ABSTRACT
This article evaluates the relationship between welfare state institutions and informal voluntary activity. In particular, we wish to investigate the extent to which welfare state arrangements influence the individual's decision to volunteer informally. Focusing on informal volunteering – a vital, yet often neglected form of social capital – we reinvestigate the hypotheses on crowding in and crowding out. Our sub-national multilevel analysis of the Swiss cantons reveals that there is no significant effect of welfare statism on informal volunteering. Cantonal differences in individual propensity to volunteer informally can rather be attributed to different cultural-linguistic backgrounds as well as to the cantonal degree of urbanisation.
1. Introduction
Recent literature on social capital has increasingly been focusing on the role the state and political institutions have in the creation of social capital (e.g., Hooghe and Stolle 2003; Rothstein and Stolle 2003; Kumlin and Rothstein 2005; Freitag 2006; Larsen 2007; Rothstein and Stolle 2008). Over the course of the last two decades, social capital researchers have paid particular attention to the influence of welfare state institutions on the formation of civil society (Wolfe 1989; Skocpol 1996; Zijderveld 1999; Rothstein 2001; Salamon and Sokolowski 2003). While one of the leading hypotheses holds that expansionary welfare states harm social capital, another view, however, claims the opposite: Rather than crowding out the civic spirit in society, the welfare state promotes civic institutions and practices (van Oorschot and Arts 2005). This well-known debate forms the starting point of our analysis. Although a growing number of studies focus on the welfare state-social capital nexus, two important issues remain to be addressed.
First, we argue that for a better understanding of the effects of welfare state arrangements on social capital one needs to move beyond general discussions of welfare state impact on social capital; instead, one should concentrate on specific forms of social capital. In this respect, we will investigate the effect of welfare state institutions on an important, yet often neglected form of social capital in this debate, namely, informal voluntary work. In general, volunteering is defined as follows: ‘Volunteering means any activity in which time is given freely to benefit another person, group or organization’ (Wilson 2000: 215; emphasis added). To date, only a handful of authors have analysed the connection between social capital and welfare statism by focusing on informal volunteering (Boje and Strandh 2005; Kääriäinen and Lehtonen 2006; Finsveen et al.2008; van der Meer et al.2008). While formal volunteering through organisations that recruit and organise volunteers is an adequate (and frequently used) indicator for social cohesion, so too is informal voluntary activity. Informal volunteering refers to activities such as helping and supporting friends, neighbours, acquaintances and relatives (outside of one's own household) that take place directly between the people involved and outside of any formally organised structure. Social networks are built equally well on these less formal and more private person-to-person kinds of social interaction, typical in the context of unpaid informal work (Boje and Strandh 2005). More importantly, recent research on volunteering has shown that it is informal volunteering in particular that bolsters social cohesion of society (Kwak et al.2004; Boje and Strandh 2005; Taylor 2005; Stadelmann-Steffen et al.2007). Having said this, we argue that a specific type of social capital is immanent within informal volunteering, as this non-paid activity encompasses both social interaction as well as the inherent value of this interaction. This special attribute of informal volunteering is well illustrated by the so-called third person criterion (Bühlmann and Schmid 1999): Voluntary vs. recreational or leisure activities are accordingly differentiated by ascertaining whether it is possible to pay a third person to conduct the activity in question. For example, exercising or a trip to the beauty salon rank among recreational or leisure activities because a third person cannot be paid to do these activities for someone else. Driving someone outside of one's own household to sports practice or to the dentist are regarded as informal volunteering in the sense of the above definition. By means of the third person criterion, a defined material or immaterial value is assigned to any voluntary social activity so that informal volunteering becomes social capital.1 Finally, with regard to previous research on welfare systems and social capital, the results concerning the crowding in or crowding out effect seem to strongly depend on the respective dimension of social capital (Van Oorschot et al.2005; Kääriäinen and Lehtonen 2006; Stadelmann-Steffen forthcoming). Social capital comprises both structural aspects (i.e., social networks) as well as cultural aspects (i.e., trust and social norms and values) (van Deth 2003). Whereas empirical results concerning social trust predominantly reject the crowding out hypothesis (Rothstein and Stolle 2003; Delhey and Newton 2005; Kumlin and Rothstein 2005), we find a variety of contradictory results with respect to the structural dimension of social capital, meaning informal and formal voluntary activities, and social contacts (see for example Curtis et al.2001; Scheepers et al.2002; van Oorschot and Arts 2005; van der Meer et al.2009). In addition to these rather inconclusive findings, empirical results on informal volunteering remain very scarce (Boje and Strandh 2005; Kääriäinen and Lehtonen 2006; Finsveen et al.2008; van der Meer et al.2008).
Second, since informal volunteering is an inherently local social phenomenon, we will confine our analysis to the sub-national level;2 more precisely, we will investigate the influence of Swiss cantonal welfare state regimes on individual informal volunteer activity. The Swiss cantons are particularly well-suited to such an analysis, as they are powerful political entities in an exceptionally decentralized federation (Lijphart 1999; Vatter 2002). Although the Swiss cantons comprise a multitude of political and economic structures, they remain entities within a single national political system and therefore have many characteristics in common that may be treated as constants (Vatter 2002; Vatter and Freitag 2007). In this sense, intra-national comparisons may prove more fruitful than comparisons at the national level. This type of comparison also lends itself quite well to the demands of a most similar systems research design (Przeworski and Teune 1970; Lijphart 1975). Th e Swiss cantons therefore present an excellent opportunity to assess the effect of welfare state policy, exhibiting considerable autonomy in social policy and therefore substantial differences in the orientation and the design of instruments and measures implemented. These differences have also been noted by Armingeon et al. (2004), who – based on Esping-Andersen (1990) – identify four types of cantonal social security systems: liberal, conservative-Catholic, social democratic and a mixed regime that shares attributes of the first two systems, labelled liberal-conservative. With this typology, they are able to illustrate that the differences in the cantonal welfare state policies in fact resemble those between national welfare regimes.
Our approach is to use hierarchical models to simultaneously illuminate individual, societal, and, in particular, political-institutional conditions for informal voluntary activities in a systematic comparative manner. In empirical social research, this methodology is seen as a tried and tested means for overcoming micro-macro dualism, since individual attitudes and behaviour are structured both by personal traits and social context (Anderson and Singer 2008). In other words, societal conditions structure individual attitudes and influence personal behaviour. The individual data for our analysis come from the Swiss Volunteering Survey 2007 (Freiwilligen-Monitor Schweiz); official statistics are used for the aggregate-level data.
The article is structured as follows: In the next section we present previous research on the topic and illustrate the theoretical links between welfare state institutions and social capital as well as the hypotheses derived from them. We then present the data and methodological references in section 3, continuing with the empirical analysis in section 4. The article closes with a summary of the most important findings and some conclusions.
