Content uploaded by Luiz de Mello
Author content
All content in this area was uploaded by Luiz de Mello on Jul 05, 2016
Content may be subject to copyright.
Journal of Economic and Social Research Vol 16(1) 2014, 1-34
Households’ Intergenerational Structure and
Participation in Associational Activities: The Role of
Interpersonal Trust
1
Luiz de Mello
*
and João Tovar Jalles
**
Abstract. Social norms in general and interpersonal trust in particular are related to
economic performance by affecting the cost of market transactions. However,
economic literature is virtually silent on how trust is transmitted among individuals.
Building on empirical work on the socialisation of social norms, this paper focuses
on the transmission of trust within the household (vertical transmission) and beyond,
through associational activities in the professional, cultural and social domains
(horizontal transmission). Using data from the latest wave of the World Values
Survey, we show that the intergenerational structure of households, which shapes
how trust is transmitted at home, affects the probability of engagement in different
associational activities outside the household. The paper also contributes to the
literature by estimating the probability of participation in associational activities and
interpersonal trust in a multivariate probit setting.
JEL Classification Codes: C35, O11, J10, Z13
Keywords: Trust, Associational Organisations, Model Selection, Multivariate Probit
1
The usual disclaimer applies. The views expressed are those of the authors and do not
necessarily reflect those of the OECD, the IMF, their member countries or their policy.
*
OECD, Secretary General Office, 2 rue Andre Pascal, 75775 Paris CEDEX 16, France.
email: luiz.demello@oecd.org
**
Corresponding Author: Center for Globalization and Governance, Nova School of Business
and Economics, 1099-032 Lisboa, Portugal. email : joaojalles@gmail.com
Luiz de Mello and João Tovar Jalles
2
1. Introduction
A large literature has emerged over the last fifteen years or so on the
economic benefits of interpersonal trust and, more generally, social norms,
including Knack and Keefer (1997), La Porta et al. (1997), Temple and
Johnson (1998) and Glaeser et al. (2000). The main argument underpinning
this literature is that trustfulness among strangers reduces transactions costs,
avoids harmful conflict, facilitates the establishment of production networks
and enables cooperation, factors which in turn improve economic
performance. However, while the economic literature has focused on the
individual and societal characteristics that are most closely related to
trustfulness or a preference for cooperation, such as educational attainment,
labour market status, age, ethnicity and religion (Alesina and La Ferrara,
2002), considerably less emphasis has been placed on the mechanisms and
channels through which social norms and interpersonal trust are actually
transmitted among individuals, a process that is often referred to as
socialisation.
A separate literature on socialisation focuses on how social norms
and interpersonal trust are transmited within the household (vertical or
within-household socialisation) and/or beyond, through (horizontal or
outside-household socialisation) associational activities in the professional,
cultural and social domains (Bisin and Verdier, 2010). This literature shows
that trustfulness is socialised through a combination of social preferences
and norms, including different types of altruism and aversion to inequity;
expectation of trustworthiness and reciprocity; and incentives, such as those
created by reputation. La Porta et al. (1997) argue that trust is more
conducive to cooperation among strangers, whereas reputation and other
forces contribute to cooperation among individuals who interact frequently,
such as family members.
2
Ashraf et al. (2006) argue that social preferences,
including altruism, are important determinants of trust, rather than the
calculative motives associated with expectations of trustworthiness and
reciprocity. This is in line with Tabellini (2008), who shows that not only
behaviour, but also values, determines the scope for interpersonal trust.
The intergenerational structure of households affects the
transmission of social values and norms within the household. Transmission
is thought to be costlier in two- than in three-generation families, where
2
Banfield (1958) discusses the “ amoral familism” in which strong family ties may induce a
social equilibrium where people exclusively care about their immediate family.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
3
socialisation efforts can be shared between parents and grandparents (Frank,
1992). Individuals in three-generation families may be less likely to engage
in different forms of horizontal socialisation (transmission outside the
household), such as participation in humanitarian organisations, sports and
arts associations, among others, while controlling for relevant individual and
household characteristics. Such vehicles for horizontal socialisation have
been used as direct metrics for social capital in the economic literature
3
,
while neglecting the relationship that exists between vertical and horizontal
socialisation in the transmission of social norms.
To shed light on the transmission of interpersonal trust, we focus on
the intergenerational structural of households and use data from the latest
wave (2005) of the World Values Survey (WVS). While previous studies
have controlled for individual and household characteristics that are known
to affect the probability of participation in associational activities outside the
household, we focus on the transmission of interpersonal trust by looking at
the interactions between vertical socialisation (based on intergenerational
household structure) and horizontal socialisation (measured as participation
in a variety associational activities outside the household).
Our empirical analysis also takes into account the fact that
individuals are free to decide on whether or not to engage in associational
activities outside the household and, if so, on which activities to engage in.
We do so by using a multivariate probit setting to model the probability of
participation in a host of associational activities and to estimate the effect of
such engagement on interpersonal trust, while taking account of the
intergenerational structure of households, as well as conventional individual
and household characteristics. We also control for cohort and country
effects in the probit regressions to account for differences in cultural traits.
On the basis of the survey-based empirical evidence, it appears that
household structure affects the probability of engagement in different
associational activities. In particular:
3
One important aspect to bear in mind throughout the paper is that a key feature in the
existing literature on social capital is its vagueness (Solow, 1999; Guiso et al., 2008). One
distinction between types of social capital opposes bonding and bridging social capital
(Granovetter, 1973; Putnam 2000). Another distinction is between formal and informal social
capital (Putnam, 2000). An ulterior distinction is between the structural and cognitive
dimensions of social capital (Krishna and Uphoff, 1999; Kassa and Parts, 2008).
Luiz de Mello and João Tovar Jalles
4
• Vertical transmission, depending on the intergenerational structure
of households, affects participation in different types of associational
activities. Individuals residing in three-generational households are more
likely to engage in humanitarian/charitable organisations and less likely to
engage in lifestyle associations, such as sport and recreational associations.
By contrast, individuals living in two-generational households are more
likely to participate in both art, music and education association and lifestyle
organisations (sports/recreational organisations).
• Not all associational activities are conducive to interpersonal trust-
building. Participation in humanitarian/charitable and art/music/educational
organisations appears to be linked to the transmission of trust, whereas
engagement in professional associations does not. Engagement with
religious or sports/recreational organisations is uncorrelated with
interpersonal trust.
• Failure to account for the correlations in the error terms of the
regressions describing the probability of engagement in the different
organisations biases the estimation of the effect of horizontal socialisation on
interpersonal trust. In particular, when the errors are assumed to be
uncorrelated, participation in religious or sports/recreational organisations
appears to be correlated with interpersonal trust, whereas engagement in
professional associations does not seem to affect the transmission of
interpersonal trust.
This paper is structured as follows. Section 2 discusses the literature
on cultural transmission and interpersonal trust. Section 3 addresses
methodological issues and defines the dataset, whereas section 4 reports the
main findings. Section 5 concludes.
