66% CGR,
respectively), as landings started decreasing in
bombayduck since 1985, while whitefish and
elasmobranch showed a drastic decrease in their
landings after 1988. A significant difference in the inter-
year growth rate was observed in ribbonfish, molluscs,

perches, seerfish and mackerel, while bombayduck
showed a significant negative trend. Similar to the
previous decade, crustaceans and bombayduck
continued to dominate and the other major landed
resource groups were: clupeids (9.31%), croakers
(6.91%), pomfrets (5.16%) and ribbonfish (5.11%).
Intensification of Fishing Longer, Deeper and
Faster: 1991-2000
During the decade 1991-2000, the mean annual
catch increased to 3.49 lakh tonnes. However,
interestingly, of the 24 groups, the mean landings of
bombayduck, pomfrets, catfish, elasmobranchs,
threadfins and goatfish decreased drastically in this
decade compared to in the previous decade. Interest
110
Vol. 29 (No.1)
January-June 2016
Fish group
Overall
Crustaceans
Bombayduck
Clupeids
Croakers
Pomfrets
Ribbonfish
Catfish
Elasmobranchs
Molluscs
Perches
Carangids
Seerfish
Mackerel
Flatfish
Eel
Lizardfish
Tunas
Whitefish
Goatfish
Threadfins
Unicorn cod
Silverbellies
Mullets
HB&FB
Mean catch
(tonnes)
CV
CGR
% of total state
landings
306175
92049
46661
28509
21161
15794
15646
12548
10999
9890
9105
8866
6353
4994
4028
2610
2505
2135
1698
1483
1370
647
297
115
105
11.04
16.91
37.31
25.09
13.20
26.63
31.60
24.65
11.54
44.18
52.68
69.77
31.72
153.37
29.11
34.50
40.11
35.61
50.76
41.18
75.52
112.85
116.66
83.97
74.07
3.16
3.80
-8.90
3.07
3.68
-2.75
10.41
1.41
-1.94
20.98
18.49
29.03
10.97
61.54
6.55
-8.96
10.66
-1.04
-5.66
5.07
15.60
31.97
-23.25
1.68
-13.71
30.06
15.24
9.31
6.91
5.16
5.11
4.10
3.59
3.23
2.97
2.90
2.07
1.63
1.32
0.85
0.82
0.70
0.55
0.48
0.45
0.21
0.10
0.04
0.03
for Tuna fishing began during the period 1990-2003,

mainly because of adoption of progressive and
innovative fishing techniques, multiday gill net and
hook and line fishing, conversion of idling Shrimp trawl
for longlining and multi-gear operation (Pillai et al.,
2007). After 1995, some landings of pelagic fishes
increased drastically probably because in the late1990s, large-sized gill nets and trammel nets began to
be employed in pelagic and mid-water fishes up to 5060 m depth. During this decade, trawling was further
extended to the depth of 90-100 m in offshore waters
and fishing effort continued to increase as the days at
sea of multi-day trawlers increased from 3-5 days/trip
to 12-15 days/trip and night trawling was also
introduced consequent to the construction of large size
trawlers (15-20 m) with more horsepower (>300 HP)
and holding capacity. This led to the increased landings

of major demersal groups like crustaceans, croakers,
molluscs, and perches. Though, the state decadal mean
catch increased during the 1990s, the majority of
resources showed a negative trend in growth, as
indicated by the CGRs (Table 6). For the first time, the
bombayduck started losing its dominance (5.75%),
while the crustaceans continued to hold the prime
position (31.88%). The ribbonfish (9.18%) emerged
as the second most important group, trailed closely by
clupeids (8.25%), mackerel (7.96%), croakers (7.49%)
and molluscs (6.48%) groups.
Beginning of Decline: 2001-2010
The decadal mean catch during 2001-2010
decreased to 3.44 lakh tonnes from 3.49 lakh tonnes in
111
Table 6. Mean fish, coefficient of variation (CV), compound growth rate (CGR) and per cent contribution of various
major fish groups in Maharashtra state during 1991-2000
Fish group
Overall
Crustaceans
Ribbonfish
Clupeids
Mackerel
Croakers
Molluscs
Bombayduck
Perches
Carangids
Pomfrets
Catfish
Elasmobranchs
Seerfish
Flatfish
Lizardfish
Tunas
Whitefish
Eel
Threadfins
Unicorn cod
Goatfish
Silverbellies
Mullets
HB&FB
Mean catch
(tonnes)
CV
CGR
% of total state
landings
349536
111429
32078
28836
27819
26173
22651
20091
13436
11204
9513
9299
7885
6652
5986
2739
2321
1466
1316
771
592
379
106
85
81
10.29
18.53
38.33
16.24
33.44
10.36
22.37
51.79
16.62
33.95
31.93
35.81
16.61
31.82
19.57
63.79
31.32
46.32
19.10
39.55
71.80
48.07
65.88
106.38
54.56
0.41
0.56
5.76
0.68
12.63
1.11
-4.29
-3.89
-3.08
-5.48
-2.01
-4.09
-1.63
1.21
-4.49
-12.41
4.44
-2.49
0.34
-2.55
-14.02
-8.99
1.35
14.79
9.97
31.88
9.18
8.25
7.96
7.49
6.48
5.75
3.84
3.21
2.72
2.66
2.26
1.90
1.71
0.78
0.66
0.42
0.38
0.22
0.17
0.11
0.03
0.02
0.02
1991-2000. Consequently, the growth trend was

negative with CGR of -4.69 per cent with significant
inter-year variations (Table 7). The decade recorded a
historical high annual landings of 4.49 lakh tonnes in
2002, after which landings from the major mechanized
gears started showing a declining trend. The landings
of commercially important groups like crustaceans,
ribbonfish, molluscs, mackerel and pomfrets, decreased
during this period affecting the total mean landings in
the state negatively.
The major landings of pelagic resources showed
an increasing trend in the case of clupeids,
bombayduck, perches, carangids, seerfish and tunas,
in which perches landings increased drastically
compared to previous decade landing. However, as
trawling became uneconomical, the number of purse
seiners increased dramatically to nearly 517 in 201011 largely by conversion of the trawlers by fitting drum
winch that increased the fishing efficiency and
subsequently led to increased landings of major pelagic
resources. The increasing export demand coupled with
good prices for crustaceans, molluscs, pomfrets on one
hand and the unrestricted expansion and relentless
intensification of fishing effort that began in the mid1990s, have led to a gradual but steady decline in many
resources, especially in the demersal fishes. That more
and more juveniles were being caught off the Mumbai
coast attested the exploitative nature of fishing as well
as the desperation to fish, irrespective of age and size
groups. A good number of mini purse seiners (ring
seiners) were introduced in Raigad and Ratnagiri
districts after 2006. Of the 24 groups, 18 groups showed
112
Vol. 29 (No.1)
January-June 2016
Fish group
Mean catch
(tonnes)
CV
CGR
% of total state
landings
344888
103965
29894
28756
27783
26774
23855
22444
13558
11117
10676
8559
8162
6713
5335
3153
2066
2035
1507
936
724
560
467
241
80
17.29
20.29
26.76
65.00
9.94
21.47
20.28
26.30
17.26
67.74
23.96
31.17
36.26
24.82
39.20
32.70
39.28
28.79
36.59
62.99
127.69
31.21
52.98
35.84
61.83
-4.69
-4.25
-4.13
-16.35
-1.56
-5.91
-3.93
-4.97
-0.18
1.57
-3.35
-8.68
-8.25
-7.71
5.26
-7.32
-9.52
-2.76
-8.09
3.24
0.13
-6.38
-11.99
6.09
6.55
30.14
8.67
8.34
8.06
7.76
6.92
6.51
3.93
3.22
3.10
2.48
2.37
1.95
1.55
0.91
0.60
0.59
0.44
0.27
0.21
0.16
0.14
0.07
0.02
Overall
Crustaceans
Clupeids
Ribbonfish
Croakers
Perches
Bombayduck
Molluscs
Carangids
Mackerel
Catfish
Elasmobranchs
Seerfish
Pomfrets
Tunas
Flatfish
Eel
Lizardfish
Threadfins
Whitefish
Silverbellies
Unicorn cod
Goatfish
HB&FB
Mullets
a negative trend in the growth rate, as indicated by the

CGR. The inter-year growth rate in perches, ribbon,
elasmobranchs and pomfrets showed a significant
difference with a negative sign.
Present Status of Stocks
In Maharashtra, the marine fish landings recorded
a steady increase since 1961, but during the decade
(2001-2010), it has shown a decreasing trend that
fluctuated widely. The growth rates decreased over the
decades as indicated by the CGRs and it became
negative during the recent decade. The resources that
have shown improvement in the percentage
contribution towards total landings during the previous
decade compared to during 1961-1970 period are
clupeids, seerfish, tunas, carangids, whitefish,
ribbonfish, perches, croakers, lizard fish, flatfish and

molluscs. Among the clupeids, golden anchovy is the
major species contributing to this fishery. Prior to 1985,
dolnet was the major gear for this fishery, but
afterwards trawlers started encroaching dolnet zone
(Khan, 2003). The decadal mean catch of seerfish in
Maharashtra increased, especially from 1980s, while
Indian seer fish landings also showed an increasing
trend during the 1980s due to the introduction and
subsequent intensification of mechanization of the craft
(Devaraj et al., 1999). Though, mean decadal catch of
carangids showed an increasing trend from 1980s
because of intensification of fishing effort, the trend in
growth rate as indicated by the CGR, showed a negative
sign during the past two decades. This is mainly
because, the early-1990s showed higher landings
113
the resources that have shown increases in the average

landings compared to the decade of 1991-2000 (other
than resources showing increase in average landings
over the five decades) were: bombayduck, silverbellies,
half beak & full beak (HB & FB), threadfins, goatfish,
catfish, elasmobranchs and eels.
Figure 1. Decadal changes in CGR with respect to

number of resource groups, 1961-2010
compared to late-1990s, similar to in the decade of

2001-2010.
When the average annual landings of different
resources in the 2001-2010 decade were compared with
those of 1961-1970, it was found that the average
landings had come down in the case of unicorncod,
mullets, eels, pomfrets, goatfish and bombayduck. The
resources that have shown increases in average decadal
landings throughout the period were clupeids, seerfish,
tunas, carangids, perches and croakers. In the decade
2001-2010, the resources that have shown a decline in
average landings are: mackerel, whitefish, ribbonfish,
lizardfish, pomfrets, flatfish, mullets, unicorncod,
crustaceans and molluscs. In the decade 2001-2010,
The majority of resources has shown a negative

trend in growth rates during the decades 1991-2000
and 2001-2010 with growth becoming negative during
the later decade. While 20 groups of fishes showed a
positive growth trend during 1961-1970 (as indicated
by CGR), the number reduced to 6 groups by 20012010 (Figure 1), thereby reversing the growth trends
in the five decades. Meanwhile, landings by the
mechanized gears during 1985-2010 (Figure 2) indicate
that landings started declining from 2002 onwards.
The 25 resource groups were classified following
the method suggested by Mohamed et al. (2010). It
was found that only two resource groups fell under the
abundant class, seven groups were under lessabundant class and 14 groups were under declining
class and 1 group each was under depleted and
collapsed classes (Table 8). This is in clear contrast
to the country level scenario wherein of the 26 resource
groups studied, 18 fell under the abundant class, 5
groups fell under less-abundant class and 1 group each
was under declining, depleted and collapsed
classes (Sathianandan et al., 2011). A comparison
between status of Indian stocks and Maharashtrian
stocks is presented in Table 8. For example, the status
of Goatfishes along the Indian coast was classified
Figure 2. Landings by different types of mechanized gears in Maharashtra during 1985-2010
114
under abundant category, but in the present study at

the state level, goatfish has come under the collapsed
class. Similarly, many groups classified as abundant
and less-abundant categories at the country level, fell
under the declining class in the case of Maharashtra.
All the important resource groups of Maharashtra
fisheries have been found under less-abundant and
declining classes.
The difference in the status of fisheries in India
and Maharashtra mainly depends on the variations in
the past three-year (2008-2010) average catch and
historical maximum catch (1961-2010) of a particular
species. For example, in Maharashtra, of clupeids, the
recent three-year average catch was 24042 tonnes and
historical maximum catch was 48303 tonnes and the
percentage of this (past three-year landing/historical
Vol. 29 (No.1)
January-June 2016
maximum*100) is 49.77 per cent which falls under the

declining category. But, at the all-India level
(Sathianandan et al., 2011), the recent three-year
average catch (2008-2010) of clupeids was 877576
tonnes and the historical maximum catch was 929404
tonnes and the percentage of this (past three-year
landing/historical maximum*100) is 94.42 per cent,
which falls under the abundant category. This is
because at the all-India level, clupeids recent average
catch is higher or nearer to the historical maximum
catch and gives a higher percentage value. The majority
of groups falling under the declining class need
caution and care to prevent further reduction. The group
unicorn cod and goatfish falling under the depleted
and collapsed classes, require immediate management
interventions for recovery of these stocks.
Table 8. Classification of different resource groups based on three-year (2008-2010) average landings
Resource
Croakers
Tunnies
Seer fish
Carangids
HB & FB
Perches
Catfish
Crustaceans
Molluscs
Clupeids
Bombayduck
Mackerel
Silverbellies
Whitefish
Ribbonfish
Threadfins
Barracudas
Lizard fish
Pomfrets
Flatfish
Elasmobranch
Eels
Mullets
Unicorn cod
Goatfish
Average
landings (t)
during
2008-10
Maximum
annual landing (t)
during
1961-2010
Year
Percentage
of total
landings
26647
7466
6818
14297
287
21381
10801
98192
19989
24042
19274
15357
732
1352
14345
859
807
1797
4963
2498
5594
1111
109
488
247
32315
10265
13256
22452
457
36351
21086
149978
31684
48303
82136
38355
3195
3146
66281
4125
1699
6670
22523
7797
14384
10091
481
5498
4180
1998
2008
2002
1989
2009
2003
1988
1998
2003
1989
1981
1996
2004
1983
2002
1976
1967
1995
1983
1995
2002
1961
1965
1965
1964
82.46
72.73
51.43
63.68
62.87
58.82
51.22
65.47
63.09
49.77
23.47
40.04
22.92
42.99
21.64
20.82
47.52
26.94
22.03
32.04
38.89
11.01
22.66
8.88
5.91
Stock status
Stock status of
of Maharashtra
India as per
Sathianandan
et al. (2011)
Abundant
Abundant
Less abundant
Less abundant
Less abundant
Less abundant
Less abundant
Less abundant
Less abundant
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Declining
Depleted
Collapsed
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Abundant
Declining
Less abundant
Less abundant
Abundant
Abundant
Abundant
Less abundant
Less abundant
Abundant
Less abundant
Collapsed
Abundant
115
Table 9. Stock contribution to mean catch of Maharashtra

Status
Abundant
Less abundant
Declining
Depleted
Collapsed
Percentage contribution to average

catch during 2001-2010
Percentage contribution to average

catch during 2008-2010
9.60
53.88
34.89
0.16
0.14
11.22
56.51
30.54
0.16
0.08
The contributions of these classified stocks to the

state average landings were also studied to understand
their relative importance. The percentage contribution
of these classified stocks to average total marine
landings of Maharashtra from 2008-2010 and 20012010 is given in Table 9. It is a worrisome sign that the
less abundant (7 groups) and declining (14 groups)
categories were contributing 56.51 per cent and 30.54
per cent (2008-2010), respectively towards state total
landing which is the sign of crisis in waiting.
Conclusions
Despite the average decadal mean catch of 344888
tonnes that contributed 12.59 per cent to the total
marine fish of the country during the previous decade
(2001-2010), the marine fishery of Maharashtra has
been facing crisis since late-1990s. However, the
growth rates of important resources (18 groups) have
declined and shown a negative trend, as indicated by
the compound growth rate during the previous decade.
Increased fishing effort from shrimp trawling by multiday fishing trips, introduction of purse seines in 1980s
and increasing pollution load are responsible for
declining growth rates. The historical highest annual
landings were recorded in the year 2002, but thereafter
the state total landings started showing a decreasing
trend and reached 3.07 lakh tonnes in 2011 (CMFRI,
2012). Among the 24 groups of resources, mackerel,
whitefish, ribbonfish, lizardfish, pomfrets, flatfish,
mullets, unicorn cod, crustaceans and molluscs have
shown a decline in their landings in recent decades.
The species which have shown increased mean
catch over the decade are clupeids, croakers, perches,
carangids and seerfish, even though their growth rate
is negative in the recent decade. The negative growth
is mainly because of having a decreasing trend in
landings of these species after 2005. However, their
decadal mean catch is high compared to in the previous

decades. Also, crustaceans and molluscs play an
important role in maintaining the landings at a higher
level over the decade even though their mean catch
during the recent decade is slightly less compared to
in the previous decade 1991-2000. Landing by different
types of mechanized gears has shown a decreasing
trend from 2002 onwards. A major cause of concern is
that most of the stocks which were abundant or lessabundant at the national level, are actually declining
in Maharashtra. It indicates that the state-wise scenario
is different from the national scenario and state-wise
understanding of marine fishery resources is very
important to formulate appropriate regulatory and
management measures.
Acknowledgments
The authors thank Dr W.S. Lakra, Director, Central
Institute of Fisheries Education, Mumbai, for his
encouragement and support. They also extend their
gratitude to the Central Marine Fisheries Research
Institute for providing data for the present study. The
authors thank the anonymous referee for his critical
comment on the paper.
References
CMFRI (Central Marine Fisheries Research Institute) (2012)
Annual Report 2011-12. Cochin. 186 p.
CMFRI (Central Marine Fisheries Research Institute) (2014)
Annual Report 2013-14. Cochin. 274 p.
Dehadarai, P.V. and Yadava, Y.S. (2004) Fisheries
development. Vol.13. In: State of the Indian FarmerA
Millennium Study. Department of Agriculture and Cooperation, Ministry of Agriculture, Government of
India, New Delhi. 173 p.
Devaraj, M., Mohamad Kasim, H., Muthiah, C. and Pillai,
N.G.K. (1999) Assessment of the exploited Seerfish
116

stocks in the Indian waters. Journal of the Marine
Biological Association of India, 41(1 & 2): 62-84.
James, P.S.B.R. (2010) Taxonomic status of marine pelagic

fishes of India, research priorities and conservation
strategies for the sustainability of their fisheries. Indian
Journal of Animal Sciences, 80(4): 39-45.
Khan, M.Z. (2003) Golden Anchovy. In: Status of Exploited
Marine Fishery Resources of India, Eds: J. Mohan and
A. A. Jayaprakash. Central Marine Fisheries Research
Institute, Cochin, pp 40-44.
Mohamed, K.S., Sathianandan, T.V., Zacharia, P.U., Asokan,
P.K., Krishnakumar, P.K., Abdurahiman, K.P., Veena
Shettigar, S. and Raveendra Durgekar, N. (2010)
Depleted and collapsed marine fish stocks along southwest coast of India A simple criterion to assess the
status. In: Coastal Fishery Resources of India:
Conservation and Sustainable Utilization, Eds: B.
Meenakumari, M.R. Boopendranath, Leela Edwin, T.V.
Sankar, Nikita Gopal and G. Ninan. Society of Fisheries
Technologists, Cochin. pp. 67-76.
Najmudeen, T.M. and Sathiadhas, R. (2008) Economic
impact of juvenile fishing in a tropical multi-gear multispecies fishery. Fisheries Research, 92(2): 322-332.
Pillai, N.G.K. (2006) Pelagic fisheries of India. In: Handbook
of Fisheries and Aquaculture. Directorate of
Information & Publications of Agriculture, Indian
Council of Agricultural Research, New Delhi. pp. 5677.
Pillai, N.G.K., Jayaprakash, A.A. and Ganga, U. (2007)
Status and scope of research on pelagic fisheries of
Vol. 29 (No.1)
January-June 2016
India. In: Status and Perspectives in Marine Fisheries

Research in India, Eds: M.M. Joseph and N.G.K. Pillai.
Central Marine Fisheries Research Institute, Cochin.
pp. 52-114.
Sathianandan, T.V., Jayasankar, J., Kuriakose Somy, Mini,
K.G. and Mathew Wilson, T. (2011) Indian marine
fishery resources: Optimistic present, challenging
future. Indian Journal of Fisheries, 58(4): 1-15.
Srinath, M. (2003) An appraisal of the exploited marine
fishery resources of India. In: Status of Exploited Marine
Fishery Resources of India, Eds: M.M. Joseph and A.A.
Jayaprakash. Central Marine Fisheries Research
Institute, Cochin. pp. 1-18.
Srinath, M. and Balan, K. (2003) Potential yield from Indian
EEZ. In: Status of Exploited Marine Fishery Resources
of India, Eds: M.M. Joseph and A.A. Jayaprakash.
Central Marine Fisheries Research Institute, Cochin.
pp. 286-290.
Vivekanandan, E., Mohamed, K.S., Kuriakose, S.,
Sathianandan, T.V., Ganga, U., Pillai, S.L. and Nair,
R.J. (2009) Status of marine fish stock assessment in
India and development of a sustainability index. In: The
Second Workshop on the Assessment of Fishery Stock
Status in the South and Southeast Asia, held at Bangkok,
Thailand, 5-9 October. pp. 1-15.
Vivekanandan, E. and Jayasankar, J. (2008) Winter school
on impact of climate change on Indian marine fisheries.
In: Winter School Lecture Notes - Part 1, held at Central
Marine Fisheries Research Institute, Cochin.
18 January-7 February.
Received: July, 2015; Accepted: November, 2015

DOI: 10.5958/0974-0279.2016.00024.0
Is Sustainable Sugarcane Initiative (SSI) Technology More

Profitable than Conventional Method for Sugarcane
Production? An Economic Analysis
Arthi K., V. Saravanakumar* and R. Balasubramanian
Department of Agricultural Economics, Tamil Nadu Agricultural University, Coimbatore-641 003, Tamil Nadu
Abstract
The study has examined profitability, sources of productivity improvement and determinants of a new
technology Sustainable Sugarcane Initiative (SSI) adoption in sugarcane cultivation in Tamil Nadu
by collecting primary data from 120 sugarcane farms during 2014-15. Although the cost of cultivation
has been found higher in SSI method vis-a-vis conventional method, the cost of production is lower due
to 26 per cent more cane yield. The cost and return analysis has indicated that sugarcane cultivation is
more profitable under SSI method than under the conventional method. The decomposition analysis has
shown that the inputs, viz. fertilizers, micro-nutrients and deployment of labour are the major sources of
productivity enhancement in the SSI method. The estimates of logit model have indicated that farmers
educational level and experience are the major determinants for adoption of SSI method in sugarcane
cultivation. The major policy options suggested to improve production and profitability of sugarcane
include provision of drip irrigation with subsidy, ensuring timely availability of critical inputs and imparting
periodical trainings to farmers on SSI method such as fertigation, wide row spacing, etc.
Key words: Sugarcane cultivation, technology impact, sustainable sugarcane initiative, economic
analysis, productivity enhancing sources, Tamil Nadu
JEL Classification: Q1, Q01, Q16, E23, O32
Introduction
In India, the sugar industry is the second largest
industry next to the textile industry in playing a vital
role in the socio-economic transformation of country
(Wagh, 2015). The sugar industry being an important
agro-based industry, provides livelihoods to about 6
million sugarcane farmers and around 7 lakh workers
who are employed in the sugar mills. India ranked
second in sugarcane production in the world, after
Brazil, with an area of 5.31 million hectares and
production of 366.8 million tonnes with productivity
of 69.1 tonnes/ha during 2014-15 (ISMA, 2015). The
Email: E-mail: sharanu2k@gmail.com
This paper is part of the masters study of first author supervised by second author.
major sugarcane-growing states in India are Uttar