2. Theory, hypotheses and previous research
This study evaluates the relationship between welfare state institutions and individual informal voluntary activity. In particular, we wish to investigate the extent to which welfare state arrangements influence individual informal volunteering activity. Viewed analytically, this inquiry forces us to take hierarchical structures into account, as the assumption is posited that a macro-level condition (welfare state) is capable of influencing actions on the micro-level (the decision to volunteer). Theoretically, interactions with one's social surroundings or the observation of one's environment can shape individual choices. A specific incentive offered by the individual's economic and socio-political surroundings can influence the individual to act in a particular manner. From the perspective of neo-institutionalism, political institutions mould individual preferences and stimulate or limit behavioural options by means of certain incentive mechanisms (Mayntz and Scharpf 1995; Hall and Taylor 1996; Immergut 1998). In other words, contextual conditions act as an opportunity structure and influence individual attitudes and personal behavioural patterns (Schofer and Fourcade-Gourinchas 2001: 810). According to comparative studies on volunteering, incentive mechanisms encompass aspects of the government systems, as well as welfare state arrangements (Rothstein and Stolle 2003; Freitag 2006). This type of institutional research perspective places a broad critical focus on incentives or possible institutionalised barriers that coincide with civic engagement.
In this context, public services can be seen as ‘secondary institutions’, as the welfare state, once created, ‘can become an instrument of establishing and constructing identities’ (Crepaz 2007: 137). Consequently, the design of welfare state arrangements can influence individual attitudes and personal behavioural patterns in ways similar to the primary institutional setting of a country. For example, welfare state programs confer resources on and create specific incentives for individuals that structure and affect important personal choices, e.g., in terms of whether or not to volunteer (Esping-Andersen 1990; Pierson 1993). Furthering this line of thought leads to competing considerations on how to theorise on the relationship between welfare state policy and informal volunteering in detail. In general, two different theoretically plausible conjectures are found within the literature.
The first approach advocates the so-called ‘crowding out’ hypothesis (Habermas 1973; Offe 1984; Wolfe 1989; Coleman 1991; Etzioni 1994; Zijderveld 1999; Fukuyama 2000; Putnam 2000; Karlson 2004). Scholars supporting this perspective argue that encompassing welfare programs suppress private obligations presumed to build strong reciprocal ties for mutual support in situations of distress. Rather than relying on private social networks, the welfare state's citizens rely on the state's responsibility to guarantee social benefits in times of need. Consequently, social engagement in private personal networks decreases: ‘[P]eople will lose their moral sense of collective and communal duties and responsibilities’ (van Oorschot and Arts 2005: 6). In other words, a well-developed welfare state reduces the need for privately organised initiatives and thus leads to a ‘crowding out’ of informal voluntary work. The fundamental assumption of this perspective holds that the public provision of social services competes with social support and help provided privately by one's fellow citizens. If the state takes on these tasks, civic engagement, and particularly informal voluntary work, becomes less necessary. The resulting lack of initiative of the civic actors induces a passive civil society. This discussion thus leads to the following working hypothesis:
Hypothesis 1: The more encompassing the welfare state arrangements are, the greater the likelihood that informal volunteer activity will be crowded out by state activities.
A second approach, the so-called ‘crowding in’ hypothesis, assumes the existence of a more balanced and complementary relationship between the welfare state and private initiatives. In this vein, it is assumed that pervasive welfare state arrangements create the ‘structural and cultural conditions’ (van Oorschot and Arts 2005: 6) that pave the way for successful creation of informal voluntary activity (Skocpol 1996; Kuhnle and Alestalo 2000; Rothstein 2001; Salamon and Sokolowski 2003; Torpe 2003): Broad welfare programs help to provide circumstances of social peace and trust that afford people social security, thereby assuaging concerns over maintaining a living. People thus have increased financial and time resources which can be devoted to developing and maintaining informal volunteer work. According to this theory, the state and civil society do not necessarily perform identical functions. As such, generous welfare state services do not replace civic activities; rather, they allow for a differentiation of civil society into pre-defined areas (Kumlin and Rothstein 2005). However, according to this perspective, the state continues to be responsible for certain welfare services and seeks to cooperate with individual and societal actors. Against this background, welfare state models that are less hierarchically and bureaucratically structured provide increased opportunities for informal voluntary activity as a non-governmental source of public goods and services capable of relieving the professional systems (Heinze and Olk 2001; Zimmer 2007). From this discussion, the following working hypothesis is formed:
Hypothesis 2: The more encompassing the welfare state arrangements are, the greater the likelihood that informal volunteer activity will be strengthened by state activities.
3. Data, variables and methodological approach
For the remainder of this paper, our primary focus will be on testing the hypotheses presented above. The dependent variable is the reported individual informal volunteering activity as indicated by the Swiss Volunteering Survey 2007: ‘In addition to formal volunteering in associations and other organisations, there are alternative opportunities for volunteering. Did you perform another type of unpaid work beyond volunteering in associations or other organisations, such as babysitting (children other than your own), neighbourly-aid, participating in any kind of projects, organising a (street) party in your neighbourhood, etc. in the last four weeks? (The work has to be for the benefit of people outside one's own household)’ (question 46, own translation). Respondents indicating informal voluntary engagement were allocated the values of 1, while all others were assigned the value 0. We proceed analogously in order to measure social volunteering. Here, only respondents who indicate informal voluntary activities that imply personal helping or personal assistance were allocated the values of 1, while all others were assigned the value 0. Informal volunteering rates vary considerably across cantons. For instance, in Geneva, Zug and Ticino, only about a quarter of the resident population reported to have engaged in informal voluntary activities, whereas front-runners Nidwalden and Appenzell Inner-Rhodes have informal volunteering rates of 48 and 55%, respectively (see Table A in the appendix). The extent to which individual propensities to volunteer informally in a canton can be attributed to differently designed welfare state institutions is addressed thoroughly in the following analysis.