2. The literature on transmission of social norms and interpersonal trust
Preferences, beliefs and social norms can be transmitted among individuals
within the household (vertical socialisation) and/or beyond it, through social
interactions (oblique/horizontal socialisation) in religious, professional,
lifestyle and educational associations, for example. The early theoretical
literature on cultural transmission and socialisation – surveyed recently by
Bisin and Verdier (2010) – is based on the evolutionary biology models by
Cavalli-Sforza and Feldman (1981) and Boyd and Richardson (1985).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
5
Vertical socialisation can occur through imperfect empathy (Bisin and
Verdier, 2001; Bisin et al., 2004), when parents are altruistic about passing
on cooperative values to their children, while also deriving utility from their
children’s cooperative behaviour. The argument is based on two strands of
literature: experimental studies based on strategic games, such as variations
of the Prisoner’s Dilemma and the Dictator’s Game, to assess cooperation
among self-interested individuals, and evolutionary games, arguing that
some individuals are “encoded” to cooperate despite the costs that
cooperation may entail. These studies suggest that norms and preferences,
such as kin altruism, reciprocity and group selection, in addition to
behavioural rules, drive individuals to cooperate. In other words, parents are
altruistic about the transmission of values and norms but evaluate their
children’s choices through the lenses of their own preferences.
By contrast, Guttman (2001a and 2001b) argues that preferences for
cooperation are transmitted across generations, because parents derive utility
and material return from investing in their children. Cooperation generates a
material payoff and cooperative children take better care of their parents.
This is the case, for instance, of deferred or indirect reciprocity: Cox and
Stark (1994) and Arrondel and Masson (2001) show that forward-looking
parents (the middle generation in families of three overlapping generations
with two dependent generations) help their own parents to set an example for
their children and secure old-age support for themselves in the form of care,
attention or financial transfers. Jellal and Wolff (2002a and 2002b) also
provide empirical evidence of intra-family intergenerational transfers that
take place through demonstration effects.
4
Oblique/horizontal socialisation is closer to the idea of social capital
formation outside the household. In turn, this type of socialisation is related
to the two major strands of literature on social capital: one is based on the
transmission of social norms among individuals (Coleman, 1988, 1990;
Bourdieu, 1986), and the other is society-centric (Putnam, 1993) and focuses
on the formation and transmission of social norms and networks. Coleman
(1988) defines interpersonal trust as a public good but others have
characterised it as a private good fraught with externalities.
5
These theories
4
Literatue on social exchange and role modelling provide alternative theories of how the
willingness to provide intergenerational assistance is transmitted between generations (for a
country experience, see Ribar and Wilhelm, 2006).
5
See Knack and Keefer (1997) for the historical genesis of the term and what the authors dub
as the “elasticity” of its meaning. See also Dasgupta (2000).
Luiz de Mello and João Tovar Jalles
6
have underpinned the early economic literature on trust and economic
performance,
6
which established such a link but did not elaborate on the
mechanisms through which attitudes to trustfulness are generated and
transmitted in different societies. This is the case, for example, of Knack and
Keefer (1997), who tested the hypothesis that group participation has a
negative impact on economic growth, as proposed in the classical argument
on interest groups by Olson (1982), versus the Putnam et al. (1993)
hypothesis that group participation impacts growth positively. Olson-type
associations are formed along professional lines and common economic
interests; therefore, they are more likely to act as a force towards
redistribution towards the interests of participants, which is detrimental to
growth.
The literature has also delved into the scope for substitution or
complementarity between vertical and oblique/horizontal socialisation.
7
In
the case of heterogeneous populations, parents in minority (ethnic or
religious) groups are shown to have strong incentives to substitute vertical
transmission for oblique/horizontal transmission as a means of passing on
their values across generations. This cultural substitution effect leads to the
preservation of cultural heterogeneity in societies, whereas cultural
complementarity would result in a homogeneisation of cultural values. Bisin
and Verdier (2000, 2010), Tabellini (2008) and Bisin et al. (2010) discuss
these substitution and complementarity effects and the conditions under
which they result in the perpetuation of heterogeneity within populations,
especially in environments where parents take account of evolving social
norms (strategic substitution).
Previous empirical literature on the intergenerational transmission of
interpersonal trust found strong evidence of vertical socialisation. Dohmen et
al. (2006) uses German survey data and experimental techniques and show
that trust attitudes correlate strongly between parents and children. Uslaner
(2008) uses US survey data (General Social Survey) from 1972 to 1998 to
show that respondents’ attitudes to interpersonal (generalised) trust correlate
strongly with their grandparents’ and depend on ethnicity.
6
A recent paper by Ozcan and Bjornskov (2011) explores the connection between trust and a
broader measure of development given by the UN Human Development Index.
7
Some main references on horizontal vs vertical social capital as well as on bridging vs
bonding include Storper (2005), Rodriguez Pose and Storper (2006) and Crescenzi et al.
(2013).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
7
3. The methodology and data
3.1. Methodological issues
The basic hypothesis to be tested is that the intergenerational structure of
households affects the scope for complementarity/substitution between
vertical and horizontal socialisation and individuals’ attitudes towards
interpersonal trust. That is, the value added of the paper is based on the idea
that households’ structure affects interpersonal trust by means of its
correlation with participation in associational activities. This hypothesis can
be tested by estimating a model as follows
8
:
ii
jikij
eCFD
112
3
1
1
*
+
′
+
′
=
∑
=
αα
, for all j,
iiii
eCDT
2221
*
+
′
+
′
=
ββ
,
where
*
ij
D
and
*
i
T
are latent variables measuring individual i’s
participation in associational activity j and interpersonal trust, respectively;
ij
D
is a dichotomous observable indicator defined as
)0(1
=
=
ij
D
if
)0(0
*
<≥
ij
D
;
ik
F
denotes individual i’s family circumstance, indexed by k,
which can be two-generational (parent-child and grandparent-parent dyads)
or three-generational;
i
C
1
and
i
C
2
are vectors of control variables; and
i
e
1
and
i
e
2
are error terms.
The intuition behind estimation of our equation system is that
attitudes towards interpersonal trust are likely to be endogenous to vertical
and horizontal transmission mechanisms and the probabilities of
participation in different associational activities are correlated. As a result,
the univariate probit estimator would produce biased parameters. The
instrumental-variable version of the conventional probit model would not be
appropriate, because the indicators of participation in associational activities
and attitudes towards interpersonal trust (defined below) are dichotomous.
Instead, we estimate our equation system using a multivariate probit model,
which has the advantage of allowing a flexible correlation structure for the
8
See Beugelsdijk and Schaik (2005) and Crescenzi et al. (2013) for alternative specifications
of the relationship under scrutiny.
Luiz de Mello and João Tovar Jalles
8
unobservable variables describing choice behaviour. The disadvantage of the
multivariate probit model is that it is computationally very demanding,
because it requires high dimensional numerical- or simulation-based
integration and integration (or simulation) of the multivariate normal density
over subsets of a Euclidean space.
We use a maximum simulated likelihood estimator, which combines
a probability simulator to determine the multivariate cumulative (normal)
density function and standard estimation procedures through the
optimisation of the objective function. On the other hand, various
simulation-based methods, as suggested by Lerman and Manski (1981),
McFadden (1989) and Pakes and Pollard (1989), have been developed for
discrete choice models: the maximum simulated likelihood, the method of
simulated moments, and the method of simulated scores.