Pradesh, Maharashtra, Tamil Nadu, Karnataka, Andhra
Pradesh, Gujarat, Bihar, Haryana, Punjab and Madhya
Pradesh and among these states, Tamil Nadu occupied
fifth position in area (after Uttar Pradesh, Maharashtra,
Karnataka and Bihar), fourth rank in production and
third in productivity in 2014-15. The area and
production of sugarcane in Tamil Nadu during 201415 was 2.55 lakh ha and 22.3 Mt, respectively and the
productivity was 104 t/ha in the year 2013-14
(Government of Tamil Nadu, 2015). The sugar industry
provides direct employment to a large number of
persons, apart from providing indirect employment to
thousands of persons in the rural areas who are involved
in cultivation, harvesting and transportation of cane
and other related services. At present, Tamil Nadu has
118
43 sugar mills, of which 16 are owned by the cooperative sector and 27 by the private sector.
The demand for sugar and by-products of sugar
industry is raising, whereas production is declining due
to reasons such as decrease in cultivable lands, climate
change effect, escalating cost of inputs such as
fertilizers and human labour and fluctuating and
insufficient cane price. The sugar industry also faces
problems such as inadequate cane supply for crushing
due to reduction in area under sugarcane, labour
scarcity for harvesting, competition from other highly
remunerative crops like rice and maize, inadequate
availability of planting material at the time of onset of
season leading to inadequate coverage of targeted area
(NABARD, 2012).
Besides, irrigation water is emerging as one of the
major constraints affecting productivity and
profitability of both farmers and millers. The excessive
ground water exploitation and wide variability in
rainfall due to climate change are debilitating the
sugarcane farmers in the country (Goud, 2011). Further,
less use of farm mechanization due to closer spacing
in the conventional method of sugarcane production
increases drudgery of human labour and its production
cost. In Tamil Nadu, the sugarcane area increased from
3.15 lakh ha in 2001-02 to 3.48 lakh ha in 2012-13,
but it was minimal during the years 2003-04, 200405, 2005-06 and 2009-10 due to poor monsoon and its
aftermath. This is the major reason for deceleration of
growth rate of sugarcane area during the recent years.
These crises call for development of alternate options
and technologies for sugarcane cultivation to make it
viable and remunerative to both farmers as well as sugar
mills.
The sustainable Sugarcane Initiative (SSI) is a new
method which improves the productivity of sugarcane
by using less resource and thereby, reduces the input
cost also. The SSI is a better method of sugarcane
cultivation than the conventional methods which are
seed - water - space-intensive. By adopting SSI method,
the productivity of cane can be enhanced by practising
drip irrigation with fertigation, maintaining optimum
plant spacing of 5ft 2ft for easy penetration of
sunlight, and profuse tillering and after the
establishment of 2 or 3 tillers, the mother shoot may
be removed just one inch above the ground to facilitate
more number of tillers. The benefits of SSI method
vary depending on how efficiently farmers use these
Vol. 29 (No.1)
January-June 2016
practices. The farmers work in a synergistic way to

save inputs and achieve higher yields per unit area.
This method has been recommended by the Tamil Nadu
Agricultural University (TNAU) for wide adoption by
farmers. However, its adoption rate by the farmers is
comparatively low, ranging from 15 to 20 per cent
(Kiruthika, 2014). Due to efficient utilization of water
by fertigation method, growing of intercrops and
getting higher outputs, it was believed that SSI method
would be profitable (Loganandhan et al., 2013; Shanthy
and Ramanjaneyulu, 2014). However, there are no
holistic studies to assess the profitability of SSI, sources
of productivity changes in SSI and determinants of SSI
adoption. Therefore, this paper has studied profitability
and adoption level of SSI with the following objectives:
(i) to work out the economics of sugarcane production
under conventional and SSI methods, (ii) to examine
the sources of productivity improvement in SSI, and
(iii) to analyse the factors determining the adoption of
SSI method.

The present study was conducted in Tamil Nadu
because of significant production and consumption of
sugarcane in the state. The multistage random sampling
technique was followed to select 120 sample
households from the districts of Villupuram and Trichy.
The Villupuram district falls under the high
productivity (110 t/ha) area, whereas Trichy falls under
low productivity (71 t/ha) area. The primary data were
collected during the year 2014-15 through a wellstructured interview schedule.
Economic Analysis
(a) Estimation of Cost and Returns
To estimate cost and returns of sugarcane under
conventional and SSI methods, the standard method
developed by the Commission on Agricultural Costs
and Prices (CACP) was followed (Raju and Rao, 1990;
Narayanamoorthy, 2013).
(b) Sources of Change in Sugarcane Productivity
To measure the sources of change in sugarcane
productivity, Cobb-Douglas production function was
used as given by Equation (1):
i2
i3
YLDi = 0 NITROi1
i1 PHOS i2 POTAS i3
i4
i5
i
MNM i4 HLAB i5 MLAB i6 ui
(1)
Arthi et al. : Is SSI Technology More Profitable than Conventional Method for Sugarcane Production?
where,
Subscript i = 1 indicates conventional method; i = 2
indicates SSI method
119
decision-maker makes rational choices in maximizing

his/her utility (Rahm and Hufman, 1984). In the present
study, the producers preference to adopt SSI and
conventional methods are represented by 1 and 0,
respectively. The logistic regression (LR) model
specified in the study (Tefera et al., 2010; Loganandhan
et al., 2013; Kiruthika, 2014) is given in Equation (2):
YLD
= Sugarcane yield (t/ha)
NITROi1
= Quantity of nitrogenous fertilizers used

(N) (kg/ha)
PHOSi2
= Quantity of phosphatic fertilizers used

(P) (kg/ha)
POTASi3
= Quantity of potassic fertilizers used (K)

(kg/ha)
MNMi4
= Micronutrients applied (kg/ha)
where,
HLABi5
= Human labour (person days/ha)
MLABi6
= Machine labour (hours/ha)
Pi = 1 indicates willingness and 0, non-willingness to

adoption of SSI method
log (Pi) = ln(p i / 1-p i ) = 0 + 1 AGE 1i +

2SQAGE1i + 3EDU3i + 4HHSIZE4i
+ 5FARMSIZE5i + 6INCOME6i
(2)
0 = Intercept-term (scale parameter)
i = Coefficients to be estimated
ui = Error-term independently distributed with zero

mean and constant variance.
AGE
i1, i2, i3, i4, i5 and i6 are the regression coefficients

of nitrogenous, phosphatic and potassic fertilizers,
micronutrients, human labour and machine labour,
respectively. The family labour was imputed and
evaluated at the prevailing wage rates of hired labour
at the village level.
EDU
Chows test (Gujarati and Porter, 2014) was

employed to identify whether the parameters governing
the production relations were different in the
conventional method and SSI method and it was
used to compute the F ratio. The computed F value
was compared with F critical value p and (n+m-2p)
degrees of freedom at appropriate level of significance;
where, n refers to the number of observations and
m refers to the number of variables. The nonsignificant F value indicated no structural difference
between conventional and SSI methods.
The effect of each independent variable on

technology adoption was determined by the
coefficients and the sign of any coefficient represented
the positive and negative effect of the variable on
adoption of SSI technology. To quantify the effect of
each variable, odds-ratio is determined. It quantifies
the probability of adoption of a new technology such
as SSI and is represented algebraically by Equation
(3):
In this study, the output decomposition model

developed by Bisaliah (1977) and Balasaheb (2013)
was used to examine the productivity difference
between conventional and SSI methods due to
technological change and input use.
c. Factors Influencing the Adoption of SSI
Technology
The logistic regression models are used to predict
the probability of occurrence of binary events to
predictor variables by assuming that an individual
= Age of household-head (in years)
SQAGE = Age squared

= Education level of household-head (in
index)
HHSIZE = Household size (No.)

FARMSIZE = Total land area owned (acres)
INCOME
= Annual income (in Rupees)
Probability of adoption of technology, P = (odds ratio

of Xi 1)*100
(3)
where, Xi is the independent variable and i = 1,2,...,6.

Economics of Sugarcane Cultivation
The economics of sugarcane cultivation under
conventional and SSI methods are presented in
Table 1.
The overall total cost of sugarcane cultivation was
worked out to be ` 179008/ha under conventional
120
Vol. 29 (No.1)
January-June 2016
Table 1. Economics of sugarcane cultivation under conventional and SSI methods

(`/ha)
Particulars
Depreciation
Interest on fixed assets
Rental value of owned land
Total fixed cost
Setts cost
Fertilizer cost
Plant protection chemicals cost
Human labour cost
Machine labour cost
Total variable cost
Overall total cost
Villupuram district
Conventional
SSI
Trichy district
Conventional
SSI
Overall
Conventional
SSI
5605
(3.12)
4228
(2.35)
20540
(11.43)
30373
(16.90)
Fixed cost
13040
(6.25)
9837
(4.71)
21034
(10.08)
43911
(21.04)
5666
(3.18)
4274
(2.40)
20742
(11.63)
30682
(17.21)
11931
(5.80)
9001
(4.38)
21271
(10.34)
42203
(20.52)
5635
(3.15)
4251
(2.37)
20641
(11.53)
30528
(17.05)
12486
(6.03)
9419
(4.55)
21153
(10.21)
43057
(20.78)
19385
(10.79)
18793
(10.46)
2793
(1.55)
93225
(51.87)
15160
(8.43)
149356
(83.10)
179729
Variable cost
18935
(9.07)
21568
(10.34)
2540
(1.22)
106726
(51.14)
15008
(7.19)
164777
(78.96)
208688
20086
(11.27)
17894
(10.04)
3037
(1.70)
91228
(51.17)
15359
(8.61)
147604
(82.79)
178286
23053
(11.21)
20044
(9.75)
2259
(1.10)
102954
(50.06)
15136
(7.36)
163446
(79.48)
205649
19736
(11.02)
18344
(10.25)
2915
(1.63)
92227
(51.52)
15260
(8.52)
148480
(82.95)
179008
20994
(10.13)
20806
(10.04)
2400
(1.16)
104840
(50.61)
15072
(7.28)
164112
(79.22)
207169
Note: Figures within the parentheses indicate percentage to the total
method and ` 207169/ha under SSI method. In this

cost, the share of total fixed cost ranged from 17 to 21
per cent and the remaining 79 to 83 per cent was
accounted by the total variable cost. The total fixed
cost component (depreciation and interest) incurred in
SSI method was more due to additional expenditure
on establishment of drip irrigation infrastructure and
its maintenance.
Among the variable cost components, the cost on
human labour was highest (50 %). The major share of
labour cost was accounted for by the harvesting
operation, which ranged from 34 to 37 per cent, under
both conventional and SSI methods. The share of wages
for harvesting was higher in the SSI method due to
harvesting higher cane yield as compared to in the
conventional method. The expenditure on fertilizers
was the second major variable cost, which accounted

for approximately 10 per cent of total costs in both the
methods. This cost included expenditure on farmyard
manure, fertilizers such as nitrogenous (urea),
phosphatic (di-ammonium phosphate, single super
phosphate) and potassic (muriate of potash); and micronutrients including humic acid, and bio-fertilizers. It
was also observed that the cost of humic acid was
higher in the SSI method (` 1098) than conventional
method (` 862). Use of humic acid is the prevalent
practice in sugarcane cultivation and it increases the
fertilizer-use efficiency (Bohme and Thi Lua, 1997).
The liquid bio-fertilizers like AzoGro, PhoSol and
PotaVit were used by the farmers that are also supplied
by the sugar factories. The cost on bio-fertilizer was
higher in SSI method (` 970) than in conventional
121
Table 2. Estimation of different cost and returns measures

Particulars
Cost A1 (`/ha)
Cost B1 (`/ha)
Cost B2 (`/ha)
Cost C1(`/ha)
Cost C2 (Total cost)
Yield (t/ha)
Price (`/tonne)
Cost of production (`/tonne)
Gross returns (`/ha)*
Net returns (`/ha)
Farm business income (`/ha)
Farm investment income (`/ha)
Family labour income (`/ha)
Benefit-cost ratio
Villupuram district
Conventionl
SSI
147736
151964
172504
159189
179729
104
2300
1421
239200
91464
91464
84239
66696
1.62
171592
181429
202463
187654
208688
132
2300
1300
303600
132008
132008
125783
101137
1.77
Trichy district
Conventionl
SSI
146256
150530
171272
157544
178286
101
2275
1448
229775
83519
83519
76505
58503
1.57
168817
177818
199089
184378
205649
125
2275
1351
284375
115558
115558
108998
85286
1.68
Overall
Conventionl
146996
151247
171888
158367
179008
102.5
2288
1434
234469
87473
87473
80353
62581
1.60
SSI
170205
179624
200776
186016
207169
128.5
2288
1325
293944
123739
123739
117347
93168
1.73
*In the study area, 70 per cent farmers cultivated pulses (black gram, green gram and cow pea) as intercrops and SSI
farmers would additionally generate the gross income of ` 16200 /ha/annum.
method (` 859). The cost of plant protection chemicals

(PPC) was higher in the conventional (` 2915) than
under SSI (` 2400) method. The expenditure incurred
on planting materials was less (` 19736) (for setts) in
conventional than SSI (` 20994) (for seedlings)
method. The planting materials were selected from the
sugarcane varieties such as Co 86032, CoV 94102 and
SI 309. However, the majority of farmers (95 %) used
planting materials from Co 86032 variety under both
the methods. The farm machineries were widely used
for ploughing and ridge formation activities to reduce
human drudgery. The share of machine labour in the
total cost was 8.52 per cent and 7.28 per cent under
conventional and SSI methods, respectively.
The SSI is referred as More with less technology,
but the present study has found inputs cost higher under
SSI than under conventional method. The reasons for
this higher cost of inputs are: (i) although the quantity
of planting materials used is less in the SSI method,
the cost of producing single-budded chips is more. That
is why the cost of planting material in SSI method is at
par with or marginally higher than the conventional
method, (ii) under SSI technology, there is water saving
of 40-70 per cent and therefore consumption of
electricity is also reduced. But electricity is available
free for agriculture in the state, therefore, it is very
hard to measure the economics of water saving, (iii) it

is observed that farmers applied same qunatity of
fertilizers in SSI as well as conventional system, (iv)
human labour cost (i.e. wages) was higher in the SSI
method due to higher yield. Mohanty et al. (2015) have
also shown that the cost of inputs in SSI method is at
par with the conventional method.
The different cost components for sugarcane
cultivation (Cost A1, Cost B1, Cost B2, Cost C1 and
Cost C2) were worked out and are presented in Table
2.
A perusal of Table 2 revealed that all cost
components, viz. Cost A1, Cost B1, Cost B2, Cost C1
and Cost C2 were higher in SSI than in conventional
method. The average yield obtained from SSI method
was 128.5 t/ha, i.e. 26 per cent more than the yield
from conventional method (102.5 t/ha). Therefore, the
overall cost of production of was lower in SSI (` 1325/t)
than conventional (` 1434/t) method. The net returns
realized from sugarcane were lower under conventional
method (` 87473/ha) than under SSI method (` 123739/
ha). The farm business income represents the financial
return on the shareholders capital invested in the farm
business. It is used while assessing the impact of new
policies or regulations on the individual farm business.
122
The overall farm business income under

conventional method (` 87473/ha) was lower than
under SSI method (` 123739/ha). It indicates that in
the wake of capital losses the SSI farmers can better
withstand the uncertainties as compared to the
conventional sugarcane growers. The overall family
labour income (` 93168/ha) was also more in the SSI
method. Farm investment income represents the return
on all capital invested in the farm business including
borrowed capital; it also incorporates returns to risk
and entrepreneurship. The overall farm investment
income was lower for conventional method (` 80353/
ha) than the SSI method (` 117047/ha). It is a general
measure of profitability of farming and it has indicated
that SSI method is more profitable than the
conventional method. The overall benefit-cost ratio was
1.60 under conventional and 1.73 under SSI method.
Summing-up, the sugarcane cultivation under SSI
method is more profitable than the conventional method
in the study areas. However, the escalating input-costs,
fluctuating output-prices, and delayed payments by
sugar mills are the major factors limiting productivity
and profitability of sugarcane cultivation.
Production Function Estimates for Sugarcane
Cultivation
To analyse the efficiency of resources used in
sugarcane production, Cobb-Douglas production
function was fitted separately for both conventional
and SSI methods, including important inputs, viz NPK,
micronutrients and human labour and machine labour
as explanatory variable. Table 3 presents the production
function estimates for conventional and SSI methods
of sugarcane cultivation.
Table 3 reveals that variations in sugarcane yield
could be explained up to 94 per cent in the conventional
method and 85 per cent in SSI method by the
independent variables included in the model. The F
value of conventional and SSI methods (150.821 and
51.520, respectively) was found statistically significant
which indicated the overall significance of the model.
In both conventional and SSI methods, the elasticity
coefficients for nitrogenous fertilizer (0.5943) and
human labour (0.9881) were found positive and
significant at one per cent level. The coefficient for
potassic fertilizer was found significant at one per cent
level in the SSI method. Since, sugarcane crop is a
Vol. 29 (No.1)
January-June 2016
Table 3. Production function estimates for sugarcane

cultivation
Variable code
CONSTANT
NITRO
PHOS
POTAS
MNM
HLAB
MLAB
R square
F value
Regression coefficient
Conventional
SSI
method
method
-5.0139***
(0.6248)
0.5943***
(0.1004)
0.0045
(0.0184)
-0.0012
(0.0387)
0.0371**
(0.0103)
0.9881***
(0.1206)
-0.0159
(0.0105)
0.94
150.821***
-2.4928***
(0.6970)
0.3519***
(0.1236)
0.1424**
(0.0505)
0.2386***
(0.0460)
-0.0650
(0.04419)
0.5763***
(0.0971)
-0.0175
(0.0268)
0.85
51.520***
Note: ***(P<0.01), **(P<0.05), *(P<0.1).

Figures within the parentheses are standard errors of
estimates.
luxurious consumer of potassic fertilizers, it was

observed that farmers apply potassic fertilizers more
than the recommended level.
Decomposition Analysis
Chow test was used to test the shift from
conventional to SSI method. The calculated value of
Chow test (which follows F distribution) was 7.71
which was more than the critical value of F (7,106),
viz. 2.79. Hence, the null hypothesis was rejected which
implied that there was a significant difference between
SSI and conventional methods. Therefore,
decomposition analysis was performed to identify the
different sources of productivity improvement under
the SSI method. The sources of output gain in the SSI
method were decomposed using production function
parameters (given in Table 3) and geometric mean level
of input (see Annexure I). The results of decomposition
analysis are given in Table 4.
The total gain in sugarcane production due to shift
from conventional method to SSI method was found
Table 4. Sources of productivity gain in SSI method of
sugarcane production
Sl. Source of productivity difference
No.
Contribution
(%)
A.
Total observed productivity gain
24.41
B.
Productivity gain due to technology

change, i.e. neutral technology and
non-neutral technology change
2.16
C.
Total productivity gain due to input-use

(1+2+3+4+5+6)
20.49
1.
Nitrogenous fertilizers (kg/ha)
5.68
2.
Phosphatic fertilizers (kg/ha)
0.48
3.
Potassic fertilizers (kg/ha)
2.46
4.
Micronutrients (kg/ha)
3.53
5.
Human labours (persondays/ha)
7.77
6.
Machine labours (hours/ha)
0.54
D.
Total estimated productivity gain (B + C) 22.65
E.
Residual factors
1.76
to be 22.65 per cent, which was mainly contributed by

the difference in the levels of input-use in these
methods. The total contribution of difference in inputuse levels to productivity gain was 20.49 per cent,
which indicated that the productivity of conventional
practices can be increased by 20.49 per cent, if the
input-use levels on these farms could be increased to
the levels of SSI method. In the total productivity gain
due to input-use, the contribution was highest of human
labour (7.77%), followed by nitrogenous fertilizers
(5.68%), micronutrients (3.53%), potassic fertilizers
(2.46%), machine labour (0.54 %) and phosphatic
fertilizers (0.48 %). Among the components of
technological change, the contribution of neutral
technological change in total productivity was
estimated as 2.16 per cent.
Thus, it can be inferred from the decomposition
analysis, that the SSI method was not able to
consolidate the gains due to introduction of a new
technology. The yield gain was largely due to the raise
in input levels in the SSI method. Hence, the extension
agencies should provide training to the farmers on the
adoption of SSI method of sugarcane production very
effectively. Besides, the yield of sugarcane is also
influenced by drip irrigation and fertigation. Hence,
these factors should be exposed and demonstrated to
123
Table 5. Parameter estimates of logit model

Variables
Coefficient
Odds ratio
CONSTANT
AGE
SQAGE
EDU
HHSIZE
FARMSIZE
INCOME
7.8051
-0.3091
0.00298*
2.5036***
-1.3429**
-0.3370*
0.0001**
0.6708
1.0028
12.2265
0.2610
0.7138
1.0001
Note: ***(P<0.01), **(P<0.05), *(P<0.1).