The central explanatory variable is the welfare state regime type. We will run three regression models, each with an alternative operationalisation of the welfare state. We will use Armingeon et al.'s (2004) indicator for welfare state regime type. The ideal welfare types applied to the Swiss cantons can be described as follows (Armingeon et al.2004: 23–4): The liberal welfare state is marked by a restricted transfer of payments. It relies primarily on market solutions to meet societal needs. People in distress can turn to residual mean-tested public social programs only as a last resort. This basic social security net provides no further guarantees beyond the poverty line. In place of providing public welfare schemes, liberal systems rather subsidise private insurance plans, which therefore become necessary for people to protect themselves against risks like unemployment, disease, poverty in old age, etc. By contrast, in the conservative-Catholic welfare state, there are generous income transfers and social services; benefits, however, are based on previous earnings and the social status the affected person had during employment. Assistance through public welfare programs is not provided until the family's capability to support itself is exhausted. This type of welfare regime encourages traditional family structures; it perpetuates the traditional model of the male breadwinner (Saint-Arnaud and Bernard 2003). The social democratic welfare regime implies comprehensive social programmes, which, by design, are universal. Programmes aim primarily to balance social inequality, irrespective of former employment status, gender, nationality, etc. Generally, the system tends to socialise the costs of raising a family and provides opportunities for individual independence (Rostila 2007: 225). See Table B in the appendix for the classification of cantons used in the following analysis.4
Furthermore, we will supplement the measurement of the aforementioned welfare state arrangements by social spending statistics: Here, we use two different indicators, namely, the cantonal welfare state's social spending per capita (understood as welfare state generosity) and the cantonal welfare state's social spending as a percentage of the GDP (conceived as welfare state priority). Social expenditures as a percentage of the GDP is the most widely used indicator of welfare efforts in empirical literature. It can be seen as a measure of the importance of welfare state policy relative to a country's available resources. The main shortcoming is the fact that the measure not only depends on public social spending, but also on the GDP, implying that high values of this indicator can be driven both by substantial social spending and a comparatively low GDP. Social expenditures per capita take into account that the effect of welfare spending depends largely on how much could actually be paid per a given population. See Table B in the appendix for the cantons’ social spending statistics.
Although we focus on the welfare state regime as the pivotal independent variable explaining the variation in informal volunteering across cantons, we will also include various controls at the individual and contextual levels of analysis. In addition to institutional arrangements affecting the incentives for individuals to volunteer informally, previous research on the determinants of volunteering show a variety of other factors that should be considered in order to avoid invalid findings. Concerning the selection of control variables at the individual level, we draw on the findings of previous research dedicated to explaining volunteerism (see Wilson 2000 for an overview). The final selection of variables is based on Verba et al. (1995). They argue that there are three reasons as to why individuals do not actively participate: ‘because they can't; because they don't want to; or because nobody asked’ (15). Therefore, a willingness to volunteer informally implies resources (such as time, money, and education), psychological engagement with collective and communal duties (socio-cultural aspects such as religiosity) and social contacts (acquaintances, family) that simplify recruitment and provide opportunities to volunteer. In order to capture the factors mentioned by Verba et al. (1995) and including standard demographics, the following individual variables should be considered as potential determinants of informal volunteering: level of education, professional status, nationality (national/foreigner), employment status, number of children, duration of residency, religious denomination, age, and gender. Data for these variables are taken from the Swiss Volunteering Survey 2007.
To take potential confounding contextual variables into account, we include the cantonal level of urbanisation, language region and religious context in our analysis. Previous research on volunteering in Switzerland shows that these variables proved to be of importance in explaining volunteering (see e.g., Bühlmann and Freitag 2007). Whereas the degree of urbanisation is supposed to have a negative impact on informal volunteering (Kriesi 2004), research on Swiss informal volunteering consistently shows that voluntary activity in the Swiss German cantons is significantly higher than in the Latin language region (cp. Stadelmann-Steffen et al.2007). We further assume that a canton's religious context (measured by the share of non-religious people) impacts informal volunteering: In regions with larger shares of a religious population, individuals are presumed to exhibit an increased propensity to volunteer informally than in areas that are less devout (Bühlmann and Freitag 2007).5 We use the values of the contextual factors measured prior to 2006 to ensure that the potential cause precedes the effect. Further details on data sources and operationalisation of variables used in the analysis can be found in Table C of the appendix.
4. Empirical results: informal volunteering in the welfare state
In this section we subject our theory-derived relationships to empirical tests, with a focus on the influence of welfare state arrangements on the development of informal volunteering. As mentioned above, we will present three different models, each with different operationalisation of the welfare state. The first and second model quantitatively measure welfare statism using cantonal social spending per capita and cantonal social spending as a percentage of the GDP, respectively. Model 3 estimates welfare statism qualitatively using the regime typology of Armingeon et al. (2004) based on Esping-Andersen (1990). As can be seen in Table 1 the results for the ‘empty models’ yield statistically significant contextual variance. In addition, the ‘likelihood ratio test versus logistic regression’ provided by STATA.10 also confirms the adequacy of using multilevel analysis for the data at hand.
| . | Informal volunteering . | Social informal volunteering . |
|---|---|---|
| Contextual variance | 0.03 (0.01) | 0.08 (0.01) |
| Likelihood Ratio test vs. logistic regression | 20.54*** | 72.83*** |
| . | Informal volunteering . | Social informal volunteering . |
|---|---|---|
| Contextual variance | 0.03 (0.01) | 0.08 (0.01) |
| Likelihood Ratio test vs. logistic regression | 20.54*** | 72.83*** |
Note: Estimated contextual variance, standard errors in brackets. Individual variance is set to 1, with standard error 0. ***P<0.001
First, we will briefly report findings at the individual level of analysis, continuing with a presentation of effects at the contextual level, paying special attention to the welfare state regime type (see Table 2).
