You are on page 1of 22

ISSN 0080-6757 Doi: 10.1111/1467-9477.

12080
C 2016 Nordic Political Science Association
V

Solid or Flexible? Social Trust from Early


Adolescence to Young Adulthood

Ali Abdelzadeh* and Erik Lundberg†

The belief that people are generally fair and trustworthy has generated plenty of scholarly
attention in recent decades, particularly in the Scandinavian countries, which are often known
for high levels of social trust. This article draws attention to the current discussion in the liter-
ature on whether social trust is a stable cultural trait marked by persistence or is based on
experiences and subject to change throughout life. Based on unique longitudinal data from
five different cohorts of young people in Sweden, ranging in age from 13 to 28 years, this arti-
cle provides an empirical contribution on how social trust develops over time. The results
show that there is a greater degree of instability in social trust between 13 and 15 years of age
than in other age groups, and that social trust appears to stabilize with age. Findings also indi-
cate that there are substantial inter-individual differences in social trust among young people
within the same age group, both in initial levels and in the rates of change over time. The arti-
cle concludes that although social trust is relatively stable it tends to crystallize in early adult-
hood, highlighting the relevance of the impressionable-years hypothesis.

Introduction
Over the last two decades, scholars have been increasingly interested in
social capital research, especially after Robert Putnam’s (1993) study of
regional government in Italy. In particular, recent research has empha-
sized the role of social trust, which is one of the components of the con-
cept of ‘social capital’. The confidence and belief that strangers (i.e.,
fellow citizens about whom we do not have any specific information) will
not take advantage of us (Uslaner 2002) is presumed to be an invaluable
resource and asset for individuals as well as for society at large. Research
shows that trust is positively related to confidence in political institutions,
democratic satisfaction and political participation (Cigler & Joslyn 2002;
Delhey & Newton 2003; Rothstein & Uslaner 2005). Thus, research on
social trust is of direct relevance for the study of political science.
In work focusing on social trust, a fundamental query in the literature
has to do with the extent to which social trust is either amenable or

* Ali Abdelzadeh, School of Humanities, Education and Social Sciences, Orebro Univer-

sity, SE-701 82, Orebro, Sweden. E-mail: abdelzadeh.ali@gmail.com

Ersta Sk€
ondal University College Stockholm, Sweden.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 207


C 2016 Nordic Political Science Association
V

resistant to change (e.g., Glanville & Paxton 2007; Uslaner 2008; Bauer
2014; Sønderskov & Dinesen 2014; Freitag & Bauer 2016). Two perspec-
tives in the literature have attracted particular attention. The cultural per-
spective stipulates that social trust is a stable and ‘sticky’ phenomenon
learned in early childhood and stable over time. The experiential perspec-
tive suggests that social trust is a product of experience, and that we con-
stantly modify it in response to changing circumstances. Existing empirical
studies lend support to both perspectives (Glanville & Paxton 2007; Uslaner
2008; Dinesen 2011).
However, although these studies have increased our knowledge, our
understanding of the malleability of social trust is still unclear. In addition,
most previous studies have been conducted with adults. Our knowledge is
particularly lacking for young people (see, however, Flanagan & Stout
2010), which is unfortunate, because youth is considered a formative and
critical period for the emergence and formation of social trust and politi-
cal orientations (Hooghe & Wilkenfeld 2008; Hatemi et al. 2009). As
regards adolescence, the ‘impressionable-years hypothesis’ represents a
variety of the cultural and experiential perspective (Osborne et al. 2011;
Dinas 2013; Sears & Brown 2013). According to this hypothesis there is a
critical period in early adulthood when political orientations and attitudes
are open to influence and when political and social attitudes and orienta-
tions are formed. Although there is evidence to support this hypothesis,
few studies have analyzed its relevance with respect to social trust (see,
however, Flanagan & Stout 2010).
This article contributes to this ongoing theoretical discussion by empiri-
cally analyzing how social trust develops among five cohorts of young peo-
ple, ranging from early adolescence to adulthood, in a longitudinal
manner. More specifically, we will answer the following research question:
How does social trust develop among young people and when, if at all,
does it stabilize? To answer this question, we draw from a two-wave longi-
tudinal dataset of five different cohorts of adolescents and young adults
(13, 16, 20, 22 and 26 years old). The target sample comprised approxi-
mately 1,000 individuals in each age group and participants came from a
medium-sized Swedish city with a total population of about 135,000, thus
filling several criteria of national representativeness.
The article is relevant for political science in various ways. First, it adds
to the understanding of the development of orientations among younger
generations of citizens, which has been proven to be politically consequen-
tial for society at large and for intergroup relations. Second, by answering
the question of when trust is most unstable and changeable, the article is
of relevance for the ongoing discussion on how social trust is produced
and reproduced (e.g., Sønderskov & Dinesen 2014). Third, information on
how social trust develops and (if at all) stabilizes is needed to understand

208 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

when social trust can be enhanced. Such knowledge has particular rele-
vance for countries where trust tends to be declining (cf. Twenge et al.
2014).
The article is structured as follows. We present our theoretical points of
departure by drawing on three perspectives on the nature and character
of social trust. This is followed by an explanation of the research design,
including descriptions of our methods and research strategy. We then
present our results, describing the stability and changeability of social trust
and its direction of change. The concluding section summarizes the results
and discusses their further implications for research on social trust.

Theoretical Perspectives on Stability and Change in


Social Trust
For decades, scholars in various fields have provided insights into to the
nature and character of social trust. Two divergent perspectives in the lit-
erature have gained particular attention (Uslaner 2008). The first is the
cultural perspective, which implies that social trust is ‘part of an enduring
culture’ and thus is relatively stable (Dinesen 2012). According to this
view, social trust is closely related to one’s individual personality and is
correlated with personality traits that are stable over the life course (Ash-
ton & Lee 2001; Dinesen et al. 2014). According to Erikson (1950), infants
develop a sense of basic trust in – or mistrust of – the external world
through a consistent and dependable relationship with their parents. It is
formed in the early and oral stages of an infant’s upbringing and tends to
persist throughout life. Thus, social trust is presumed to remain a rela-
tively persistent trait that changes only slowly as a result of later experien-
ces (Rosenberg 1956; Allport 1961).1 In sum, social trust constitutes a
social fact or a ‘sticky’ structure that is very stable over time (Uslaner
2002; Tr€agårdh 2013).
In line with the cultural perspective, empirical studies support the claim
that social trust is a relatively stable phenomenon (e.g., Claibourn &
Martin 2000; Uslaner 2002; Flanagan & Stout 2010; Bekkers 2012). Using
American data from the 1972–1996 General Social Survey, Uslaner (2008)
found that social trust seems to endure over many generations. Further,
making use of a three-generation, longitudinal study design (Youth–Parent
Socialization study), it has been shown that the level of social trust
expressed by a high school senior in 1965 was highly predictive of trust
reported in 1982 (i.e., 17 years later). Further, in a three-wave longitudinal
study, Bekkers (2012) found that social trust is relatively stable over a
four-year period.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 209