Before estimating our equation system, we apply the Bayesian
Model Averaging (BMA) approach for model selection to all associational
activities and interpersonal trust. This is because there is no single model for
assessing the links among household structure, participation in associational
activities and attitudes to interpersonal trust. To facilitate model selection,
BMA treats parameters and models as random and attempts to summarise
the uncertainty about the model in terms of a probability distribution over
the space of possible models. The method is used to average the posterior
distribution for the parameters under all possible models, where the weights
are the posterior model probabilities. To evaluate the posterior model
probability, BMA uses the Bayesian Information Criteria (BIC) to
approximate the Bayes factors that are needed to compute the posterior
model probability.
9
3.2. Data
We follow the literature and use the conventional measure of interpersonal
trust and associational activity available from WVS. WVS is a repeated
cross-section survey that reports data for an increasing number of countries
(over 80 countries in the 2005 wave) since 1981.
10
The survey provides
responses to questions related to attitudes to interpersonal trust and has
therefore been used extensively in the empirical literature on social capital.
Interpersonal trust is coded using question V23 (“Generally speaking, would
9
See Raftery (1995), Sala-i-Martin et al. (2004), and Malik and Temple (2009) for more
discussion.
10
See Inglehart et al. (2004) for more information on the WVS.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
9
you say that most people can be trusted or that you need to be very careful in
dealing with people?”) as taking the value of “1” if respondents selected
option 1 (“Most people can be trusted”) as their answer, and “0” otherwise.
Associational activity is coded using questions V24-V33 (“Now I am going
to read off a list of voluntary organisations. For each one, could you tell me
whether you are an active member, an inactive member or not a member of
that type of organisation?”) as taking the value of “1” if respondents selected
option 2 (“Active member”) as their answer to the questions related to
participation in each of the associational activities listed as admissible
answers: church or religious organisation (V24); sport or recreational
organisation (V25); art, music or educational organisation (V26); labour
union (V27); political party (V28); environmental organisation (V29);
professional association (V30); humanitarian or charitable organisation
(V31); consumer association (V32); other (V33); and “0” otherwise.
Family structure is defined as follows. Two- and three-generational
families are identified by coding the responses to question V240 (“Do you
live with your parents?”) as taking the value of “1” if respondents selected
option 1 (“yes”) as their answer, and “0” otherwise; and by coding the
responses to question V56 (“Have you had any children?”) as taking the
value of “1” if respondents selected options 1-8 (1 to 8 children) as their
answer, and “0” otherwise. WVS does not allow for exploring the different
links that may exist among family members, as in the case of relations by
blood, legal and informal ties, adoption and other forms of kinship. Neither
does it allow for distinguishing homogamous unions from other structures,
which affect socialisation mechanisms (Bisin and Verdier, 2000).
Respondents living in two-generational households are those who answered
yes to the former question and selected option “0” to the latter. Those living
in three-generational households selected options “1-8” when answering the
latter question.
The sets of control variables include the conventional
personal/household characteristics that are associated with participation in
associational activities and interpersonal trust, such as gender, ethnicity, age,
educational attainment
11
and labour-market status. In addition to these
controls, indicator variables are used to identify whether or not respondents
are farmers or agricultural workers, because cooperation is required across
generations living in the same household, which creates a sense of lasting
11
The literature also considers the role of education as a socialisation device (Gradstein and
Justman, 2002 and 2005).
Luiz de Mello and João Tovar Jalles
10
obligation across the generations (Furstenberg, 2005), and hence patterns of
socialisation within the household and beyond. Finally, the regressions
include country and cohort effects.
Country effects are important because of
the geographical diffusion of cultural traits (Rendine et al., 1986), group size
(Bisin and Verdier, 2010), and the effects of heterogeneity on patterns of
socialisation (Tabellini, 2008; Bisin et al., 2010).
12
Cohort effects are
important because of attitudinal changes in social norms and traits across
generations.
A preliminary look at the raw data shows that the most popular
associational activities involve church or religious organisations (about 22%
of respondents); sport and recreational organisations (13.5%); art, music and
educational organisations (9.5%); humanitarian or charitable organisations
(7.5%); and professional associations or trade unions taken together (10.3%).
Without controlling for other determinants of attitudes towards interpersonal
trust, individuals that declared themselves to participate actively in
humanitarian/charitable, sport/recreational, art/music/education and
professional organisations tend to be more trustful of strangers than their
counterparts who do not participate in these organisations. But this is not the
case of respondents who declared themselves to be actively involved in
church/religious organisations.
4. Empirical results
4.1. Model selection
We begin with the BMA analysis for a set of 14 regressors. The results,
reported in Tables 1 and 2, show that being married and head of household
are associated with greater interpersonal trust. Age and educational
attainment (secondary and tertiary level) also affects interpersonal trust
positively, although labour market status does not seem to be correlated
strongly with interpersonal trust. In addition, there is strong correlation
among the different associational activities, implying that participation in
one type of organisation increases the probability of participation in the
12
We are aware of the literature that also highlights the regional nature of such dimension,
such as Duranton and Rodriguez Pose (2009). Within countries heterogeneity in family
structures is a well-known feature in a number of countries and it is often adopted as a
variable to explain differences in economic performance across regions (see, e.g., Beugelsdijk
and Smulders, 2003; Levin and Cross, 2004; Ruef, 2002; and Alesina and Giuliano, 2010).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
11
others (Table 2).
13
As for the relationship between personal characteristics
and the probability of participation in associational activities, the BMA
results show that being male decreases the probability of engagement in
church/religious associations, but it increases the probability of participation
in sport/recreational activities, as well as in professional associations and
labour unions. The effects of educational attainment are more complex and
vary across the different associational activities.
[insert Tables 1 and 2]
4.2. Baseline results
Our equation system was estimated for the most popular associational
activities on the basis of survey responses. Age and country effects are
omitted for reasons of parsimony. The results reported in Table 3 show that
individuals living in three-generational households are more likely to
participate in humanitarian and charitable organisations and are less likely to
participate in sport or recreational organisations. In other words, results
suggest that deferred or indirect reciprocity, as well as demonstration effects,
may also motivate individuals living in three-generational dyads to
participate in humanitarian associations, which is a mechanism for
interpersonal trust-building outside the household. Individuals living in two-
generational households are more likely to participate in both art, music and
educational organisations and sport or recreational organisations. In turn,
individuals who participate in art, music and educational organisations as
well as in humanitarian and charitable organisations are also more trustful of
strangers. Participation in sport or recreational organisations does not appear
to affect attitudes to trustfulness. Following the literature, the results
reported confirm that some household structures are associated with a higher
propensity of participation in associational activities, suggesting
complementarity between vertical and horizontal mechanisms of
transmission of attitudes towards interpersonal trust.
As far as the coefficient estimates of control variables are concerned, being
married and head of household enhance trustfulness but is not strongly
13
The “reference organisation” for each column 2-6 is defined as follows: col.2 includes
humanitarian, sport and art organisations; col.3 includes humanitarian, church and art
organisations; col. 4 includes humanitarian, church and sport organisations; col. 5 includes
church, sport and art organisations; col. 6 includes humanitarian, church, sport and art
organisations.