Descriptive statistics of the variables are given in AnnexureII.
the farmers for more adoption of SSI method and hence

for realizing more yield and economic returns.
Factors Influencing Adoption of SSI Method
To identify the factors that influence adoption of
SSI method in sugarcane cultivation, the estimated
parameters used from the logit model were age,
education, household size, total area and annual
income. The parameter estimates of logit regression
are given in Table 5. A perusal of Table 5 revealed that
the educational level of farmers was positive and
significant at one per cent level with coefficient value
as 2.50, depicting a positive influence on willingness
to adopt SSI technology in sugarcane cultivation. The
household size showed a negative value for coefficient
(-1.34) and significance at five per cent level. The
annual income was positive and significant at five per
cent level.
A comparison of odds ratio showed that if age,
household size and total area increase by one unit, the
probability of SSI adoption would be reduced by -32.9
per cent, -73.90 per cent and -28.62 per cent,
respectively. The coefficients of age square, education
and annual income have shown that with increase of
one unit, the probability of being an SSI adopter would
increase by 0.28 per cent, and was 1122.65 per cent
i.e. 0.01 per cent, respectively. It was concluded that
education is a major factor influencing the adoption of
SSI method at one per cent significant level. Similar
studies conducted by Kiruthika (2014) and Tripp and
Ali (2001) have also reported that education has a
positive impact on the adoption level of technology.
124
Vol. 29 (No.1)
January-June 2016
Conclusions and Policy Recommendations
References
In India, the sugar industry is facing the problem

of cane supply, which is due to reduction in area under
sugarcane, labour scarcity, timely availability of
planting material, less use of farm mechanization and
escalation in input costs. This warrants adoption of new
and improved methods in sugarcane cultivation so as
to make it viable and remunerative and to increase the
supply. Therefore, this study has evaluated the
profitability of sugarcane production under
conventional method and a new method, viz.,
Sustainable Sugarcane Initiative (SSI). It has also
identified the factors that influence the adoption of SSI
method and constraints faced by the farmers in SSI
adoption.
Balasaheb, T.D. (2013) Quantification of Yield Gaps in

Different Planting Types of Sugarcane in Maharashtra.
Indira Gandhi Institute of Development Research,
Mumbai.
Although the total cost of cultivation has been

found higher in SSI method than conventional method,
the yield obtained from the SSI method was more and
therefore, overall gross returns and net returns were
higher in the SSI method. The study has concluded
that sugarcane cultivation is more profitable under SSI
than conventional method. The increase in productivity
under SSI over conventional method has been found
24.41 per cent. The study has found the contribution
of productivity gain by input-use as 20.49 per cent and
technological change as 2.16 per cent. It has concluded
that SSI is more with less technology in terms of
water and seed and it is a positive technological change
(implying more output with same inputs) in term of
human labour, fertilizer and machine labour.
The logit regression analysis has revealed
education and experience to be most influential factors
for willingness on adoption of SSI method in sugarcane
cultivation. Therefore, efforts should be made to
increase the educational level and knowledge status of
farmers about the latest technologies such as SSI,
fertigation, intercropping, bio-fertilizers, etc.
Development of necessary infrastructural facilities,
ensuring timely availability of critical inputs (thereby
input outlets) and quality planting materials, periodical
trainings on SSI method and effective policy for
deploying labour from MGNREGA are the major
policy options suggested to improve sugarcane
production and farm profitability.
Bisaliah, S. (1977) Decomposition analysis of output change

under new production technology in wheat farming:
Some implications to returns on investment. Indian
Bohme, M. and Thi Lua, H. (1997) Influence of mineral
and organic treatments in the Rhizosphere on the growth
of tomato plants. Acta Horticulturae, 450: 161- 168.
Chouhan, S. Singh, S.R.K, Pande, A.K. and Gautam, U.S.
(2013) Adoption dynamics of improved sugarcane
cultivation in Madhya Pradesh. Indian Research
Journal of Extension Education, 13(2): 26-30.
Ganesh Kumar, B. (2003) A study to measure the
technological change in dairy farming in Tamil Nadu.
Productivity, 44 (1): 97-104.
Goud, Vinod (2011) First National Seminar on Sustainable
Sugarcane Initiative (SSI), A Methodology to Improve
Cane Productivity. Tamil Nadu Agricultural University,
Coimbatore, Tamil Nadu. 24- 25 August.
GoTN (Government of Tamil Nadu) (2015) Seasonal and
Crop Report, Tamil Nadu - 2013-14. Department of
Economics and Statistics, Chennai.
Gujarati, D.N. and Porter, D.C. (2014) Basic Econometrics.
Fifth Edition. McGraw Hill. pp.1-915.
ISMA (Indian Sugar Mills Association) (2015) Indian
Sugar: Statistical Abstract-Advanced Estimates. New
Delhi.
Kiruthika, N. (2014) Determinants of adoption of drip
irrigation in sugarcane cultivation in Tamil Nadu.
American International Journal of Research in
Humanities, Arts and Social Sciences, 5(2): 143-146.
Loganandhan, N., Gujja, Biksham, Goud, Vinod and
Natarajan, U.S. (2013) Sustainable sugarcane initiative
(SSI): A methodology of more with less. Sugar
Technology, 15(1): 98-102.
Mohanty, M., Das, P. P. and Nanda S.S. (2015) Technology
for enhanced cane production and economic returns in
real farming situations under East Coast climatic
conditions of India. Sugar Technology, 17(2). (DOI:
10.1007/s12355-014-0311-8)
Acknowledgements
Murthy, S.R.S. (2010) Economics of Sugarcane Production

and Processing. National Bank for Agriculture and
Rural Development. Mumbai.
The authors thank the anonymous referee for his

helpful suggestions.
NABARD (National Bank for Agriculture and Rural

Development) (2012) AgriSri SSI- Sustainable
Sugarcane Initiative. Natural Resource Management
Centre, Kolkata.
Narayanamoorthy, A. (2013) Profitability in crops
cultivation in India: Some evidence from cost of
cultivation survey data. Indian Journal of Agricultural
Economics, 68(1): 104-121.
125
Shanthy, R.T. and Ramanjaneyulu, S. (2014) Socioeconomic performance analysis of sugarcane cultivation
under sustainable sugarcane initiative method. Indian
Research Journal of Extension Education, 14(3): 9398.
Pusappa, K.N. (2013) Economics of Sugarcane Cultivation

in Andhra Pradesh. Department of Economics, Andhra
University,Visakhapatnam. pp. 1-13.
Tefera, T. and Melesse Giorgis, W. (2010) Determinants of

coffee husk manure adoption: A case study from
southern Ethiopia. Indian Journal of Agricultural
Economics, 65(1): 159-172.
Rahm, R.M. and Hufman, W.E. (1984) The adoption of

reduced tillage: The role of human capital and other
variables. American Journal of Agricultural Economics,
66: 405-413.
Thennarasu, R. and Banumathy, V. (2011) Economics of

sugarcane production using ecofriendly technology in
Cuddalore district, Tamil Nadu. Indian Journal of
Raju, V.T. and Rao, D.V.S. (1990) Economics of Farm

Production and Management. Oxford and IBH
Pulishing Co. Pvt. Ltd, New Delhi. 185 p.
Tripp, R. and Ali, A. (2001) Farmer Access to Natural Pest

Control Products: Experience from an IPM Project in
India, Overseas Development Institute, Agricultural
Research and Extension Network, Paper No.113,
London, U.K.
Rao, R. (2012) Efficiency, yield gap and constraints analysis

in irrigated vis--vis rainfed sugarcane in North Coastal
Zone of Andhra Pradesh. Agricultural Economics
Research Review, 25(1): 167-171.
Rao, R. (2014) Input use and cost of cultivation of sugarcane
A study in Telangana region of Andhra Pradesh. IOSR
Journal of Economics and Finance, 5(5): 67-74.
Wagh, A.D (2015) Investment in Indian sugar industry: A

perspective. Journal of Contemporary Research, 3(1):
111-117.
Received: December, 2015; Accepted: May, 2016
126
Vol. 29 (No.1)
January-June 2016
Annexure I: Geometric mean levels of inputs used in SSI method

Variable
Conventional method
SSI method
Per cent change in input use/Output
101.90
539.59
137.49
219.56
52.14
361.10
8.58
127.38
605.70
141.19
238.45
88.54
413.29
10.47
25.00
12.25
2.69
8.60
69.81
14.45
22.02
Yield (t/ha)
N fertilizer (kg/ha)
P fertilizer (kg/ha)
K fertilizer (kg/ha)
Micronutrients (kg/ha)
Human labour (person days/ha)
Machine labour(hours/ha)
Annexure II: Descriptive statistics for the factors affecting adoption of SSI method
Factor
Age (years)
Age square
Education (index*)
Household size (No.)
Total area (acres)
Annual income (`)
Mean value
Minimum value
Maximum value
Standard deviation ()
47.41
2377.05
4.46
4.75
8.46
312350
28
784
1
3
1.25
137281
71
5041
6
9
32
826048
11.39
1116.67
1.60
1.51
5.29
174831
Note : * Weighted average of formal education received by the household members (illiterate = 0, primary = 1, middle = 2,
secondary = 3, higher secondary = 4, graduate = 5 and post-graduate = 6).

DOI: 10.5958/0974-0279.2016.00025.2
Effect of Change in Indian Rice Price on Nepalese Rice Market:

A Partial Equilibrium Model
Bhawani Mishraa*, Krishna L. Poudelb and Dilli Raj Mishrac
University of Missouri, Office of Social and Economic Data Analysis, Columbia, USA
University of Missouri, Department of Agricultural and Applied Economics, Columbia, USA
c
Tribhuvan University, Kailali Multiple Campus, Department of Management, Dhangadhi, Nepal
a
Abstract
Using a partial equilibrium model, this study has investigated cross-border price effect on the rice demand
and supply in Nepal. Due to bigger market and close proximity, Indian rice price is likely to affect
Nepalese rice price. This study has examined the effect of Indian rice price on Nepalese rice demand, and
supply by making projections for the next 10 years period using the baseline data from 1992 to 2013. The
model has estimated that the per capita rice demand will decrease from 130 kg/capita to
128 kg/capita between 2013 and 2023. The model has also introduced 3 per cent price shock (increase) in
Indian rice price, a trend in past five years, to examine the scenario of rice import and export for Nepal.
Due to increase in price in the Indian market, Nepalese rice price would also increase which will give an
incentive to the farmers to produce more rice and Nepal will stop importing rice from India by the
year 2019.
Key words: Partial equilibrium model, Nepalese rice market, rice import, rice export, Nepal
Introduction
Rice is the staple food of nearly half of the worlds
7.2 billion inhabitants (IRRI, 2014). Almost 40 per cent
of the worlds rice is harvested in South Asia (Gumma
et al., 2011). In Nepal, rice is grown on about 1.42
million hectares, which is almost 46 per cent of land
cultivated in the year 2012-13 (MoAD, 2013). As a
result, rice is a major source of food and plays a major
role in food security of Nepalese households (Prasad
et al., 2011). Despite an increasing trend, the rice
production is still not sufficient to meet the demand of
many Nepalese households (MoAD, 2013). Therefore,
rice is imported both formally and informally in Nepal
through porous border between India and Nepal. The
price of Indian rice plays an important role in the rice
Email: mishrab@umsystem.edu
trade between these two countries. Nepalese rice

market is an open market with no tariff; however, rice
is mostly bought formally by the state-owned
corporation to meet its demand in the food deficit
districts (Tobias et al., 2012). The Indian policy allows
exporting of only fine-grade rice, but recently, India
has changed the policy by allowing export of other
rice grades with quantitative restrictions (Tobias et al.,
2012). But, the export ban or quantitative restrictions
on rice has no noticeable effect on quantities imported
in Nepal due to a porous border between these two
countries (World Bank, 2010). Due to the bigger market
size, rice price shocks originated from India are likely
to affect the Nepalese rice market. Therefore, it is
essential to examine the factual situation. If the Indian
rice price affects the Nepalese rice market, then what
would be the possible scenarios on the demand and
import of rice in future under the current trend or
128
possible price shock scenario. This study has examined

these questions. The results of this study are likely to
help to the policy makers to understand the price effect
of export and import on the Nepalese rice market.
Additionally, rice being a major food commodity for
Nepalese households, examining cross border price
effect on rice demand and supply would provide
insights into the household food security in Nepal
(Sanogo and Amadou, 2010).
Study Framework
The unequal sizes of Indian and Nepalese rice
markets make a perfect example of international trade
theory between small and large country case studies.
Under this framework, the price of a large country is
determined by the demand and supply of rice in its
domestic market. But, the large countrys price is
exogenous for a small country and affects the small
countrys price. Therefore, large countrys price plays
a role in determining the demand and supply of a small
country. After trade with no restrictions between the
two countries, the Indian rice price equals the Nepalese
rice price (Figure 1).
Vol. 29 (No.1)
January-June 2016
supply of rice. Therefore, these factors should be

included in estimating the model and making future
projections.

The information on rice price was collected from
Food and Agriculture Organization (FAO). Rice
production, supply, and distribution data were collected
from the Ministry of Agricultural Development, Nepal;
and United States Department of Agriculture, Foreign
Agricultural Service (USDA-FAS). Macroeconomic
variables such as GDP and population were collected
from the World Bank and from United Nations,
Department of Economic and Social Affairs,
respectively.
where, PNep is Nepalese domestic rice price, PInd is

Indian domestic rice price, and PT is equilibrium price
after trade occurs. In addition, income, current
production level, and population affect the demand and
This study has used a partial equilibrium model to

examine the effect of change in Indian rice price on
Nepalese rice demand and supply. A partial equilibrium
model is a useful tool to study sector specific policy.
The Indian domestic price model acts as a connector
between Nepalese rice model and Indian rice model.
In order to project the effect of Indian rice price, many
auxiliary regression models were estimated with
exogenous explanatory variables. This method was
used by the Food and Agricultural Policy Research
Institute (FAPRI), University of Missouri, Columbia,
for baseline projections. Due to amount or quality of
data, auxiliary regression models were used rather than
other methods, such as 2SLS. The following model
components were used to estimate the effect of Indian
rice price on Nepalese rice demand and supply.
Domestic Rice Supply
This equilibrium price relationship can be

expressed by Equation (1):
PNep = PInd = PT
(1)
Production Supply = f (lagged price, lagged

production, trend)
Beginning Stock = Ending stocks of lagged time
QSInd
PT
QDNep
Q
Q
Figure 1. Equilibrium rice price after trade between
India and Nepal
Note: In figure QSInd denotes quantity supply of India, QDNep
denotes quantity demand in Nepal, and QT denotes quantity
at the equilibrium price PT.
T
Total Supply = Production + Beginning Stocks +

Import
Domestic Rice Demand
Domestic per capita demand=
f (domestic rice price, income, trend)
Ending Stock = Beginning stock + Production
Consumption Export
Total Demand = Domestic consumption + Ending
stocks + Export
Mishra et al. : Effect of Change in Indian Rice Price on Nepalese Rice Market
Domestic Rice Price
129
Table 1. Coefficient estimates of auxiliary models
Domestic Rice Price = f (Real rice price in India

with exchange rate)
Model Closure
Imports = Total demand Production Beginning
stocks
Export = Beginning stock + Production
Consumption Ending stock
The partial equilibrium models equate the supply
and demand in one or more markets so that the markets
clear at their equilibrium price levels.

The partial equilibrium model can provide answer
to two objectives. One, it makes a projection of baseline
data such as price, import, demand, and supply of rice
for Nepal. Two, it also examines the 3 per cent Indian
rice price shock effect on Nepalese rice price, demand
and supply. In order to make projection, we estimated
the demand, supply and market closure equations. The
model also included trend variable to capture secular
trend in addition to usual variables income, price, and
production in the previous year. Coefficients of those
regression models showed the expected signs (Table
1). In the supply equation, own price supply elasticity
was taken from the other study (World Bank, 2010).
Most of the variables used in the model had
baseline up to the year 2013 and the year 2014 was the
first year of the projection. Due to unavailability of
data, some of the variables like Indian rice price from
2009 were projected earlier than 2013. The model
estimates that total rice demand will increase by 1.1
Demand Equation
Coefficient
Intercept
Trend
GDP per capita
Real domestic rice price (Nepal)
0.152589*
-0.001056
0.000023
-0.000247**
Supply Equation@ (ln production supply)

Intercept
ln (Real lagged price Nepal)
ln (lagged prod)
Trend
Market Closure (Domestic rice)
Intercept
Indian domestic rice price
0.681477
0.300000S
0.418948
0.003690*
5529.865374*
0.261917*
Notes: SOwn price supply elasticity is the lower end value

and was taken from the other study.
@
Supply equation has been estimated in natural
logarithmic form.
* and ** denote significance at 5 per cent and 10 per
cent levels, respectively.
Source: Authors estimates
per cent in Nepal (Table 2). Similarly, real producers

rice price in India will grow by 0.2 per cent from 2013
to 2023. However, during the past five years, it was
around 3 per cent. Under the current trend, there is no
prospect of rice export until 2023 to India, despite a
higher price in India.
Based on the current trend, the rice production will
continue to grow from 3361 thousand tonnes to 3603
thousand tonnes during the projection period, 2014 to
2023 (Figure 2). Rice production is likely to increase
Table 2. Summary of average estimate of rice demand, export, import, supply, and producer price
Particulars
GDP per capita (constant US$ 2001)
Total demand (000 Mt)
Exports (000 Mt)
Total supply (000 Mt)
Milled production (000 Mt)
Imports (000 Mt)
Real Indian producer rice price (constant US$ 2001)
2003 to 2013
Average
2014
2015
2014 to 2023
Average
292
3041
0
3041
2896
145
310
315
3640
0
3640
3293
347
341
321
3680
0
3680
3340
340
342
344
3821
0
3821
3464
356
344
Source: USDA, World Bank, FAO, and authors estimate (2014)
130
Vol. 29 (No.1)
January-June 2016
Figure 2. Rice production (milled) in Nepal: 1992-2023

Source: USDA and authors estimates (2014)
only slowly during the projection period. Due to higher

Indian rice price, Nepalese farmers have a strong
incentive to produce more rice. The increase in rice
demand in Nepal is mostly due to the population
pressure.
Based on the current trends, the domestic rice
demand will decrease from 129 kg to 128 kg per capita
by 2023 despite the increasing total demand shown in
Table 2 (Figure 3). The total demand is increasing and
per capita demand is decreasing which indicates that
population growth or increase in real income might
have contributed to the increasing total demand.
Nepal will continue to import rice from India under

the current scenario of population growth, income, and
shortfall in domestic rice production. However, a higher
Indian rice price gives an incentive to the Nepalese
farmers to produce more rice. Under the current
scenario, there will be an increase in import of rice
from India till 2021 and then there would be a decrease
in import due to decreasing trend of per capita rice
demand and increase in domestic rice production due
to price increase as shown earlier (Figure 4). The trend
also shows that rice export will be zero in the projection
period 2013 to 2023 under the current scenario.
Figure 3. Per capita demand of rice in Nepal, 1992-2023

131
Figure 4. Rice import from India, 1992-2023

Source: FAO and authors estimates
As Sanogo (2008) explained, the price fluctuation

in India is transmited, in both short- and long-run,
across the border. Therefore, Nepals rice import and
potential export are likely to depend largely on Indian
rice price. The trend shows that Indian nominal rice
price is increasing by 15 per cent per year on average
during the projection period (2014-2023), whereas
domestic rice price is increasing by 8 per cent per year
on an average in the same period. Despite a higher
price signal from the Indian market, which is an
incentive for more rice production at the domestic
market, Nepal still imports rice from India. This could
be the combination of lag in the adjustment of
production, increasing population pressure, and income
in Nepal. The price differential between India and
Nepal is also due to market inefficiencies and relative
transportation costs in Nepal (Sanogo, 2008).
In addition to the current trend, we have also

examined the effect of 3 per cent price increase in the
Indian rice market. Presetnly, the real domestic price
is increasing at a slower rate of about 1 per cent per
year, but after the 3 per cent increase in rice price in
India, the domestic price will also rise sharply on an
average by 10 per cent per year (Figure 5). This
provides a strong evidence of strong price transmission
signal from India to Nepal and is likely to have a large
impact on the Nepalese rice market.
After imposing a 3 per cent shock in Indian rice
price, the domestic demand of rice in Nepal will
decrease from 130 kg/capita to 123 kg/capita between
2013 and 2023 (Figure 6). However, under the current
trend, the per capita demand decreases from 130 kg/
capita to 128 kg/capita only.
Figure 5. Real domestic rice price (NRs 2001) after 3 per cent shock on the Indian rice price
Source: FAO and authors estimates
132
Vol. 29 (No.1)
January-June 2016
Figure 6. Per capita demand of rice in Nepal after 3 per cent shock on Indian rice price
Figure 7. Rice import from India after 3 per cent price shock in Indian rice market
Source: FAO and authors estimates (2014)
Due to the effect of 3 per cent increase in Indian

rice price, the demand from Nepalese consumers would
reduce and there would be an incentive to produce more
rice. Consequently, the import of rice from India is
projected to be zero by the year 2019 (Figure 7).
Conclusions
The paper has measured the cross border price
effect on rice demand and supply using partial
equilibrium model. Any change in the Indian rice price
is transmitted to Nepal and thereby affects the demand
and supply scenario in Nepalese rice market. The
analysis has been done under two scenarios: (a) current
trend, and (b) 3 per cent increase in the Indian rice
price. To examine the effect of Indian rice price,
projections of rice demand, supply, import, and export

have been made for the next 10 years using the baseline
data from 1992 to 2013. The study has found that under
the current trend, domestic price will increase slowly
in the projection period. Under the current trend
scenario, there will not be rice export but import of
rice will also increase until 2021 and then it will
decrease. The current trend has also shown that Indian
rice price is increasing at a higher rate than Nepalese
rice price in Nepal. The study has also introduced a 3
per cent price shock to examine the change in domestic
rice price. The results show that 3 per cent increase in
Indian rice price would increases 10 per cent price in
the domestic market. Due to this effect, farmers have a
strong incentive to produce more rice and Nepal is
likely to stop importing rice from India from the year

2019. The per capita demand has also shown a
decreasing trend during the projection period. Thus,
the study has estimated the future export, import,
demand, and supply scenarios of rice in Nepal.
References
FAO (Food and Agriculture Organization of United Nations)
(1999) Report of the Expert Consultation on Bridging
the Rice Yield Gap. FAO Regional Office for Asia and
the Pacific, Bangkok.
FAO (Food and Agriculture Organization of United Nations)
(2014) Retrieved from FAOSTAT: http://
faostat3.fao.org/home/ on 7-10-2014
Gumma, K.M., Nelson, A., Thenkabail, S.P. and Singh, A.N.
(2011) Mapping rice areas of South Asia using MODIS
multi-temporal data. Journal of Applied Remote
Sensing, 5(1): 1-26.
IRRI (International Rice Research Institute) (2014) The New
Green Revolution: A Bigger Rice Bowl. Manila
Karkee, M. (2008) Nepal Economic Growth Assessment
Agriculture. USAID, Kathmandu, Nepal.
MoAD (Ministry of Agricultural Development) (2013)
Statistical Information on Nepalese Agriculture: 2013.
Kathmandu, Nepal.
Prasad, K.S., Pullabhotla, H. and Ganesh-Kumar, A.
(2011) Supply and Demand for Cereals in Nepal 2010-
133
2030. Discussion paper 01120. IFPRI, New Delhi

Office.
Sanogo, I. (2008) Spatial integration of the rice market:
Empirical evidence from mid-west and far-west Nepal,
and the Nepalese-Indian border. Asian Journal of
Agriculture and Development, 4(1): 139-156.
Sanogo, I. and Amadou, M.M. (2010) Rice market
integration and food security in Nepal: The role of crossborder trade with India. Food Policy, 35(4): 312-322.
Tobias, M., Valera, G.H., Mottaleb, A.K. and Mohanty, S.
(2012) Handbook on Rice Policy for Asia, Eds: H.G.
Valera, K.A. Mottaleb and S. Mohanty. International
Rice Research Institute, Los Baos.
United Nations, Department of Economic and Social Affairs
(2014) Retrieved from http://esa.un.org/wpp/excel-data/
population.htm on 7-9-2014.
USDA (United States Department of Agriculture) (2014)
Production, Supply, and Distribution. Retrieved from
http://apps.fas.usda.gov/psdonline/psdQuery.aspx on 711-2014.
World Bank (2010) Food Price Increases in South Asia
National Responses and Regional Dimensions.
Washington, DC.
World Bank (2014) World Development Indicators.
Retrieved from http://data.worldbank.org/country/nepal
on 9-12-2014.
Received: December, 2015; Accepted: March, 2016

DOI: 10.5958/0974-0279.2016.00026.4
Export of Rice from India: Performance and Determinants

Anup Adhikari, M.K. Sekhon* and Manjeet Kaur
Department of Economics and Sociology, Punjab Agricultural University, Ludhiana-141004, Punjab
Abstract
The study has examined the growth performance and identified determinants of rice exports from India
with special reference to basmati rice during the period 1980-81 to 2012-13. Compound growth rate,
instability index and Markov chain analysis, were applied to estimate trend, and instability and the
project export to different countries. The time series data were made stationary before estimating the
determinants of Indian rice export. The study has observed that rice contributed substantially to the
national income during the study period. The higher growth observed in value of basmati rice export
(15.87%) was due to higher growth in unit value than quantity of export (7.55%) during 1980-81 to
2012-13. The growth rate of unit value of rice export was higher in period I (13.48%) than period II
(5.06%). The growth rates in export of non-basmati rice in terms of quantity, export earnings and unit
value were 10.87 per cent, 17.74 per cent and 6.20 per cent, respectively during the study period. The
instability index has been found highest for quantity (43.37 %) in case of basmati rice and value (141.36%)
in case of total rice during the entire period. The UAE has been found to be a highly preferred market for
Indian basmati rice and Nigeria for Indian non-basmati rice, as indicated by the probability of retention
of their previous shares. The study has projected that during 2013-14 the major markets for Indian basmati
rice would be Iran and Saudi Arabia, whereas for Indian non- basmati rice, the major markets would be
Nigeria and South Africa. The estimated regression model has shown that export price, international
price, lagged production, domestic consumption, and exchange rate are the major determinants of rice
export from India. In order to sustain in the international market, Indian export price needs to be competitive
besides improvement in quality and sanitary standards.
Key words: Rice, export destinations, performance, instability index, Markov Chain Analysis, export
determinants
JEL Classification: F17, Q17
Introduction
International trade plays an important role in the
economic development of a country. The participation
of India in international trade is largely confined to
primary products, especially of the agricultural sector.
Indian trade policy for agricultural commodities is
guided by the twin objectives of ensuring national food
Email: sekhonmk@yahoo.co.in
This paper is drawn from the M.Sc. thesis of the first author
submitted to the Department of Economics and Sociology,
Punjab Agricultural University, Ludhiana
security and building export markets for enhancing the

farmers income. In 2011, India was on the tenth
position in agriculture and food exports in the world
and had 2.07 per cent share in global export of
agricultural and food products. Rice is one of the most
important exportable agricultural commodities from
India. The steady increase in production, availability
of buffer stock and rising demand for basmati rice in
the international market has helped India in increasing
rice export. In 1990-91, India was able to export only
505 thousand tonnes rice with a total values of ` 459.63
crore and this amount increased to ` 33858.19 crore in
136
2012-13. The major markets for export of Indian rice

were: United Arab Emeritus, Saudi Arabia, Iran,
Kuwait and Iraq (for basmati rice).
The new multilateral trade regime coupled with
the policy changes adopted by the most nations aiming
towards globalization, provided new opportunities as
well as posed several challenges for expanding trade
in agricultural products. India too has been able to
derive significant benefits from the changed global
environment. Since global trade environment is highly
dynamic and export is influenced by both micro and
macro policies, it is important to understand the
product-specific dynamics to improve export
performance of agricultural commodities. The global
markets for Indian rice are highly dynamic and the
barriers to trade are being lowered gradually all around
the world (Singh, 2001). The export of rice is also
related with the buffer stock of rice held by the
government. Because of comfortable buffer stock, India
became a major exporter of rice in 2012. There is a
strong demand for Indian rice in the international
markets. The increasing consumer demand for rice and
Indias strength for production of basmati as well as
non-basmati rice, coupled with liberal export policy,
and large public stock have created ample scope for
rice export. In recent years, the African countries have
also shifted to Indian non-basmati rice because of price
competitiveness (Chandrashekhar, 2013).
The future performance of any product in
international markets can be judged in the light of its
past performance. Therefore, evaluation of past
performance of rice is necessary to work out its export
potential, challenges and opportunities. The present
study has analysed major markets for Indian rice, in
terms of its future share. Rice export from India is
determined by various factors and therefore, reliable
estimates of determinants of export are essential for
the formulation of appropriate policies. The present
study has examined the growth performance and
instability of rice, important destinations, with
projections for export in future and has identified the
factors affecting export of rice.