C 2016 Nordic Political Science Association
V

In line with these findings, recent twin studies have argued and provided
some support for the idea that social trust may have a genetic component
(e.g., Shikishima et al. 2006; Hiraishi et al. 2008; Oskarsson et al. 2016).
For example, Hatemi et al. (2009), using longitudinal data on twins, have
shown that political attitudes accumulate markedly between the ages of 9
and 17. Recently, Oskarsson et al. (2016) in a study on monozygotic twin
pairs, concluded that the relationship between education and social trust
largely is driven by common genetic factors. Thus, this provides more evi-
dence in line with the cultural perspective.
In contrast, the experiential perspective states that social trust is not so
much a social fact or personality trait as a product of our environment
and experience, and that we modify it in response to changing circumstan-
ces (Putnam 2000). According to Hardin (1992), individuals constantly
modify and update their trustful and distrustful attitudes and feelings due
to changes in society. Research in political psychology refers to this as the
‘lifelong openness model’, which stipulates that attitudes remain open to
influence during one’s life and that individuals constantly modify attitudes
in response to new information and experience. Dinesen (2012) lends sup-
port to this perspective when analyzing how first-generation immigrants
from low-trust countries such as Turkey and Poland are affected by
migrating to high-trust countries in Northern Europe. He concludes that
the destination country context has clear effects on levels of trust of immi-
grants and that immigrants adjust their levels of trust in line with those of
the natives in the new country. Likewise, Glanville and Paxton (2007) con-
clude that trust among adults can be affected by changes in the social
environment and is not wholly determined by past socialization or innate
characteristics. In addition, studies have found that external factors in the
environment influence social trust. Laurence (2015) found that job dis-
placement between the ages of 33 and 50 appears to significantly influence
individuals’ generalized trust negatively, even several years after the event
occurred. Likewise, ethnic fractionalization correlates negatively with
social trust but this effect appears to depend on the quality of government
institution (e.g., Alesina & La Ferrara 2002; Dinesen & Sønderskov 2015).
Moreover, political institutions that are fair, trustful and efficient have a
positive influence on social trust (e.g., Rothstein & Stolle 2003;
Sønderskov & Dinesen 2014; 2016).
The experiential perspective has been criticized for being too extreme,
as certain discontinuities, such as the crystallization of one’s identity and
critical events in life or work, can influence one’s attitudes (Sigel 1989). In
this regard, a growing body of studies shows that political orientations and
behaviour are shaped early in life (Jennings & Niemi 1981; Flanagan &
Sherrod 1998; Torney-Purta et al. 2001). As a challenge to the cultural and
experiential perspective, the impressionable-years hypothesis claims that

210 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

there is a critical period during adolescence when attitudes and political


orientations are particularly open to influence (Osborne et al. 2011; Dinas
2013; Sears & Brown 2013). Thus, political orientations tend to be incom-
pletely crystallized when the individual enters early adulthood, and gains
may be seen during this period. Once individuals enter late adulthood, the
crystallization of political orientations and attitudes begin to show ‘modest
gains thereafter’ (Sears & Brown 2013, 75). The claim is that early adult-
hood is the critical period when people are becoming aware of the social
and political world around them and when identity and the self, as well as
political and social attitudes, are formed. At the same time, it is possible
that political and social orientations that are subject to strong information
flows and are regularly practiced may become stronger or change through-
out life. If social trust is influenced by individuals’ day-to-day interactions,
there may be no sharp discontinuity during adolescence.
Research provides some support for the impressionable-years hypothe-
sis. For example, using a two-wave panel design and focusing on three
cohorts of adolescents (early, middle and late), Flanagan and Stout (2010)
found that social trust was more stable by late adolescence than in early
and middle adolescence. Moreover, using a Youth–Parent Socialization
panel study Jennings and Stoker (2004) found that generalized trust levels
at age 17 are strongly related to trust levels at age 34. This could be inter-
preted as social trust crystallizing during late adulthood. Further, other
studies have shown that during adolescence people actively develop their
political values (Abramson 1972; Hooghe & Wilkenfeld 2008), and that
many political orientations are highly stable among adults (Krosnick &
Alwin 1989; Stolle & Hooghe 2004; Prior 2010; Sherrod et al. 2010).
Three theoretical perspectives relevant for broadening the view on how
social trust develops among adolescents have been presented. However,
despite considerable efforts, additional research is needed before we can
build a comprehensive picture of the malleability of social trust from early
adolescence to young adulthood – a period in life characterized by maxi-
mum change and maximum concentrated learning of political ideas and
societal norms (Niemi & Hepburn 1995; Verba et al. 1995). Previous
research suffers from two important limitations. First, many empirical
studies, especially in the social capital tradition, have based their findings
and conclusions on cross-sectional data rather than longitudinal data. Due
to their cross-sectional nature, many have not been able to provide
answers to the question of stability and change in social trust over time. A
second limitation of previous research on this topic is its strong focus on
adult cohorts (Stolle & Hooghe 2004). As we have pointed out, there are
a number of good reasons to focus on adolescence. Therefore, this article
provides empirical insights into how social trust develops among young
people and when (if at all) it stabilizes.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 211


C 2016 Nordic Political Science Association
V

Method
Data and Measurement
Participants in the study came from a medium-sized Swedish city with a
total population of about 135,000. We have a two-wave longitudinal data-
set from five different cohorts of adolescents and young adults (13, 16, 20,
22 and 26 years old; see Table 1). The first wave of data collection took
place in the spring of 2010/11, and the second approximately two years
later. Data from the two younger cohorts (13 and 16 years old) comes
from 13 schools within the city: ten of the city’s 14 junior high schools
(Grundskolan, in Swedish), and three of its seven senior high schools
(Gymnasieskolan, in Swedish). All seventh graders in the junior high
schools and first year students in the senior high schools were targeted in
the study. The 13 schools were strategically selected in order to represent
both public and private schools and the students’ diverse socioeconomic
and ethnic backgrounds. The students filled out the questionnaires during
regular school hours in their classrooms. To meet the possible objection
that the presence of teachers might influence the responses of the stu-
dents, we ensured that no teachers were in the classrooms during data col-
lection. Trained research assistants distributed the questionnaires and
informed the students that participation in the study was voluntary; the
students were assured of the confidentiality of their responses. No student
was paid for participating in the study, but each class received a contribu-
tion to its class fund.
For the older cohorts (20, 22 and 26 years old), the target sample (i.e.,
all respondents who were supposed to take part in the study) comprised
approximately 1,000 individuals in each age group, who were randomly
selected from a list of all 20–26 year olds living in the city, provided by
the county. In both waves, all participants in the older cohorts were given
the option to fill in the questionnaire in paper form or in an equivalent
online version. The questionnaire was posted to the target sample along
with information about the study and a personalized web link to the
online version of the questionnaire. The participants were informed that
their involvement in the study was voluntary, and they were assured of
the confidentiality of their responses. Participants received a gift certificate
worth 250 SEK (about e27) for taking part in the study.
The motive for choosing the city was related to its national representa-
tiveness with regard to demographic characteristics. According to national
statistics for 2010, the city was similar to the Swedish average in terms of,
for instance, annual income (225,242 SEK per person, compared with
229,056 SEK per person for the whole country), rate of unemployment
(9.5 percent, compared with 8.4 percent for the whole country) and