| . | Social spending per capita . | Social spending as % of GDP . | Welfare regime type . |
|---|---|---|---|
| Fixed effects . | Log odds (standard error) . | ||
| Individual level | |||
| Resources | |||
| Education (ref: middle) | |||
| Education none completed | −0.721***(0.198) | −0.720***(0.198) | −0.720***(0.198) |
| Education low | 0.002(0.090) | 0.001(0.090) | 0.002(0.090) |
| Education high | 0.064(0.063) | 0.064(0.063) | 0.066(0.063) |
| Professional status | 0.298***(0.062) | 0.298***(0.062) | 0.298***(0.062) |
| Employment status (ref: full-time) | |||
| Part-time | 0.373***(0.073) | 0.374***(0.073) | 0.375***(0.073) |
| Not working | 0.476***(0.117) | 0.476***(0.117) | 0.475***(0.117) |
| Foreigner | −0.441***(0.087) | −0.441***(0.087) | −0.435***(0.087) |
| Social integration | |||
| Children | 0.347***(0.070) | 0.348***(0.070) | 0.348***(0.070) |
| Duration of residency | 0.184*(0.078) | 0.183*(0.078) | 0.191*(0.078) |
| Socio-cultural aspects | |||
| Religiosity | 0.025(0.080) | 0.024(0.080) | 0.024(0.080) |
| Socio-demographics | |||
| Age | 0.019(0.010) | 0.024(0.080) | 0.019(0.010) |
| Ageˆ2 | −0.000**(0.000) | −0.000**(0.000) | −0.000**(0.000) |
| Sex | −0.385***(0.063) | −0.385***(0.063) | −0.385***(0.063) |
| Contextual level | |||
| Welfare Effort | −0.000(0.000) | −7.594(10.593) | |
| Welfare regime type (ref: social democratic) | |||
| Liberal | 0.118(0.096) | ||
| Conservative | 0.136(0.090) | ||
| Urbanisation | −0.004*(0.002) | −0.004*(0.002) | −0.002(0.002) |
| Language Region | 0.003*(0.001) | 0.003*(0.001) | 0.004***(0.001) |
| Religious Context | 1.589(0.873) | 1.289(0.727) | 1.209(0.671) |
| Constant | −0.954***(0.278) | −0.928**(0.294) | −1.235***(0.309) |
| Random effect | |||
| Contextual variance % of variance explained | 1.60e-13 (4.14e-08) 100% | 7.15e-12 (8.91e-07) 100% | 1.30e-20 (8.88e-12) 100% |
| N (# of cantons) Deviance | 5951(26) 7695.484 | 5951(26) 7695.784 | 5951(26) 7693.930 |
| . | Social spending per capita . | Social spending as % of GDP . | Welfare regime type . |
|---|---|---|---|
| Fixed effects . | Log odds (standard error) . | ||
| Individual level | |||
| Resources | |||
| Education (ref: middle) | |||
| Education none completed | −0.721***(0.198) | −0.720***(0.198) | −0.720***(0.198) |
| Education low | 0.002(0.090) | 0.001(0.090) | 0.002(0.090) |
| Education high | 0.064(0.063) | 0.064(0.063) | 0.066(0.063) |
| Professional status | 0.298***(0.062) | 0.298***(0.062) | 0.298***(0.062) |
| Employment status (ref: full-time) | |||
| Part-time | 0.373***(0.073) | 0.374***(0.073) | 0.375***(0.073) |
| Not working | 0.476***(0.117) | 0.476***(0.117) | 0.475***(0.117) |
| Foreigner | −0.441***(0.087) | −0.441***(0.087) | −0.435***(0.087) |
| Social integration | |||
| Children | 0.347***(0.070) | 0.348***(0.070) | 0.348***(0.070) |
| Duration of residency | 0.184*(0.078) | 0.183*(0.078) | 0.191*(0.078) |
| Socio-cultural aspects | |||
| Religiosity | 0.025(0.080) | 0.024(0.080) | 0.024(0.080) |
| Socio-demographics | |||
| Age | 0.019(0.010) | 0.024(0.080) | 0.019(0.010) |
| Ageˆ2 | −0.000**(0.000) | −0.000**(0.000) | −0.000**(0.000) |
| Sex | −0.385***(0.063) | −0.385***(0.063) | −0.385***(0.063) |
| Contextual level | |||
| Welfare Effort | −0.000(0.000) | −7.594(10.593) | |
| Welfare regime type (ref: social democratic) | |||
| Liberal | 0.118(0.096) | ||
| Conservative | 0.136(0.090) | ||
| Urbanisation | −0.004*(0.002) | −0.004*(0.002) | −0.002(0.002) |
| Language Region | 0.003*(0.001) | 0.003*(0.001) | 0.004***(0.001) |
| Religious Context | 1.589(0.873) | 1.289(0.727) | 1.209(0.671) |
| Constant | −0.954***(0.278) | −0.928**(0.294) | −1.235***(0.309) |
| Random effect | |||
| Contextual variance % of variance explained | 1.60e-13 (4.14e-08) 100% | 7.15e-12 (8.91e-07) 100% | 1.30e-20 (8.88e-12) 100% |
| N (# of cantons) Deviance | 5951(26) 7695.484 | 5951(26) 7695.784 | 5951(26) 7693.930 |
Note: *P<0.05; **P<0.01; ***P<0.001.
Overall, individual findings are consistent with previous results on informal volunteering in Switzerland (Nollert and Huser 2007; Stadelmann-Steffen et al.2007). The estimations confirm that educational attainment does not substantively influence informal engagement. Nevertheless, a higher professional status (currently or in the past) is associated with more frequent informal volunteerism than a lower status. Concerning employment status, non-working people and part-time workers tend to engage more frequently in informal volunteer activities than full-time workers.6 Affirming former studies’ findings, foreigners participate less than Swiss nationals. Furthermore, social integration reveals to be an important factor for being involved in informal voluntary activities: Having school-aged children at home as well as a relatively long duration of residency positively affects informal volunteering. In corroboration with prior research, the relationship between informal volunteering and age is inversely u-shaped, with women being more active than men. Finally, there is no statistically significant effect of individual religiosity on informal volunteering.7
We will now address empirical findings at the contextual level of analysis. First, we can state that results of each of the three models, irrespective of the operationalisation of the welfare state, point consistently to the same conclusion regarding the influence of welfare state policy on informal volunteering. Although the signs of the regression coefficient of the welfare state variables uniformly indicate a crowding out of informal volunteering by welfare statism, none of the coefficients reach statistical significance. Consequently, we conclude that there is no meaningful effect of the characteristic shaping of welfare state arrangements on informal volunteering in Switzerland – neither confirming the argument of welfare state optimists, nor that of pessimists.
In further analyses, we tested the effect of welfare state effort exclusively on social informal volunteering, meaning informal voluntary activities that incorporate personal helping or personal assistance. As mentioned above, a crowding out effect should therefore be most probable for this specified area of informal volunteering since a potential trade-off between public and private services, is most probable if the activities compared (of the state and civil society) can replace each other. However, as can be seen in these re-specified models (Table 3), neither the welfare state effort nor the welfare regime types are significantly associated with social informal volunteering.