The study is based on the time series data on export
quantity, value and unit value which were compiled
from various sources for a period of 33 years (1980-
Vol. 29 (No.1)
January-June 2016
81 to 2012-13). For analysis, the total period was subdivided into two periods, viz. Period-I: Pre-WTO
period (1980-81 to 1994-95); and Period-II: Post-WTO
period (1995-96 to 2012-13). The division of period
was based on the assumption that liberalization as well
as formation of WTO helped in boosting up rice export.
The data on export quantity and value were collected
from Agricultural Statistical Compendium, and
APEDA. Similarly, data on production, import, stock,
and international price of rice were collected from
various sources like indiastat.com website, hand book
of Indian Economy, pink sheet of World Bank, etc.
The global data on exports and imports were collected
from various issues of Rice Year Book, USDA websites.
To analyze trend, the compound growth rates for export
of basmati rice and total rice in terms of quantity, value
and unit value realizations were worked out for the
both sub-periods and the overall period. The Instability
Index was computed using Cuddy-Della Valle Index,
I = CV *(1-R2)0.5(Singh and Byerlee, 1990), where,
CV is the coefficient of variation and R2 is the corrected
coefficient of determination of the log linear function.
The trade directions of export were analyzed using
the first order Markov chain approach. Central to the
Markov chain analysis is the estimation of transitional
probability matrix Pij. The elements Pij of the matrix P
indicates the probability that export will switch from
the ith country to the jth country with passage of time
(Dent, 1967; Lee et al., 1970; Gillet, 1976). The
diagonal elements of the matrix measure the probability
that the export share of a country will be retained.
Hence, the examination of the diagonal elements will
indicate the preference of an importing country for a
particular countrys exports.
In the context of the present study, the structural
changes were treated as a random process with selected
importing countries. The average export to a particular
country was considered to be a random variable which
depended only on the past exports to that country, and
can be denoted algebraically by Equation (1):
(1)
where, Ejt denotes exports from India to the jth country
during the year t, Eit-1 denotes exports from India to
the ith country during the period t-1, Pij is the probability
that exports will shift from the ith country to the jth
Adhikari et al. : Export of Rice from India: Performance and Determinants
country, ejt is the error- term which is statistically

independent of Eit-1, t is the number of years considered
for the analysis, and r is the number of importing
countries. The transitional probabilities Pij which can
be arranged in a (c x r) matrix have the following
properties:
O Pij 1
137
export to these countries in the previous period (t-1)

with the transitional probability matrix. Multiple
regression was carried out, using ordinary least square
(OLS) estimation procedure, in the statistical software
E-Views.
Determinants of Rice Exports
The factors affecting the export of rice were
identified using log-log linear type of function:
The minimum absolute deviation (MAD)

estimation procedure was employed to estimate the
transitional probability, which minimizes the sum of
absolute deviations (Fisher, 1967; Wagner, 1959). The
conventional linear programming technique was used,
as this satisfies the properties of transitional
probabilities of non-negativity restrictions and row sum
constraints in estimation. The linear programming
formulation is stated as:
Min OP* + Ie
Subject to, XP* + V = Y,
GP* = 1, P*e 0
where,
0 = Vector of zeroes,
P* = Vector in which probability Pij is arranged,
I = Appropriate dimensioned column vector of unit,
e = Vector of absolute error (|U|),
Y = Vector of export to each country,
X = Block diagonal matrix of lagged values of Y,
V = Vector of errors, and
G = Grouping matrix to add the row elements of P as
arranged in P* to unity.
After calculating the transitional probability
matrix, the expected shares of export were calculated
by Equation (2):
(j=1,2,3, r)
(2)
where, Yjt is the predicted proportions of the jth

countrys share at time t, Yt-1 is the observed
proportion of the ith countrys share at time t-1, and
Pij is the estimated transitional probability matrix.
Thus, the expected export shares of each country
during period t were obtained by multiplying the
lnQT = b 0 +b 1 lnEXP+b 2 lnINT +b 3 lnLGP+

b4 lnDMC+b5lnEXC +
(3)
where, QT= Total rice export from India (000 t), EXP
= Export prices (`/t), INT= International price (Thai
5% $/t), LGP= Lagged production of rice (million
tonnes), DMC = Domestic consumption of rice (000
tonnes), EXC = Exchange rate with dollar (`/$), =
Error-term, and b 1, ,b 5 are the regression
coefficients and b0 is a constant.
The export prices and international prices for rice
have been represented by their respective unit values.
The monthly unit value of international price of rice
was derived from the international price given by pink
sheet of World Bank and was converted according to
the fiscal year. The international price of rice is
generally represented by the Thai rice, fob Bangkok.
Due to nonavailability of data on domestic
consumption, it was computed as per expression (4):
Domestic consumption = Production + Import +
Stock change Export
(4)
While dealing with time series data, the first step
is to check whether the underlying time series is
stationary or not. In a time series, the set of possible
values at a particular point of timet is denoted by Yt,
and a time series is denoted by {Y (t), t T}. For
stationary, Yt must fulfill the following three
characteristics:
(i) E(Yt) =
( i.e. Mean is constant)
(ii) Var(Yt)= E(Yt- )2 = 2 (i.e. Variance is constant)

(iii) k =E[(Yt - ) (Yt+k - ] (i.e. Covariance is
constant)
To check unit root in the data, augmented DickeyFuller (ADF) test was used adopting the regression
Equation (5):
138
(5)
where, t is the white noise error, Yt-1 = (Yt-1 Yt-2),
where represents the first difference, m represents
the number of lagged difference. These lags are
included to make error- term white noise in above
equation. 1 is the intercept and t represents time trend.
The null hypothesis of ADF is H0: =0; There is unit
root, and the alternative hypothesis is H1: <0; There
is no unit root.
CUSUM Test and CUSUM of Squares Test
The CUSUM test is based on the cumulative sum
of the equation errors in regression. The E-Views
represent graphically the cumulative sum of errors
together with critical lines of 5 per cent. The equation
parameters are not considered stable if the whole sum
of recursive errors gets outside the two critical lines.
CUSUM of Squares test is calculated similarly and
interpreted as CUSUM test, with the difference that
instead of recursive errors, we use recursive doubled
errors.

Scenario of World Rice Trade
India ranks second in the production of rice after
China, and accounted for 22.12 per cent of global rice
production in 2012-13. The international rice market
is characterized by high volatility with inelastic supply
and demand. The global trade of rice is very low in
comparison to its production. Only seven per cent rice
production is traded and global rice exports are highly
concentrated with the top five exporters accounting for
85 per cent of global net trade (Wailes and Chavez,
2012).
The global export of rice increased from 11.70 Mt
in 1990 to 38.66 Mt in 2013. Although, China and India
together accounts for half of the world rice production,
Thailand was a major player till 2011 on rice export
front. When the Government of India removed ban on
export of non-basmati rice in September, 2011, India
emerged as a leading exporter of rice with market share
of 25.73 per cent, followed by Vietnam (19.38 %) and
Thailand (17.45 %) in 2012. Indias share in world
export of rice has increased from 25.73 per cent in 2012
to 27.16 per cent in 2013. The other exporters such as
Vol. 29 (No.1)
January-June 2016
Vietnam and Thailand shared 17.59 per cent and 17.33

per cent, respectively in the world export of rice. The
country-wise rice export and imports, given in
Annexure I, show that India was a major exporter and
China was the largest importer of rice in the world,
followed by Nigeria and Iran during the year 2013.
The international rice market is quite thin and
oligopolistic in nature. The entry of a big player like
India can affect the world price significantly
(Bhattacharyya and Pal, 2000). Protectionist measures
taken by the Indian government during 2008 like ban
on export of non-basmati rice and imposition of export
tariff on basmati rice to increase domestic supply and
lower domestic price reduced the rice supply in world
markets. The period from mid-2007 to mid-2008 is
referred to as world rice crisis period, which resulted
in rise in world price of rice. Deliberated trade policy
based on scientific facts and figures will retain Indias
market share in the global rice trade.
Indias Production and Exports of Rice during
1980-81 to 2012-13
India has been exporting substantial quantities of
rice to various parts of the world since 1980-81. In
2012-13, the total production of rice was around 104
Mt, of which around 10.1 Mt was exported (Figure 1).
In 1980-81, the total quantity of rice (basmati +
non-basmati) exported from India to different countries
was 7.27 lakh tonnes, which rose to 49.1 lakh tonnes
in 1995-96. After a lot of fluctuations, the export of
total rice started increasing again from 2009-10 and it
reached 101.5 lakh tonnes in 2012-13. There was not
much earning from exports of rice during 1980s and
1990s but it increased gradually during the past five
years. In terms of value, the rice export was of ` 223.86
crore in 1980-81 which increased to ` 1164.40 crore
in 2008-09 ( Table 1). But in between 1980-81 and
2008-09, it showed a fluctuating pattern in export
earnings which may be due to fluctuating domestic
production, inconsistent export policies, frequent bans
on export of non-basmati rice, currency devaluation,
and volatility in the international market. In the year
2000-01, the earning from rice exports recorded the
lowest value due to the fall in international prices of
rice. The exports of rice gradually increased from 200809 and reached ` 33,858.19 crore in 2012-13. The
139
Figure 1. Indias production and export of rice during 1980-81 to 2012-13

Table 1. Share of Indian rice export to total agricultural exports: 1980-81 to 2012-13
Year
1980-81
1985-86
1990-91
1995-96
2000-01
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
Total rice
export
(Quantity,
000 tonnes)
Total rice
export
(value,
in crore `)
Unit value
(/tonne)
Agricultural
export
(in crore `)
Per cent share

of total rice
export in
agricultural export
727.35
245.01
505.11
4914.01
1534.47
4088.16
4747.92
6469.28
2488.29
2156.32
2471.35
7175.89
10147.89
223.86
196.32
459.63
4568.12
2943.39
6221.25
7035.88
11754.39
11164.4
11254.42
11585.92
24108.74
33858.19
3077.71
8012.61
9099.60
9296.11
19181.80
15217.73
14818.87
18169.55
44867.76
52192.72
46880.94
33596.86
33364.77
2375.70
3271.50
6012.76
20397.74
28657.37
49216.96
62411.42
79039.72
85951.67
89341.50
117483.60
187609.30
230141.13
9.42
6.00
7.64
22.40
10.27
12.64
11.27
14.87
12.99
12.60
9.86
12.85
14.71
Source: GoI (2013a); GoI (2013b); indiastat.com; Bansil (1992)
percentage share of total rice export in total agricultural

export had increased from 9.42 per cent in 1980-81 to
12.60 per cent in 2009-10, but decreased to 9.86 per
cent in 2010-11. The decrease in percentage terms was
due to higher export of agricultural products, other than
rice. In the year 2012-13, the share of rice export in
total agricultural export was 14.71 per cent.
Growth of Rice Export from India

The annual compound growth rates (ACGR) for
export of basmati rice and total rice in terms of quantity,
value and unit value during the period 1980-81 to 201213 are presented in Table 2. During period 1980-1994,
the quantity of rice export recorded a positive but non-
140
Vol. 29 (No.1)
January-June 2016
Table 2. Compound growth rates of quantity, value and unit value of basmati rice and total rice export from India,
1980-81 to 2012-13
(in per cent)
Period
Quantity
Basmati rice
Value
Total rice (basmati and non-basmati)

Quantity
Value
Unit value
Unit value
I (1980-81 to 1994-95)
0.22ns
(2.06)
13.74**
(1.93)
13.48**
(0.84)
4.07ns
(3.17)
14.96**
(2.84)
10.46**
(1.05)
II (1995-96 to 2012-13)
12.11**
(0.93)
17.78**
(1.72)
5.06**
(0.99)
3.72ns
(2.34)
12.34**
(1.85)
8.31**
(1.65)
Overall (1980-81 to 2012-13)
7.55**
(0.73)
15.87**
(0.69)
7.74**
(0.49)
10.87**
(1.25)
17.74**
(0.94)
6.20**
(0.60)
Note: Figures within the parentheses indicate standard error in respective values
** and * denote significance at 1 per cent, and 5 per cent levels, respectively.
significant compound growth due to lower domestic

production and restriction of minimum export price
till 1991. In contrast, the export value and unit value
registered a positive growth rate of 13.74 per cent and
13.48 per cent per annum. The export of basmati rice
in quantity grew at the rate of 12.11 per cent per annum
during the period 1995-2012. The increase in quantity
of rice export was due to a strong demand for basmati
rice in the international markets and comfortable
production in the country. The value of export of
basmati rice also increased at a compound growth rate
of 17.74 per cent per annum. The export performance
of basmati rice in terms of quantity and value was better
in period II.
During the overall period 1980-2012, the total rice
export in terms of quantity, value, and unit value grew
at a positive and significant growth rate of 10.87 per
cent, 17.74 per cent, and 6.20 per cent per annum,
respectively, which may be due to higher demand of
basmati rice in the international market, consistent
policies for export of basmati rice, higher international
price, and increased domestic production making
comfortable stock of rice in the central pool (Gangwar
and Rai, 1995; Shende et al., 1998).
Instability in Rice Export from India
The higher export instability in terms of quantity
of rice during period I may be due to more variability
in quantum exported than during period II. The
consistent export policy, international demand for and
domestic production of basmati rice during period II

may be the other factors. Similarly, variability in
quantum of export and unit price were the dominant
factors for the variability in export earnings. The export
earnings showed a higher stability during period I than
period II. The unit value realization exhibited more
instability during period II than period I and overall
period, may be due to the impact of devaluation of
Indian rupee after 1991.
The instability index of basmati rice and total rice
exports in terms of quantity, value and unit value is
presented in Table 3. The quantity exported registered
instability of 46.23 per cent, 53.23 per cent and 50.68
per cent during period I, period II and overall period.
The export value recorded instability of 41.22 per cent,
43.53 per cent, and 37.40 per cent during period I,
period II and overall period, respectively. The quantity
exported has been more stability during period I than
period II and overall period. However, the export value
was found more unstable during period II than period
I and overall period, while unit value realization was
more stable during period I as compared to period II
and overall period.
The variability in rice exports was due to the
changes in the composition of rice varieties. The export
instability of rice in terms of quantity as well as value
may be due to increasing domestic and international
rice demand, volatility in world prices, changes in
export policy of India and political situation in
international markets.
141
Table 3. Instability index of rice export from India: 1980-81 to 2012-13

Period
I (1980-81 to 1994-95)
II (1995-96 to 2012-13)
Overall (1980-81 to 2012-13)
CV(%)
Basmati rice
Instability index
Total rice
Instability index
CV(%)
33.16
69.77
95.91
Quantity
33.16
19.49
43.37
46.23
53.23
97.71
46.23
53.23
50.68
Value
I (1980-81 to 1994-95)
II (1995-96 to 2012-13)
Overall (1980-81 to 2012-13)
I (1980-81 to 1994-95)
II (1995-96 to 2012-13)
Overall (1980-81 to 2012-13)
71.92
101.98
152.98
34.49
34.43
33.52
76.41
88.40
141.36
41.22
43.53
37.40
55.09
35.60
63.00
Unit Value
11.42
21.60
20.41
42.61
60.52
74.91
13.87
36.56
34.57
Major Export Destinations of Indian Basmati Rice

The major export destinations of Indian rice vary
according to variety of rice. Basmati rice is mainly
exported to Iran, Saudi Arabia, Kuwait, Iran, UK, USA,
UAE and Iraq. The major destinations of Indian basmati
rice and their shares illustrated in Table 4 show that
nearly 75 per cent of the total export of basmati rice
from India was concentrated in countries like Saudi
Arabia, UK, Kuwait, UAE, Iran and Iraq. Among the
major importing countries, Iran accounted for the
highest share (33.3%) in total export of basmati rice
from India, followed by Saudi Arabia (18.85 %), UAE
(6.76 %), Kuwait (5.46 %) and Iraq (5.55 %) in the
year 2012-13. Since 2008, when a new rice variety,
Pusa 1121 was notified as basmati rice, Iran has been
a major buyer of this rice variety (Sidhu et al., 2014).
The preference of Iran for this variety of basmati rice
was due to its lower cost vis--vis of other varieties
making it affordable for every class of consumer in
Iran.
Rice is not produced in Saudi Arabia and the
country imports a substantial quantity of basmati rice
from India to meet its strong consumer demand. Also,
the Government of Saudi Arabia does not impose any
tariff on rice import providing Saudi traders the
opportunity to re-export the imported rice to the nearby
countries. There has been a shift in export destinations
of Indian basmati rice over the period of six years
(2007-2012).
Trade Directions of Basmati Rice Export from

India
The directions of trade of basmati rice export to
different destinations was examined by estimating the
transitional probability matrix using Markov chain
analysis and the values of transitional probability
matrix (TPM) are presented in Table 5. There are six
major countries that imported Indian basmati rice in
large quantities and rest of the countries were pooled
under others category. The diagonal elements in the
TPM provide information on the probability of
retention of trade, while row elements indicate the
probability of loss in trade on account of competing
countries. The column elements indicate the probability
of gain in trade from the competing countries.
A close look at Table 5 indicated that UAE was
the most stable market among the major importers of
Indian basmati rice, as exhibited by the highest
probability of retention at 0.8226, which means that
UAE had retained its original export share of 82.26
per cent during the period 2000-01 to 2012-13. Iran
and Saudi Arabia imported larger quantities of basmati
rice than UAE; but UAE was the most stable market.
Saudi Arabia had the probability of retention 0.7517,
and it retained its original export share of 75 per cent.
Similarly, Iran had retained its original export share of
72 per cent. This implies that Saudi Arabia and Iran
were also the stable importers of Indian basmati rice.
Kuwait, UK and other countries have shown lower
142
Vol. 29 (No.1)
January-June 2016
Table 4. Share of basmati rice exported from India to major countries: 2007-08 to 2012-13
(Value in crore `)
Year
Saudi
Arabia
Kuwait
UK
USA
UAE
Iran
Iraq
Others
Total
2007-08
2038.35
(46.92)
3102.81
(32.74)
3295.47
(30.26)
3132.94
(27.59)
3380.88
(21.88)
3659.08
(18.85)
401.68
(9.25)
733.94
(7.74)
1030.14
(9.46)
1091.64
(9.61)
1362.92
(8.82)
1059.68
(5.46)
291.08
(6.70)
431.04
(4.55)
195.89
(1.80)
351.77
(3.10)
629.46
(4.07)
849.98
(4.38)
142.10
(3.27)
272.60
(2.88)
153.50
(1.41)
250.32
(2.20)
503.88
(3.26)
561.69
(2.89)
689.83
(15.88)
2786.20
(29.40)
3094.65
(28.42)
2839.76
(25.01)
3432.79
(22.22)
1311.20
(6.76)
20.33
(0.47)
980.76
(10.35)
2053.09
(18.85)
2033.96
(17.91)
2843.21
(18.40)
6463.50
(33.30)
16.81
(0.39)
34.48
(0.36)
35.90
(0.33)
165.31
(1.46)
672.87
(4.36)
1076.67
(5.55)
744.39
(17.13)
1135.22
(11.98)
1030.46
(9.46)
1488.93
(13.11)
2623.60
(16.98)
4427.58
(22.81)
4344.57
(100.00)
9477.05
(100.00)
10889.10
(100.00)
11354.63
(100.00)
15449.61
(100.00)
19409.38
(100.00)
2008-09
2009-10
2010-11
2011-12
2012-13
Note: Figures within the parentheses indicate per cent share of each importing country.
Source: GoI (2013b)
Table 5. Transitional probability matrix of Indian basmati rice export, 2000-01 to 2012-13
Country
Saudi Arabia
Kuwait
UK
UAE
Iran
Iraq
Others
Saudi Arabia
Kuwait
UK
UAE
Iran
Iraq
Others
0.7517
0.0000
0.5327
0.0000
0.0000
0.0000
0.3570
0.1055
0.3831
0.0000
0.0171
0.1005
0.0000
0.0000
0.0498
0.0015
0.4673
0.0000
0.0000
0.0000
0.0680
0.0000
0.1156
0.0000
0.8226
0.0000
0.0000
0.0350
0.0000
0.0000
0.0000
0.1603
0.7209
1.0000
0.0000
0.0000
0.0018
0.0000
0.0000
0.0766
0.0000
0.0000
0.0930
0.4980
0.0000
0.0000
0.1021
0.0000
0.5400
probability of retention, viz. 0.3831, 0.4673 and 0.5400,

respectively. On the contrary, Iraq has shown zero
probability of retention, indicating that Iraq was an
unstable importer of Indian Basmati rice. The major
gainer among the importers of Indian basmati rice over
the study period was Iran which had a transfer
probability of 1.000 from Iraq and 0.1603 from UAE.
In a similar manner, Saudi Arabia gained 53.27 per
cent market share from UK and 35.70 per cent from
others.
In addition to having high probability of retention,
UAE was also likely to gain 11.56 per cent market
share from Kuwait. On the other hand, UAE was likely
to lose 16.03 per cent market share to Iran and 1.71
per cent to Kuwait. Iraq, which is an unstable market
for Indian basmati rice, is likely to gain 7.66 per cent