212 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


Table 1. Information about the Sample: Response Rates and Sociodemographic Characteristics

Number of respondents
(target sample in Response
Year of parentheses) rates (%) Number of Nationality
the first respondents Gender (born outside
Cohort Year of birth survey T1 T2 T1 T2 to both waves (female) Sweden)
V

13 years 1997/98 2010 904 (960) 843 (954) 94.2 88.4 789 50.7% 9.4%

Scandinavian Political Studies, Vol. 40 – No. 2, 2017


16 years 1993/94 2010 892 (1052) 740 (916) 84.8 80.8 610 50.9% 8.8%
20 years 1991/92 2011 605 (980) 600 (987) 61.7 60.8 464 58.7% 7.0%
22 years 1988/89 2010 539 (932) 560 (976) 57.8 57.4 403 60.6% 8.3%
26 years 1985/86 2011 606 (990) 596 (989) 61.2 60.3 477 54.5% 10.9%

213
C 2016 Nordic Political Science Association
C 2016 Nordic Political Science Association
V

percentage of immigrants (5.7 percent, compared with 6.7 percent for the
whole country). For both younger and older cohorts, the time interval
between the two measurement points was approximately two years. Fur-
ther information about the samples is presented in Table 1. As can be
seen, the response rates ranged from 57 to 94 percent, and the samples
are representative in relation to gender in the sense that they reflect the
overall population of interest.
There are of course both advantages and disadvantages with the chosen
data and sampling technique. The non-random sample used in the analysis
is quite common in these types of longitudinal studies (cf. Lazarsfeld et al.
1944; Kiousis & McDevitt 2008; Gastil & Xenos 2010; Mattila et al. 2011;
Persson 2012). However, this sampling strategy does have some potential
limitations. Perhaps most importantly, we have to be careful when gener-
alizing our findings to young people living in other contexts. On the other
hand, the longitudinal design, the excellent participation rates (see Table
1), on-site administration and monitoring of questionnaire completion, the
five different cohorts of young people and the similarity of the sociodemo-
graphic indicators to the national averages are obvious strengths.
Moving on to measurements, to indicate their levels of social trust the
respondents were asked: ‘If you think about people in general, to what
extent do you think that the following statements apply to you?’ The two
statements were: (a) ‘Most people can be trusted’, and (b) ‘Most people
are fair and do not take advantage of you’ (Flanagan et al. 2007; Flanagan
& Stout 2010). The correlation between the two statements ranged from
r 5 0.64 to r 5 0.78, indicating high internal consistency. Participants
responded to the two statements on a five-point scale, ranging from 1
(‘Doesn’t apply at all’) to 5 (‘Applies perfectly’). These two items of social
trust are similar to measures used in many studies, such as the World Val-
ues Survey, Eurobarometer and the European Social Survey.2 As the
overall purpose of the research programme from which the data comes
has been to understand social and political socialization processes among
young people, the dataset employed includes a number of other constructs
focusing on young people in different everyday life contexts, such as fam-
ily and school. Here, however, we only make use of those constructs that
are most relevant to the aim and objectives of this study.

Research Strategy
To examine the stability and change of social trust longitudinally, we per-
formed a number of statistical analyses. First, to examine the stability
coefficients – that is, the correlation between social trust at time-point 1
(T1) and time-point 2 (T2) – we conducted four types of correlations for
each cohort: Kendall’s tau-b, Spearman’s rho rank correlations, Pearson’s

214 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

product moment correlation and a corrected value of Pearson’s correla-


tion. Disattenuated correlations (i.e., the correction of correlations) is a
well-known statistical procedure used to correct parameter estimates for
the attenuating effects of random error of measurement (Wiley & Wiley
1970). To correct for measurement error or unreliability, we applied the
correction for attenuation formula (see Cohen & Cohen 1983), where the
variables were rescaled using information about the observed measure’s
variance and reliability.
Second, to examine the correlations between the initial levels and the
change levels in social trust and address the question about the direction
of change between T1 and T2, we modeled a series of latent change mod-
els (LCMs; McArdle & Nesselroade 1994) using maximum likelihood esti-
mation in Mplus 6 (Muthen & Muthen 199822010). These models, also
known as ‘latent growth curve models’ or ‘latent curve analysis models’
(McArdle & Nesselroade 1994), represent an optimal tool for studying
change over time and focus when using a longitudinal dataset, mainly for
patterns of growth and decline. More specifically, in the current study the
individual growth or decline of social trust was a function of a latent inter-
cept and a latent slope. The latent intercept reflects the average initial
value of social trust at the start of the longitudinal change process. The
latent slope indicates the average individual change rate between the two
time points.
Finally, we examined changes of social trust at individual level using turn-
over tables, which are simple contingency tables spanned by the same vari-
able observed on two measurements occasions (Meiser et al. 1997; Agresti
2013). The reason for using such an approach is to gain a more detailed
understanding of changes in social trust taking place at the individual level
(from T1 to T2). Moreover, when using correlations and latent change mod-
els it is possible that opposing changes taking place at a lower level (i.e., at
the individual level) cancel each other out (cf. Madsen 2004). In other
words, by using turnover tables we gain a more comprehensive overview of
stability and change in social trust between T1 and T2 at an individual level.
More specifically, we tested whether the proportion of young people
increasing in social trust was equivalent in proportion to those loosing social
trust (Engel & Reinecke 1996; Bijleveld et al. 1998; Agresti 2013). This is
referred to as the ‘quasi-symmetry model’ and allows us to detect a change
from T1 to T2 in the distribution across the response categories (i.e., low,
medium and high level of social trust). To evaluate the distance between
the observed distribution of social trust in the quasi-symmetry model, we
used the overall goodness-of-fit statistic likelihood ratio G2 (Von Eye &
Mun 2013). A non-significant value of G2 for the quasi-symmetry model
indicates that the proportion of youths increasing in social trust is same as
the proportion of those decreasing in social trust from T1 to T2.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 215


