Skip to main content


Log in

Culture on the Rise: How and Why Cultural Membership Promotes Democratic Politics

  • Published:
International Journal of Politics, Culture, and Society Aims and scope Submit manuscript


Selectively using Tocqueville, many social scientists suggest that civic participation increases democracy. We go beyond this neo-Tocquevillian model in three ways. First, to capture broader political and economic transformations, we consider different types of participation; results change if we analyze separate participation arenas. Some are declining, but a dramatic finding is the rise of arts and culture. Second, to assess impacts of participation, we study more dimensions of democratic politics, including distinct norms of citizenship and their associated political repertoires. Third, by analyzing global International Social Survey Programme and World Values Survey data, we identify dramatic subcultural differences: the Tocquevillian model is positive, negative, or zero in different subcultures and contexts that we explicate.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others


  1. Data from World Values Survey of national samples of citizens in each country. Question: A066. “Please look carefully at the following of voluntary organizations and activities and say…which if any do you belong to? Education, Arts, Music or Cultural Activities.” In Canada, a study on citizens' preferences regarding federal spending points in the same direction, by finding that one of the few items that show significant change between 1994 and 2010 is support for “arts and culture,” which climbed from 15 to 30 %. See

  2. Most studies on the arts are case studies whose authors have not sought to explore the broader and the political implications of arts participation: see, e.g. Lloyd (2006). Since finding these results in the World Values Survey, we have looked for other studies that might have reported similar results and found them inconsistent. An exception is Stolle and Rochon (1998), even though it is limited to three case-studies.

  3. The main source of the data used in our statistical analysis is the World Values Survey (1999–2004 wave). Since our main variable of interest—belong to education, arts, music, or cultural activities—is not included in the fifth wave of the WVS (2005–2007), we do not use data from that wave. A slightly different model was implemented on vote (on the social participation bloc, we only had information for religious and cultural groups): in this case, we used the International Social Survey Programme 2004 data. We also run the same model using as moderator variables: political cultures, i.e., new political culture, class/party politics, and clientelism and cultural traditions. We are using secondary data sources of already existent data. This implies that we are not only limited to a specific period of time, geographical scope, but we are also limited in terms of substantive information provided by each item surveyed.

  4. We thus restrict our analysis to democratic countries. Our list of 42 democratic countries is based on the Polity Score. Details of the indicators that constitute the index and the criteria for the classification of countries, according to the information, are available at

  5. “Representative democracy” is an index composed of the following variables: “voted in last election” and “political action: attend political meetings or rally” (source: International Social Survey Programme 2004).

  6. “Social trust” is an index composed by the variables: most people can be trusted; do you think people try to take advantage of you ((1) “can't be too careful,” (2) “most people can be trusted”). Trust in political institutions corresponds to the variable confidence in the government (1 “none at all” to 4 “a great deal”) (source: World Values Survey 1999–2004; see, e.g., Rothstein and Stolle 2008).

  7. “Protest” is an index composed by the following variables: political action—sign petition; joining boycotts; attending lawful demonstrations; joining unofficial strikes; and occupying buildings and factories. They have three-point scale: 1, “would never do”; 2, “might do”; and 3, “have done” (source: World Values Survey 1999–2004).

  8. The “duty-based” norm is an index composed by the following WVS variables: give authorities information to help justice; future changes: greater respect for authority; national goals: maintaining order in nation; and also by ISSP 2004 variables: good citizen: always vote in elections, never try to evade taxes, always obey laws, and serve in the military.

  9. The “engagement” norm is an index composed by the following WVS variables: politics important in life; reasons to help: in the interest of society; discuss political matters with friends and also by ISSP 2004 variables: good citizen: keep watch in government; active in associations; understand other opinions; choose products with ethical concerns; and help less privileged in the country/in the world.

  10. The “solidarity” norm is an index composed by the following WVS variables: importance of eliminating big income inequalities; reasons for voluntary work: solidarity with poor and disadvantaged; and ISSP 2004 variables: rights in democracy: government respect minorities; access to adequate standard of living; and tolerance of disagreement.