market share from Iran.
Actual and Estimated Shares of Indian Basmati
Rice Export to Importing Countries
The actual and estimated shares of basmati rice
exported from India to different countries (in
percentage term) were shown in Table 6. A comparison
of this proportion during the study period revealed that
the observed proportions of export shares were
consistent with the estimated shares of export, which
were derived from the Markov chain process. However,
differences have been observed in some years, which
could be due to limitation of the model that the present
estimates depend only on the previous year
227.35
(10.50)
196.11
(10.65)
172.79
(8.38)
137.69
(6.91)
237.00
(8.39)
237.93
(7.82)
306.87
(10.99)
401.68
9.25
733.94
7.74
1030.14
9.46
1091.64
(9.61)
1362.92
(8.82)
1059.68
(5.46)
204.55
(9.44)
188.91
(10.26)
178.64
(8.66)
167.22
(8.39)
260.95
(9.24)
270.85
(8.90)
258.33
(9.25)
382.75
(8.81)
754.74
(7.96)
1001.58
(9.20)
1001.71
(8.82)
1223.27
(7.92)
Kuwait
Actual Estimated
303.70
(14.02)
193.26
(10.49)
205.16
(9.95)
219.39
(11.01)
265.40
(9.40)
222.66
(7.32)
196.23
(7.03)
291.08
6.70
431.04
4.55
195.89
1.80
351.77
(3.10)
629.46
(4.07)
849.98
(4.38)
227.23
(10.49)
165.81
(9.00)
185.61
(9.00)
187.33
(9.40)
243.31
(8.62)
237.77
(7.81)
201.27
(7.21)
298.32
(6.87)
452.63
(4.78)
337.55
(3.10)
440.16
(3.88)
677.06
(4.38)
UK
Actual Estimated
90.53
(4.18)
49.97
(2.71)
84.33
(4.09)
104.88
(5.26)
143.97
(5.10)
159.57
(5.24)
305.21
(10.93)
689.83
15.88
2786.20
29.40
3094.65
28.42
2839.76
25.01
3432.79
(22.22)
1311.20
(6.76)
116.35
(5.37)
75.37
(4.09)
108.46
(5.26)
118.39
(5.94)
166.39
(5.89)
184.79
(6.07)
310.92
11.13
644.91
14.84
2426.01
25.60
2706.14
(24.85)
2523.02
(22.22)
3090.79
(20.01)
UAE
Actual Estimated
Note: Figures within the parentheses indicate percentage to total export of basmati rice.
2012-13
2011-12
2010-11
2009-10
2008-09
2007-08
2006-07
2005-06
2004-05
2003-04
2002-03
2001-02
1098.78
(50.73)
1058.81
(57.49)
1052.21
(51.03)
1067.96
(53.58)
1587.43
(56.22)
1664.42
(54.70)
1240.96
(44.43)
2038.35
(46.92)
3102.81
(32.74)
3295.47
(30.26)
3132.94
(27.59)
3380.88
(21.88)
3659.08
(18.85)
2000-01
1146.86
(52.95)
1017.18
(55.23)
1095.22
(53.11)
1084.95
(54.44)
1544.49
(54.69)
1635.28
(53.74)
1286.16
(46.05)
2003.87
(46.12)
3064.76
(32.34)
3004.40
(27.59)
3163.51
(27.86)
3993.47
(25.85)
Saudi Arabia
Actual Estimated
Year
0.00
(0.00)
12.32
(0.67)
0.11
(0.01)
0.26
(0.01)
2.44
(0.09)
13.98
(0.46)
46.17
(1.65)
20.33
(0.47)
980.76
(10.35)
2053.09
(18.85)
2033.96
(17.91)
2843.21
(18.40)
6463.50
(33.30)
14.51
(0.67)
16.89
(0.92)
15.11
(0.73)
17.00
(0.85)
24.89
(0.88)
36.63
(1.20)
82.85
(2.97)
142.04
(3.27)
1188.07
(12.54)
2011.93
(18.48)
2086.70
(18.38)
3272.67
(21.18)
Iran
Actual Estimated
Table 6. Actual and estimated shares of Indian basmati rice export from India, 2000-01 to 2012-13
0.00
(0.00)
0.00
(0.00)
1.51
(0.07)
0.00
(0.00)
0.05
(0.00)
0.98
(0.03)
0.65
(0.02)
16.81
(0.39)
34.48
(0.36)
35.90
(0.33)
165.31
(1.46)
672.87
(4.36)
1076.67
(5.55)
0.40
(0.02)
1.29
(0.07)
0.31
(0.02)
0.26
(0.01)
0.60
(0.02)
1.49
(0.05)
4.07
(0.15)
2.26
(0.05)
76.39
(0.81)
159.02
(1.46)
157.66
(1.39)
220.10
(1.42)
Iraq
Actual Estimated
445.58
(20.57)
331.29
(17.99)
546.00
(26.48)
462.85
(23.22)
587.56
(20.81)
743.52
(24.43)
696.71
(24.95)
886.49
(20.40)
1407.82
(14.86)
1183.96
(10.87)
1739.25
(15.32)
3127.48
(20.24)
4989.27
(25.71)
456.04
(21.06)
376.31
(20.43)
478.77
(23.22)
417.87
(20.97)
583.22
(20.65)
676.24
(22.22)
649.18
(23.24)
870.42
(20.03)
1514.46
(15.98)
1668.49
(15.32)
1981.87
(17.45)
2972.24
(19.24)
Others
Actual Estimated
2165.94
(100.00)
1841.76
(100.00)
2062.11
100.00
1993.03
(100.00)
2823.85
(100.00)
3043.06
(100.00)
2792.80
(100.00)
4344.57
(100.00)
9477.05
(100.00)
10889.10
(100.00)
11354.63
(100.00)
15449.61
(100.00)
19409.38
(100.00)
Total
Actual
(in crore `)

143
144
Vol. 29 (No.1)
January-June 2016
Table 7. Projected exports of Indian basmati rice to major importing countries: 2013-14 to 2017-18
(Value in crore `)
Year
2013-14
2014-15
2015-16
2016-17
2017-18
Saudi Arabia
Kuwait
UK
UAE
Iran
Iraq
Other countsries
4984.81
(25.68)
5744.79
(29.60)
6290.35
(32.41)
6734.15
(34.70)
7105.83
(36.61)
1463.86
(7.54)
1707.68
(8.80)
1787.80
(9.21)
1806.46
(9.31)
1804.72
(9.30)
920.09
(4.74)
967.24
(4.98)
1017.77
(5.24)
1069.93
(5.51)
1118.37
(5.76)
1375.72
(7.09)
1448.66
(7.46)
1531.84
(7.89)
1610.20
(8.30)
1677.81
(8.64)
5946.15
(30.64)
5003.67
(25.78)
4297.05
(22.14)
3729.27
(19.21)
3278.57
(16.89)
496.79
(2.56)
457.88
(2.36)
386.14
(1.99)
332.17
(1.71)
288.73
(1.49)
4221.97
(21.75)
4079.46
(21.02)
4098.43
(21.12)
4127.20
(21.26)
4135.36
(21.30)
Note: Figures within the parentheses indicate per cent to respective values
Source: GoI (2013b)
observations and the exports also depend on sudden

policy changes, leading to abrupt increase or decrease
in exports to a country.
Projection of Indian Basmati Rice Export to Major
Importing Countries
Table 7 shows the export of Indian basmati rice to
different countries which was computed using the
transitional probability matrix. It was projected that
during 2013-14, the major markets for Indian basmati
rice would be Iran (30.64 %), Saudi Arabia (25.68 %)
and others (21.75 %). The projected exports to Saudi
Arabia have exhibited an increasing trend in both
absolute value and percentage to total export. The
reason behind increasing projected exports to Saudi
Arabia is that it is likely to gain 53.27 per cent market
share from UK and 35.70 per cent from others.
Similarly, the projected value of basmati export to
Kuwait has shown an increasing trend. The projected
market share is likely to increase marginally from 7.54
per cent to 9.30 per cent from 2013-14 to 2017-18. In
the case of Iran and Iraq, the projected value has shown
a decreasing trend, in both absolute and relative to total
export from India. The reason for showing a decreasing
trend by Iraq was that it had totally lost its market share
to Iran, although it had gained small market shares from
Iran and Kuwait. The projected exports to Iran have
shown a decline from ` 5946 crore in 2013-14 to
` 3278 crore in 2017-18, i.e. 30.64 per cent to 16.89
per cent during the foresaid period.
Determinants of Rice Export from India

Difference in comparative advantage, geographical
and political proximity and degree of trade barriers are
the major determinants of destinations for exports of
any food commodity (Kumar et al., 2008). The factors
determining rice export from India were studied by
using multiple regression of log-log form where
quantity exported was regressed against export price,
exchange rate, international price, domestic
consumption and lagged production. The data used for
regression analysis were for 33 years (1980-81 to 201213). Since all variables are stationary at I (1), as shown
in Annexure II, OLS estimation was employed and
coefficients are given in Table 8.
Table 8. Estimates of determinants of rice export from
India
Variable
Constant
D (Export price)
D (Exchange rate)
D (International price)
D (Domestic consumption)
D (Lagged production)
Coefficient
0.15**
-2.00***
1.08ns
1.20***
-0.95ns
0.53ns
Standard
error
0.06
0.21
0.64
0.25
0.64
0.64
R-squared=0.8153, Adjusted R-squared=0.7798, DW=2.44, F-statistics=22.958, Prob(F-statistics)=0.0000

Note: D=Difference; ** and *** indicate significance at 5
per cent and 1 per cent levels, respectively. ns= Nonsignificant
145
Table 9. Estimates of determinants of rice export from India by model I and model II
Dependent variable: D (Quantity)
Model I
Coefficient
Constant
D (Export price)
D (Exchange rate)
D (International price)
D (Domestic consumption)
D (Lagged production)
Std. error
0.16**
0.06
-2.02***
0.20
1.21*
0.62
1.23***
0.24
-1.30**
0.47
R-squared=0.8104
Adjusted R-squared=0.7823
D-W=2.37
F-statistics=28.854
Model II
Coefficient
Std. error
0.13**
0.06
-1.98***
0.21
1.02ns
0.66
1.15***
0.25
1.16**
0.49
R-squared=0.7994
Adjusted R-squared=0.7697
D-W=2.55
F-statistics=26.9033
Prob (F-statistics) = 0.0000
Note: D=Difference ;*, **, and *** indicate level of significance at 10 per cent, 5 per cent and 1 per cent level respectively.
Table 8 shows that export price and international

price are significant at 1 per cent level of significance.
However, other three variables, viz. exchange rate,
domestic production, and lagged production, are nonsignificant. So, domestic consumption and lagged
production were treated separately and two models
were formulated.
In model I, quantity exported was the dependent
variable and export price, exchange rate, international
price and domestic consumption were independent
variables. In model II, export price, exchange rate,
international price and lagged production were
independent variables and the results are shown in
Table 9.
CUSUM Test and CUSUM of Square Test
The CUSUM test is based on the cumulative sum
of the recursive residuals. The test finds parameter
instability, if the cumulative sum goes outside the area
between the two critical lines. Both these tests have
suggested stability of parameters, as revealed by Figure
2 (a-d).
Finally, both the functions were tested by various
econometric tests to achieve best, linear and unbiased
coefficient. The results of various tests are depicted in
Table 10.
Both the models have shown that export of rice
from India is highly price sensitive. The negative and
significant coefficient of Indian export price indicates
that it has a negative impact on exports of rice. In order

to grab international markets, Indian export price has
to be competitive. Increase in export price means that
exports become expensive in the world market. Other
empirical studies drawing similar results were of
Abolagba et al. (2010), Bilal and Rizvi, (2013) and
Shende et al. (1999). However, Kumar et al. (2008)
have found a positive relationship between Indian
export price on export of cucumber and gherkin
products from India, but the result was non-significant,
indicating that export price has not played any
significant role in its export.
The international price of rice is significant with
high positive elasticity, which indicates that
international price has a positive impact on exports of
rice from India. This may be due to the fact that when
price of Thai rice increases, the importer shifts towards
Indian rice due to price advantage, indicating the role
of government to set price just below Thai 5%, in such
a way that price becomes competitive in the
international markets. The exchange rate has a positive
impact on rice export from India and is significant at
10 per cent level (model I). When the exchange rate of
rupee with dollar increases (devaluation of rupee
occurs), then exports from India become cheaper for
the importing countries, and hence export increases.
Thus, a positive impact of exchange rate is assumed
on rice exports from India. Haleem et al. (2005), Kumar
et al. (2008) and Shende et al. (1999) have also revealed
the positive and significant effect of exchange rate.
146
(a)
CUSUM
January-June 2016
(b)
CUSUM
5% significance
(c)
CUSUM
Vol. 29 (No.1)
5% significance
10% significance
(d)
CUSUM
10% significance
Figure 2. CUSUM test and CUSUM of square test, (a) CUSUM 5% significance model I , (b) CUSUM 10%
significance Model II, (c) CUSUM 5% significance, and (d) CUSUM 10% significance
The domestic consumption of rice shows a negative

and significant effect on its export from India. When
domestic consumption increases, smaller surpluses are
available for exports, and ultimately its export
decreases. Abolagba et al. (2010) and Bilal and Rizvi
(2013) have endorsed the same view. But, in recent
years, a drastic change has been seen in the food habits
of people. The consumption of cereals is decreasing in
India because of increasing preferences for processed
foods, vegetables and high protein food items (Chopra
et al., 2014). The coefficient of lagged production has
been found positive and statistically significant at 5
per cent level of significance. This means that rice
export is more influenced by the previous years

production. The positive and significant influence of
lagged production has been seen not only on export of
rice (Gangwar and Rai, 1995 ; Sekhar, 2003) but also
on export of onion, banana and black pepper ( Hema
and Kumar, 2007), indicating increase in export
earnings with increase in production.
Conclusions
Rice contributes substantially to the national
income through exports of its basmati as well as nonbasmati rice varieties. The study has revealed that
Indian rice exports had a fabulous performance during
147
Table 10. Econometric test for assumptions of Model I and Model II

Hypothesis
Ho= The variance of error-terms are constant

H1= The variance of error terms are not constant
Ho= There is no serial correlation
H1= There is a serial correlation
Ho= The error-term are normally distributed.

H1= The error-terms are not normally distributed.
Ho= The variance of error-terms are constant
H1= The variance of error-terms are not constant
Ho= There is no serial correlation
H1= There is a serial correlation
Ho= The error-term are normally distributed.

H1= The error-terms are not normally distributed.
Name of test
Fstatistics
Observed
R- squared
R-square
Prob.
chi-square
for observed
Decision
1.677
0.7948
The variance of
error-terms
are constant
3.5066
0.1732
No serial
correlation
1.5847ns
0.457
Error-terms are
normally
distributed
1.116
0.8917
The variance of
error-terms
is constant
3.746
0.1536
No serial
correlation
0.595
Error-terms are
normally
distributed
Model I
Breusch0.3733ns
PaganGodfrey
Breusch1.538ns
Godfrey Serial
Correlation
LM Test
JarqueBera
Model II
Breusch0.2439ns
PaganGodfrey
Breusch1.657ns
Godfrey Serial
Correlation
LM Test
JarqueBera
1.038ns
Note: NS=Non-significant
the study period 1980-81 to 2012-13. It has registered

positive growth rates in terms of quantity, value and
unit value for basmati rice and total rice. The instability
index has been found higher for export quantity (43.37
%) of basmati rice than its value (33.52 %) and unit
value (20.41 %). India faces competition in export of
non-basmati rice from Thailand and Vietnam in
international market due to better quality of rice from
these countries, and from Pakistan for basmati rice.
The study has also revealed that export of rice from
India is highly price sensitive. Export price,
international price, lagged production, domestic
consumption, and exchange rate are the major
determinants of rice export from India. In order to
sustain in the international market, Indian export price
needs to be competitive besides meeting quality and
sanitary and phyto-sanitary standards.
Acknowledgements
The authors are grateful to the anonymous referee
for the valuable suggestions. The views expressed are
those of the authors only.
References
Abolagba, E.O., Onyekwere, N.C., Abonkpolor, B.N. and
Umar, H.Y. ( 2010) Determinants of agricultural
exports. Journal of Humane Ecology, 29(3): 181-184.
All India Rice Exporters Association (2013) Retrived from
www.airea.net.
Bansil, P.C. (1992) Agricultural Statistical Compendium,
Vol. I, Food grains, Part II. Techno Economic Research
Institute, New Delhi.
Bilal, M. and Rizvi, S.B.H. (2013) Determinants of rice
exports: An empirical analysis of Pakistan. Journal of
Global and Scientific Issues, 1(4): 5-16.
148
Bhattacharyya, B. and Pal, P. (2000) Food security in India

in the context of Agreement on Agriculture.
Proceedings World Trade Organisation vis-a-vis
Punjab Agriculture, May 27, pp: 35-42, Punjab
Agricultural, University, Ludhiana.
Chandrashekhar, G. (2013) India returns to world rice market
with a bang. The Hindu, Business Line, May 9.
Chopra, Shruti, Toor, M.S. and Vatta, K. (2014) Pattern of
changing consumption in India: An analysis.
Indian journal of Agricultural Development and Policy,
24(2): 51-58.
Dent, W.T. (1967) Application of Markov analysis to
international wool flows. Review of Economics and
Statistics, 49(2): 613-616.
Fisher, W.D. (1967) Note on curve fitting with minimum
deviations by linear programming. Journal of the
American Statistical Association, 56: 359-363.
Gangwar, A.C. and Rai, K.N. (1995) Emerging agricultural
export opportunities in India. The Bihar Journal of
Agricultural Marketing, 3(4): 370-377.
Gillet, B.E. (1976) Introduction of Operation Research
A Computer Oriented Algorithmic Approach. McGrawHill Inc., New York, USA.
GoI (Government of India) (2013a) Agricultural Statistics
at a Glance 2013. Ministry of Agricultural, New Delhi.
GoI (Government of India) (2013b) Agricultural and
Processed Food Products Export Development
Authority (APEDA). Ministry of Commerce and
Industry, New Delhi. Retrieved from http://
agriexchange.apeda.gov.in.
Haleem, U., Mushtaq, K., Abbas, A. and Shiekh, A.D.
(2005) Estimation of export supply function for citrus
fruit in Pakistan. The Pakistan Development Review,
44 (Part-II): 659-672.
Hema, M. and Kumar, R. (2007) Production and trade
performance of major spices grown in western ghat
region in India. Indian Journal of Agricultural
Marketing, 21(1): 83-95.
Kumar, N.R., Rai, A.B. and Rai, M. (2008) Export of
cucumber and gherkin from India: Performance,
destinations, competitiveness and determinants.
Agricultural Economics Research Review, 21(1): 130138.
Lee, T.C., Judge, G.G. and Zeellenner, A. (1970) Estimating
the Parameters of the Probability Model from Aggregate
Vol. 29 (No.1)
January-June 2016
Time Series Data. North-Holland Publishing Company,

Amsterdam.
Mackinnon, J.G. (1996) Numerical distribution functions
for unit root and co integration tests. Journal of Applied
Economics,11: 601-618.
RBI (Reserve Bank of India) (2013) Hand Book of Indian
Economy, 2012-13. Retrived from: http://
r b i d o c s . r b i . o rg . i n / r d o c s / P u b l i c a t i o n s / P D F s /
FHB160913FLS.pdf.
Sekhar, C.S.C. (2003) Agricultural trade liberalization
Likely implication for rice sector in India. Indian
Journal of Agricultural Economics, 58: 42-63.
Shende, N.V., Deoghare, B.L. and Shende, T. K. (1998) Rice
exports: Present status and future prospects. Indian
Journal of Agricultural Marketing, 12(1&2): 3-5.
Shende, N.V., Bhole, B.D. and Shende, P.V. (1999) Export
performance of India in tea, coffee and tobacco. Indian
Journal of Agricultural Marketing, 13: 78-81.
Sidhu, J.S., Singh, J. and Kumar, R. (2014) Role of market
intelligence in agriculture: A success story of basmati
cultivation in Punjab. Indian Journal of Economics, 10:
26-31.
Singh, A.J. and Byerlee, D. (1990) Relative variability in
wheat yields across countries and over time. Journal
of Agricultural Economics, 14(1): 21-32.
Singh, K. (2001) Performance and Prospects of Agricultural
Exports of India. Ph.D dissertation, Punjab Agricultural
University, Ludhiana, Punjab.
USDA (United States Department of Agriculture) (2013)
Rice Year Book, various issues. Retrieved from
www.usda.gov
Wailes, E.J. and Chavez, E.D. (2012) World Rice Outlook.
International rice baseline with deterministic and
stochastic projections, 2012-2021. Department of
Agricultural Economics and Agribusiness, Division of
Agriculture, University of Arkansas, Fayetteville,
AR72701.
Wagner, H.H. (1959) Linear programming for regression
analysis. Journal of the American
St a t i s t i c a l
Association, 54: 206-12.
www.indiastat.com
www.siteresources.worldbank.org/INTPROSPECTS/.../
PINK_DATA.xlsx World Bank Commodity Price
Data (Pink Sheet)
Received: March, 2015; Accepted: February, 2016
149
Annexure 1. World rice trade showing major exporters and importers: 1990-2013
(000 tonnes)
Country
India
Vietnam
Thailand
Pakistan
USA
Others
World Total
China
Nigeria
Iran
Phillipines
Indonesia
Others
World Total
1990
1995
0.51
(4.36)
1.67
(14.27)
3.94
(33.68)
0.90
(7.69)
2.42
(20.68)
2.25
(19.23)
11.70
(100.00)
4.18
(20.10)
2.32
(11.15)
5.89
(28.32)
1.59
(7.64)
2.99
(14.38)
3.83
(18.41)
20.80
(100.00)
0.06
(0.51)
0.22
(1.88)
0.87
(7.44)
0.54
(4.62)
0.08
(0.68)
9.94
(84.96)
11.70
(100.00)
1.96
(9.42)
0.45
(2.16)
1.58
(7.60)
0.28
(1.35)
3.01
(14.47)
13.52
(65.00)
20.80
(100.00)
2000
2005
Exporters
1.45
4.69
(6.37)
(16.20)
3.37
5.17
(14.79)
(17.86)
6.55
7.27
(28.75)
(25.11)
2.03
3.03
(8.91)
(10.47)
2.85
3.86
(12.51)
(13.33)
6.54
4.92
(28.71)
(16.99)
22.78
28.95
(100.00)
(100.00)
Importers
0.28
0.61
(1.23)
(2.11)
1.25
1.78
(5.49)
(6.15)
1.10
1.20
(4.83)
(4.15)
0.90
1.89
(3.95)
(6.53)
1.50
0.50
(6.58)
(1.73)
17.75
22.98
(77.92)
(79.38)
22.78
28.95
(100.00)
(100.00)
Note: Figures within the parentheses are percentage of the total.