C 2016 Nordic Political Science Association
V

Results
Stability and Change in Social Trust
In this first part of the results section, we present the findings on stability
and change in social trust over time and cohorts. As shown in Figure 1,
the mean level of social trust is already relatively high (3.29, on a 1–5
scale) at age 13, but has decreased significantly at age 15 to 2.92 (Cohen’s
d 5 0.44). The aggregate level of social trust remains comparatively stable
and low between ages 15 and 18 (Cohen’s d 5 0.03) The low level of social
trust during this life stage might not be so surprising given the fact it co-
occurs with constant social, political, biological and psychological develop-
ments shaping adolescents’ perceptions of themselves and other people
(Elliott & Feldman 1990; Niemi & Hepburn 1995; Hooghe & Wilkenfeld
2008). Social trust increases, however, from 2.89 at age 18 to 3.32 at age
28, which is a statistically significant difference (Cohen’s d 5 0.48). By con-
trast, the mean level of social trust remained quite stable from the ages of
24 and 28 years (Cohen’s d 5 0.12). Overall, the mean level of social trust
seems to be relatively stable across different cohorts and times.
To further examine the stability of social trust, we analyzed the correla-
tions for each cohort between two time points (T1 and T2), which repre-
sent the stability coefficients. As can be seen in Table 2, social trust seems

Figure 1. Means of Social Trust by Age.

5.00

4.50

4.00

3.50 3.29 3.29 3.32


3.21 3.15 3.21
3.09
2.92 2.96 2.98
2.89
3.00

2.50

2.00

1.50

1.00
13 14 15 16 17 18 20 22 24 26 28
Age

Note: Mean levels of social trust are computed using all the information available; thus they also include
young people who participated only in the second wave.

216 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

Table 2. Correlations of Social Trust over Time and Percentages of Youth Providing the
Identical Answers or Moving No More Than One Category Two Years Later

Age 13–15 Age 16–18 Age 20–22 Age 22–24 Age 26–28

Overall measure of social


trust
Pearson’s r 0.382 0.497 0.593 0.571 0.595
Corrected r 0.484 0.603 0.689 0.661 0.687
Kendall’s Tau-b rs 0.304 0.413 0.491 0.450 0.494
Spearman’s rs 0.379 0.503 0.595 0.546 0.592
Item 1: ‘Most people can
be trusted’
Pearson’s r 0.396 0.451 0.589 0.560 0.568
Identical scores (item 1) 45.5 47.9 50.9 46.3 53.3
Identical plus 1/–1 point 88.2 88.7 91.6 89.8 90.9
(item 1)
Item 2: ‘Most people are
fair and do not take
advantage of you’
Pearson’s r 0.283 0.442 0.500 0.498 0.527
Identical scores (item 2) 36.4 45.0 46.1 46.6 52.2
Identical plus 1/–1 point 80.9 88.1 88.0 90.5 91.2
(item 2)

Note: All correlations were significant at p < 0.001.

to increase in stability across the cohorts, irrespective of the type of corre-


lation taken into consideration. For example, corrected correlations
increased from 0.484 among 13 year olds to 0.687 among 26 year olds.
The only exception is cohort four (i.e., ages 22–24), where the stability
coefficient is not increasing compared with the previous cohort (i.e., ages
20–22). However, when comparing the two correlation coefficients (i.e.,
0.593 and 0.571), we found no statistical difference between the two
cohorts (z 5 0.49, p 5 0.62), indicating that the level of stability for cohort
four (ages 22–24) is not significantly lower than that among 20 year olds.
On the other hand, we found that cohort one (ages 13–15) had a signifi-
cantly lower level of stability in social trust than all four other cohorts.
Moreover, Pearson’s r for late adolescents (ages 16–18) was statistically
different from correlations for cohorts three (ages 20–22) and five (ages
26–28), indicating that social trust becomes more stable over two years in
these two older cohorts. The level of stability of social trust was also
reflected in the identical score: younger cohorts show lower percentages of
identical scores between T1 and T2 than older cohorts. In this context, it is
worth mentioning that prior to the analysis undertaken in this section, we
controlled for the possible confounding effect of different levels of educa-
tion among the older cohorts (i.e., 20, 22, 26) on the stability of social
trust. The results showed that there were no substantial changes in the

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 217


C 2016 Nordic Political Science Association
V

results when taking into consideration the effect of education on the cor-
relations between social trust at two measurement points. Thus, the pat-
terns of results remained intact. All in all, regardless of which type of
measure was used to examine the stability pattern of social trust across
cohorts, the results showed a greater degree of instability between 13 and
15 years of age and an increase in stability after late adolescence (ages
16–18).
In a second step, we analyzed each item of social trust separately to see
if there were any differences between the two in terms of stability. As the
lower part of Table 2 shows, the overall correlation coefficients for
the first item (‘Most people can be trusted’) are stronger than those of the
second item of social trust (‘Most people are fair and do not take advant-
age of you’). For example, when comparing the coefficients (i.e., r 5 0.396
and r 5 0.283) for the two items in the youngest cohort, we found a statis-
tically significant difference (z 5 2.43, p < 0.05), indicating that the second
item of trust is less stable. For the three youngest cohorts, these differen-
ces were also reflected in the percentages of identical scores. On the
whole, there seemed to be some differences between the two items of
social trust with respect to stability and change.

What Direction Does the Change Take?


After addressing the question of when social trust stabilizes, we proceeded
to analyze the direction of change in social trust. To this end, we used
LCMs (McArdle & Nesselroade 1994), where measures of social trust at
two time points were included as manifest variables and served as indica-
tors of two latent factors: the intercept, which reflects the mean level of
social trust at the first data occasion; and the slope, which is the amount
of linear change occurring between the two time points. The results from
the LCMs models for each cohort are reported in Table 3. The mean
slope is negative and statistically significant for the two younger cohorts

Table 3. Latent Change Models for Social Trust

Cohort Mean intercept Mean slope Variance intercept Variance slope

(1) 13–15 years 3.344 (0.031)** 20.418 (0.037)** 0.722 (0.038)** 0.984 (0.052)**
(2) 16–18 years 3.024 (0.034)** 20.111 (0.037)* 0.724 (0.041)** 0.824 (0.047)**
(3) 20–22 years 3.173 (0.043)** 20.005 (0.039)ns 0.860 (0.056)** 0.706 (0.046)**
(4) 22–24 years 3.105 (0.045)** 0.163 (0.041)** 0.824 (0.058)** 0.691 (0.049)**
(5) 26–28 years 3.308 (0.040)** 0.045 (0.035)ns 0.780 (0.051)** 0.600 (0.039)**

Note: Entries are unstandardized coefficients (standard errors in parentheses). *p < 0.01;
**p < 0.001.