  11. In the case of the ethnic/civic norm axis (identity and civic norms), we only have information in the World Values Survey in one variable. In the absence of other options, we maintain it in our analysis in these circumstances. In the ISSP 2004, there was no information available on this normative dimension. The WVS variable is: how proud of nationality (civic norm: not very/not at all proud).

  12. Both NPC and CP are statistical indexes composed by World Values Survey items (fourth wave, described below). The different political cultures are multidimensional phenomena so a single indicator cannot measure them adequately. The means of the NPC and CP indexes were calculated across all respondents. In the analysis, the filtering criterion was inclusion of the observations that scored above the average value. The results from the regression estimates were then compared to each dominant political culture. For clientelism, due to the lack of available survey data, the measure was the index provided by Worldwide Governance Indicators (in this case, all respondents received the corresponding national figure). We also analyze potential effects of cultural or civilizational traditions on political behavior and beliefs (a relation extensively studied by S. Eisenstadt). These traditions correspond to the historically dominant religious culture associated with each country (Protestantism, Catholicism, Orthodox Christianity, and Eastern religions). Each country in our sample was classified by its dominant cultural tradition and, in each regression analysis, the countries that did not belong to the tradition were filtered out.

  13. In a recent article, Amy C. Alexander and Christian Welzel similarly point out that “empowering socioeconomic conditions” are conducive to make “people capable of practicing democracy”: see Alexander and Welzel (2011).

  14. Class/party politics is an index composed by the WVS variables: work orientations: compared with leisure; materialism orientations; and society aimed: extensive welfare vs. low taxes.

  15. Clientelism is measured by Worldwide Governance Indicators, which include control of corruption; rule of law; regulatory quality; government effectiveness; political stability and absence of violence; and voice and accountability. More information is available here:,,menuPK:1740542~pagePK:64168427~piPK:64168435~theSitePK:1740530,00.html.

  16. New political culture is an index composed by the WVS variables: being with people with different ideas; choose products with environmental concerns; and post-materialism four-item scale (maintain order; greater democracy; curb inflation; greater freedom of speech), in which items 1 and 3 express a materialist orientation, whereas 2 and 4 indicate post-materialist values.

  17. In the case of cultural traditions (Protestant, Catholic, Orthodox Christianity, and Eastern religions), each respondent is linked to each one of these four types depending on the dominant cultural tradition in his/her country.

  18. In the case of WVS, the variables accounting for “voluntary organizations membership” are as follows: belong to community, religious, arts, and professional voluntary organizations: 1 “not mentioned,” 2 “belong”; membership in religious and cultural voluntary organizations in the case of ISSP 2004. We compare cultural organizations with community, religious, and professional associations because these were the main types of social organizations chosen by Putnam to illustrate his claims in Bowling Alone (see Putnam 2000, pp. 48–92).

  19. To account for the hierarchical nature of our data, we employed multilevel regression analysis using, for the national level, United Nations Development Programme data for the Gini index, education gross enrollment ratio, cultural trade as percentage of the GDP (see, and WVS data (membership in voluntary organizations and socio-demographic variables as controls) for the individual level. The statistical tests applied to the model used WVS data (1999–2004 wave). A slightly different model was implemented on voting as ISSP only has information regarding religious and arts groups (not on community and professional voluntary associations). Except for this difference, the same model was applied to voting. As Fig.  2 indicates, we included socioeconomic variables as direct effects and as moderator variables (one at a time in seven separately estimated models), three political cultures (“new political culture,” class/party politics, clientelism) and four cultural traditions (Protestantism, Catholicism, Orthodox Christianity and Eastern religions).

  20. One other concern relates to the fact that in each country, the levels of culture membership are a relatively small part of the national sample. Our solution was not to conduct single country analyses, but to group countries by cultural type, which raised the Ns.