Source: USDA (2013)
2010
2011
2012
2013
2.23
(7.01)
6.73
(21.17)
9.05
(28.47)
4.00
(12.58)
3.87
(12.17)
5.91
(18.59)
31.79
(100.00)
4.64
(12.70)
7.00
(19.16)
10.65
(29.15)
3.41
(9.33)
3.25
(8.89)
7.59
(20.77)
36.54
(100.00)
10.25
(25.73)
7.72
(19.38)
6.95
(17.45)
3.50
(8.79)
3.33
(8.36)
8.09
(20.31)
39.83
(100.00)
10.50
(27.16)
6.80
(17.59)
6.70
(17.33)
3.50
(9.05)
3.27
(8.46)
7.89
(20.41)
38.66
(100.00)
0.37
(1.16)
2.00
(6.29)
1.52
(4.78)
2.40
(7.55)
1.15
(3.62)
24.35
(76.60)
31.79
(100.00)
0.58
(1.59)
2.55
(6.98)
1.87
(5.12)
1.20
(3.28)
3.10
(8.48)
27.25
(74.58)
36.54
(100.00)
2.90
(7.28)
3.40
(8.54)
1.75
(4.39)
1.50
(3.77)
1.96
(4.92)
28.32
(71.10)
39.83
(100.00)
3.20
(8.28)
2.60
(6.73)
2.15
(5.56)
1.10
(2.85)
0.65
(1.68)
28.96
(74.91)
38.66
(100.00)
150
Vol. 29 (No.1)
January-June 2016
Annexure II. Stationary (unit root) test for variables

Variable
Linear graph
QT
Intercept
Trend and intercept
Intercept
Trend and intercept
Intercept
Trend and intercept
Intercept
Trend and intercept
Intercept
Trend and intercept
Intercept
Trend and intercept
EXC
EXP
INT
DC
LGP
At level
Test statistic
Critical values
(99%)
(95%)
-0.91
-3.21
-2.44
-0.93
-1.80
-3.36
-1.27
-2.01
-3.91**
-3.30
-1.53
-4.85**
-3.65
-4.28
-3.65
-4.27
-3.66
-4.28
-3.65
-4.27
-3.67
-4.27
-3.66
-4.27
-2.95
-3.56
-2.96
-3.56
-2.96
-3.56
-2.96
-3.56
-2.96
-3.56
-2.96
-3.55
At first difference
Test statistic
Critical values
(99%)
(95%)
-6.20**
-6.24**
-3.94**
-4.20**
-4.91**
-4.83**
-4.75**
-5.12**
-6.85**
-5.12**
-9.92**
-9.81**
-3.66
-4.28
-3.66
-4.28
-3.67
-4.29
-3.66
-4.28
-3.67
-4.29
-3.66
4.28
-2.96
-3.56
-2.96
-3.56
-2.96
-3.57
-2.96
-3.56
-2.96
-3.57
-2.96
3.56
Notes: Critical values have been taken from Mackinnon (1996)

All variables are in log form
Optimum lag selection is 8 on basis of Schwartz Information Criterion (SIC) default set by E-Views.
** denotes significance at 1 per cent level.
Decision
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%
I(1) at 1%

DOI: 10.5958/0974-0279.2016.00027.6
Research Note
Crop Diversification in Gadag District of Karnataka

N.D. Basavaraja, T.M. Gajananab* and M. Satishkumara
Department of Agricultural Economics, Gandhi Krishi Vigyan Kendra,
University of Agricultural Sciences, Bengaluru - 560 065, Karnataka
b
ICAR-Indian Institute of Horticultural Research, Bengaluru -560 089, Karnataka
a
Abstract
The nature, extent and determinants of crop diversification have been analysed in the Gadag district of
Karnataka over space and time. The area under vegetables, fruits and pulse crops has registered a higher
(7-11%) growth compared to cereals, oilseeds, fibre and other crop groups (<1%). Over the years, the
share of cereal crop groups has decreased significantly from 32.53 per cent to 28.81 per cent and that of
fruits and vegetables has increased considerably from 0.10 per cent to 0.25 per cent for fruits and from
4.66 per cent to 7.80 per cent for vegetable crops. The transition probability matrix has indicated that
over the years, horticultural crop groups have retained a higher share (92%) in terms of area under crops.
The Northern Dry Zone (Gadag taluk) has been found to be more diversified with Entropy index of 0.99
and household crop richness of 3.20 compared to 0.55 and 1.90 in the Northern Transitional Zone (Shirahatti
taluk), respectively. The major factors influencing crop diversification have been identified as size of
landholding, gross irrigated area, and net return realised per farm.
Key words: Crop diversification, horticulture, entropy, crop richness, probability matrix, Karnataka
JEL Classification: Q16, R11
Introduction
In India, the agricultural sector contributes 13.7
per cent to the national gross domestic product (GDP).
This sector also provides employment to 52 per cent
of the total work force of the country. In dry land
agriculture, diversification serves as a sole source of
combating risk against climate and weather vagaries.
Crop diversification in India is generally viewed as a
shift from the traditionally grown less-remunerative
crops to more remunerative crops. The crop
diversification ensures security for food, nutrition,
income and employment to a wider section of the
society and hence, has a significant bearing on GDP
Email: tmgajanana@yahoo.com, gajanana@iihr.res.in
This paper forms part of the M.Sc. thesis submitted by the
first author to the University of Agricultural Sciences,
Bengaluru, under the guidance of the second author.
of the nation. Gopalappa (1996), based on his study in

Andhra Pradesh, reported that there was scope to
increase income through crop diversification. Acharya
et al. (2011) have reported that crop diversification
contributes to increased cropping intensity, higher
employment, commercialization of farming, reduction
in migration of male members and involvement of
women in income-generating activities. However, the
study has analysed crop diversification at the state level
and there is a need to have micro level evidences of
crop diversification.
The present study has analysed the nature and
extent of crop diversification at the micro level under
two distinct agro-climatic conditions of the state. It has
also identified the factors that influence crop
diversification in this area. It was hypothesised that
crop diversification is more in the northern dry zone
(NDZ) than in the northern transitional zone (NTZ).
152

To analyse the nature and extent of diversification,
Gadag, which is one of the districts in Karnataka that
are undergoing diversification, was selected. From the
district, two taluks belonging to different zones were
selected. The primary data were collected from 30
sample farmers from each of the selected taluks
representing different zones such as northern dry zone
(NDZ) and northern transitional zone (NTZ).
The district Gadag was formed (carving out of
Dharwad district) in 1997 and the secondary data
pertaining to area under important crop groups were
obtained for the period from 1998-99 to 2011-12 from
the Directorate of Economics and Statistics and
Directorate of Horticulture, Government of Karnataka.
The primary data on socio-economic characteristics,
cropping pattern, factors influencing diversification,
etc. were collected from the sample farmers of Gadag
(NDZ) and Shirahatti (NTZ) taluks for the agricultural
year 2012-13. The growth in area under different crop
groups in Gadag and Shirahatti taluks of Gadag district
were analysed through the standard compound growth
function. Different indices, viz. Herfindahl index,
Simpson Index, Margaleff Index and Entropy Index
were used for assessing crop diversification.
Herfindahl Index (HI) It is the sum of square of
the proportion of acreage under each crop to the total
cropped area and is given by Equation (1):
(1)
where, Pi represents acreage proportion of the ith crop
in total cropped area.
As diversification increases, the sum of square of
the proportion of activities decreases and so also the
indices (HI). The Herfindahl index takes the value of
one when there is specialization and approaches zero
when there is diversification. Since the index measures
concentration; it is transformed by subtracting from
one, i.e., 1 HI. The transformed value of HI avoids
confusion on comparing it with other indices.
Simpson Index It is the most suitable index for
measuring diversification of crops in a particular
geographical region and is calculated by Equation (2):
Simpson Index (SI) = 1 Pi2
(2)
where, Pi = Ai / Ai is the proportion of the ith activity

in acreage. If SI is near zero, it indicates that the zone
Vol. 29 (No.1)
January-June 2016
or region is near to the specialization in growing of a

particular crop and if it is close to one, then the zone is
fully diversified in terms of crops.
Magalef Index (MI) It is the number of crops
normalized for area planted and has a lower limit of
zero when only one variety is grown. The MI is
constructed as per expression (3):
CD = S-1/ln A
(3)
where, CD is the crop diversity maintained by a farmer;

A is the total area planted to all crops by a farmer; S is
the total number of crops maintained by the farmer.
Entropy Index (EI) It is a direct measure of
diversification having a logarithmic character and is
given by Equation (4):
(4)
where, P i represents acreage proportion of the ith crop
in total cropped area.
The Entropy index increases with diversification.
The Entropy index approaches zero when the farm is
specialized and Pi equals one (perfect specialization)
and takes a maximum value when there is perfect
diversification. The upper limit of Entropy Index is
determined by the base of logarithms and the number
of crops. The upper value of the index can exceed one,
when the number of crops is higher than the value of
the logarithms base, and it is less than one when the
number of crops is lower than the base of logarithm.
Markov Chain Analysis To assess the dynamism
in direction of area under crops during 1998 to 2012,
transitional probabilities were calculated based on
linear programming (LP) approach using LINDO
software. To know the shift in cropping pattern,
different crop groups like cereals, pulses, sugar crops,
oilseeds, fibre and horticulture crops were considered.
Markov chain analysis develops a transitional
probability matrix P, whose elements Pij indicate the
probability (share) of crop group switching from the
ith crop group to the jth crop group over time. Its diagonal
elements represent retention share of respective crop
group in terms of area under crops. This can be
algebraically expressed as Equation (5):
Ejt = [Eit -1] Pij + ejt
i=1,...,n
(5)
Basavaraj et al. : Crop Diversification in Gadag District of Karnataka
where,
Ejt = Area under crop to the jth crop group in year t
Eit-1 = Area under crop of ith crop group during the
year t-1
Pij = The probability of shift in area under ith crop
group to jth crop group
ejt = The error-term statistically independent of
Eit-1, and
n = The number of crop groups.
The transitional probabilities Pij arranged in (m n)
matrix have the following properties:
Pij = 1 and 0 Pij 1
153
The per cent position of each rank was converted

to scores by referring to the tables given by Garret and
Woodworth (1969). Then, for each factor, the scores
of individual respondents were summed up and divided
by the total number of respondents for whom scores
were gathered. The mean scores for all the factors were
ranked, following the decision criterion, that is higher
the score, more important is the benefit and problem
to farmers.
Multiple Linear Regression Analysis It was used
to examine the factors influencing crop diversification
in the study area, through Equation (7):
Y=a0+a1X1+a2X2+a3X3+u
i=1,...,n
The transitional probability matrix (T) based on
LP framework is estimated using Minimization of
Mean Absolute Deviation (MAD).
Min, OP* + I e
(7)
where, Y is the crop diversification (Index or crop

richness), X1 is the size of holding (acre), X2 is the
income of family (`), X3 is the irrigated land (acre), u
is the random error-term and a0 is the constant term.
ST
Growth in Area under Crop Groups in Gadag

District
X P* + V = Y
GP* = 1
P* > 0
where, P* is the transitional probability matrix, 0 is
the zero vector, I is an appropriately dimensional
vector of areas, and e is the vector of absolute errors.
Garret Ranking Technique In this study, Garret
ranking technique was used to assess the benefits of
practising crop diversification and constraints in crop
diversification.
The respondents were asked to rank (in the order
of severity) the factors- benefits and constraints and
these ranks were converted to scores by referring to
Garret table.
The growth in area under different crop groups over

the period of 14 years from 1998-99 to 2011-12 was
analyzed using the exponential growth function. The
results (Table1) showed a positive growth in area
0.43 per cent in cereals, 6.58 per cent in pulses, 22.84
per cent in sugar crops and 0.53 per cent for spice crops
Table 1. Growth in area of different crop groups in the
Gadag district, 1998-99 to 2011-12
(in per cent)
Crop group
Gadag
taluk
Shirahatti
taluk
Gadag
district
Rij= Rank given for the ith item by the jth individual,
and
Cereals
Pulses
Sugar crops
Spices
Fruits
Vegetables
Oilseeds
Fibres
Medicinal and aromatic
Flowers
-0.967
5.63*
10.09
-0.471
7.82*
4.00*
-1.01
-1.72
25.57*
-4.19
0.72
5.72*
21.37*
-3.911
5.17*
0.94
-0.057
0.58
-17.96
0.78
0.43
6.58*
22.84*
0.53
10.67*
6.7*
-0.91
-0.55
-10.03
-1.18
Nj=Number of items ranked by the jth individual
Note: * denotes significance at 5 per cent level.
The order of merit given by the respondents was

changed into ranks by using formula, (6):
(6)
where,
154
Vol. 29 (No.1)
January-June 2016
Table 2. Changes in share of crop groups in total cropped area in Gadag district, TE 2000-01 to 2011-12
(Per cent)
Period
TE 2000-01
TE 2003-04
TE 2006-07
TE 2009-10
TE 2011-12
Cereals
Pulses
Sugar
Spices
32.53
30.96
26.63
28.23
28.81
16.39
18.98
25.87
25.31
26.89
0.04
0.10
0.02
0.10
0.90
3.48
3.28
3.23
3.07
3.15
Crop group
Fruits Vegetables Oilseeds
Fibres
0.10
0.16
0.17
0.23
0.25
13.87
11.83
9.13
9.66
12.73
annually. The area under fruits and vegetables increased

at the rate of 10.67 per cent and 6.7 per cent per annum,
respectively. But, in the case of oilseeds, fibre crops,
medicinal and aromatic and flower crops, the growth
rate during 1998-99 to 2011-12 was negative with the
values at level being -0.91 per cent, -0.55 per cent,
-10.03 per cent and -1.18 per cent per annum,
respectively.
At the taluk level also, the area under commercial
crops like sugar cane and horticultural crops increased.
Among the two taluks under study, the growth in area
under fruits, vegetables and medicinal & aromatic crops
was found to be more in Gadag taluk than in Shirahatti
taluk.
Changes in Cropping Pattern in Gadag District
The changes in the cropping pattern for the period
1998-2012 were analysed for the Gadag district. The
results (Table 2) indicated that the share in area of
cereals, oilseeds and fibre crops had a decline in Gadag
district. The area share of vegetables, pulses and fruits
increased during the study period. Over the years, the
area share of cereal crops group decreased significantly
from 32.53 per cent in TE 2003-04 to 28.81 per cent in
TE 2011-12. However, cereals still occupied the major
share of the cropped area, indicating the farmers
awareness about their value in food security besides
livelihood security. For the pulse crops, the area share
increased from 16.39 to 26.89 per cent. It may be due
to the awareness among the farmers about protein
requirement (nutritional security) and its availability
from pulses. It is a healthy sign as India relies heavily
on the imports of pulses to meet its domestic
requirements. It is also to be noted that in order to
4.66
5.51
6.00
7.87
7.80
28.80
29.09
28.85
25.47
19.41
Medicinal
Flowers
and aromatic
crops
0.06
0.02
0.01
0.01
0.02
0.07
0.07
0.08
0.06
0.05
address the supply side constraints and to enhance the

seed replacement rates, the Government of India has
launched the National Food Security MissionPulses
(NFSM -pulses) and Accelerated Pulses Production
Program (A3P) in the major pulses-growing states of
Madhya Pradesh, Rajasthan, Maharashtra, Uttar
Pradesh and Karnataka (Rimal et al., 2015). In
Karnataka, both the programs have been in existence
since 2007-08 and the accelerated pulses production
program (A3P) was launched with a target of 38000
ha, under pulses in the districts of Gadag, Raichur,
Bijapur, Bidar and Dharwad to increase the productivity
of tur, gram, horsegram, blackgram, greengram,
cowpea and avare (Anon, 2010). As a result, there has
been an increase in production of pulses in these
districts which also includes Gadag (Venkatesh, 2013).
The positive impact of A3P on increased pulses
productivity and production was also observed in other
pulses- growing major states of Madhya Pradesh,
Rajasthan, Maharashtra, Uttar Pradesh and Karnataka
(Rimal et al., 2015).
The area share of medicinal & aromatic crops, fibre
crops, flower crops and spice crops was less in the total
cropped area (except spices) and decreased over the
years. In the case of fruits and vegetables, their share
to the total cropped area increased significantly from
0.10 per cent and 4.66 per cent to 0.25 per cent and
7.80 per cent, respectively. This clearly reflects the role
of high-value horticultural crops and the emphasis laid
on this sector by way of promotion of this sector by
the Horticulture Department through implementation
of its schemes like Comprehensive Horticulture
Development Programme (CHDP), National
Horticulture Mission (NHM), etc.
155
Table 3. Transitional probability matrix for area under different crop groups in Gadag district, 1998-99 to 2011-12
Crop group
Cereals
Pulses
Oilseeds
Fibres
Horticulture
Cereals
Pulses
Oilseeds
Fibres
Horticulture
0.48
0.37
0.00
0.56
0.00
0.00
0.49
0.10
0.00
0.06
0.32
0.00
0.46
0.00
0.02
0.19
0.03
0.00
0.44
0.00
0.00
0.11
0.44
0.00
0.92
Transitional Probability Matrix for Area under

Different Crop Groups in Gadag District
The change in the area under crops in the Gadag
district was analysed using Markov chain, capturing
the shift in the area under crops over a period of 14
years. Cereals, pulses, oilseeds, fibre and horticultural
crop groups were considered for the analysis. In the
horticultural crop group, vegetables, fruits, flowers,
medicinal & aromatic and spice crops were included.
The results (Table 3) revealed that among the five crop
groups, horticulture group retained a higher share
(92%) over the years and it gained even the meagre
shares from fibre and oilseed groups, and this indicates
that over the years, farmers are shifting towards
horticultural crops group. The cereals and pulses groups
retained a higher share compared with fibre and
oilseeds crop groups, but lesser compared to
horticultural crops. Fibre and oilseed groups retained
only a meagre share over the years. The analysis
indicates an overall shift towards horticultural crops
at the cost of other crops.
Crop Diversification Index
Table 4 shows the average values of Simpson Index
for different crop groups in Gadag district, Gadag taluk
and Shirahatti taluk for 14 years (1998-2012). The
calculated average values of Simpson Index for
different crop groups were 0.77 for Gadag district, 0.80
Table 4. Indices for area under different crop groups of
Gadag district, 1998-99 to 2011-12
Indices
Gadag
taluk
Shirahatti
taluk
Gadag
district
Simpson Index
Entropy Index
Herfindahl Index
0.80*
1.71*
0.20*
0.75
1.56
0.25
0.77
1.59
0.23
for Gadag taluk and 0.75 for Shirahatti taluk.

Bhattacharya (2008) has indicated that Simpson
Diversity Index moved up from 0.52 in 1997-98 to
0.59 in 2004-05 implying a gradual shift in cropping
pattern towards high-value crops like flowers, fruits
and vegetables in West Bengal.
Entropy Index for Area under Different Crop
Groups of Gadag District
The Entropy Index increases with increase in
diversification and vice versa. The results have clearly
shown that Gadag district, Gadag taluk and Shirahatti
taluk are diversifying over the years and among the
two study taluks, Gadag taluk has diversified more than
Shirahatti taluk. The results are in line with the findings
of Kalankar (2003), who studied agricultural output
growth and diversification in Maharashtra and reported
the decline in Herfindhal Index by 12.16 per cent and
increase in Entropy Index by 5.27 per cent over the
period 1961-64 to 1995-98.
Herfindahl Index for Area under Different Crop
Groups of Gadag District
The Herfindahl Index would decrease with increase
in diversification. The results (Table 4) indicated that
values of Herfindahl Index was less for both the taluks
as well as the Gadag district, thereby implying
diversification. Among the two taluks, Gadag taluk
having lower index value (0.20) is more diversified
than Shirahatti taluk (0.25).
Farm Level Crop Diversification in Gadag and
Shirahatti Taluks
An attempt was made to analyse crop
diversification at the farm level based on the crops
grown in the selected taluks, viz. Gadag and Shirahatti.
The crops wheat, maize, green gram, cotton, groundnut,
chilli, onion, sunflower, mango, sapota, banana,
156
Table 5. Crop diversification indices for different crops

at farm level
Indices
Gadag
talik
Shirahatti
taluk
Simpson Index (HH)

Simpson Index (Community)
Margalef Index
Entropy Index
Crop richness (community)
Household richness
0.58*
0.89*
1.18*
0.99*
15
3.20
0.31*
0.83*
0.44*
0.55*
10
1.90
sorghum, brinjal, tomato and bengalgram were

considered for analysis using Simpson, Entropy and
Margalef indices. The results are presented in Table 5.
All the indices have indicated significantly higher
values in both the taluks and among the taluks, Gadag
(NDZ) has been observed to be more diversified than
Shirahatti taluk (NTZ).
The sample farmers in the Gadag taluk cultivated
15 crops, but Shirahatti farmers cultivated only 10
crops. On an average, each household of Gadag taluk
maintained about 3 crops, whereas in Shirahatti taluk,
the average number of crops per household was only
2. This reveals that Gadag (NDZ) is more diversified
even at the farm level. Thus, analysis of both secondary
and primary data indicated that NDZ is more diversified
than NTZ.
Vol. 29 (No.1)
January-June 2016
Crop Diversification and Size of Landholding

There existed a strong association (r = 0.67 and
0.78) between size of landholding and crop
diversification. The size of holding appeared to be the
major driving force for crop diversification towards
commercial crops like Bt-cotton and horticultural crops
like onion, chilli, brinjal, tomato, sapota, mango and
banana (Table 6). The relationship between crop
diversification and different farm sizes (marginal, small
and large) was analysed for both Gadag and Shirahatti
taluks. The results (Table 7) reveal that, in both the
taluks, crop diversification increased with increase in
landholding-size and this was because farmers with
large holdings can grow more number of crops, whereas
small size of landholding was the constraint for growing
a larger number of crops. Kiru et al. (2014) have also
reported that an increase in the size of landholding will
better enable a farmer to diversify. The results also
reveal that among the two taluks, Gadag (3 crops)
experienced more crop diversification compared to
Shirahatti taluk (2 crops).
Crop Diversification and Irrigation in Gadag Taluk
As Gadag taluk showed a higher diversification,
an attempt was made to study the impact of factors
like extent of irrigation. It was observed that the number
of crops grown by irrigated farmers was more (15) than
by non-irrigated farmers (11). This was also true in the
case of household crop richness. Further, the average
Table 6. Correlation between crop richness and size of holding

Particulars
Crop richness
Size of holding
Gadag taluk
Crop richness
Size of holding
1.00
0.78
1.00
Shirahatti taluk
Crop richness
Size of holding
1.00
0.67
1.00
Table 7. Crop diversification and size of holding in Gadag district

Farmers
category
Marginal
Small
Large
All
Gadag taluk
Average landholding
Crop richness of
(acres)
household
1.57
4.00
13.81
8.67
2.14
2.57
3.94
3.20
Shirahatti taluk
Average landholding
Crop richness of
(acres)
household
1.93
4.13
17.67
10.38
1.14
1.63
2.40
1.90
157
Table 8. A comparison between irrigated and nonirrigated farms in Gadag taluk
Table 9. Factors determining crop diversification in

Gadag taluk
Particulars
Variable
Irrigated
farmer
Size of landholding (acre) 12.80 (27.1* )

Crop richness (HH)
3.76
No. of crops grown
15
Gross return (` /acre)
28785
Cost of cultivation (`/acre)
14563
Net returns (`/acre)
14222
Non-irrigated
farmer
5.50
2.76
11
24391
11844
12547
Note: * denotes percentage of irrigated land
size of holding was larger for irrigated farmers (12.8

acres). The larger size of holding, irrigation facilities
and the associated number of high-value crops resulted
in high returns (` 28785/acre) for irrigated farmers.
This result is in line with Birthal et al. (2006), who
have indicated that the extent of irrigation is the
significant factor responsible for crop diversification.
Factors Influencing Crop Diversification in Gadag
Taluk
The factors influencing crop diversification in
Gadag taluk were identified using regression analysis
wherein household richness was regressed on factors
like size of holding, farm income and gross area
irrigated (as a linear measure for extent of irrigation).
The size of holding, family income and area under
irrigation were found to have positive effect on crop
diversification. The average landholding in Gadag taluk
was 8.67 acres and the average farm income was
` 122352/acre (Table 9). The area under irrigated land
was found to influence crop diversification positively
and this result was in contrast with Joshi et al. (2004)
who reported a negative regression coefficient between
irrigation and crop diversification while studying the
pattern of agricultural diversification in South Asia.
At farm level, the major constraints in crop
diversification in Gadag district were captured using
Garret ranking technique. The price of crop was
regarded as the major factor influencing diversification
because farmers expect high income from crop
enterprise. The availability of irrigation water was the
second most important factor responsible for crop
diversification. In Gadag taluk, sample farmers could
grow 15 crops because of water availability. Further,
Intercept
Gross irrigated area
Income of family
Size of holding
F value
Adj. R2
Coefficient
t-stat
1.826771***
0.034202*
0.000005**
0.080910***
24.05
0.71
7.96
1.44
2.20
2.80
Note: *, ** and *** denote significance at 10 per cent, 5

per cent and 1 per cent levels, respectively.
Table 10. Constraints for crop diversification in Gadag
district
Constraint
Price of crop
Water availability
Risk yield and price of crop
Market availability
Cost of cultivation
Productivity
Input availability
Income of family
Garret score
Ranking
68.57
56.10
50.80
49.87
46.27
45.80
44.20
40.00
I
II
III
IV
V
VI
VII
VIII
the irrigated farmers could grow crops like transplanted

chilli, brinjal, banana, mango, sapota, cotton (Bt), onion
and tomato.
The market availability is an important factor for
diversification. The Gadag district has well organised
marketing infrastructure with 21 sub-markets. In
particular, 7 markets are present in Gadag taluk itself.
Costlier inputs resulting in high cost of cultivation were
regarded as another important factor influencing crop
diversification (Table 10). Input availability and
economic status of family also influence crop
diversification. The hypothesis that economic variables
like size of holding, irrigated area, household income,
access to market, etc. are the causes behind crop
diversification is proved valid with our results.