218 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

(ages 13–16 and 16–18), indicating a decrease in social trust across the two
years. Moreover, the variance estimates pointed to considerable inter-
individual differences in both cohorts, as reflected in both initial levels
(variance intercept) and rates of change (variance slope). With regard to
the three older cohorts, only in cohort four was there a significant and
positive mean slope change, indicating an increase in social trust over the
two years. However, variance estimates of both intercept and slope
revealed substantial inter-individual differences within each cohort. In
sum, there is an overall decrease in social trust among adolescents aged 13
to 18. The decrease is much stronger for the youngest cohort (aged
13–15). There is no significant change in social trust for the older cohorts,
except in cohort four, where we see an increase.

Individual Direction of Change in Social Trust


Following the research strategy outlined above, we used turnover tables to
show the direction of the inter-individual variations in social trust. For
easier inspection of the tables, we recoded our two items of social trust
into a trichotomy: High (including respondents who, concerning state-
ments about social trust, reported ‘applies perfectly or quite well’);
Medium (those who reported ‘kind of applies’); and Low (those who
reported ‘doesn’t apply at all or so well’).
The results of these turnover tables are presented in Tables 4 and 5. As
shown, the quasi-symmetry model adequately fits the data in all cohorts.
And this applies for both measures of social trust. So, when taking into
consideration the different distributions of social trust at T1 and T2 (i.e.,
the marginal distributions), the probability of changes can be assumed to
be equal in both directions. Put differently, the proportion of young peo-
ple reporting a decrease in social trust (i.e., moving from high to low lev-
els) is almost the same as that of those reporting an increase (i.e., moving
from low to high levels). In the same way, the proportion of young people
moving from medium to low levels is very similar to that of those moving
from low to medium levels. The same logic also applies to movements
between any other combinations of categories. At the same time, even
these results show lower stability of social trust among the younger
cohorts. For example, Table 4 shows that about 20 percent of the partici-
pants in the first cohort (ages 13–15) remains in the ‘high’ category from
T1 to T2, compared with 36 percent for the fifth cohort (ages 26–28). All
in all, despite the increased stability of social trust with age, the results
from the turnover tables indicate that changes are taking place in young
people’s levels of social trust at the individual level.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 219


V

220
Table 4. Turnover Tables for the First Item of Social Trust and Goodness-of-fit Statistic Likelihood Ratio G2

‘Most people
can be trusted’ Age 13–15 Age 16–18 Age 20–22 Age 22–24 Age 26–28

T2
T1 High Medium Low High Medium Low High Medium Low High Medium Low High Medium Low
C 2016 Nordic Political Science Association

High 19.8 17.3 7.7 17.9 9.4 3.4 30.6 8.0 2.6 27.9 11.2 1.2 36.2 10.3 1.7
[44.1] [38.6] [17.3] [58.3] [30.5] [11.2] [74.3) [19.4] [6.3] [69.1] [27.8] [3.1] [75.1] [10.3] [3.5]
Medium 8.3 17.0 12.8 10.7 17.6 14.3 10.3 17.7 7.3 14.4 13.2 5.7 10.3 16.2 4.8
[21.7] [44.6] [12.8] [25.1] [41.3] [33.6] [29.3] [50.0] [20.7] [43.3] [39.6] [17.2] [32.9] [51.7] [15.4]
Low 1.1 5.5 10.5 3.3 8.4 15.1 2.4 9.1 12.1 4.5 9.2 12.7 3.8 6.9 9.7
[6.5] [32.3] [61.3] [12.3] [31.3) [56.4] [10.1] [38.5] [51.4] [17.0] [34.9] [48.1] [18.6] [34.0] [47.4]
G2 Quasi-symmetry 0.70 (1) ns 0.77 (1) ns 1.14 (1) ns 0.87 (1) ns 0.70 (1) ns

Notes: Numbers in the upper half of the table are percentages. Row percentages in brackets. Degrees of freedom in parentheses. *p < 0.05;
**p < 0.01; ***p < 0.001.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017


Table 5. Turnover Tables for the Second Item of Social Trust and Goodness-of-fit Statistic Likelihood Ratio G2

‘Most people are fair


and do not take
advantage of you’ Age 13–15 Age 16–18 Age 20–22 Age 22–24 Age 26–28

T2
T1 High Medium Low High Medium Low High Medium Low High Medium Low High Medium Low

High 19.4 19.1 9.1 18.8 9.2 4.3 26.0 11.0 3.9 26.6 11.0 1.3 37.9 11.5 2.5
[40.7] [40.1] [19.2] [58.2] [28.6] [13.3] [63.5] [27.0] [9.5] [68.4) [28.4] [3.2] [73.0] [22.2] [4.8]
V

Medium 9.4 17.3 11.5 8.9 21.6 12.9 10.0 17.3 8.2 13.3 17.8 5.8 10.7 13.8 5.0

Scandinavian Political Studies, Vol. 40 – No. 2, 2017


[24.6) [45.3] [30.1] [20.5] [49.8) [29.7] [28.0] [48.8] [23.2] [36.1] [48.3] [15.6] [36.2] [46.8) [17.0]
Low 1.7 5.0 7.6 2.6 8.7 13.0 3.0 8.7 11.9 4.8 9.5 10.0 2.7 7.5 8.2
[11.7) [35.0] [53.4] [10.8) [35.8) [53.4] [12.8] [36.7] [50.7] [19.6] [39.2] [41.2] [14.8] [40.9] [44.3]
G2 Quasi-symmetry 0.16 (1) ns 0.02 (1) ns 0.18 (1) ns 1.12 (1) ns 0.23(1) ns

Notes: Numbers in the upper half of the table are percentages. Row percentages in brackets. Degrees of freedom in parentheses. *p < 0.05;
**p < 0.01; ***p < 0.001.