Luiz de Mello and João Tovar Jalles
12
associated with participation in the different associational activities. Gender
matters for participation but not for attitude towards interpersonal trust: men
are more likely to participate in sport or recreational organisations and
professional associations and labour unions and less so than women to
engage in religious or art, music and education associations. Age is a
powerful determinant of participation in lifestyle activities (sports and arts)
but not of interpersonal trust. The empirical findings also show that ethnicity
is a powerful determinant of attitudes towards interpersonal trust as well as
participation in associational activities which is in accordance to the
literature discussed above. Educational attainment is strongly correlated with
participation, especially in the case of lifestyle organisations. Occupational
effects suggest that farmers/agricultural workers are less likely to participate
in lifestyle organisations, such as sports and recreational, art, musical or
education; and more likely to participate in the case of church or religious,
professional association or labour union ones.
The null hypothesis that the error terms of the participation
equations are uncorrelated is rejected at classical levels of significance
(Table 4). However, the null hypothesis of no correlation between the error
terms of the participation equations and that of the interpersonal trust
equation could not be rejected, except for professional association and labour
union.
[insert Tables 3 and 4]
Failure to account for the correlations in the error terms of the
regressions describing the probability of engagement in the different
organisations biases the estimation of the effect of horizontal socialisation on
interpersonal trust. To assess the magnitude of the bias, we re-estimated our
equation system by probit for the individual associational activities and
interpersonal trust regressions. The results, reported in Table 5, show that the
estimated coefficients for most if not all controls are similar in magnitude.
Individuals living in two-generational households are more likely to
participate in sport or recreational organisations and art, music or education
organisations (coefficients are slightly higher in magnitude than in the case
of the multivariate probit estimation). Individuals living in three-generational
households are more (less) likely to participate humanitarian or charitable
organisation (sport or recreational organisation); again, the coefficients are
slightly higher in magnitude than in the multivariate probit regressions.
As for the determinants of interpersonal trust, when the error terms are
assumed to be uncorrelated, participation in church or religious
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
13
organisations, as well as sport or recreation organisations, increases
interpersonal trust (positive and statistically significant coefficients). On the
other hand, belonging to professional associations or labour unions does not
seem to reduce interpersonal trust, as opposed to the case when the
regressions are estimated by multivariate probit.
[insert Table 5]
4.3. Sensitivity analysis
We begin with a sensitivity exercise where we re-run our equation system
excluding one regressor at a time. With this we aim to assess the importance
of each regressor in affecting the magnitude, size and statistical significance
of the main variables of interest. Table 6 includes the estimated coefficients
for two- and three-generational households for the five associational
activities reported in columns 2-6 of Table 3 for ease of reference.
14
The
results show that the effects of intergenerational structure are robust in the
case of sport or recreational organisations. When marital status is omitted,
individuals living in two-generational households appear to be less likely to
participate in the church or religious organisations. Moreover, the results are
quite robust for participation in art, music and educational activities,
humanitarian and charitable organisations, and professional organisations
and labour unions.
We also assessed the robustness of the baseline results for attitudes
to interpersonal trust by omitting one regressor at a time. The results
reported in Table 7 show that participation in church or religious
organisations does not correlate with interpersonal trust, a finding that is
robust throughout the exercise. When “gender” is excluded participation in
sport or recreational activities now appears to be positively associated with
trust. In the cases of art, music and educational organisations, as well as
professional organisations and labour unions, the positive and negative
effects on interpersonal trust, respectively, disappear when either “gender”
and “age” are omitted.
We then use factor analysis and combine the most relevant
associational activities (the ones with statistically significant coefficients in
Table 3) in a single regressor in our equation system . The first factor of the
Principal Components Analysis explains 90% of the total variance in the
standardised data. The first principal component is normalised in such a way
14
Other coefficient estimates are omitted for reasons of parsimony but these are available
from the authors upon request.
Luiz de Mello and João Tovar Jalles
14
that high values indicate higher participation in associational activities. Our
standardised index, PCA, can be written as:
15
humanprofpolpartylabourartPCA 41.044.031.034.037.0
+
+
+
+
=
This aggregate index is then used as the main dependent
variable/regressor in our equation system, which is estimated using three-
stage least squares. The results, reported in Table 8, show that participation
in associational activities is strongly correlated with interpersonal trust,
which is in line with previous literature, but there does not seem to be a
statistically significant link between participation in organisations and the
intergenerational structure of households. Being married and head of
household enhance trustfulness but the latter is not strongly associated with
participation in associational activities (column 2). Similarly, gender matters
for participation but not for explaining attitudes towards interpersonal trust.
Age does not appear to be a powerful determinant of participation in lifestyle
activities nor of interpersonal trust. Ethnicity remains strongly correlated
with both participation and trust. Educational attainment is strongly
correlated with participation, and tertiary education seems particularly
relevant for fostering interpersonal trust. Occupational effects are important,
and farmers/agricultural workers are less likely to participate in associational
organisations.
[insert Tables 6, 7 and 8]
5. Conclusions and discussion
In this paper we use data from the latest WVS wave to explore the role of
interpersonal trust in the relationship between the intergenerational structure
of households and participation in organisational activities.
Generally speaking, we find that the probability that individuals
participate in a variety of associational activities, including lifestyle,
humanitarian and professional associations, depends on whether they live in
two- or three-generational households. Participation in some of these
15
A likelihood ratio test was used to examine the “sphericity” case, allowing for sampling
variability in the correlations. This test comfortably rejects sphericity at the 1% level with a
Kaiser-Meyer-Olkin measure of sampling adequacy equal to 0.70.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
15
associations in turn correlates strongly with attitudes to interpersonal trust.
While previous studies have controlled for individual and household
characteristics that are known to affect the probability of participation and
attitudes to interpersonal trust, this paper contributes to the literature by
highlighting the importance of intergenerational household structure.
Another contribution of our analysis is that the regressions are estimated by
multivariate probit, which allows for correlated errors among the
regressions, in recognition of the fact that the probability of participation in a
given association is affected by whether or not individuals participate in
other associations.
By exploring how interpersonal trust is transmitted within the
household and among individuals through associational activities, our
analysis contributes to the debate on the relevance of social norms.
Ultimately, mechanisms for the transmission of social norms are
economically important because they explain why certain social norms are
persistent within societies: if non-cooperative norms that are associated with
poorer economic performance are perpetuated across generations, and are
reflected in how individuals engage in associational activities outside the
household, so do the root causes of economic underperformance.
Luiz de Mello and João Tovar Jalles
16
References
Alesina, A. and E. La Ferrara (2002) “Who Trusts Others?” Journal of
Public Economics, 85, 207-234.
Alesina, A. and Giuliano, P. (2010) “The Power of Families.” Journal of
Economic Growth, 15(2), 93-125.
Arrondel, L. and A. Masson (2001) “Family Transfers Involving Three
Generations.” Scandinavian Journal of Economics, 103, 415-43.
Ashraf, N., Bohnet, I., and Piankov, N. (2006) ”Decomposing Trust and
Trustworthiness.” Experimental Economics, 9, 193-208.