The study has revealed a higher growth rate in the
area under horticultural crops and pulses than in cereals,
oilseeds, fibre and other crop groups. Over the years,
though share of cereal crop groups has decreased
158
significantly, it still occupies the major share of the

cropped area reflecting farmers concern for food
security. The share of fruits and vegetables in the total
cropped area has increased significantly. This result is
amply supported by the transitional probability matrix,
indicating that among five crop groups, horticultural
crop group has retained the highest share of 92 per
cent, reflecting its stability. Simpson Index and Entropy
Index for different crop groups from 1998 to 2012 have
been found relatively higher for the Gadag taluk
compared to Shirahatti taluk, indicating thereby that
NDZ is more diversified than NTZ.
The size of holding is the major determinant of
crop diversification. However, investment on
infrastructure like irrigation by the government would
help even the small farmers enabling them to be a part
of the diversification process. Labour both human
and bullock is becoming scarce, costly and also not
available at the right time. Hence, mechanization in
the form of tractor and associated machineries/
implements is required by the farmers as its nonavailability is reflected as one of the major constraints
hindering diversification, especially for small farmers.
Efforts may be made to develop custom hiring centre
for tractors and other machineries required as the small
farmers cannot afford to invest in costly equipments.
Crop diversification ensures livelihood security to
the farmers and as market development has direct
bearing on crop diversification, there is a need to
develop more markets in the study area. At present,
support mechanism is not operating for horticultural
crops and as stable price acts as an incentive for
growing a number of crops, this may be ensured by
market intervention or price support even for
horticultural crops. The program A3P has increased
the production of pulses and the farmers need to be
made aware about this program. Crop diversification
should address food security, nutritional security as well
as livelihood security and this calls for enabling the
farmers to grow cereals and pulses in some area instead
of totally diversifying in favour of high-value crops
like cotton (Bt) and horticultural crops. Conducting
awareness generation campaigns, intensifying the role
of extension service providers, promoting researchbased recommendations and technologies on the farm
would strengthen the crop diversification process in
the study area.
Vol. 29 (No.1)
January-June 2016
Acknowledgements
The authors thank the anonymous referee for the
useful comments which helped in improving the paper.
References
Acharya, S.P., Basavaraja, H., Kunnal, L.B., Mahajanashetti,
S.B. and Bhat, A.R.S. (2011) Crop diversification in
Karnataka: An economic analysis. Agricultural
Economics Research Review, 24 (1): 351-357.
Anonymous (2010) Scheme to increase pulses output
launched. The Hindu, 11 August.
Bhattacharya, R., (2008) Crop Diversification: A search for
an alternative income of the farmers in the state of West
Bengal in India. International Conference on Applied
Economics: 83-93
Birthal, P.S., Jha, A.K., Joshi, P.K. and Singh, D.K. (2006)
Agricultural diversification in North-East region of
India: Implication for growth and equity. Indian Journal
of Agricultural Economics, 61 (3): 328-340.
Garret, E. Henry and Woodworth, R.S. (1969) Statistics in
Psychology and Education. Vaklis, Feffer and Simons
Pvt. Ltd, Bombay. pp. 329.
GoI (Government of India) (2010) Operational Guidelines
for Accelerated Pulses Production Program (A3P).
National Food Security Mission, Ministry of
Agriculture, New Delhi.
Gopalappa, D.V. (1996) Crop diversification and income
levels in Karimnagar district of Andhra Pradesh. Indian
Joshi, P.K., Gulati, A., Birthal, P.S. and Tewari, L. (2004)
Agriculture diversification in South Asia: Patterns,
determinants and policy implications. Economic and
Political Weekly, 39(18): 2457-2467.
Kalankar, S.S. (2003) Agricultural development and sources
of output growth in Maharashtra State. Artha Vijnana,
45(3/4): 297-324.
Kiru, Sichoongwe, Lawrence, Mapemba, Gelson Tembo and
Davies, Ngongola (2014) The determinants and extent
of crop diversification among smallholder farmers: A
case study of southern province Zambia. Journal of
Agricultural Science, 6(11): 150-159.
Rimal, N.S., Kumar, Shiv, Singh, D.R., Chahal, V.P. and
Shaloo (2015) Sources of growth in pulses production
in India. Agricultural Economics Research Review,
28(1): 91-102.
Venkatesh, M. (2013) Rice & pulse production register big
growth in Karnataka. The New Indian Express, 8
January.
Received: January, 2015; Accepted: February, 2016

DOI: 10.5958/0974-0279.2016.00028.8
Research Note
Impact of Integrated Farming of Water Chestnut and

Cat Fish on Livelihood of Farmers in Seasonal
Waterlogged Areas of Odisha
Souvik Ghosha*, S. Roy Chowdhuryb, R.K. Mohantyb and P.S. Brahmanandb
Institute of Agriculture, Visva-Bharati, Sriniketan 731 236, West Bengal
ICAR-Indian Institute of Water Management, Bhubaneswar 751 023, Odisha
a
Abstract
The paper has assessed the impact of integration of water chestnut cultivation with cat fish culture on
livelihood of farmers in the seasonal waterlogged areas of Balasore district in Odisha. It has revealed an
increase in production of water chestnut by 1.99 t/ha with an additional harvest of catfish to the extent of
0.78 tonnes, resulting in more than 50 per cent increase in the income of farmers. The study has observed
that before adoption of technology, four out of five types of assets measuring the changes in livelihood of
farm families were below the average level (barring social assets), which increased considerably after
adoption of technology with highest gain in financial assets (41%), followed by physical assets (35%),
social assets (31%) and human assets (29%). About 86 per cent farmers could be brought to above
average level of living with the change in farming situation on adoption of water chestnut cultivation
integrated with cat fish farming in Odisha. The developed package of practices for integrated farming of
water chestnut and catfish has been adopted by the CARE India under their dissemination programme
in tribal districts of Odisha.
Key words: Water chestnut, cat fish, integrated farming, adoption, waterlogged area, Odisha
Introduction
Water chestnut (Trapa bispinosa Roxb.) (singhara
or pani phal) is one of the few neglected but
economically important aquatic crops grown in
different parts of India. It is a natural crop in the areas
where water stagnation above the ground persists for
more than six months in a year. The 8 million ha
shallow low land ecosystem in the low-lying areas of
the country, of which 5.8 Mha is in eastern India,
provides ideal environment for the cultivation of this
crop, mainly during kharif season (Roy Chowdhury et
al., 2006).
Email: souvik.ghosh@visva-bharati.ac.in
In the eastern region of the country, the state of

Odisha has about 0.085 Mha waterlogged area, where
there is potential for diversification of agriculture
toward remunerative crops like water chestnut (Roy
Chowdhury et al., 2004a). Due to its aquatic habitat,
crop has resurrection ability despite exposure to brief
submergence or flash flood. The crop gradually adjusts
itself with the rise in water level to keep its leaf crown
afloat (Roy Chowdhury et al., 2003). This makes the
crop flood-resilient, especially in the low-lying areas.
Water chestnut fruits are generally consumed as raw
or after boiling. Following sun drying of mature fruit,
nut-flour is also used as a source of non-cereal
carbohydrate diet (Alam et al., 2001; Roy Chowdhury
et al., 2004b). A significant portion of the nut is
processed for use as flour for food or for textile sizing.
160
Vol. 29 (No.1)
January-June 2016
The low-lying waterlogged areas are suitable for

the auto recruitment of native fish species and are
natural habitat of captured fisheries (Kunda et al.,
2008). Therefore, integration of fish and aquatic crops
like water chestnut is economically lucrative under the
waterlogged ecosystem. The farmers are hesitant to
grow fishes in the isolated water bodies due to risk of
theft but integration of water chestnut with fish farming
offers a surface cover protection besides diversifying
livelihood options. The air-breathing fishes like
catfishes fit well in such a situation due to their
additional respiratory organ as well as habitat
preference and nature of growth. Due to favourable
factors like higher yield, production-size index and
performance index, air-breathing fish culture along
with water chestnut is advisable (Roy Chowdhury et
al., 2005).
equipment), and financial capital (savings, credit, and

income from employment, trade and remittances). The
rural livelihoods are complex and wide-ranging
(Ashley et al., 2003). The crop diversification, farm
sector diversification and livelihood diversification
influence the rural economy. Therefore, rural livelihood
diversification through appropriate and advanced
agriculture holds the key for development of rural
economy (Mehta, 2009).
A livelihood depends on people, their capabilities

and means of living including food, income and assets.
A livelihood is sustainable when it maintains or
enhances the assets on which it depends. Many of the
definitions of livelihood security currently in use are
derived from the work of Chambers and Conway
(1992). The idea of livelihood embodies three
fundamental attributes: (i) possession of human
capabilities (such as education, skills, health, and
psychological orientation); (ii) access to tangible and
intangible assets; and (iii) the existence of economic
activities. The interaction between these attributes
defines the livelihood strategy of a household. The
sustainable livelihoods are achieved through access to
a range of livelihood resources (natural, economic,
human, financial and social) which are combined in
the pursuit of different livelihood strategies, viz.
agricultural intensification, livelihood diversification,
migration, etc. (Scones, 1997). People and their access
to assets are at the heart of livelihoods approaches.
The impact assessment of technology on farming

situation and livelihood of farmers was carried out
covering a randomly selected sample of 35 farmers
adopting integration of water chestnut (WCN)
cultivation and aquaculture in the Balasore district of
Odisha for more than five years. The impact on the
farming situation of farmers on adoption of a
technology was realized through a comparison of
farming pattern, acreage, production, cost of cultivation
and gross income before and after the adoption of the
technology. The impact on livelihoods was measured
through finding comparative position of physical,
social, financial, human and natural assets of the
farmers before and after adoption of the intervention.
In the Department for International Development

(DFID) framework (1999), five categories of assets or
capitals have been identified, which are: human capital
(skills, knowledge, health and ability to work), social
capital (social resources, including informal networks,
membership of formalized groups and relationships that
facilitate co-operation), natural capital (natural
resources such as land, soil, water, forests and
fisheries), physical capital (basic infrastructure, such
as roads, water and sanitation, schools, communication
avenues and producer goods including tools and
In this backdrop, the present study was conducted

to study the impact of integrated farming of water
chestnut and cat fish on farming practices as well as
livelihood of farmers in seasonal waterlogged areas of
Odisha
Materials and Methods
The physical assets included type of housing,

sanitation, conveyance, electric, cooking and
communication facilities. The social assets mainly
referred to the recognition, social and political
participation, active involvement in developmental
works, common services used and group membership
pattern. The financial assets were measured on the basis
of parameters like sources of income, kinds of savings
and investments, lending and borrowing. The human
assets involved language competencies, education/
literacy level, management skill and mobility. The
natural assets were the possession of natural resources
of farm family, viz. farm size, irrigated land, livestock
holding, poultry and fishpond. All the above-mentioned
parameters under five types of assets were measured
on the basis of responses of farmers on a 5-point
continuum scale (minimum and maximum values being
Ghosh et al. : Impact of Integrated Farming of Water Chestnut and Cat Fish on Livelihood of Farmers
1 and 5, respectively) during the interview schedule

survey and focus group discussions. Overall, standard
of living of farmers was assessed on the basis of their
assets holding before and after adoption of the
technology; thus, the value of overall standard of living
ranged from 5 to 25.
Standard of living of farmer adopting the
technology, Li= (Pi+Si+Fi+Hi+Ni)
where, i indicates the number of farmers who adopted
the technology; i = 1,2, ..,35
Pi (Physical assets) = PAij/j,
where, j (=1,2, .,7) indicates the parameters
measuring physical assets, viz. (1) No. of rooms in
house, (2) Type of roof of the house, (3) Sanitary /
Latrine condition, (4) Type of vehicles-owned, (5)
Electric power usage, (6) Cooking facilities, and (7)
Telephone connectivity.
Si (Social assets) = SAik/k,
where, k (=1,2, .,4) indicates the parameters
measuring social assets, viz. (1) Respect/Recognition
in village, (2) Participation in local political issues, (3)
Use of common facilities at the locality, and (4)
Membership in common bodies / clubs / groups.
Fi (Financial assets) =FFil/l,
where, l (=1,2, .,4) indicates the variables measuring
financial assets, viz. (1) Sources of income (agriculture,
agricultural labour, livestock, fish farming, business,
interest from loan given, salary, etc.), (2) Kinds of
savings (bank, post office, chit fund, group fund, etc.),
161
(3) Kinds of investment (insurance, deposits in bank/

finance company, bonds, etc.), and (4) Lending.
Hi (Human assets) = HHim/m,
where, m (=1,2, .,4) indicates the variables
measuring human assets, viz. (1) Communication
ability, (2) Education/ literacy, (3) Management skills
(ability to
manage agriculture, livestock, fish
farming, business, marketing, etc.), and (4) Travel /
mobility.
Ni (Natural assets) = NNin/n,
where, n (=1,2, .,6) indicates the variables measuring
natural assets, viz. (1) Landholding, (2) Irrigation
sources, (3) Livestock holding, (4) Poultry birds
holding, (5) Fish pond, and (6) Any other resource.

Table 1 presents the information on farming
practices in the study area and adoption of farming
practices and their integration of water chestnut
cultivation and aquaculture in the Balasore district of
Odisha. It reveals that planting of water chestnut is
undertaken during early-June to early-July. The depth
of water between 0.5 m and 1.5 m is considered
favourable for cultivation and higher yield of water
chestnut. A layer of 15-20 cm soft mud reach in organic
matter at the bottom of water body favours better
growth of water chestnut with less turbidity and near
neutral pH. The farmers plant 3-4 young seedlings
loosely tied at the bottom in a knot in the mud bottom
of water with gentle push by toe maintaining the
Table 1. Impact of integration of water chestnut and cat fish (Magur) on farming practices in seasonal waterlogged
areas of Odisha (Farm size = 3.301.39 acre, N=35)
Farming
Farming situation before technology adoption
practice
Area
(acre)
MeanSD
Production
(t)
MeanSD
Paddy
1.650.87
(52.72%)
1.530.95
(62.09%)
1.150.69
(60.00%)
1.120.63
(56.25%)
5.232.31
(44.17%)
0.220.15
(68.18%)
Water
chestnut
Fish
Cost of
cultivation
(`/acre)
MeanSD
Farming situation after technology adoption
Gross income
(`/acre)
MeanSD
Area
(acre)
MeanSD
Production
(t)
MeanSD
41652409
81184551
(57.84%)
(56.06%)
197949473 5073524451
(47.86%)
(48.19%)
23241626
62354308
(69.97%)
(69.09%)
1.630.86
(52.76%)
1.660.93
(56.02%)
1.590.93
(58.49%)
1.490.70
(46.98%)
6.452.70
(41.86%)
0.580.27
(46.55%)
Cost of
cultivation
(`/acre)
MeanSD
Gross income
(`/acre)
MeanSD
68914106 140576460
(59.58%)
(45.96%)
2400011277 6814325079
(46.99%)
(36.80%)
49142331 217579983
(47.44%)
(45.88%)
Notes: SD stands for standard deviation. The figures within the parentheses indicate coefficient of variance.
162
spacing of about 1.5 m 1.5 m. About 4400 to 4500

bundles of seedlings (3-4 seedlings in each bundle)
are required for 1 ha area. The farmers apply compost
manure of about 7-8 t/ha to the water body during the
last week of May or early June, before the arrival of
monsoons. Although N:P:K fertilizers are
recommended @ 40:40:60 kg/ha in three splits, the
farmers in general do not strictly follow the
recommended dose.
The environment suitable for water chestnut
cultivation is also conducive for rearing of air-breathing
fish, and farmers prefer to grow Magur (Clarias
batrachus) species because of its relatively high market
price (` 40/kg approx.) along with water chestnut. The
identical habitat preference of water chestnut and
Magur has provided an opportunity to the farmers for
the integrated farming with better income. The farmers
release the air-breathing fish (Magur) during midAugust when the water chestnut crop establishes itself
with full canopy growth. The release of fish during the
period of crop establishment may affect the crop
growth. The farmers release the fingerlings of about
15-25 g mean body weight @ 7500-10000/ha after the
establishment of the crop. Even though fish feed @ 3
per cent of mean body weight twice-a-day is
recommended, the expensive nature of fish feed, does
not allow the farmers to follow the recommendation
entirely, they rather provide the feed as per their
affordability and availability to reduce the cost of
production compromising with the poor growth.
The farmers perceive that the fish (Magur) gets
natural food from this integrated eco-system as the
water chestnut provides huge detritus food, weed
associated fauna, benthic organism and insects that
reduce 25-30 per cent feed requirement. The coproduction of Magur and water chestnut also reduces
the problem of water chestnut beetle (Galerucella
birmanica) as Magur being carnivorous consumes it.
Thus, the coproduction system of Magur and water
chestnut provides congenial environment for the fish
in terms of natural food availability, water quality,
benthic population structure, and lower degree of
cannibalism. It also results in an increase in gross and
net water productivity. The harvesting of water chestnut
fruits is generally done during the months of November
and December, after which the crop gradually
decomposes and fish is also harvested.
Vol. 29 (No.1)
January-June 2016
The status of farming before and after adoption of

the technology is presented in Table 1. It is revealed
that integration of water chestnut (WCN) cultivation
and fish (Magur) was adopted by the farmers in about
50 per cent area, while paddy crop was grown in rest
of the area, though with low yield. Although there was
not much change in the acreage of paddy, its production
had increased. As the low lying areas are suitable for
auto recruitment of native fish species and natural
habitat of captured fisheries, few farmers used to
capture fishes in a portion (1.15 acre out of 1.53 acre)
of water chestnut growing area with relatively low fish
yield.
The adoption of integration of Magur fish with
water chestnut in most of the areas (1.59 acre out of
1.66 acre) provided more than double fish yield. The
average production of water chestnut was increased
by about 1.16 t/ha (from 8.55 t to 9.71 t) with additional
fish harvest of 0.43 t/ha. The integration of aquaculture
with water chestnut increased the average income of
farmers by ` 33,000/acre as compared to pre-adoption
situation. The total average income of the farmers from
the farming was increased by more than 50 per cent
after adoption of this technology. The variations among
the farmers with respect to acreage, production, cost
of cultivation and gross income were reduced after
adoption of the technology, as reflected from the values
of coefficient of variance.
The gross and net economic index of water
productivity of the above co-production system
evaluated by Roy Chowdhury et al. (2005), revealed
the gross water productivity to be ` 5.20, ` 6.00 and
` 12.90 per cubic metre of water for water chestnut,
fish, water chestnut+fish, respectively; the net water
productivity reported to be ` 2.80, ` 3.40 and ` 7.90
per cubic metre of water for water chestnut, fish, water
chestnut+fish, respectively. After the complete harvest
of fish and water chestnut in December January, the
available water can be used for supplementary
irrigation to crops like green gram, black gram in the
adjacent fields that would further increase the gross
and net water productivity up to ` 19.70 and ` 10.90
per cubic metre of water, respectively. However, in the
present study none of the selected farmers had total
holding at one place due to fragmentation; and
therefore, they grow water chestnut + fish (Magur) in
the seasonal waterlogged areas and cultivate paddy in
the fields not affected by waterlogging problem. Thus,
163
Figure 1. Change in average level of different types of assets measuring livelihood of farmers after adoption of
water chestnut and cat fish (Magur) integrated farming in seasonal waterlogged areas of Odisha
they could not use the left-over water after harvest of

water chestnut + fish by December-January to raise
the pulse crops for providing supplementary irrigations.
water resource and livestock holding) was due to the

fact that their growth generally requires a longer time
as compared to other types of assets.
The variables under five types of assets measuring

the changes in livelihood of farmers were assessed on
the basis of their responses on a 5-point continuum
scale (minimum and maximum values being 1 and 5,
respectively) and the mean values were derived for each
type of asset. It is evident from the Figure 1 that there
was improvement in all the five types of assets
measuring the changes in the livelihood of farm
families during post-adoption period. Four out of five
types of asset holdings, barring the social asset, were
found to be below average before adoption of the
integrated farming of water chestnut and cat fish
(Magur) culture technology. The gain was found
maximum in the financial assets (41%), followed by
physical assets (35%), social assets (31%) and human
assets (29%). Except natural asset gain (7%), all other
assets holding of farm families increased considerably
and came at above-average level. The high
improvements in financial and physical assets indicate
the betterment of living as well as economic conditions.
The overall standard of living of farmers was

assessed through summing up of the mean values of
all five types assets holdings of sample farmers before
and after adoption of the technology. It is presented in
Figure 2. A perusal of Figure 2 reveals that living
standard of only 3 out of 35 farmers was above the
average level prior to adoption of the technology, but
after adoption of technology, 30 out of 35 farmers came
above the average level of living. The mean value of
overall standard of living of all the technology-adopting
farmers derived through addition of the mean values
of five assets, indicated it to be in the range of 10.60 to
15.65 during pre-adoption and it increased to 13.55 to
20.95 during post-adoption period (minimum and
maximum values being 5 and 25, respectively).
The increased income on adoption of technology

even motivated the farmers to invest and intervene
further leading to the growth in physical and financial
assets. Social recognition was also reflected with higher
mean values of both social and human assets holdings
of the farmers. The low growth in natural assets (land,
Being a dynamic process, the livelihood

diversification depends on many factors having spatial
and temporal variations. This process of change varies
from farmer to farmer and over the space and time
(Ghosh et al., 2011). Therefore, the adoption of any
technology is not exclusive, but one of the factors
influencing the changes in livelihood of farmers. The
rural livelihoods are also wide-ranging (Ashley et al.,
2003). Both crop diversification and farm sector
diversification lead to livelihood diversification
influencing the rural economy; therefore, the adoption
164
Vol. 29 (No.1)
January-June 2016
Figure 2. Overall standard of living of farmers before and after adoption of water chestnut and cat fish (Magur)
integrated farming in seasonal waterlogged areas of Odisha
of appropriate agricultural technology holds the key

for development of rural economy (Mehta, 2009).
Concluding Remarks
There is market demand for both cat fish (Magur)
and water chestnut; therefore, growing water chestnut
in combination with Magur could provide good income
to the farmers of seasonal waterlogged areas in the
Balasore district of Odisha. Moreover, options of postharvest processing of nut to flour could potentially
avoid distress sale of excess harvest as well as provided
better market price. Growing fishes in isolated water
bodies has always been vulnerable to theft/poaching
and farmers are hesitant to invest in fisheries away from
their homestead. The integration of water chestnut with
it could offer a surface cover protection besides adding
income. This shallow waterlogged areas of eastern
India, where surface drainage is not possible, and water
stagnates with depth of more than 0.50 m for a period
of about six months, this technology is farmer-friendly
and a cost-effective option.
In this study, the impact assessment of technology
on farming situation and livelihood of farmers has been
carried out covering a sample of 35 farmers adopting
integration of water chestnut (WCN) cultivation and
cat fish culture. The smothering effect given by water
chestnut crop over water body could deter the pilferage
of fish cultivated below. The potential of this

technology has been reflected through the growth of
overall farming system and provision of better earning
and living to the small and marginal farmers of the
waterlogged ecosystem. The integration could
supplement fish feed requirement. Under the ICARCARE collaboration on Dissemination of Inland
Water Management Technologies, the developed
package of practices for water chestnut cultivation has
been adopted by The CARE India under their
dissemination program in three tribal districts
(Bolangir, Phulbani and Gajapati) of Odisha. The
successful implementation of the technology has led
to the spread of water chestnut cultivation technology
in the tribal districts of Odisha.
References
Alam, M.S., Singh, D.S. and Sehgal, V.K. (2001)
Optimization of screening in a water chestnut
decorticator. Journal of Agricultural Engineering, 38:
29-34.
Ashley, C., Start, D., Slater, R. and Deshingkar, P. (2003)
Understanding Livelihoods in Rural India: Diversity,
Change and Exclusion. Policy Guidance Sheets
produced by the Overseas Development Institute for
the Livelihood Options Study, funded by the
UK Department for International Development (DFID).
Chambers, R. and Conway, G.R. (1992) Sustainable Rural
Livelihoods: Practical Concepts for the 21st Century.
IDS Discussion Paper 296, Institute of Development
Studies, Brighton, UK.
DFID (Department for International Development) (1999)
Sustainable Livelihoods Guidance Sheets. Eldis
Document Store (http://www.eldis.org).
Ghosh, Souvik, Kumar, Ashwani, James, B.K., Roy
Chowdhury, S., Brahmanand, P.S., Mohanty, R.K. and
Kar, G. (2011) Impact assessment of the technologies
on the farming and livelihood of farmers. Research
Bulletin No. 52, Directorate of Water Management
(Indian Council of Agricultural Research),
Bhubaneswar, Odisha. 56p
Kunda, M., Azim, M.E., Wahab, M.A., Dewan, S., Roos,
N. and Thilsted, S.H. (2008) Potential of mixed
culture of freshwater prawn and self-recruiting small
species mola in rotational rice-fish/prawn culture
systems in Bangladesh. Aquaculture Research, 39: 506517.
Roy Chowdhury, S., Kumar, Ashwani, Sahoo, N., Kundu,
D.K., Anand, P.S.B. and Reddy, G.P. (2006) Growth
environment and production physiology of water
chestnut under shallow waterlogged condition and
swamp taro in marshy land. Research Bulletin No. 37,
Water Technology Centre for Eastern Region (Indian
Council of Agricultural Research), Bhubaneswar,
Odisha. 25p.
165
Roy Chowdhury, S., Mohanty, R.K., Anand, P.S.B., Sahoo,