221
C 2016 Nordic Political Science Association
C 2016 Nordic Political Science Association
V

Discussion
The aim of the study was to contribute to a better understanding of how
social trust develops and when (if at all) social trust stabilizes between ado-
lescence and young adulthood. Using unique longitudinal data from five
different cohorts, ranging in age from 13 to 28 years, our study contributes
in several ways to increasing the knowledge on social trust in general, and
its malleability in particular. The results show that social trust appears to
stabilize as individuals mature. Regardless of type of measure, the results
showed a greater degree of instability between 13 and 15 years of age and
an increase in stability after late adolescence (ages 16–18). However, we
also showed that there are indeed substantial inter-individual differences in
social trust among young people within the same age group, both in initial
levels and in the rates of change over biannual time points. In addition,
the results indicated that the proportion of youth whose trust decreases is
equal to that of youth whose trust increases over time.
The finding that social trust appears to stabilize as individuals mature
and shows an increase in stability after late adolescence tallies well with
the impressionable-years hypothesis that political orientations tend to be
incomplete during early adulthood and crystallize when individuals enter
late adulthood (Sears & Brown 2013, 75). In addition, it supports
Flanagan and Stout and other scholars who found that social trust was
more stable by late adolescence than in early and middle adolescence
(cf. Jennings & Stoker 2002; 2004; Flanagan & Stout 2010) and a growing
number of studies showing that various political behaviours and attitudes,
such as political trust, political interest and tolerance, are shaped early
in life (e.g., Flanagan & Sherrod 1998; Torney-Purta et al. 2001; Hooghe
& Wilkenfeld 2008; Miklikowska 2016; Russo & Stattin 2016).
However, it is important to note that the findings do not wholly elimi-
nate the relevance of a cultural perspective. It could be argued that the
changes in trust during adolescence reported above are relatively modest,
supporting the view that social trust is a relatively stable and ‘sticky’ atti-
tude (Uslaner 2002). We have also shown that although social trust
appears to develop in line with the impressionable-years hypothesis, it is
not entirely fixed for the cohorts of interest for the current study, lending
support to the experiential perspective. Indeed, more research is needed
to explore what can be regarded as stable in relation to social trust and
how much impact respondents’ ambivalence has on the changes taking
place. That being said, our findings point to the significance of the
impressionable-years hypothesis for understanding the development of
adolescents’ social trust.
Finally, the results also revealed differences between the two measures
of social trust with regard to their malleability. As became apparent from

222 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

the results, the second measure, asking whether most people are fair and
do not take advantage, seemed to be more amenable to change than the
first measure (i.e., ‘most people can be trusted’). These systematic changes
occurring between two measurement occasions have not been captured in
earlier studies, probably due to the use of correlation coefficients in esti-
mating continuity and stability. It is possible that the response distribu-
tions shift upward or downward, while stability coefficients nonetheless
remain high (Madsen 2004). This finding implies that in order to better
understand stability in social trust, correlations alone are not a sufficient
means for estimating stability; rather they should be complemented with
other statistical tools, such as contingency tables.
This study has some limitations and strengths that warrant attention.
One potential limitation concerns the use of only two waves of data for
each cohort, meaning that we were not able to examine how social trust
develops over a longer period of time. However, the information about
stability and change across a wide age range offers some indication of
how and when social trust might develop as young people grow older. A
second limitation concerns the use of only two items to measure social
trust. Ideally, a construct with multiple measures of the same phenomenon
is preferable to single-item scales. However, it should be added that the
items used in the current study are similar to those used in numerous
studies and international surveys. A final potential limitation concerns
whether findings from our sample can be generalized to a wider popula-
tion (Wallman Lundåsen & Wollebæk 2013). The data used in this study
come from a community-based sample in a medium-sized Swedish city. It
might therefore be argued that findings cannot be generalized from this
particular context. Ideally, randomly selected samples drawn from the gen-
eral population are to be preferred. Nonetheless, the population of the
city from which the samples were drawn is, on average, very similar to the
national population on most basic demographic characteristics (e.g.,
income, parents’ immigrant status and unemployment). At the same time,
the research project from which the data in the current study were
obtained employs a community-based approach, since it is more suitable
and practically feasible given the longitudinal design of the project. Bear-
ing this in mind, the data used in the study seem to be representative of
the general population on several relevant characteristics for the relevant
age group, and have adequate validity. Related to that, it might be argued
that because of the different institutional arrangements and contextual set-
tings in democratic societies, the findings from the specific Swedish case
cannot be generalized to other countries. Ideally, of course, the findings
should be replicable in other democratic societies. At the same time, while
the purely empirical findings in the current study may differ from those in
other countries, it is reasonable to assume that the theoretical

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 223


C 2016 Nordic Political Science Association
V

contributions are not context-specific and dependent to the same degree.


To clarify further, even if, for example, levels of social trust are very likely
to vary across democratic societies, there is good reason to believe that
the patterns of results regarding the stability of social trust might be simi-
lar in various political settings.
The study also has several strengths. First and foremost, by adopting a
longitudinal design and covering a broad period range of young people’s
ages (i.e., 13 to 28), we have been able to add new insights to the research
on social trust about when it starts to stabilize during adolescence and
young adulthood. Knowing when social trust starts to stabilize facilitates
the selection of different explanatory factors believed to influence the var-
iations in the variable in question. Our results suggest that if our aim is to
fully understand how social trust is produced and reproduced, we must
direct our attention to the study of young people – perhaps as young as
13 years old, since our results showed a greater degree of instability
between 13 and 15 years of age. However, given the importance of social
trust for both individuals and society at large, this knowledge is not only
vital for a better understanding of the origins of social trust, it is also of
value for voluntary associations and governmental institutions that aim to
boost social trust among young people through various programmes and
interventions. With better knowledge of stability and change in social trust
across different age groups, governments and nongovernmental actors
have a better chance of strengthening social trust, particularly among the
younger generations where levels of social trust appear to be significantly
lower than in the population as a whole.

ACKNOWLEDGEMENTS
This study was made possible by access to data from the Political Socialization Programme, a
longitudinal research programme at YeS (Youth & Society) at Orebro€ University, Sweden.
Responsible for the planning, implementation and financing of the collection of data in this pro-
ject were Professors Erik Amnå, Mats Ekstr€ om, Margaret Kerr and Håkan Stattin. The data
collection was supported by grants from the Bank of Sweden Tercentenary Foundation. The
authors of this article would like to thank the anonymous reviewers for valuable comments on
earlier drafts. Note that the authors’ names are in alphabetical order and they share equal
responsibility for the article.

NOTES
1. A closely related argument as to why social trust is stable notes that trust is affected and
reinforced by our interactions with other people. First, people who trust others are more
likely to be trusted in turn. Second, while people who trust others perceive lots of evi-
dence to support their view that people can be trusted in general, distrusters are less
likely to become more trusting in their interactions since they suspect that others have
self-serving strategic or selfish motives behind their interactions with others (Rotter
1967; Bekkers 2012).
2. In some cases the stem question is the same, while the items may have different response
scale formats.