Banfield, Edward C. (1958) “The Moral Basis of a Backward Society.” New
York: The Free Press.
Beugelsdijk, S., and Smulders, S., (2003) “Bridging and bonding social
capital: Which type is good for economic growth?” Paper presented
at the European Regional Science Association, Jyvaskila, Finland.
Beugelsdijk, S. and A. B. T. M. van Schaik. (2005) “Social capital and
growth in European regions: An empirical test.” European Journal
of Political Economy, 21 (2), 301-24.
Bisin , A. and Verdier, T. (2010) “The Economics of Cultural Transmission
and Socialisation.” NBER Working Paper, no. 16512, National
Bureau of Economic Research, Cambridge, MA.
Bisin A. and Verdier, T. (2000) “Beyond the Melting Pot: Cultural
Transmission, Marriage and the Evolution of Ethnic and Religious
Traits.” Quarterly Journal of Economics, 115, 955-88.
Bisin A. and Verdier, T. (2001) “The Economics of Cultural Transmission
and the Dynamics of Preferences.” Journal of Economic Theory, 97,
289-319.
Bisin A., Pattachini, E., Verdier, T., and Zenou, Y. (2010) Are Muslim
Immigrants Different in terms of Cultural Integration.” Journal of
the European Economic Association, 6, 445-56.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
17
Bisin A., Topa, G., and Verdier, T. (2004) “Cooperation as a Transmitted
Cultural Trait.” Rationality and Society,16, 477-507.
Bourdieu, P. (1986) “The Forms of Capital, in J. G. Richardson
(ed.).”Handbook of Theory and Research for the Sociology of
Education, Greenwood Press, New York, N.Y.
Boyd, R. and Richardson, P.J. (1985) “Culture and the Evolutionary
Process.” Chicago University Press, Chicago, IL.
Cavalli-Sforza, L.L. and Feldman, M. (1981) “Cultural Transmission and
Evolution: A Quantitative Approach.” Princeton University Press,
Princeton, N.J.
Coleman, J. (1988) “Social Capital in the Creation of Human Capital.”
American Journal of Sociology, 94, 312.
Coleman, J.S. (1990) “Foundations of Social Theory.” Belknap, Cambridge,
MA.
Cox, D. and Stark, O. (1994) “Intergenerational Transfers and the
Demonstration Effect.” Economics Department Working Papers in
Economics, Boston College, Boston, MA.
Crescenzi, R., Gagliardi, L. and Percoco, M. (2013) “Social capital and the
innovative performance of Italian provinces.” Environment and
Planning A, 45(4), 908 – 929.
Dasgupta, P. (2000) “Economic Development and the Idea of Social
Capital.” in Serageldin, I. and Dasgupta, P. (eds.), Social Capital: A
Multifaceted Perspective, World Bank, Washington D.C.
Dohmen, T., Falk, A., Huffman, D., and Sunde, U. (2006) “The
Intergenerational Transmission of Risk and Trust Attitudes.” IZA
Discussion Paper, No. 2380, Institute for the Study of Labour,
Bonn.
Luiz de Mello and João Tovar Jalles
18
Duranton, G. and Rodriguez Pose, A. (2009) “Family types and the
persistence of Regional Disparities in Europe.” Economic
Geography, 85(1), 23–47.
Frank, R.H. (1992) “Melding Sociology and Economics: James Coleman's
Foundations of Social Theory.” Journal of Economic Literature, 30,
147-70.
Furstenberg, F.F. (2005) “Banking on Families: How Families Generate and
Distribute Social Capital.” Journal of Marriage and Family, 67,
809-21.
Glaeser, E.L., Laibson, D.I., Scheinkman, J.A., and Souter, C.L. (2000) ”
Measuring Trust.” Quarterly Journal of Economics, 115. 811-46.
Gradstein, M. and Justman, M. (2002) “Social Cohesion and Growth.”
American Economic Review,. 92, 1192-1204.
Gradstein, M. and Justman, M. (2005) “The Melting Pot and School
Choice.” Journal of Public Economics, 89, 871-96.
Granovetter, M. (1973) “The Strength of Weak Ties.” American Journal of
Sociology 78(6), 1360-1380.
Guiso, L., Sapienza, P., and Zingales, L. (2008) “Long Term Persistence.”
NBER Working Paper 14278.
Guttman, J. M. (2001a) “Self-Enforcing Reciprocity Norms and
Intergenerational Transfers: Theory and Evidence.” Journal of
Public Economics, 81, 117-51.
Guttman, J. M. (2001b) “Families, Markets and Self-Enforcing Reciprocity
Norms.” Annales d’Economie et de Statistique, 63-64, 89-110.
Inglehart, R., Basanez, M., Diez-Medrano, J., Halman, L., and Luijkx, R.
(2004) “Human Beliefs and Values.” University of Michigan Press,
Ann Arbor, MI.
Jellal, M. and Wolff, F.-C. (2002a) “Cultural Evolutionary Altruism: Theory
and Evidence.” European Journal of Political Economy, 18, 241-62.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
19
Jellal, M. and Wolff, F.-C. (2002b) “Altruistic Bequests with Inherited
Tastes.” International Journal of Business and Economics, 1, 95-
113.
Kaasa, A. and Parts, E. (2007) “Individual-level determinants of social
capital in Europe: Differences between country groups.” University
of Tartu (Estonia), mimeo.
Krishna, A. and Uphoff, N. (1999) “Mapping and measuring social capital: a
conceptual and empirical study of collective action for conserving
and developing watersheds in Rajasthan, India.” Social Capital
Initiative, Working Paper #13, Washington, DC: The World Bank.
Knack, S. and Keefer, P. (1997) “Does Social Capital Have an Economic
Payoff? A Cross-Country Investigation.” Quarterly Journal of
Economics, 112, 1251-88.
La Porta, R., Silanes, F.L. de., Schleifer, A., and Vishny, R. (1997) “Trust in
Large Organisations.” American Economic Review, 87, 333-38.
Lerman S. R., and Manski, C. (1981) “On the Use of Simulated Frequencies
to Approximate Choice Models” In C. Manski and D. McFadden
(eds.) Structural Analysis of Discrete Data with Econometric
Applications, 305-319, Cambridge, Massachusetts: MIT Press
Levin, D., and Cross, R. (2004) “The strength of weak ties you can trust: The
mediating role of trust in effective knowledge transfer.”
Management Science, 50, 1477–1490.
Malik, A., and Temple J. (2009) “The geography of output volatility.”
Journal of Development Economics, 90(2), 163-178.
McFadden, D. L. (1989) “A Method of Simulated Moments of Estimation of
Discrete Choice Models without Numerical Integration.”
Econometrica 57, 995-1026.
Olson, M. (1982) “The Rise and Decline of Nations.” New Haven,
Connecticut, Yale University Press.
Luiz de Mello and João Tovar Jalles
20
Ozcan, B. and Bjornskov, C. (2011) “Social Trust and Human
Development.” Journal of Socio-Economics, 40, 753-762.
Pakes, A., and Pollard, D. (1989) “Simulation and the Asymptotics of
Optimization Estimators.” Econometrica, 57, 1027-1057.