N. and Kumar, Ashwani (2005) Integration of water
chestnut and fish: An ideal package of practice.
Extension Folder, Water Technology Centre for Eastern
Region (Indian Council of Agricultural Research),
Bhubaneswar, Odisha. 6p.
Roy Chowdhury, S., Sahoo, N., Mohanty, R.K., Jena, S.K.
and Verma, H.N. (2004a) Performance of water chestnut
in farmers field in coastal area in Odisha. Journal of
Indian Society of Coastal Agricultural Research, 22:
267-269.
Roy Chowdhury, S., Sahoo, N. and Verma, H.N. (2004b)
Changes in fruit size, fruit dry matter and carbohydrate
composition at different stages in developing water
chestnut fruit. Journal of Aquatic Plant Management,
42: 99-102.
Roy Chowdhury, S., Sahoo, N. and Verma, H.N. (2003)
Growth behavior and yield of five water chestnut
varieties under waterlogged condition. Indian Journal
of Plant Physiology, 8: 369-371.
Mehta R. (2009) Rural Livelihood Diversification and its
Measurement Issues: Focus India. Wye City Group on
Rural Statistics and Agricultural Household Income,
Second Annual Meeting, 11-12 June. FAO, Rome.
Scones, Ian (1997) Sustainable Rural Livelihoods: A
Framework for Analysis. IDS Working Paper 72.
Institute of Development Studies, Brighton, UK.
Received: September, 2015; Accepted February, 2016

Book Review
Genetically Modified Crops and Agricultural
Development by Matin Qaim, Palgrave Macmillan US,
Year: 2015, e-Book ISBN: 978-1-137-40572-2,
Hardcover ISBN: 978-1-137-40571-5, DOI: 10.1057/
9781137405722, Number of Pages: 206.
Matin Qaims book titled Genetically Modified
Crops and Agricultural Development is a very central
and significant volume that brings forward Qaims
advanced study on the impacts of genetically modified
crops (GMCs). He unmistakably exposes the
hollowness of the numerous arguments against genetic
modification and gives compelling proof that GMCs
are, in fact, contributing majorly in convalescing human
lives. It can be said without a doubt that this book
represents one of the first balanced and evidence-based
accounts of GMCs.
The book begins by offering a general introduction
of the main issues and general beliefs that are associated
with GMCs and then moves into focusing on plant
breeding related to agricultural development. Qaim
notes that there are three overarching goals of
agricultural development; namely: sufficient food,
improvement of the livelihoods of those associated with
the agricultural sector and lastly, sustainability. He
addresses the progression of plant breeding and brings
to light the similarities in the main beliefs that are
fundamental to both conventional as well as GM plant
breeding. The book reveals that crop traits targeted
through genetic engineering are not completely
different from those pursued by conventional breeding.
Qaim argues that the two approaches involve
significant genetic modification of crops and gives
examples as to how crops have historically evolved
through human interventions and he strongly
emphasizes that GMCs stand for another step forward
in this ever-changing process. He states that although
conventional breeding has indeed made considerable
advances over the past forty to fifty years and has been
a driving factor of the green revolution enabling yield
increases in many Asian countries but it is the advances

in molecular biology that have provided tools for
research to go beyond yield considerations alone;
thereby giving rise to investigations of the various
options that are not readily available through
conventional breeding. This includes new modes of
tolerance to soil salinity, drought, flooding, high
temperatures and so on as well as resistance to certain
pests and diseases. Qaim believes that although GM
adoption patterns are geographically uneven, the GM
crops are among the fastest-adopted agricultural
technologies in human history. He also touches upon
the topic of the changing roles of public and private
sector organizations with respect to the developments
in the seed sector.
Chapter 3 in the book gives a detailed report of
the potential of GMCs and the risks associated with
them. Also discussed are the current GM techniques,
breeding objectives and their potentials to address
agronomic and nutritional constraints. Qaim further
includes a critical review of the arguments presented
by those opposed to GM. He addresses the possible
environmental, health, and social risks of GMCs. The
book infers that since the commercialization of GMCs,
their consumption has not led to any legitimate case of
health problems and that this has also been verified by
the World Health Organization. Qaim even points out
that Bt technology in particular, actually leads to
significant reductions in the use of chemical pesticides.
Despite this evidence that there are no health risks from
consuming GMCs, the book acknowledges that antiGM groups are yet to alter their stance on the issue. In
fact, Qaim argues that the meticulous and mandatory
testing regimes that are required for GMCs before they
can be registered for release should also be applied to
other non-GM crops derived from techniques such as
radiation and chemical treatments. Additionally, Qaim
presents ample examination of environmental risks
from GMCs raised by anti-GM groups. He makes an
important conclusion that humans have in reality bred
168
food crops from the wild into the farm environment

over the years; and these specific food crops are no
longer adapted to prosper in natural bio-diverse
ecosystems.
The next chapter of this book is based on the
adoption and impacts of commercialized GMCs and
makes an effort to tackle the controversial issues that
cause a big divide between those supporting and those
opposing GMCs. Qaim starts off by describing the
current status and impacts of GMCs grown
commercially and then summarises various publicfunded studies in order to compare the performance of
GM and non-GM crops in similar agro-ecological
regions. He makes use of literature that has looked at
the effects of adoption of GMCs on crop yields, costs
of production, etc., using cross-sectional data, repeated
surveys, panel data and various other methodologies.
In addition, the book also provides a case study of Bt
cotton adoption in India. The overall conclusions drawn
show that, although there existed local differences, the
overall average yield increase of GMCs - due to less
pest damage was approximately 21 per cent and
pesticide-use was reduced by around 37 per cent. The
book also mentions (without measurements) the reality
of financial benefits derived from improved health of
farmer as well as consumers, reduction in water
pollution among others.
New and future applications, regulations and
complexities of GMCs are addressed in the subsequent
chapters of Qaims book. An overview of the research
and development of GMCs is provided in Chapter 5.
The book reveals that many GMCs, including crops
with virus and fungal resistance, have already been
developed and tested and could be commercialized
within the next few years. Furthermore, aspects related
to GM food crops with higher contents of
micronutrients are discussed.
The regulatory issues of GM crops are reviewed
in Chapter 6, such as food safety regulations, food
labelling, and intellectual property rights. He analyses
the reasons as to why commercialization of GMCs is
Vol. 29 (No.1)
January-June 2016
still restricted to just two traits (insect resistance and

herbicide tolerance) among other analysis. The book
further argues that the principal constraint for
commercialization of GMCs is the anti-GM perceptions
which influence government policy as well as lead to
negative public views. Several of these popular antiGM narratives are explored and refuted in Chapter 7.
Qaim states that, with the exception of Spain, most
countries in the EU have a cap on GMCs and in some
cases, a permanent ban on their development making
it impossible for publicly funded institutions to consider
commercialization of GMCs that they have developed.
Added to this, the exceedingly high cost to register a
product has predictably resulted in these multinational
corporations being the sole marketers of GMCs.
The final chapter in the book summarises that
future GM crop applications may produce much bigger
benefits and development if the blockade of public
resistance can be overcome. He says that issues such
as equal access to suitable seed technologies by poor
farmers cannot be solved by banning GMOs. The book
concludes that, given the global challenges ahead,
sustainable agricultural development and food security
will not be possible without harnessing the potentials
of plant biotechnology, including GMCs.
Matin Qaim argues that the way forward for GMCs
is by building trust in the various stakeholders (i.e. the
public, government and media) through honesty, a
moral obligation to not misinform the consumer and
evidence. He emphasises that, in his personal opinion,
the world is better off with GM crops than without;
and that future challenges of agricultural development
can only be properly addressed if we harness all
promising areas of science responsibly.
Dr A.V. Manjunatha
Agricultural Development and Rural
Transformation Centre
Institute for Social and Economic Change
Nagarabhavi P.O. Bangalore-560072, Karnataka


Guidelines for Submission of Papers/Abstracts
1.
The journal publishes research articles, review

articles, research notes and communications in
basic and applied research on economic aspects
of agriculture and rural development.
Comprehensive review articles in the area of
agricultural economics (including livestock,
horticulture and fisheries), conference/symposia
proceedings and book reviews are also published
in the Journal. To encourage the young researchers,
recent abstracts of M.Sc. and Ph.D. theses in
agricultural economics are also published.
2. The journal is managed by the eminent economists
under the domain of Agricultural Economics
Research Association (India). The authors
submitting papers to Agricultural Economics
Research Review should be members of this
Association.
3. Paper submission
The soft copy of the manuscript may be
submitted at aeraindia@gmail.com
At the time of manuscript submission a
processing fee of Rs 500/- (Rupees five hundred
only) per paper has to be paid by the author(s).
It is applicaple from 1st January, 2015.
For papers which are accepted for publication,
at least the corresponding/first author should
be a member of AERA (India). It is application
from January-June 2015 issue of Agricultural
Economics Research Review.
The authors should submit the following
declarations:
(a) The paper entitled ........ has not been and
will not be submitted in full or in part to any
other journal till the decision of AERR on it.
(b) On publication in AERR, the copyrights on
the full contents of the paper will be of
Agricultural Economics Research Association
(India).
The processing fee and membership fee may be
sent through Demand Draft (D/D) or multicity
cheque payable to Agricultural Economics

Research Association (India), New Delhi and
should be sent to
The Managing Editor
F-4, A Block
Dev Prakash Shastri Marg
Pusa, New Delhi - 110 012
4.
Authors should include the complete source of

research article, like project, along with sponsor,
M.Sc./Ph.D. thesis, etc. They should mention the
title of the Project/Thesis also.
5.
The length of papers should not be more than 20

(double space) typed pages, including tables,
diagrams and appendices.
6.
Name(s) and affiliation(s) of the author(s) with email address(es) should be provided on a separate
page along with the title of the article.
7.
Only essential mathematical notations may be

used. All statistical formulae should be neatly
typed. Footnotes should be numbered
consecutively in plain arabic superscripts.
8.
Abstracts of recent M.Sc. theses should not exceed

one typed page and those of Ph.D. theses, two
typed pages. The authors must mention at the top
of the abstracts the degree for which the thesis
was submitted and the year of award, like M.Sc.
(2008), Ph.D. (2008). The name of major advisor
should also be included.
9.
References: Only cited works should be included

in reference list. The citation of references should
be in the following order: author(s) name(s); year;
title of article; name of journal; volume; number;
and pages. Please follow the style of citations as
in the latest issue of this journal. Papers not
submitted in the standard format, as suggested
above will not be considered for publication.
170
The reference list should be alphabetized and not

numbered. To avoid ambiguity, the title of journal
should be given in full.
The following examples, as typical enteries,
provide guidance for enteries in the reference list.
Research Paper: Rosegrant, M.W. and Pingali,
P.I. (1994) Policy and technology for rice
productivity growth in Asia. Journal of
International Development, 6(6): 6656-88.
Book: Rosegrant, M.W. and Hazell, P.B.R. (2000)
Transforming Asian Economy: The Unfinished
Revolution. Oxford University Press, Hong Kong.
Chapter in a Book or Paper in a published
proceedings: Kumar, P. (2001) Agricultural
performance and productivity. In: Indian
Agricultural Policy at the Crosswords, Eds: S.S.
Acharya and D.P. Chaudhri. Rawat Publication,
New Delhi, pp. 353-476.
Paper in a Conference/Symposium: Kapoor,
B.C. (2000) Managing in the face of not-sodeveloped and organized environment. In:
Proceedings of National Symposium on
Management and Development, held at Institute
of Public Administration, Jaipur, 23-25 July.
Thesis: Behera, Sumanta (2004) Impact of
Technological Change and Economics of Cocoon
Production in Mandya District of Andhra Pradesh,
MSc (Seri Technol) Thesis, submitted to
University of Mysore.
Units: Use SI units; a few examples are given
below:
Hectare
ha
Milligram
mg
Rupees
`
Million hectares Mha
Litre
L
Tonne
t
Millilitre
mL Million tonnes
Mt
Gram
g
Metre
m
Kilogram
kg
Centimetre
cm
Please note that no full stop is used after the
abbreviation of units.
10. The journal Agricultural Economics Research
Review is available on-line, please contact at the
following addresses for on-line services:
(i) Diva Enterprises Pvt. Ltd.
B-9, Basement, Block-A
Naraina Vihar, New Delhi 110 028, India
http:// www.indianjournals.com
Vol. 29 (No.1)
January-June 2016
(ii) AgEcon Search

Waite Library, Department of Applied
Economics, University of Minnesota, 1994
Buford Ave, St. Paul MN 55108-6040, USA
http:// www.agecon search
(iii) EBSCO Information Services
10 Estate Street, Ipswich MA 01938 USA
E-mail: epalmquist@ebsco.com
11. Membership Tariffs/Subscription to Agricultural
Economics Research Review* (Effective from 1st
January, 2015)
Particulars
Print copy*
Epublication*
Print +
Online*
(Epublication)
National
Institutional
Membership
` 3000/-
` 3371/-
` 4500/-
Individual
Membership
` 700/-
` 700/-
` 1050/-
Student
Membership
` 500/-
` 500/-
` 750/-
Life
Membership
` 5000/-
` 5000/-
` 5000/-
US $ 150
US $ 225
US $ 300
Individual
Membership
US $ 75
US $ 225
US $ 300
Life
Membership
US $ 500
US $ 500
US $ 500
Overseas
Institutional
Membership
* For print copy of journal, please send subscription to

AERA (India) at the address given below.
* For E-publication and Print + E-publication, please
send subscription to Diva Enterprises Pvt. Ltd., B-9,
Basement, Block-A, Naraina Vihar, New Delhi 110 028
The annual subscription includes two regular issues of the
journal and a Conference Number.
Address for correspondence:

Secretary/Treasurer
Agricultural Economics Research Association
F-4, A Block, National Agricultural Science Centre
Complex, Dev Prakash Shastri Marg, Pusa
New Delhi - 110 012
Email: aeraindia@gmail.com
Webpage: www.aeraindia.in
Correspondence may preferably be done
through E-mail.
Regd. No. F2 (A/67)/89
Statement of Ownership and other Particulars

AGRICULTURAL ECONOMICS RESEARCH REVIEW
Form IV (See Rule 8)
Place of Publication
New Delhi
Periodicity of Publication
Half Yearly
Printers Name
M/s Cambridge Printing Works

B-85, Naraina Industrial Area, Phase II
New Delhi 110 028
Publishers Name
Dr. Suresh Pal
Nationality
Indian
Address
Commission for Agriculture Costs & Prices

Government of India, New Delhi - 110 001
Editors Name
Dr. P.S. Birthal
Nationality
Indian
Address
Institute of Development Studies

Jaipur - 302 004, Rajasthan
Name and Address of Individuals

who own the newspaper and
partners or shareholders holding
more than one per cent of the
total capital
Agricultural Economics Research

Association (India)
F-4, A Block
DPS Marg, Pusa, New Delhi - 110 012
I, Dr. Suresh Pal, hereby declare that the particulars given above are true to the best of my knowledge and belief.
Sd/(Suresh Pal)
Signature of Publisher
AGRICULTURAL ECONOMICS RESEARCH ASSOCIATION (INDIA)
EXECUTIVE COMMITTEE
President
: Dr. P.G. Chengappa, ICAR National Professor, Institute for Social and Economic
Change, Bangalore - 560 072, Karnataka
Conference President : Dr. P. Kumar, Former Professor, Division of Agricultural Economics, ICAR-Indian
(AERA Conference 2016)
Agricultural Research Institute, New Delhi - 110 012
Vice Presidents
: Dr. R.K.P. Singh, Former Professor, RAU, 101, Shivam Heritage, Digha Road,
Patna - 800 001, Bihar
Dr. M.H. Wani, Rajiv Gandhi Chair Professor, Divison of Agricultural Economics
and Marketing, Sher-e-Kashmir University of Agricultural Science and Technology
of Kashmir, Shalimar Campus, Srinagar - 191 121, Jammu & Kashmir
Dr. M. Krishnan, Principal Scientist and Head, Fisheries Economics, Extension
and Statistics Division, ICAR-Central Institute of Fisheries Education,
Mumbai - 400 061, Maharashtra
Secretary
: Dr. Suresh Pal, Member, Commission for Agriculture Costs & Prices, Government
of India, New Delhi - 110 001
Treasurer
: Dr. Anjani Kumar, Research Fellow, International Food Policy Research Institute,
NASC Complex, Dev Prakash Shastri Marg, New Delhi - 110 012
Joint Secretaries
: Dr. Raka Saxena, Senior Scientist, ICAR-National Institute of Agricultural

Dr. Suresh S. Patil, Professor and Head, Department of Agricultural Economics,
University of Agricultural Sciences, Raichur - 584 102, Karnataka
Members
: Dr. Madhusudan Bhattarai, International Food Policy Research Institute, NASC

Complex, Dev Prakash Shastri Marg, New Delhi - 110 012
Dr. Subhash Chand, Principal Scientist, ICAR-National Institute of Agricultural
Dr. N.R. Kumar, Principal Scientist, ICAR-Central Institute of Fisheries Education,
Versoa, Mumbai - 400 061, Maharashtra
Dr. S.N. Mishra, Research Wing, Orissa University of Agriculture and Technology,
Bhubaneswar - 751 003, Odisha
Dr. Swadesh Prakash, Senior Scientist/Economist, ICAR-Central Institute of
Fisheries Education, Mumbai - 400 061, Maharashtra
Dr. V. Saravanakumar, Directorate of Planning and Monitoring (DPM), Tamil
Nadu Agricultural University, Coimbatore-641 003, Tamil Nadu
Dr. Siddayya, Associate Professor, Centre for Natural Resource Management,
National Institute of Rural Development & Panchayati Raj, Rajendranagar,
Hyderabad - 500 030, Telangana
Dr. Ram Singh, College of Post Graduate Studies, Central Agricultural University,
Umiam, Barapani, Shiilong - 793 103, Meghalaya

AERRVol29 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AERRVol29 1

Uploaded by

Copyright:

Available Formats

AGRICULTURAL ECONOMICS RESEARCH REVIEW

Dr. Naveen P. Singh, Principal Scientist, ICAR-National Institute of Agricultural

Agricultural Economics Research Review

Agricultural Economics Research Association (India)

Agricultural Economics Research Review

Agricultural Economics Research Association (India) 2016

Financial support from:

Agricultural Economics Research Review

Institutional Synergies in Processing and Value Addition: Role of a Producers Organisation

Impact of Nonfarm Activity on Rural Income Inequality in South 24 Parganas District of

Changing Cropping Pattern, Agricultural Diversification and Productivity in Odisha

Crisis of Sustainability or Perils of Ill-managed Open Access Fisheries? Analysis of

Export of Rice from India: Performance and Determinants

Guidelines for Submission of Papers/Abstracts

Divya Kumari, H.V.

Nayak, Dinesh Kumar

Mishra, Dilli Raj

Agricultural Economics Research Review

Patterns and Drivers of Dairy Development in India: Insights

and the ownership rate exceeds 70 per cent among the

Agricultural Economics Research Review

products (Mishra and Roy, 2011). Milk and milk

different types of bovines owned by households

India has the worlds largest bovine population,

Ideally, one would like to have data on yield and

In this paper, we study trends and patterns in dairy

Data and Key Assumptions

Dairying in India: A Smallholders Enterprise

Kishore et al. : Patterns and Drivers of Dairy Development in India

Table 1. Ownership of in-milk bovines by operational landholding size class in 2002-03

Marginal9 (< 1 ha)

Source: Estimated by authors from GoI (2006).

when the rural poor tended dry bovines of the rich

Table 2. Growth in marginal holdings and their share

Per cent share

Source: Estimated by authors

households, HER is higher for households with larger

Source: Estimated by authors using data from NSSO

Agricultural Economics Research Review

Table 4. Regression showing household level

Bovine stock is more equitably distributed than

Notes: *** p<0.01, ** p<0.05, * p<0.1

land-constrained households in the non-farm sectors

Kishore et al. : Patterns and Drivers of Dairy Development in India

A probit analysis (Table 5) of more than 16,000

Table 5. Characteristics associated with higher

Note: *** p<0.01, ** p<0.05, * p<0.1

Agricultural Economics Research Review

Per cent rural

Average number of bovines owned per household

for milk and milk products, India needs to reduce its

districts in northern states of Punjab and Haryana have

Kishore et al. : Patterns and Drivers of Dairy Development in India

Ratio of number of buffaloes to cows

machines allow even smallholders to use these for

they account for only 24 per cent of crossbred cattle

Table 9. Population shares of different types of bovines in India: 1997 to 2007

Agricultural Economics Research Review

data is important to understand the causes of such

where, Yit is the dependent variable, HER of district i

where, Yist is the herd efficiency ratio in district i in

Population Density The literature predicts that

Kishore et al. : Patterns and Drivers of Dairy Development in India

Notes: * p<0.01, p<0.05, * p<0.1

Note: * p<0.01, p<0.05, * p<0.1

Note: * p<0.001, p<0.01, * p<0.05