224 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

REFERENCES
Abramson, P. R. 1972. ‘Political Efficacy and Political Trust among Black Schoolchildren: Two
Explanations’, Journal of Politics, 34, 1243–69.
Agresti, A. 2013. Categorical Data Analysis. Hoboken, NJ: John Wiley & Sons.
Alesina, A. & La Ferrara, E. 2002. ‘Who Trusts Others?’, Journal of Public Economics, 85,
207–34.
Allport, G. W. 1961. Pattern and Growth in Personality. Oxford: Holt, Reinhart & Winston.
Ashton, M. C. & Lee, K. 2001. ‘A Theoretical Basis for the Major Dimensions of Personality’,
European Journal of Personality, 15, 327–53.
Bauer, P. C. 2014. ‘Negative Experiences and Trust: A Causal Analysis of the Effects of Vic-
timization on Generalized Trust’, European Sociological Review, DOI: 10.1093/esr/jcu096.
Bekkers, R. 2012. ‘Trust and Volunteering: Selection or Causation? Evidence from a 4 Year
Panel Study’, Political Behavior, 34, 225–47.
Bijleveld, C. C. J. H., Van der Kamp, L. J. T., Mooijaart, A., Van der Kloot, W. A., Van der
Leeden, R. & Van der Burg, E. 1998. Longitudinal Data Analysis: Designs, Models and
Methods. London: Sage.
Cigler, A. & Joslyn, M. R. 2002. ‘The Extensiveness of Group Membership and Social Capital:
The Impact on Political Tolerance Attitudes’, Political Research Quarterly, 55, 7–25.
Claibourn, M. P. & Martin, P. S. 2000. ‘Trusting and Joining? An Empirical Test of the Recip-
rocal Nature of Social Capital’, Political Behavior, 22, 267–91.
Cohen, J. & Cohen, P. 1983. Applied Multiple Regression/Correlation Analysis for the Behav-
ioral Sciences. Mahwah, NJ: Lawrence Erlbaum.
Delhey, J. & Newton, K. 2003. ‘Who Trusts? The Origins of Social Trust in Seven Societies’,
European Societies, 5, 93–137.
Dinas, E. 2013. ‘Opening “Openness to Change” Political Events and the Increased Sensitivity
of Young Adults’, Political Research Quarterly, 66, 868–82.
Dinesen, P. T. 2011. ‘A Note on the Measurement of Generalized Trust of Immigrants and
Natives’, Social Indicators Research, 103, 169–77.
Dinesen, P. T. 2012. ‘Does Generalized (Dis)Trust Travel? Examining the Impact of Cultural
Heritage and Destination-country Environment on Trust of Immigrants’, Political Psychol-
ogy, 33, 495–511.
Dinesen, P. T. & Sønderskov, K. M. 2015. ‘Ethnic Diversity and Social Trust Evidence from
the Micro-context’, American Sociological Review. DOI: 10.1177/0003122415577989.
Dinesen, P. T., Nørgaard, A. S. & Klemmensen, R. 2014. ‘The Civic Personality: Personality
and Democratic Citizenship’, Political Studies, 62, 134–52.
Elliott, G. R. & Feldman, S. S. 1990. ‘Capturing the Adolescent Experience’, in Feldman, S. S.
& Elliott, G. R., eds, At the Threshold: The Developing Adolescent. Cambridge, MA: Har-
vard University Press.
Engel, U. & Reinecke, J. 1996. Analysis of Change: Advanced Techniques in Panel Data Anal-
ysis. Berlin: Walter de Gruyter.
Erikson, E. H. 1950. Childhood and Society. New York: Norton.
Flanagan, C. A. & Sherrod, L. R. 1998. ‘Youth Political Development: An Introduction’, Jour-
nal of Social Issues, 54, 447–56.
Flanagan, C. A. & Stout, M. 2010. ‘Developmental Patterns of Social Trust between Early and
Late Adolescence: Age and School Climate Effects’, Journal of Research on Adolescence,
20, 748–73.
Flanagan, C. A., Syvertsen, A. K. & Stout, M. 2007. Civic Measurement Models: Tapping Ado-
lescents’ Civic Engagement. College Park, MD: Center for Information and Research on
Civic Learning and Engagement.
Freitag, M. & Bauer, P. C. 2016. ‘Personality Traits and the Propensity to Trust Friends and
Strangers’, Social Science Journal. DOI: 10.1016/jsoscij201512002.
Gastil, J. & Xenos, M. 2010. ‘Of Attitudes and Engagement: Clarifying the Reciprocal Rela-
tionship between Civic Attitudes and Political Participation’, Journal of Communication, 60,
318–43.
Glanville, J. L. & Paxton, P. 2007. ‘How Do We Learn to Trust? A Confirmatory Tetrad Anal-
ysis of the Sources of Generalized Trust’, Social Psychology Quarterly, 70, 230–42.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 225


C 2016 Nordic Political Science Association
V

Hardin, R. 1992. ‘The Street-level Epistemology of Trust’, Analyse & Kritik, 14, 152–76.
Hatemi, P. K., Funk, C. L., Medland, S. E., Maes, H. M., Silberg, J. L., Martin, N. G. & Eaves,
L. J. 2009. ‘Genetic and Environmental Transmission of Political Attitudes over a Life
Time’, Journal of Politics, 71, 1141–56.
Hiraishi, K., Yamagata, S., Shikishima, C. & Ando, J. 2008. ‘Maintenance of Genetic Variation
in Personality through Control of Mental Mechanisms: A Test of Trust, Extraversion and
Agreeableness’, Evolution and Human Behavior, 29, 79–85.
Hooghe, M. & Wilkenfeld, B. 2008. ‘The Stability of Political Attitudes and Behaviors across
Adolescence and Early Adulthood: A Comparison of Survey Data on Adolescents and
Young Adults in Eight Countries’, Journal of Youth and Adolescence, 37, 155–67.
Jennings, M. K. & Niemi, R. G. 1981. Generations and Politics: A Panel Study of Youth Adults
and Their Parents. Princeton, NJ: Princeton University Press.
Jennings, M. K. & Stoker, L. 2002. ‘Generational Change, Life Cycle Processes and Social
Capital’. Paper presented at the workshop ‘Citizenship on Trial: Interdisciplinary Perspec-
tives on the Political Socialisation of Adolescents’, McGill University.
Jennings, M. K. & Stoker, L. 2004. ‘Social Trust and Civic Engagement across Time and Gen-
erations’, Acta Politica, 39, 342–79.
Kiousis, S. & McDevitt, M. 2008. ‘Agenda Setting in Civic Development Effects of Curricula
and Issue Importance on Youth Voter Turnout’, Communication Research, 35, 481–502.
Krosnick, J. A. & Alwin, D. F. 1989. ‘Aging and Susceptibility to Attitude Change’. Journal of
Personality and Social Psychology, 57, 416–25.
Laurence, J. 2015. ‘(Dis)Placing Trust: The Long-term Effects of Job Displacement on Gener-
alised Trust over the Adult Lifecourse’, Social Science Research, 50, 46–59.
Lazarsfeld, P. F., Berelson, B. & Gaudet, H. 1944. The People’s Choice: How the Voter Makes
Up His Mind in a Presidential Campaign. New York: Duell, Sloan & Pearce.
Madsen, D. 2004. ‘Stability Coefficient’, in Lewis-Beck, M. S., Bryman, A. & Liao, T. F., eds,
The Sage Encyclopedia of Social Science Research Methods. Thousand Oaks, CA: Sage.
Mattila, M., Wass, H., S€ oderlund, P., Fredriksson, S., Fadjukoff, P. & Kokko, K. 2011. ‘Person-
ality and Turnout: Results from the Finnish Longitudinal Studies’, Scandinavian Political
Studies, 34, 287–306.
McArdle, J. J. & Nesselroade, J. R. 1994. ‘Using Multivariate Data to Structure Developmen-
tal Change’, in Cohen, S. H. & Reese, H. W., eds, Life-span Developmental Psychology:
Methodological Contributions. Mahwah, NJ: Lawrence Erlbaum.
Meiser, T., Von Eye, A. & Spiel, C. 1997. ‘Loglinear Symmetry and Quasi-symmetry Models
for the Analysis of Change’. Biometrical Journal, 39, 351–368.
Miklikowska, M. 2016. ‘Like Parent, Like Child? Development of Prejudice and Tolerance
towards Immigrants’, British Journal of Psychology, 107, 95–116.
Muthen, L. K. & Muthen, B. O. 199822010. Mplus User’s Guide. Los Angeles, CA: Muth en
& Muthen.
Niemi, R. G. & Hepburn, M. A. 1995. ‘The Rebirth of Political Socialization’. Perspectives on
Political Science, 24, 7–16.
Osborne, D., Sears, D. O. & Valentino, N. A. 2011. ‘The End of the Solidly Democratic South:
The Impressionable-years Hypothesis’. Political Psychology, 32, 81–108.
Oskarsson, S., Dinesen, P. T., Dawes, C. T., Johannesson, M. & Magnusson, P. 2016. ‘Educa-
tion and Social Trust: Testing a Causal Hypothesis Using the Discordant Twin Design’.
Political Psychology. DOI: 10.1111/pops.12343.
Persson, M. 2012. ‘Does Type of Education Affect Political Participation?
Results from a Panel Survey of Swedish Adolescents’, Scandinavian Political Studies, 35,
198–221.
Prior, M. 2010. ‘You’ve Either Got It or You Don’t? The Stability of Political Interest over
the Life Cycle’, Journal of Politics, 72, 747–66.
Putnam, R. D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton,
NJ: Princeton University Press.
Putnam, R. D. 2000. Bowling Alone: The Collapse and Revival of American Community. New
York: Simon & Schuster.
Rosenberg, M. 1956. ‘Misanthropy and Political Ideology’, American Sociological Review, 21,
690–5.

