One of the vital ingredients of an effective reward management strategy is the development and maintenance of a competitive pay and

benets position that enables organizations to attract and retain high-quality people. This can only be done if data on the levels of pay in the external market are systematically monitored by a process of market rate analysis. arket rate analysis is the process of collecting and comparing data on the rates and benefits provided for similar !obs in other organizations and the rates at which pay is increasing elsewhere. "t is conducted by means of surveys that review published data and#or collect data from various sources. THE PROCESS OF MARKET RATE ANALYSIS The process of market rate analysis consists of the following steps as described in the rest of this chapter$ %ecide that it is necessary to collect market rate data by means of surveys for a defined purpose or purposes such as those listed above. %ecide on the !obs for which market rate data will be collected. "dentify potential sources of market rate data and select the most appropriate ones. &repare information on the !obs to be surveyed. 'nalyse and interpret the data from the various sources. &resent the outcome of the analysis with proposals for action. The validity and reliability of ar!et rate data This is based on three factors$ (ob matching ) the extent to which good !ob matching has taken place. *ample frame ) the extent to which the sample of organizations from which the data have been collected is fully representative of the organizations with which comparisons need to be made in such terms as sector+ technology or type of business+ size and location. Timing ) the degree to which the information is up to date or can be updated reliably. ,y its very nature+ the information in published surveys+ upon which many people rely+ can soon become out of date. "ndeed+ the moment surveys are produced they are out of date ) pay levels may have changed and people may have moved in or out since the date of the survey. -hilst it is not possible to overcome this completely+ as data must be gathered and analysed+ surveys that aim to have as short a time as possible between data collection and the publication of results are likely to be of more use than those with longer lead times. .stimates can be made of likely movements since the survey took place+ but they are mainly guesswork.

SO"RCES OF MARKET #ATA The main sources are$ Online pay data/ 0eneral published national surveys/ 0eneral local surveys/ *ector and industry surveys/ Occupational surveys/ anagement consultants1 databases including online services/ ,espoke surveys conducted by the organization/ &ay clubs/ &ublished data in !ournals/ Official sources/ 2ecruitment consultants and agencies/ 'nalysis of recruitment data/ (ob advertisements

*pecial *urvey &ay 5lubs &ublished %ata 2ecruitment %ata (ob 'dvertisement

$ *urveys conducted by organisations 3e.g.$ *ector and industry surveys+ and occupational surveys4 0roup of employers who exchange data on pay levels %ata on settlement and pay levels from 5*O or &2, report or 2emuneration orders &ay data derived from analysis of pay levels for recruitment &ay data from !ob advertisements

0rade and pay structures provide a logically designed framework within which an organization1s pay policies can be implemented. They enable the organization to determine where !obs should be placed in a hierarchy+ dene pay levels and the scope for pay progression and provide the basis upon which relativities can be managed+ equal pay achieved and the processes of monitoring and controlling the implementation of pay practices take place. ' grade and pay structure can also serve as a medium through which the organization communicates the career and pay opportunities available to employees. 02'%. *T265T62.* ' grade structure consists of a sequence or hierarchy of grades+ bands or levels into which groups of !obs that are broadly comparable in size are placed. There may be a single structure+ which is dened by the number of grades or bands it contains 'lternatively the structure may be divided into a number of career or !ob families consisting of groups of !obs where the essential nature and purpose of the work are similar but the work is carried out at different levels. The main types of graded structures as described in this chapter are$ 7arrow-graded structures+ which consist of a sequence of narrow grades 3generally 89 or more4. They are sometimes called multi-graded structures. ,road-graded structures+ which have fewer grades 3generally six to nine4. ,road-banded structures+ which consist of a limited number of grades or bands 3often four to ve4. *tructures with six or seven grades are often described as broadbanded even when their characteristics are typical of broad grades. 5areer family structures+ which consist of a number of families 3groups of !obs with similar characteristics4 each divided typically into six to eight levels. The levels are described in terms of key responsibilities and knowledge+ skill and competence requirements. Therefore the different levels define career progression routes within and between career families. There is a common grade and pay structure across all the career families. (ob family structures+ which are similar to career families except that pay levels in each family may differ to re:ect market rate considerations 3this is sometimes referred to as market grouping4. The structure is therefore more concerned with market rate relativities than mapping careers. The number of levels in families may also vary. 5ombined structures+ in which broad bands are superimposed on career#!ob families or broad bands are divided into families. &ay spines+ consisting of a series of incremental ;pay points1 extending from the lowest- to the highest-paid !obs covered by the structure.

PAY STR"CT"RES ' grade structure becomes a pay structure when pay ranges+ brackets or scales are attached to each grade+ band or level. "n some broad-banded structures as described later in this chapter+ reference points and pay zones may be placed within the bands and these define the range of pay for !obs allocated to each band. &ay structures are defined by the number of grades they contain and+ especially in narrow- or broad-graded structures+ the span or width of the pay ranges attached to each grade. *pan is the scope the grade provides for pay progression and is usually measured as the difference between the lowest point in the range and the highest point in the range as a percentage of the lowest point. &ay structures define the different levels of pay for !obs or groups of !obs by reference to their relative internal value as determined by !ob evaluation+ to external relativities as established by market rate surveys and+ sometimes+ to negotiated rates for !obs. They provide scope for pay progression in accordance with performance+ competence+ contribution or service.