  21. This is a particularly sensitive issue as we focus on a broad population of cases - democratic countries - in diverse political and/or cultural contexts. See, e.g., King et al. (2004).

  22. We are aware of the potential reciprocity bias that exists in the general model we are testing—our general model analyzes if social participation impacts democratic politics, but the inverted relation might also hold. We tested this possibility and, in fact, the levels of political participation, trust, and the adherence to norms of citizenship predict membership in voluntary organizations. The variation explained by these models—measured by the adjusted R 2—is inferior to our model of interest.

  23. In addition, in order to test for the possible contamination of the measurement of arts participation by educational organizations we have run OLS regressions using the Citizenship, Democracy and Involvement Survey and World Values Survey data. We compared the results of our predictors of interest - only for the US (the CID questionnaire was only applied in the US): ‘member: cultural or hobby activities organization’ (CID, 2005) and ‘belong to education, arts, music or cultural activities’ (WVS, 1999-2004). The findings show that the results for the model with the item without reference to education (CID) do not diverge significantly from derived from the WVS: both are positive and significant predicting, in the case, protest.

  24. In Table 4, a clear pattern emerges from the results: cultural membership is a significant predictor of political participation, attitudes, and norms of citizenship in most contexts analyzed. Note also the coefficients for arts participation predicting Protest in Table  1 (new political culture context) and Table  2 (class politics context). In Table  3 , we present the results for the representative democracy component. Cultural membership is the only indicator that shows significant regression coefficients throughout the contexts.

  25. We also implemented a test of between-subjects effects to see if each context had a different effect on each of our dependent variables. Result: they indicated a significant context effect (for example, impact of different political cultures on protest: F (2, 52,979) = 31.86, p < 0.001).

  26. Except in the case of clientelism.

  27. Clientelism is a context constructed not from information obtained by means of cross-national surveys (which did not include “clientelist” items), as in the case of the other contexts, but from national information gathered by the World Bank.

  28. Standardized regression coefficients (arts membership in clientelist context predicting solidarity norm of citizenship): β = 0.098 (p < 0.001).

  29. Standardized regression coefficients (arts membership in clientelist context predicting protest, engaged norm of citizenship): β = −0.029 (p < 0.001), β = −0.037 (p < 0.001).

  30. NPC contexts show significantly higher impacts of cultural membership than of the other types of apolitical associations on protest activities. This holds true for the three political cultures, NPC, class politics, and clientelist, where cultural membership is a better predictor of protest than any other type of organizational participation (religious, community, or professional).

  31. Using religion as a prime historical indicator of traditional basic values and culture is classic in social science, from Max Weber's works on sociology of religion (1958, 1964) to Talcott Parsons (1951), to Henri Mendras (1971), and even Daniel Bell (1973).

  32. Due to data availability, this last context is analyzed only for three countries (India, Japan, and South Korea). Future waves of the World Value Survey and similar cross-national surveys should try to enlarge the number of countries from this part of the world.

  33. “More educated” refers here to the positive and statistically significant regression coefficient correspondent to the direct effect of the variable predicting the dependent variables (concerning the highest level of education attained by the respondent).

  34. The complete list of major Catholic countries included in our analysis is Argentina, Austria, Belgium, Canada, Chile, Croatia, Czech Republic, El Salvador, France, Hungary, Ireland, Luxembourg, Malta, Mexico, Peru, Poland, Portugal, Slovakia, Slovenia, Spain, and Uruguay.

  35. Standardized regression coefficients (cultural membership predicting voting): β = −0.028 (p < 0.001).

  36. Standardized regression coefficients (cultural membership predicting protest): β = 0.057 (p < 0.001).

  37. Standardized regression coefficients (cultural membership predicting duty and engaged norm): β = −0.047 (p < 0.001) and 0.025 (p < 0.001).

  38. “Cultural determinants” refer to structural features of each context, in the sense of Raymond Boudon's “operative” definition of structure (see Boudon 1971).