| . | Social spending per capita . | Social spending as % of GDP . | Welfare regime type . |
|---|---|---|---|
| Fixed effects . | Log odds (standard error) . | ||
| Individual level | |||
| Resources | |||
| Education (ref: middle) | |||
| Education none completed | −0.852***(0.246) | −0.852***(0.246) | −0.843***(0.246) |
| Education low | −0.064(0.099) | −0.064(0.100) | −0.061(0.100) |
| Education high | 0.012(0.070) | 0.011(0.070) | 0.012(0.070) |
| Professional status | 0.185** (0.068) | 0.184** (0.068) | 0.184** (0.068) |
| Employment status (ref: full-time) | |||
| Part-time | 0.270***(0.080) | 0.271***(0.080) | 0.270***(0.080) |
| Not working | 0.402**(0.129) | 0.403**(0.129) | 0.399**(0.129) |
| Foreigner | −0.391***(0.100) | −0.391***(0.100) | −0.395***(0.100) |
| Social integration | |||
| Children | 0.304***(0.077) | 0.304***(0.077) | 0.305***(0.077) |
| Duration of residency | 0.246**(0.089) | 0.245**(0.089) | 0.240**(0.089) |
| Socio-cultural aspects | |||
| Religiosity | 0.008(0.089) | 0.007(0.089) | 0.009(0.089) |
| Socio-demographics | |||
| Age | 0.028*(0.011) | 0.028*(0.011) | 0.028*(0.011) |
| Ageˆ2 | −0.000**(0.000) | −0.000**(0.000) | −0.000**(0.000) |
| Sex | −0.501***(0.070) | −0.500***(0.070) | −0.499***(0.070) |
| Contextual level | |||
| Welfare Effort | −0.000(0.000) | −10.072(11.805) | |
| Welfare regime type (ref: social democratic) | |||
| Liberal | −0.067(0.104) | ||
| Conservative | −0.128(0.100) | ||
| Urbanisation | −0.002(0.002) | −0.002(0.002) | −0.003(0.002) |
| Language Region | 0.006***(0.001) | 0.006***(0.001) | 0.006***(0.001) |
| Religious Context | 0.918(0.969) | 0.620(0.804) | 0.200(0.748) |
| Constant | −2.049***(0.315) | −2.003***(0.333) | −1.940***(0.346) |
| Random effect | |||
| Contextual variance % of variance explained | 3.21e-13 (5.28e-08) 100% | 1.32e-12 (1.08e-07) 100% | 8.75e-15 (7.94e-09) 100% |
| N (# of cantons) | 5951 (26) | 5951 (26) | 5951 (26) |
| Deviance | 6657.28 | 6657.37 | 6656.33 |
| . | Social spending per capita . | Social spending as % of GDP . | Welfare regime type . |
|---|---|---|---|
| Fixed effects . | Log odds (standard error) . | ||
| Individual level | |||
| Resources | |||
| Education (ref: middle) | |||
| Education none completed | −0.852***(0.246) | −0.852***(0.246) | −0.843***(0.246) |
| Education low | −0.064(0.099) | −0.064(0.100) | −0.061(0.100) |
| Education high | 0.012(0.070) | 0.011(0.070) | 0.012(0.070) |
| Professional status | 0.185** (0.068) | 0.184** (0.068) | 0.184** (0.068) |
| Employment status (ref: full-time) | |||
| Part-time | 0.270***(0.080) | 0.271***(0.080) | 0.270***(0.080) |
| Not working | 0.402**(0.129) | 0.403**(0.129) | 0.399**(0.129) |
| Foreigner | −0.391***(0.100) | −0.391***(0.100) | −0.395***(0.100) |
| Social integration | |||
| Children | 0.304***(0.077) | 0.304***(0.077) | 0.305***(0.077) |
| Duration of residency | 0.246**(0.089) | 0.245**(0.089) | 0.240**(0.089) |
| Socio-cultural aspects | |||
| Religiosity | 0.008(0.089) | 0.007(0.089) | 0.009(0.089) |
| Socio-demographics | |||
| Age | 0.028*(0.011) | 0.028*(0.011) | 0.028*(0.011) |
| Ageˆ2 | −0.000**(0.000) | −0.000**(0.000) | −0.000**(0.000) |
| Sex | −0.501***(0.070) | −0.500***(0.070) | −0.499***(0.070) |
| Contextual level | |||
| Welfare Effort | −0.000(0.000) | −10.072(11.805) | |
| Welfare regime type (ref: social democratic) | |||
| Liberal | −0.067(0.104) | ||
| Conservative | −0.128(0.100) | ||
| Urbanisation | −0.002(0.002) | −0.002(0.002) | −0.003(0.002) |
| Language Region | 0.006***(0.001) | 0.006***(0.001) | 0.006***(0.001) |
| Religious Context | 0.918(0.969) | 0.620(0.804) | 0.200(0.748) |
| Constant | −2.049***(0.315) | −2.003***(0.333) | −1.940***(0.346) |
| Random effect | |||
| Contextual variance % of variance explained | 3.21e-13 (5.28e-08) 100% | 1.32e-12 (1.08e-07) 100% | 8.75e-15 (7.94e-09) 100% |
| N (# of cantons) | 5951 (26) | 5951 (26) | 5951 (26) |
| Deviance | 6657.28 | 6657.37 | 6656.33 |
Note: *P<0.05; **P<0.01; ***P<0.001.
Relating the results to the meagre body of research on informal volunteering in the welfare state, we are able to add strength to the general finding that there is no clear story to be told on this relationship. In line with the analyses provided by Boje and Strandh (2005), Finsveen et al. (2008), Kääriäinen and Lehtonen (2006), and van der Meer et al. (2008), we detect no significant effects of social spending expenditures or welfare state regimes on informal volunteering in our sub-national analysis.
With regard to the control variables, our analysis shows that the individual's propensity to volunteer informally is influenced by the degree of urbanisation and the cultural background shaped by the linguistic region in which one resides. These socio-structural and cultural effects dominate the shaping of individual informal volunteerism such that there is no significant role for welfare state institutions to play. The higher the degree of urbanisation, the lesser the extent of informal volunteering. Furthermore, Swiss-Germans’ informal voluntary activity is significantly higher than that of the French-speaking and Italian-speaking Swiss. We have to note though that basing calculations on the welfare state typology, the effect of urbanisation loses statistical significance; the influence of the linguistic region however remains robust.8 Notably, the German-speaking cantons are those cantons with the most liberal institutions of social welfare politics, while in the French- and Italian-speaking cantons conservative and social democratic welfare statism prevails. It is therefore a very complex task to neatly disentangle the cultural aspects (language region) from the institutional feature (welfare state).9 Nevertheless, together these cantonal characteristics can completely explain cantonal differences in informal volunteering.10
5. Conclusion
The interplay between a state's civil society and its welfare state politics is currently receiving a great deal of attention in academic debates. The turning point in this debate is the question of the extent to which state-organised social services can replace or, conversely, promote (informal) voluntary activity by members of society (Dahlberg 2005). Although most studies have found an influence of welfare state politics on civic engagement or trust, the direction of the effect of this relationship has been assessed differently depending on the indicators examined (cp. for example Boje and Strandh 2005; Daatland and Lowenstein 2005; Scheepers and Grotenhuis 2005; van Oorschot and Arts 2005; van Oorschot et al.2005; Kääriäinen and Lehtonen 2006 Stadelmann-Steffen forthcoming).
In their empirical analyses, researchers focus on different aspects of civil society or social capital, sometimes focusing on the structural aspects, foremost social networks and sometimes on the cultural facet, mainly social trust. With regard to social trust, the investigations clearly corroborate the assumption that a comprehensive welfare state provides optimal conditions for a deep trusting society, explicitly rejecting the crowding out of social trust as one important component of the social capital concept. Nevertheless, no paralleled level of clarity exists concerning the network aspect of social capital. In this domain, either the conflicting reasoning found in the theoretical accounts gets carried over to the empirical results or researchers find no effect whatsoever.
In view of the strong theoretical argument suggesting a systematic influence of political institutions such as the welfare state on informal volunteering on the one hand, and the inconclusiveness of results on the other, we investigated this matter – this time with an alternatively elaborated research design and high quality data on informal volunteering in Switzerland. Rather than focusing on cross-national comparisons, we have ‘scaled down’ (Snyder 2001) to the sub-national level. Using multilevel analysis, we have investigated the influence of welfare state commitment in the Swiss cantons on a fundamental dimension of individual social capital – informal volunteering.