226 Scandinavian Political Studies, Vol. 40 – No. 2, 2017


C 2016 Nordic Political Science Association
V

Rothstein, B. & Stolle, D. 2003. ‘Introduction: Social Capital in Scandinavia’, Scandinavian


Political Studies, 26, 1–26.
Rothstein, B. & Uslaner, E. M. 2005. ‘All for All: Equality, Corruption and Social Trust’,
World Politics, 58, 41–72.
Rotter, J. B. 1967. ‘A New Scale for the Measurement of Interpersonal Trust’, Journal of Per-
sonality, 35, 651–65.
Russo, S. & Stattin, H. 2016. ‘Stability and Change in Youths’ Political Interest’, Social Indica-
tors Research. DOI: 10.1007/s11205-016-1302-9.
Sears, D. O. & Brown, C. 2013. ‘Childhood and Adult Political Development’, in Sears, D. O.,
Huddy, L. & Jervis, R., eds, Oxford Handbook of Political Psychology. Oxford: Oxford Uni-
versity Press.
Sherrod, L. R., Torney-Purta, J. & Flanagan, C. A. eds. 2010. Handbook of Research on Civic
Engagement in Youth. Chichester: Wiley.
Shikishima, C., Hiraishi, K. & Ando, J. 2006. ‘Genetic and Environmental Influences on Gen-
eral Trust: A Test of a Theory of Trust with Behavioral Genetic and Evolutionary Psycho-
logical Approaches’, Japanese Journal of Social Psychology, 22, 48–57.
Sigel, R. S. 1989. Political Learning in Adulthood: A Sourcebook of Theory and Research.
Chicago, IL: University of Chicago Press.
Sønderskov, K. M. & Dinesen, P. T. 2014. ‘Danish Exceptionalism: Explaining the Unique
Increase in Social Trust over the Past 30 Years’, European Sociological Review, 80, 550–73.
Sønderskov, K. M. & Dinesen, P. T. 2016. ‘Trusting the State, Trusting Each Other? The
Effect of Institutional Trust on Social Trust’, Political Behavior, 38, 179–202.
Stolle, D. & Hooghe, M. 2004. ‘The Roots of Social Capital: Attitudinal and Network Mecha-
nisms in the Relation between Youth and Adult Indicators of Social Capital’, Acta Politica,
39, 422–41.
Torney-Purta, J., Lehmann, R., Oswald, H. & Schulz, W. 2001. ‘Citizenship and Education in
Twenty-eight Countries: Civic Knowledge and Engagement at Age Fourteen’. Amsterdam:
International Association for the Evaluation of Educational Achievement.
Tr€agårdh, L. 2013. ‘The Historical Incubators of Trust in Sweden: From the Rule of Blood to
the Rule of Law’, in Reuter, M., Wijkstrom, F. & Uggla, B. K., eds, Trust and Organizations:
Confidence across Borders. Basingstoke: Palgrave Macmillan.
Twenge, J. M., Campbell, W. K. & Carter, N. T. 2014. ‘Declines in Trust in Others and Confi-
dence in Institutions among American Adults and Late Adolescents, 1972–2012’. Psycholog-
ical Science, 25, 1914–23.
Uslaner, E. M. 2002. The Moral Foundations of Trust. Cambridge: Cambridge University
Press.
Uslaner, E. M. 2008. ‘Where You Stand Depends Upon Where Your Grandparents Sat: The
Inheritability of Generalized Trust’, Public Opinion Quarterly, 72, 725–40.
Verba, S., Schlozman, K. L. & Brady, H. E. 1995. Voice and Equality: Civic Voluntarism in
American Politics. Cambridge, MA: Harvard University Press.
Von Eye, A. & Mun, E.-Y. 2013. Log-linear Modeling: Concepts, Interpretation and Applica-
tion. Chichester: Wiley.
Wallman Lundåsen, S. & Wollebæk, D. 2013. ‘Diversity and Community Trust in Swedish
Local Communities’, Journal of Elections, Public Opinion and Parties, 23, 299–321.
Wiley, D. E. & Wiley, J. A. 1970. ‘The Estimation of Measurement Error in Panel Data’,
American Sociological Review, 35, 112–17.

Scandinavian Political Studies, Vol. 40 – No. 2, 2017 227


Copyright of Scandinavian Political Studies is the property of Wiley-Blackwell and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.

You might also like