  39. The complete list of major Protestant countries in our analysis is Australia, Estonia, Finland, Germany, Iceland, the Netherlands, New Zealand, Norway, Switzerland, Great Britain, and the USA.

  40. Still, it has to be said that the sources of political trust are conservative and hierarchical, i.e., low income, right-wing self-positioning, and belonging to a religious organization most increase trust.

  41. Standardized regression coefficients (cultural membership predicting protest activities, social trust, and engagement norm of citizenship): β = 0.029 (p < 0.001), β = 0.056 (p < 0.001), and β = 0.023 (p < 0.001).

  42. Standardized regression coefficients (cultural membership predicting duty and solidarity norms of citizenship): β = −0.057 (p < 0.001), β = −0.056 (p < 0.001).

  43. The importance of arts and culture in contemporary urban policy is discussed in the very recent work of Grodach and Silver (2012, p. 13).


  • Achterberg, P. (2006). Considering cultural conflict. Class politics and cultural politics in Western societies. Maastricht: Shaker.

    Google Scholar 

  • Alexander, A., & Welzel, C. (2011). Measuring effective democracy: the human empowerment approach. Journal of Comparative Politics, 43, 271–289.

    Article  Google Scholar 

  • Arendt, H. (1982). Lectures on Kant's political philosophy. Chicago: The University of Chicago Press.

    Google Scholar 

  • Belfiore, E., & Bennett, O. (2008). The social impact of the arts. An intellectual history. London: Palgrave-Macmillan.

    Book  Google Scholar 

  • Bell, D. (1973). The coming of post-industrial society. New York: Basic Books.

  • Boshcken, H. (2003). Global cities, systemic power, and upper-middle-class influence. Urban Affairs Review, 38, 808–830.

    Article  Google Scholar 

  • Boudon, R. (1971). The uses of structuralism. London: Heinemann.

    Google Scholar 

  • Bourdieu, P. ([1979] 1984). Distinction. Cambridge: Harvard University Press.

  • Chan, T. W. (2013). Understanding cultural omnivores: social and political attitudes. Available in:

  • Clark, T., Hoffman-Martinot, V. (1998). The new political culture. Boulder, CO: Westview Press.

  • Clark, T. L., Seymour, M. (2001). The breakdown of class politics. A debate on post-industrial stratification. Baltimore, MA: John Hopkins University Press.

  • Coleman, J. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95–120.

    Article  Google Scholar 

  • Dalton, R. (2007). Democratic challenges, democratic choices: the erosion of political support in advanced industrial democracies. Oxford: Oxford University Press.

    Google Scholar 

  • de Tocqueville, A. (1945). Democracy in America (p. 205). New York: Vintage Books.

    Google Scholar 

  • Denters, B., Gabriel, O., & Torcal, M. (2007). Norms of good citizenship. In J. van Deth, J. Montero, & A. Westholm (Eds.), Citizenship and involvement in European democracies. A comparative analysis (pp. 88–108). London: Routledge.

    Google Scholar 

  • Dewey, J. ([1937] 2008). Democracy and educational administration. In J. A. Boydston (Ed.), The later works of John Dewey 1925–1953. Volume 11 1935-1937 (Essays, reviews, Trotsky inquiry, miscellany, and liberalism and social action, Vol. 11, pp. 207–225). Carbondale: Southern Illinois University.

  • DiMaggio, P. (1996). Are art-museum visitors different from other people? The relationship between attendance and social and political attitudes in the United States. Poetics, 24, 161–180.

    Article  Google Scholar 

  • Eisenstadt, S. (2008). The great revolutions and the civilizations of modernity. Leiden: Brill.

    Google Scholar 

  • Evans, G. (1999). The end of class politics? Class voting in comparative context. Oxford: Oxford University Press.

    Book  Google Scholar 

  • Fleury, L. (2013). Sociology of culture and cultural practices. Lanham: Lexington Books.