Putnam, R. (1993) “Making democracy work: civic tradition in modern Italy.
“Princeton: Princeton University Press.
Putnam, R. (2000) “Bowling Alone : The Collapse and Revival of American
Community New York.” Simon and Schuster
Putnam, R., Leonardi, R., and Nanetti, R. (1993) “Making Democracy
Work.” Princeton, New Jersey, Princeton University Press.
Raftery, A. (1995) “Bayesian Model selection in social research, in
Sociological Methodology.” ed. Peter Marsden, Blackwells,
Cambridge.
Rendine, S., Piazza, A., and Cavalli-Sforza, L.L. (1986) “Simulation and
Separation by Principle Components of Multiple Demic Expansions
in Europe.” American Naturalist, 128, 681-706.
Ribar, D. and Wilhelm, M. (2006) “Exchange, role modelling and the
intergenerational transmission of elder support attitudes: evidence
from three generations of Mexican-Americans.” Journal of Socio-
Economics, 35, 514-531.
Rodríguez-Pose, A. and Storper, M. (2006) “Better rules or stronger
communities? On the social foundations of institutional change and
its economic effects.” Economic Geography, 82(1), 1–25.
Ruef, M., (2002) “Strong Ties, Weak Ties and Islands: Structural and
Cultural Predictors of Organizational Innovation.” Industrial and
Corporate Change 11, 427–49.
Sala-i-Martin, X., Doppelhofer, G., and Miller, R. (2004) “Determinants of
long-run growth: A Bayesian Averaging Classical Estimates
(BACE) approach.” American Economic Review, 94(4), 813-835.
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
21
Solow, R., (1999) “Notes on Social Capital and Economic Performance. in
Dasgupta and Serageldin, eds. Social Capital: A Multifaceted
Perspective.” Washington: The World Bank.
Storper, M. (2005) “Society, community and economic development.”
Studies in Comparative Economic Development, 39(4), 30–57.
Tabellini, G. (2008) “The Scope for Cooperation: Values and Incentives.”
CESifo Working Paper, No. 2236, CESifo, Munich.
Temple, J., and Johnson, P. (1998) “Social Capability and Economic
Growth.” Quarterly Journal of Economics, 113(3), 965-990
Uslaner, E.M. (2008) “Where you Stand Depends upon Where Your
Grandparents Sat: The Inheritability of Generalised Trust.” Public
Opinion Quarterly, 72, 725-40.
Luiz de Mello and João Tovar Jalles
22
Table 1: BMA analysis: Interpersonal trust (Dependent variable:
Interpersonal trust)
Church or
religious
organisation
Sport or
recreational
organisation
Art, music or
education
organisation
Humanitarian
or charitable
organisation
Professional
association
or labour
union
Participation
in reference
organisation
-,* +,* +,* +,* +
Married +,* +*, +,* +,* +,*
Head of
household +,* +,* +,* +,* +,*
Gender (1 =
Male) + - + + +
Age +,* +,* +,* +,* +,*
Ethnicity +,* +,* +,* +,* +,*
Primary
education + + + + +
Secondary
education +,* +,* +,* +,* +,*
Tertiary
education +,* +,* +,* +,* +,*
Full-time
employed + + + + +
Part-time
employed + + + + +
Self-employed
- - - - -
Unemployed - - - - -
Farmer + + + + +
Agricultural
workers +,* +,* +,* +,* +,*
No. of
observations 20872 20742 20676 20561 20567
Source: Author’s estimations.
Note: The “reference organization” for each column 2-6 is defined as follows: col.2 includes
humanitarian, sport and art organizations; col.3 includes humanitarian, church and art organizations; col.
4 includes humanitarian, church and sport organizations; col. 5 includes church, sport and art
organizations; col. 6 includes humanitarian, church, sport and art organizations. Statistical significance at
the 1, 5, and 10 percent levels is denoted by respectively (***), (**) and (*).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
23
Table 2: BMA analysis: Associational activities
(Dependent variable: Participation in associational activities)
Church or
religious
organisation
Sport or
recreational
organisation
Art, music
or education
organisation
Humanitarian or
charitable
organisation
Professional
association
or
labour union
Participation
in reference
organisation
+,* +,* +,* +,* +,*
Two-
generation
family - +,* +,* - +
Three-
generation
family - -,* - + +
Married - - - +,* +,*
Head of
household -,* + + +,* +
Gender (1 =
Male) -,* +,* - + +,*
Age -,* -,* - - -
Ethnicity -,* - -,* + -
Primary
education + - - - -
Secondary
education -,* +,* + + -
Tertiary
education -,* +,* +,* + +
Full-time
employed -,* - - +,* -
Part-time
employed -,* + +,* +,* -
Self-
employed + - - + +
Unemployed - - - + +
Farmer +,* - - - +
Agricultural
workers - -,* -,* - -
No. of
observations 20218 20077 20020 19910 19946
Source: Author’s estimations.
Note: See footnote Table 1 for details on “refrence organization”. Statistical significance at the 1, 5, and
10 percent levels is denoted by respectively (***), (**) and (*).
Luiz de Mello and João Tovar Jalles
24
Table 3: Socialisation and interpersonal trust: Multivariate probit analysis
Church or
religious
organisation
Sport or
recreational
organisation
Art, music
or education
organisation
Humanitarian
or charitable
organisation
Professional
association
or labour
union
Interpersonal
trust
Participation
in reference
organisation
Church or
religious
organisation
0.05
(0.056)
Sport or
recreational
organisation
-0.02
(0.057)
Art, music or
educational
organisation
0.17***
(0.060)
Humanitarian
or charitable
organisation
0.13**
(0.062)
Professional
association or
labour union
-0.14**
(0.059)
Two-
generation
household
0.01 0.15*** 0.18*** -0.05 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
household
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.051) (0.058) (0.055)
Married 0.18*** 0.02 -0.03 0.05 0.09*** 0.06**
(0.029) (0.031) (0.033) (0.036) (0.034) (0.026)
Head of
household -0.01 -0.01 -0.02 -0.10** 0.05 0.09**
(0.043) (0.043) (0.048) (0.048) (0.044) (0.035)
Gender (1 =
Male) -0.21*** 0.34*** -0.05* -0.04 0.10*** 0.02
(0.025) (0.026) (0.028) (0.030) (0.028) (0.023)
Age -0.00 -0.02*** -0.02*** -0.00 0.01 0.00
(0.004) (0.005) (0.005) (0.005) (0.006) (0.004)
Ethnicity -0.00*** -0.00*** -0.00*** -0.00*** -0.00*** 0.00***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
25
Source: Author’s estimations.
Note: The multivariate probit model is estimated by maximum simulated likelihood with robust standard
errors (in parentheses) and includes country and cohort effects (not reported). See footnote Table 1 for
details on “refrence organization”. Statistical significance of coefficient estimates at the 1, 5, and 10
percent levels is denoted by respectively (***), (**) and (*).