Particularly in Switzerland, recent findings have shown that it is the informal kind of unpaid work that is the most important and dominant form of volunteering and, as such, an indispensable element of the social capital keeping society together (Stadelmann-Steffen et al.2007). Similarly to Boje and Strandh (2005), Finsveen et al. (2008), Kääriäinen and Lehtonen (2006), and van der Meer et al. (2008), we do not find any conclusive impact of welfare state size on informal voluntary activity. Our sub-national analysis of the Swiss welfare statism thus confirms these cross-national investigations of welfare state effects. Individual propensity to volunteer informally in the Alpine Republic is steered rather by socio-structural and cultural conditions than by welfare institutions. Regardless of the characteristic shaping of the welfare state, individuals living in the countryside simply engage more frequently in voluntary activities. The same holds true for people from a Swiss-German cultural background. In short, we found neither a crowding in nor a crowding out effect of the cantonal welfare state programmes on informal volunteering activity.
These findings could indicate that the theoretical importance attributed to the impact of welfare state size on social capital is, at least concerning informal volunteering in Switzerland, overemphasized. However, we should not hastily conclude that there is no effect at all of welfare statism on informal volunteering. Welfare state arrangements may crowd out voluntary engagement in some instances, while boosting engagement in other contexts (Stadelmann-Steffen forthcoming), thereby preventing a clear result in statistical terms to be found in the present analysis. Before we abandon the theoretically strong idea that welfare state arrangements could have a substantive impact on informal volunteer behaviour, we should first refine the analysis on this topic: It is necessary to further theoretically and empirically single out which aspects of welfare state arrangements, which institutions and which kind of welfare spending are directly connected to individual willingness and capability to volunteer informally in specified areas. Moreover, we should elaborate on the measurement of welfare state efforts in order to ensure more reliable empirical analyses in this field. The measurements currently used in the investigations on social capital in welfare states are indeed very rough measurements and are far from providing an encompassing picture of comprehensive welfare states. The persistent urgency to reform the welfare systems, coupled with the indispensable role that informal volunteering plays in keeping society together, serves as a motivational push to overcome these shortcomings in further research.
Footnotes
An informal voluntary activity is profitable for the supplier as well as for the beneficiary: The recipient benefits from the volunteer's charitableness, whose dedication can, in turn, be regarded as an investment in social capital, thereby generating an obligation on the part of the beneficiary which may be honoured in the future. This rule is valid particularly if norms of reciprocity are observed and if there is a high degree of interpersonal trust between the involved persons.
Informal voluntary activity (helping and supporting people around us like friends, neighbours, acquaintances) is virtually per definition an activity that takes place in the proximate social surroundings. We are well aware, however, that – at least theoretically – there are possible situations in which this type of volunteering takes place beyond one's immediate community. We are grateful to the anonymous reviewers for this comment; however, looking at the overall empirical results on informal volunteering in the Swiss Volunteering Study (2007), we conclude that it is correct to define informal volunteering in Switzerland as a truly local phenomenon (Stadelmann-Steffen et al. 2007). In Switzerland even the formal voluntary activity is predominantly performed locally: The Swiss Volunteering Survey (2007) shows that over 80% of the Swiss indicate that their formal voluntary activity is locally oriented (ibid.).
We should repeat that the focus in this article is on one specific aspect of informal social capital, namely informal volunteering. Therefore, we do not consider other contributions that address informal social capital without explicit mention of informal volunteering (see e.g., Scheepers et al.2002 and van der Meer et al.2009).
As mentioned above, Armingeon et al. (2004) distinguish four different types of welfare state regimes: liberal, conservative, social democratic and liberal-conservative. Concerning the detailed procedure for the classification of cantonal welfare regimes, we refer to the description given by Armingeon et al. (2004). For simplicity's sake, we combine the conservative and liberal-conservative categories in our empirical analysis. This procedure is straightforward as the liberal and the social democratic welfare states are the most clearly distinguished types of welfare states and typically seen as the classic antipodes in welfare state typologies whereas the conservative type is in itself a hybrid as is the case with the liberal-conservative in the Swiss cantons. Furthermore, results do not differ if we do not collapse the two hybrids in one category. Additionally, Armingeon et al. (2004) investigate four different sub-regimes of the cantonal welfare regimes, namely, employment, education, taxation and social security. We will limit our analysis to the field of social security because, at least theoretically, this field is the most directly connected to engagement in informal volunteering, which mainly covers informal assistance.
It is widely accepted that religious people generally share an ethic that boosts the propensity to volunteer, compared to non-religious p eople (Wilson and Musick 1997). Ideally, we would base our measurement of the religious context (the share of religious people) on considerations such as church attendance, frequency of prayer, or similar more direct measurements of religiosity. Unfortunately such data is not available.
We had insufficient data on occupation and income and we therefore could not directly include these variables in our models. The educational level and the professional status represent the best proxies available for the socioeconomic status. Further data measuring social class (cp. Davis Smith 1998) were not available in the Swiss Volunteering Survey 2007.
In models not presented in the article, we checked if the results changed by introducing further individual controls sometimes used to explain voluntary activity, such as trust, political interest, and contact frequency with friends; the inclusion of these controls did not however change our results. Against this background, we decided to keep the parsimonious model, excluding variables being not explicitly a root cause of informal volunteering. We thank the anonymous reviewer for pointing out this matter.
We tested the sensitivity of the empirical findings by including other controlling factors such as cantonal GDP, cantonal direct democracy, cantonal autonomy, share of Protestants and Catholics, respectively, age distribution of the cantonal populations. However, the results do not change in any substantive way: While the welfare state variables remain statistically insignificant, the language region and/or the degree of urbanisation consistently show statistically significant effects on informal volunteering. Furthermore, results do not change if we exclude Basle-Town and/or Geneva – both exhibiting remarkably high social spending figures per capita – from the analysis.
However, tests of multi-collinearity at the contextual level of analysis show that, according to common standards, multi-collinearity is not an issue for our models: variance inflation factors (VIFs) on the contextual level do not exceed 2.65, 2.8 or 2.61 for the models 1, 2, and 3, respectively.
Compared to the null model, the contextual variables account for 100 percent of variance at the cantonal level.