2upees ' narrow+ multi-graded structure

2upees 7arrow-graded+ broad-graded and broad-banded structures 2upees ' career family structure

2upees ' !ob family structure 2upees ' pay spin

#e$idin% on %rade bo&ndarie' ' grade boundary is the point between one grade and the next higher or lower grade+ which is dened as a rate of pay and a points score if a points evaluation scheme is used to determine grades. The boundaries therefore dene the span of a grade in those terms. 'n analytical !ob evaluation exercise will produce a rank order of !obs according to their !ob evaluation scores. ' decision then has to be made on where the boundaries that will define grades should be placed in the rank order. *o far as possible+ boundaries should divide groups or clusters of !obs that are significantly different in size so that all the !obs placed in a grade are clearly smaller than the !obs in the next higher grade and larger than the !obs placed in the next lower grade. <ixing grade boundaries is one of the most critical aspects of grade structure design following an analytical !ob evaluation exercise. "t requires !udgement ) the process is not scientic and it is rare to nd a situation where there is one right and obvious answer. "n theory+ grade boundaries could be determined by deciding on the number of grades in advance and then dividing the rank order into equal parts. ,ut this would mean drawing grade boundary lines arbitrarily and the result could be the separation of groups of !obs that should properly be placed in the same grade. The best approach is to analyse the rank order to identify any signicant gaps in the points scores between ad!acent !obs. These natural breaks in points scores will then constitute the boundaries between clusters of !obs that can be allocated to ad!acent grades. ' distinct gap between the highest rated !ob in one grade and the lowest rated !ob in the grade above will help to !ustify the allocation of !obs between grades. "t will therefore reduce boundary problems leading to dissatisfaction with gradings when the distinction is less well dened. &rovisionally+ it may be decided in advance when designing a conventional graded structure that a certain number of grades is required+ but the gap analysis will conrm the number of grades that is appropriate+ taking into account the natural divisions between !obs in the rank order. =owever+ the existence of a number of natural breaks cannot be guaranteed+ which means that !udgement has to be exercised as to where boundaries should be drawn when the scores between ad!acent !obs are close. "n cases where there are no obvious natural breaks+ the guidelines that should be considered when deciding on boundaries are as follows.

(&ideline' for de$idin% on %rade bo&ndarie' (obs with common features as indicated by the !ob evaluation factors are grouped together so that a distinction can be made between the characteristics of the !obs in different grades ) it should be possible to demonstrate that the !obs grouped into one grade resemble each other more than they resemble !obs placed in ad!acent grades. The grade hierarchy should take account of the organizational hierarchy+ ie !obs in which the !ob holder reports to a higher level !ob holder should be placed in a lower grade+ although this principle should not be followed slavishly when an organization is over-hierarchical with+ perhaps+ a series of one-over-one reporting relationships. The boundaries should not be placed between !obs mainly carried out by men and !obs mainly carried out by women. The boundaries should ideally not be placed immediately above !obs in which large numbers of people are employed. The grade width in terms of !ob evaluation points should represent a significant step in demand as indicated by the !ob evaluation scheme.

The $on'ideration' to be ta!en into a$$o&nt )hen de$idin% on the n& ber of %rade' The range and types of roles to be covered by the structure. The range of pay and !ob evaluation points scores to be accommodated. The number of levels in the organizational hierarchy 3this will be an important factor in a broad-banded structure4. %ecisions on where grade boundaries should be placed following a !ob evaluation exercise that has produced a ranked order of !obs ) this might identify the existence of clearly defined clusters of !obs at the various levels in the hierarchy between which there are significant differences in !ob size. The fact that within a given range of pay and responsibility+ the greater the number of grades the smaller their width+ and vice versa ) this is associated with views on what is regarded as the desirable width of a range+ taking into account the scope for progression+ the size of increments in a pay spine and equal pay issues. The problem of ;grade drift1 3un!ustified upgrading in response to pressure+ lack of promotion opportunities or because !ob evaluation has been applied laxly4+ which can be increased if there are too many narrow grades.

The fa$tor' affe$tin% de$i'ion' on the )idth of %rade' or band' >iews on the scope that should be allowed for performance+ contribution or career progression within grades. .qual pay considerations ) wide grades+ especially extended incremental scales+ are a ma!or cause of pay gaps between men and women simply because women+ who are more likely to have career breaks than men+ may not have the same

opportunity as men to progress to the upper regions of the range/ male !obs may therefore cluster towards the top of the range while women1s may cluster towards the bottom. %ecisions on the number of grades ) the greater the number the smaller the width. %ecisions on the value of increments in a pay spine ) if it is believed that the number of increments should be restricted+ for equal pay or other reasons+ but that the number of grades should also be limited+ then it is necessary to increase the value of the increments. "n a broad-banded structure+ the range of market rates and !ob evaluation scores covering the !obs allocated to the band.