Primary
education 0.02 -0.08** -0.18*** -0.07 -0.08* -0.02
(0.033) (0.040) (0.044) (0.043) (0.042) (0.032)
Secondary
education -0.01 0.18*** 0.21*** 0.16*** 0.13*** 0.02
(0.035) (0.036) (0.039) (0.044) (0.042) (0.032)
Tertiary
education 0.08** 0.29*** 0.48*** 0.33*** 0.45*** 0.19***
(0.042) (0.039) (0.040) (0.045) (0.041) (0.036)
Full-time
employed -0.16*** -0.00 -0.00 0.03 0.39*** -0.02
(0.034) (0.036) (0.038) (0.043) (0.040) (0.031)
Part-time
employed -0.07 0.03 0.13*** 0.13** 0.27*** -0.04
(0.050) (0.051) (0.052) (0.056) (0.053) (0.043)
Self-
employed -0.04 -0.06 0.02 0.09** 0.22*** 0.04
(0.037) (0.042) (0.045) (0.046) (0.045) (0.036)
Unemployed
-0.01 -0.09 -0.08 0.10 0.06 -0.15***
(0.054) (0.066) (0.069) (0.065) (0.071) (0.051)
Farmer 0.05 -0.13*** -0.17*** -0.08 -0.02 0.20***
(0.040) (0.047) (0.050) (0.052) (0.050) (0.040)
Agricultural
workers 0.19*** -0.25*** -0.41*** 0.06 0.08* 0.01
(0.042) (0.048) (0.056) (0.048) (0.047) (0.038)
No. of
observations
18,900 18,900 18,900 18,900 18,900 18,900
Luiz de Mello and João Tovar Jalles
26
Table 4 : Multivariate probit estimations: Variance-covariance matrix
Church
or
religious
organisation
Sport
or
recreational
organisation
Art,
music
or educational
organisation
Humanitarian
or
charitable
organisation
Professional
association
or labour
union
Interpersonal
trust
Church or
religious
organisation 1.00
Sport or
recreational
organisation 0.2 *** 1.00
Art, music or
educational
organisation 0.25 *** 0.50 *** 1.00
Humanitarian
or charitable
organisation 0.35 *** 0.31 *** 0.43 *** 1.00
Professional
association or
labour union 0.23 *** 0.29 *** 0.39 *** 0.42 *** 1.00
Interpersonal
trust 0.01 0.04 -0.01 0.01 0.05 **
1.00
Source: Author’s estimations.
Note: Refers to the model reported in Table 1. Statistical significance at the 1, 5, and 10 percent levels is
denoted by respectively (***), (**) and (*).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
27
Table 5 : Bivariate probit estimations: Associational activities and
interpersonal trust
(Dependent variables: participation in associational activities and
interpersonal trust)
Church
Or
religious
organisati
on
Sport
or
recreatio
nal
organisat
ion
Art, music
or
education
organisatio
n
Humanita
rian
or
charitable
organisati
on
Professio
nal
associatio
n or
labour
union
Interpersona
l trust
Participati
on in
reference
organisatio
n
Church or
religious
organisatio
n
0.05*
(0.031)
Sport or
recreational
organisatio
n
0.08**
(0.034)
Art, music
or
educational
organisatio
n
0.12***
(0.039)
Humanitari
an
or
charitable
organisatio
n
0.13***
(0.041)
Professiona
l
association
or labour
union
-0.03
(0.036)
Two-
generation
household
0.01 0.18*** 0.18*** -0.05 -0.05
(0.039) (0.040) (0.044) (0.050) (0.048)
Three-
generation
household
-0.03 -0.14*** -0.03 0.17** 0.04
(0.049) (0.050) (0.058) (0.073) (0.054)
Married 0.16*** 0.03 -0.04 0.04 0.10*** 0.05*
(0.030) (0.032) (0.034) (0.036) (0.033) (0.027)
Luiz de Mello and João Tovar Jalles
28
Head of
household -0.03 -0.02 -0.00 -0.07 0.02 0.08**
(0.044) (0.045) (0.050) (0.049) (0.045) (0.035)
Gender (1
= Male) -0.19*** 0.35*** -0.06** -0.04 0.11*** 0.04
(0.026) (0.027) (0.029) (0.031) (0.028) (0.024)
Age 0.00 -0.03*** -0.02*** 0.00 0.01 0.00
(0.004) (0.005) (0.005) (0.006) (0.005) (0.004)
Ethnicity -0.00*** -0.00*** -0.00** -0.00** -0.00*** 0.00***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Primary
education 0.04 -0.07 -0.17*** -0.07* -0.06 -0.03
(0.034) (0.041) (0.045) (0.042) (0.041) (0.033)
Secondary
education 0.02 0.16*** 0.22*** 0.21*** 0.13*** 0.06*
(0.039) (0.040) (0.045) (0.054) (0.043) (0.034)
Tertiary
education 0.09** 0.26*** 0.47*** 0.34*** 0.42*** 0.20***
(0.042) (0.041) (0.042) (0.047) (0.041) (0.037)
Full-time
employed -0.17*** -0.00 -0.04 -0.00 0.38*** -0.05
(0.034) (0.036) (0.039) (0.043) (0.039) (0.031)
Part-time
employed -0.09* 0.07 0.11** 0.12** 0.27*** -0.03
(0.049) (0.051) (0.052) (0.057) (0.053) (0.044)
Self-
employed -0.03 -0.03 0.00 0.07* 0.21*** 0.05
(0.036) (0.042) (0.044) (0.045) (0.043) (0.036)
Unemploye
d -0.04 -0.06 -0.08 0.09 0.06 -0.14***
(0.056) (0.065) (0.068) (0.065) (0.069) (0.051)
Farmer 0.03 -0.11** -0.10** -0.07 -0.04 0.20***
(0.038) (0.045) (0.047) (0.050) (0.046) (0.039)
Agricultura
l workers 0.18*** -0.25*** -0.41*** 0.04 0.07 0.03
(0.044) (0.048) (0.055) (0.047) (0.044) (0.042)
No. of
observation
s
20,188 19,900 19,991 19,881 20,353 20,320
Source: Author’s estimations.