Appendix
| Canton . | Informal volunteering rate . |
|---|---|
| Geneva | 25.05% |
| Zug | 25.52% |
| Ticino | 26.44% |
| Vaud | 32.89% |
| Neuchâtel | 34.27% |
| Appenzell Outer-RhodesR | 34.62% |
| Valais | 35.22% |
| Jura | 35.63% |
| Grison | 35.87% |
| Zurich | 37.03% |
| Basle-Town | 37.21% |
| Schaffhausen | 38.16% |
| Argovia | 38.83% |
| Fribourg | 39.18% |
| Basle-Country | 39.56% |
| Glarus | 39.61% |
| Solothurn | 40.82% |
| Berne | 40.88% |
| Schwyz | 41.28% |
| Uri | 41.52% |
| Lucerne | 42.28% |
| Obwalden | 43.24% |
| St Gall | 43.52% |
| Thurgovia | 43.78% |
| Nidwalden | 47.71% |
| Appenzell Inner-Rhodes | 54.60% |
| Canton . | Informal volunteering rate . |
|---|---|
| Geneva | 25.05% |
| Zug | 25.52% |
| Ticino | 26.44% |
| Vaud | 32.89% |
| Neuchâtel | 34.27% |
| Appenzell Outer-RhodesR | 34.62% |
| Valais | 35.22% |
| Jura | 35.63% |
| Grison | 35.87% |
| Zurich | 37.03% |
| Basle-Town | 37.21% |
| Schaffhausen | 38.16% |
| Argovia | 38.83% |
| Fribourg | 39.18% |
| Basle-Country | 39.56% |
| Glarus | 39.61% |
| Solothurn | 40.82% |
| Berne | 40.88% |
| Schwyz | 41.28% |
| Uri | 41.52% |
| Lucerne | 42.28% |
| Obwalden | 43.24% |
| St Gall | 43.52% |
| Thurgovia | 43.78% |
| Nidwalden | 47.71% |
| Appenzell Inner-Rhodes | 54.60% |
Source: Swiss Volunteering Survey 2007. Proportions are weighted with an individual weight for cantonal sampling effects, household size, and post stratification (according to age, sex, nationality, educational attainment) based on Swiss census of population 2000 (Stadelmann-Steffen et al. 2007).
| Canton . | Welfare regime type . | Social spending per capita . | Social spending as % of GDP (*1000) . |
|---|---|---|---|
| Zurich | Social democratic | 307.55 | 4.83 |
| Berne | Liberal | 427.65 | 9.89 |
| Lucerne | Liberal-conservative | 274.83 | 6.51 |
| Uri | Liberal-conservative | 182.75 | 4.10 |
| Schwyz | Liberal-conservative | 167.70 | 3.37 |
| Obwalden | Liberal-conservative | 139.58 | 3.70 |
| Nidwalden | Liberal-conservative | 197.40 | 3.18 |
| Glarus | Liberal | 82.78 | 1.38 |
| Zug | Conservative | 364.15 | 4.33 |
| Fribourg | Conservative | 268.03 | 6.91 |
| Solothurn | Liberal-conservative | 178.55 | 3.99 |
| Basle-Town | Social democratic | 1070.68 | 12.43 |
| Basle-Country | Social democratic | 239.63 | 4.63 |
| Schaffhausen | Social democratic | 200.05 | 3.86 |
| Appenzell Outer-Rhodes | Liberal | 105.23 | 2.40 |
| Appenzell Inner-Rhodes | Liberal | 175.80 | 4.02 |
| St Gall | Liberal | 212.28 | 4.83 |
| Grison | Liberal | 223.33 | 5.02 |
| Argovia | Liberal | 200.98 | 4.14 |
| Thurgovia | Liberal | 131.50 | 3.00 |
| Ticino | Social democratic | 474.50 | 12.11 |
| Vaud | Conservative | 652.15 | 13.04 |
| Valais | Conservative | 243.98 | 6.61 |
| Neuchâtel | Conservative | 584.10 | 13.13 |
| Geneva | Social democratic | 1238.80 | 21.98 |
| Jura | Liberal | 342.48 | 9.35 |
| Canton . | Welfare regime type . | Social spending per capita . | Social spending as % of GDP (*1000) . |
|---|---|---|---|
| Zurich | Social democratic | 307.55 | 4.83 |
| Berne | Liberal | 427.65 | 9.89 |
| Lucerne | Liberal-conservative | 274.83 | 6.51 |
| Uri | Liberal-conservative | 182.75 | 4.10 |
| Schwyz | Liberal-conservative | 167.70 | 3.37 |
| Obwalden | Liberal-conservative | 139.58 | 3.70 |
| Nidwalden | Liberal-conservative | 197.40 | 3.18 |
| Glarus | Liberal | 82.78 | 1.38 |
| Zug | Conservative | 364.15 | 4.33 |
| Fribourg | Conservative | 268.03 | 6.91 |
| Solothurn | Liberal-conservative | 178.55 | 3.99 |
| Basle-Town | Social democratic | 1070.68 | 12.43 |
| Basle-Country | Social democratic | 239.63 | 4.63 |
| Schaffhausen | Social democratic | 200.05 | 3.86 |
| Appenzell Outer-Rhodes | Liberal | 105.23 | 2.40 |
| Appenzell Inner-Rhodes | Liberal | 175.80 | 4.02 |
| St Gall | Liberal | 212.28 | 4.83 |
| Grison | Liberal | 223.33 | 5.02 |
| Argovia | Liberal | 200.98 | 4.14 |
| Thurgovia | Liberal | 131.50 | 3.00 |
| Ticino | Social democratic | 474.50 | 12.11 |
| Vaud | Conservative | 652.15 | 13.04 |
| Valais | Conservative | 243.98 | 6.61 |
| Neuchâtel | Conservative | 584.10 | 13.13 |
| Geneva | Social democratic | 1238.80 | 21.98 |
| Jura | Liberal | 342.48 | 9.35 |
Sources: Welfare regime types from Armingeon et al. (2004); social spending statistics (averages 2001–2004): Eidgenössische Finanzverwaltung, Öffentliche Finanzen der Schweiz, various years, Federal Statistical Office Switzerland. List of cantons in official order.