Note: The bivariate probit model s are estimated by maximum likelihood with robust standard errors (in
parentheses) and includes country and cohort effects (not reported). See footnote Table 1 for details on
“refrence organization”. Statistical significance of coefficient estimates at the 1, 5, and 10 percent levels
is denoted by respectively (***), (**) and (*).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
29
Table 6: Sensitivity analysis: Participation in associational activities
(Dependent variables: Participation in associational activities)
Dependent Variables
Excluded
variable Relevant
regressors Church
or
religious
organisation
Sport
or
recreation
al
organisatio
n
Art, music or
education
organisation
Humanit
arian
or
charitabl
e
organisa
tion
Professi
onal
associati
on or
labour
union
memory
(Table 3) Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.051) (0.058) (0.055)
Married Two-
generation
family
-0.07** 0.15*** 0.18*** -0.06 -0.07
(0.036) (0.036) (0.039) (0.045) (0.043)
Three-
generation
family
0.03 -0.11** -0.06 0.13** 0.04
(0.045) (0.047) (0.049) (0.056) (0.053)
Head of
household
Two-
generation
family
0.00 0.16*** 0.19*** -0.01 -0.03
(0.034) (0.035) (0.038) (0.042) (0.042)
Three-
generation
family
-0.02 -0.14*** -0.09** 0.06 0.02
(0.042) (0.045) (0.046) (0.052) (0.051)
Gender (1
= Male)
Two-
generation
family
-0.06 0.23*** 0.16*** -0.06 0.00
(0.037) (0.038) (0.041) (0.047) (0.046)
Three-
generation
family
0.01 -0.18*** -0.06 0.11* -0.03
(0.046) (0.048) (0.050) (0.058) (0.055)
Age
Two-
generation
family
0.01 0.18*** 0.20*** -0.04 -0.03
(0.038) (0.038) (0.042) (0.047) (0.046)
Three-
generation
family
-0.04 -0.14*** -0.07 0.12** 0.01
(0.046) (0.048) (0.050) (0.058) (0.055)
Ethnicity Two-
generation
family
0.01 0.17*** 0.19*** -0.03 -0.03
(0.036) (0.037) (0.040) (0.046) (0.045)
Three- 0.02 -0.12*** -0.06 0.12** 0.02
Luiz de Mello and João Tovar Jalles
30
generation
family
(0.043) (0.045) (0.048) (0.055) (0.052)
Primary
education
Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.04 -0.11** -0.06 0.12** 0.01
(0.046) (0.048) (0.050) (0.058) (0.055)
Secondary
education
Two-
generation
family
0.01 0.15*** 0.18*** -0.04 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.051) (0.058) (0.055)
Tertiary
education
Two-
generation
family
0.01 0.16*** 0.19*** -0.03 -0.00
(0.038) (0.039) (0.042) (0.048) (0.046)
Three-
generation
family
-0.05 -0.13*** -0.08 0.10* -0.02
(0.046) (0.048) (0.050) (0.058) (0.055)
Full-time
employed
Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.03
(0.038) (0.039) (0.042) (0.048) (0.046)
Three-
generation
family
-0.04 -0.11** -0.05 0.12** 0.00
(0.046) (0.048) (0.050) (0.058) (0.055)
Part-time
employed
Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.03
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.050) (0.058) (0.055)
Self-
employed
Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.03
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.051) (0.058) (0.055)
Unemploy
ed Two-
generation
family
0.01 0.15*** 0.18*** -0.05 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation -0.05 -0.11** -0.05 0.12** 0.01
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
31
family
(0.046) (0.048) (0.051) (0.058) (0.055)
Farmer Two-
generation
family
0.01 0.15*** 0.17*** -0.05 -0.03
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.050) (0.058) (0.055)
Agricultura
l workers
Two-
generation
family
0.01 0.15*** 0.17*** -0.05 -0.02
(0.038) (0.039) (0.042) (0.048) (0.047)
Three-
generation
family
-0.05 -0.11** -0.05 0.12** 0.01
(0.046) (0.048) (0.050) (0.058) (0.055)
Source: Author’s estimations.
Note: The multivariate probit model is estimated by maximum simulated likelihood with robust standard
errors (in parentheses) and includes all the controls of the baseline regression, as well as country and
cohort effects (not reported). The first row “memory”refers to the results in Table 3. Statistical
significance of coefficient estimates at the 1, 5, and 10 percent levels is denoted by respectively (***),
(**) and (*).
Luiz de Mello and João Tovar Jalles
32
Table 7: Sensitivity analysis: Interpersonal trust
(Dependent variable: Interpersonal trust)
Participation in associational activities
Excluded
variable Church or
religious
organisation
Sport or
recreational
organisation
Art, music
or
education
organisation
Humanitarian
or charitable
organisation
Professional
association
or labour
union
memory
(Table 3) 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Married -0.05 -0.05 0.13** 0.17*** -0.13**
(0.057) (0.058) (0.059) (0.063) (0.061)
Head of
household -0.01 0.04 0.10* 0.14** -0.06
(0.051) (0.053) (0.054) (0.056) (0.056)
Gender (1 =
Male) 0.05 0.11* 0.08 0.12* 0.04
(0.056) (0.058) (0.059) (0.064) (0.062)
Age 0.08 0.01 -0.00 0.11* -0.07
(0.055) (0.058) (0.060) (0.063) (0.061)
Ethnicity -0.04 0.05 0.08 0.06 -0.12**
(0.053) (0.053) (0.056) (0.059) (0.058)
Primary
education 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Secondary
education 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Tertiary
education 0.05 -0.03 0.18*** 0.13** -0.12**
(0.056) (0.057) (0.060) (0.062) (0.059)
Full-time
employed 0.04 -0.02 0.17*** 0.13** -0.13**
(0.056) (0.057) (0.060) (0.062) (0.059)
Part-time
employed 0.05 -0.02 0.17*** 0.13** -0.13**
(0.056) (0.057) (0.060) (0.062) (0.059)
Self-
employed 0.05 -0.02 0.17*** 0.13** -0.13**
(0.056) (0.057) (0.060) (0.062) (0.059)
Unemployed 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Farmer 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Agricultural
workers 0.05 -0.02 0.17*** 0.13** -0.14**
(0.056) (0.057) (0.060) (0.062) (0.059)
Source: Author’s estimations.
Note: The multivariate probit model is estimated by maximum simulated likelihood with robust standard
errors (in parentheses) and includes all the controls of the baseline regression, as well as country and
cohort effects (not reported). The coefficients reported are those for the regressors on associational
activities presented in the equation where interpersonal trust is the dependent variable are reported
(columns 2-6 of Table 3). The first row “memory”refers to the results in Table 3. Statistical significance
of coefficient estimates at the 1, 5, and 10 percent levels is denoted by respectively (***), (**) and (*).
Households’ Intergenerational Structure and Participation in Associational Activities: The
Role of Interpersonal Trust
33
Table 8: Bivariate probit analysis: PCA-based regressor
Dependent variables
Associational Activities Interpersonal trust
Associational activity 0.02***
(0.005)
Two-generation family 0.02
(0.016)
Three-generation family 0.00
(0.019)
Married 0.03** 0.02**
(0.012) (0.007)
Head of household 0.00 0.02**
(0.015) (0.010)
Gender (1 = Male) 0.02** 0.01
(0.010) (0.007)
Age -0.00 0.00
(0.002) (0.001)
Ethnicity -0.00*** 0.00***
(0.000) (0.000)
Primary education -0.05*** -0.00
(0.014) (0.009)
Secondary education 0.08*** 0.01
(0.014) (0.009)
Tertiary education 0.22*** 0.06***
(0.016) (0.011)
Full-time employed 0.07*** -0.01
(0.013) (0.009)
Part-time employed 0.09*** -0.01
(0.019) (0.013)
Self-employed 0.04*** 0.01
(0.016) (0.010)
Luiz de Mello and João Tovar Jalles
34
Unemployed 0.04* -0.04***
(0.022) (0.015)
Farmer -0.02 0.06***
(0.017) (0.012)
Agricultural workers -0.03* 0.00
(0.017) (0.011)
No. of observations 18,680 18,680
Source: Author’s estimations.
Note: The bivariate probit models are estimated by maximum likelihood with robust standard errors (in
parentheses) and includes country and cohort effects (not reported). Statistical significance of coefficient
estimates at the 1, 5, and 10 percent levels is denoted by respectively (***), (**) and (*).