| Variable . | Hypothesis . | Operationalisation/Source* [Question**] . |
|---|---|---|
| Dependent variable | ||
| Informal volunteering | Dummy: 1=individual performs informal unpaid voluntary work (informal assistance, e.g., babysitting (children other than your own), neighbourly-aid, participating in any kind of projects, organising a (street) party in your neighbourhood); 0=individual does not volunteer informally [Q46]. | |
| Social informal volunteering | Dummy: 1=individual performs informal unpaid voluntary work related to personal helping or personal assistance; 0=no social informal volunteering [Q46, Q48]. | |
| Independent variables–Individual level | ||
| Level of education | The probability to volunteer is greater for well educated than for less educated individuals. | 0=no level of education completed; 1=low educational achievement (‘Sekundarstufe I’); 2=middle educational achievement (‘Sekundarstufe II’); 3=high educational achievement (‘Tertiärstufe’) [SD21]. |
| Professional status | The probability to volunteer is greater for individuals with a high professional status. | 0=no superior; 1=superior [SD17]. |
| Employment | Due to time constraints, part-time workers’ and non-active individuals’ probability to volunteer is greater than that of full-time workers. | 1=full-time; 2=part-time; 3=non-active [SD14]. |
| Children | Parent's probability to volunteer is higher than that of childless people. | 0=childless; 1=at least one child under 15 years old [SD11, SD12]. |
| Foreigner | Foreigners volunteer less frequently than native Swiss. | 0=native Swiss; 1=foreigner [SD05]. |
| Duration of residency | The longer the duration, the greater the probability to volunteer. | 0=max 3 years; 1=more than 3 years/very long [SD03]. |
| Religiosity | Religious people's probability to volunteer is greater than that of the non-religious. | 0=non-religious; 1=religious [SD22]. |
| Age | An inverse u-shaped relationship between age and volunteering is expected: Engagement first increases and then decreases as an individual ages. | Age (in years); Age squared (in years) [SD01]. |
| Sex | Women's probability for informal volunteering is higher than that of men. | 0=woman; 1=man |
| Independent variables –Contextual level | ||
| Welfare regime | 1) A developed welfare state crowds out informal volunteering.2) A developed welfare state creates conditions that support informal voluntary activity. | 1=liberal; 2=conservative/mixed; 3=social democratic. Source: Armingeon et al. (2004). |
| Cantonal social spending as % of GDP (average 2001–2004) | See (1) and (2) above. | Source: Swiss Federal Statistical Office, Eidgenössische Finanzverwaltung: Öffentliche Finanzen der Schweiz, various years. |
| Cantonal social spending per capita (average 2001–2004) | See (1) and (2) above. | Source: Swiss Federal Statistical Office, Eidgenössische Finanzverwaltung: Öffentliche Finanzen der Schweiz, various years. |
| Urbanisation | The probability to volunteer is lower in urban than it is in rural areas. | Proportion of inhabitants in urban regions. Source: BADAC, 2001. |
| Language region | Individuals from the German speaking part of Switzerland are more likely to do unpaid voluntary work than individuals living in the Latin part of Switzerland. | Proportion of German-speaking people in the respective canton. Source: Swiss Federal Statistical Office, ‘Regional portraits’ (key data), 2000. |
| Religious context | The higher the proportion of non-religious people, the lower the probability to volunteer. | Proportion of non-religious people. Source: Swiss Federal Statistical Office, ‘Regional portraits’, 2000. |
| Variable . | Hypothesis . | Operationalisation/Source* [Question**] . |
|---|---|---|
| Dependent variable | ||
| Informal volunteering | Dummy: 1=individual performs informal unpaid voluntary work (informal assistance, e.g., babysitting (children other than your own), neighbourly-aid, participating in any kind of projects, organising a (street) party in your neighbourhood); 0=individual does not volunteer informally [Q46]. | |
| Social informal volunteering | Dummy: 1=individual performs informal unpaid voluntary work related to personal helping or personal assistance; 0=no social informal volunteering [Q46, Q48]. | |
| Independent variables–Individual level | ||
| Level of education | The probability to volunteer is greater for well educated than for less educated individuals. | 0=no level of education completed; 1=low educational achievement (‘Sekundarstufe I’); 2=middle educational achievement (‘Sekundarstufe II’); 3=high educational achievement (‘Tertiärstufe’) [SD21]. |
| Professional status | The probability to volunteer is greater for individuals with a high professional status. | 0=no superior; 1=superior [SD17]. |
| Employment | Due to time constraints, part-time workers’ and non-active individuals’ probability to volunteer is greater than that of full-time workers. | 1=full-time; 2=part-time; 3=non-active [SD14]. |
| Children | Parent's probability to volunteer is higher than that of childless people. | 0=childless; 1=at least one child under 15 years old [SD11, SD12]. |
| Foreigner | Foreigners volunteer less frequently than native Swiss. | 0=native Swiss; 1=foreigner [SD05]. |
| Duration of residency | The longer the duration, the greater the probability to volunteer. | 0=max 3 years; 1=more than 3 years/very long [SD03]. |
| Religiosity | Religious people's probability to volunteer is greater than that of the non-religious. | 0=non-religious; 1=religious [SD22]. |
| Age | An inverse u-shaped relationship between age and volunteering is expected: Engagement first increases and then decreases as an individual ages. | Age (in years); Age squared (in years) [SD01]. |
| Sex | Women's probability for informal volunteering is higher than that of men. | 0=woman; 1=man |
| Independent variables –Contextual level | ||
| Welfare regime | 1) A developed welfare state crowds out informal volunteering.2) A developed welfare state creates conditions that support informal voluntary activity. | 1=liberal; 2=conservative/mixed; 3=social democratic. Source: Armingeon et al. (2004). |
| Cantonal social spending as % of GDP (average 2001–2004) | See (1) and (2) above. | Source: Swiss Federal Statistical Office, Eidgenössische Finanzverwaltung: Öffentliche Finanzen der Schweiz, various years. |
| Cantonal social spending per capita (average 2001–2004) | See (1) and (2) above. | Source: Swiss Federal Statistical Office, Eidgenössische Finanzverwaltung: Öffentliche Finanzen der Schweiz, various years. |
| Urbanisation | The probability to volunteer is lower in urban than it is in rural areas. | Proportion of inhabitants in urban regions. Source: BADAC, 2001. |
| Language region | Individuals from the German speaking part of Switzerland are more likely to do unpaid voluntary work than individuals living in the Latin part of Switzerland. | Proportion of German-speaking people in the respective canton. Source: Swiss Federal Statistical Office, ‘Regional portraits’ (key data), 2000. |
| Religious context | The higher the proportion of non-religious people, the lower the probability to volunteer. | Proportion of non-religious people. Source: Swiss Federal Statistical Office, ‘Regional portraits’, 2000. |
*All individual variables are taken from the Swiss Volunteering Survey 2007.
**Number of question according to the Swiss Volunteering Survey 2007.
References
Birte Gundelach is research fellow at the Cluster of Excellence ‘Cultural Foundations of Integration’ at the University of Konstanz, Germany. Central research interests include social capital, civic engagement, and the relationship between ethnic diversity, multiculturalism, and social cohesion.
Markus Freitag is professor of comparative politics at the Department of Politics and Management, University of Konstanz, Germany. He studied political science and economics at the Universities of Heidelberg and Bern (PhD and Habilitation) and was assistant professor at the Humboldt-University of Berlin. He has published on trust, social capital, direct democracy, comparative public policy and voter participation in the British Journal of Political Science, Comparative Political Studies, European Journal of Political Research, European Union Politics, and European Sociological Review.
Isabelle Stadelmann-Steffen is post-doctoral researcher at the Department of Politics and Management, University of Konstanz, Germany. She studied political science and economics at the Universities of Bern and Geneva and earned her PhD in Bern. She has published on comparative public policy, welfare state policy, civic engagement and direct democracy in journals such as European Union Politics, Governance, Journal of Social Policy, Non-Profit and Voluntary Sector Quarterly, and Political Studies.