    Google Scholar 

  • Grodach, C, Silver, D. (2013) The politics of urban cultural policy. Global perspectives. Abingdon: Routledge.

  • Habermas, J. (1994). Struggles for recognition in the democratic constitutional state. In A. Gutmann (Ed.), Multiculturalism (pp. 107–148). Princeton, NJ: Princeton University Press.

  • Habermas, J. (2001). Constitutional democracy? A paradoxical union of contradictory principles? Political Theory 29, 766–781.

    Google Scholar 

  • Inglehart, R. (1977). The silent revolution: changing values and political styles among Western publics. Princeton: Princeton University Press.

    Google Scholar 

  • Jackman, S. (2008). General lessons on measurement. In B.-S. Jennifer, B. Henry, & C. David (Eds.), The Oxford handbook of political methodology. Oxford: Oxford University Press.

    Google Scholar 

  • King, G., Murray, C., Salomon, J., & Tandon, A. (2004). Enhancing the validity and cross-cultural comparability of measurement in survey research. American Political Science Review, 98, 191–207.

    Article  Google Scholar 

  • Lewin-Epstein, N., & Levanon, A. (2005). National identity and xenophobia in an ethnically divided society. International Journal on Multicultural Societies 7, 90–118.

    Google Scholar 

  • Lloyd, R. (2006) Neo-Bohemia: Art and Commerce in the Postindustrial City. New York: Routledge.

  • Matarasso, F. (1997). Use or ornament? On the social impact of participation in the arts. Stroud: Comedia.

    Google Scholar 

  • Matarasso, F. (2010) Full, free and equal: on the social impact of participation in the arts. In

  • Mendras, H. (1971). The vanishing peasant. Innovation and change in French agriculture. Cambridge: MIT Press.

    Google Scholar 

  • Newton, K. (2001). Trust, social capital, civil society and democracy. International Political Science Review, 22, 201–214.

    Article  Google Scholar 

  • Norris, P., & Inglehart, R. (2004). Sacred and secular. Religion and politics worldwide (p. 44). Cambridge: Cambridge University Press.

    Book  Google Scholar 

  • Novak-Leonard, J., & Brown, A. (2011). Beyond attendance: a multi-modal understanding of arts participation. Washington DC: National Endowment for the Arts.

    Google Scholar 

  • Parsons, T. (1951). The social system. New York: Free Press.

  • Passeron, J.-C., Mayol, P., & Macé, É. (2003). Culture(s): entre fragmentation et recompositions. Paris: CNDP.

    Google Scholar 

  • Peterson, R., & Simkus, A. (1992). How musical tastes mark occupational status groups. In M. Lamont & M. Fournier (Eds.), Cultivating differences: symbolic boundaries and the making of inequality (pp. 152–186). Chicago: University of Chicago Press.

    Google Scholar 

  • Pettit, P. (2000). Republicanism. A theory of freedom and government (p. 271). Oxford: Oxford University Press.

    Google Scholar 

  • Putnam, R. (1995). Bowling alone: America’s declining social capital. Journal of Democracy 6, 65–78.

    Google Scholar 

  • Rancière, J. (2004). Disagreement: politics and philosophy. Minneapolis: University of Minnesota Press.

    Google Scholar 

  • Rosenblum, N. (1998). Membership and morals. The personal uses of pluralism in America. Princeton, NJ: Princeton University Press.

  • Rothstein, B., & Stolle, D. (2008). The state and social capital: an institutional theory of generalized trust. Comparative Politics, 40, 441–459.

    Article  Google Scholar 

  • Savage, M., et al. (2009). Culture, class, distinction. Abingdon: Routledge.

    Google Scholar 

  • Stolle, D., & Rochon, T. (1998). Are all associations alike? Member diversity, associational type, and the creation of social capital. American Behavioral Scientist, 42, 47–65.

    Article  Google Scholar 

  • Tilly, C. (2006). Regimes and repertoires (pp. 30–59). Chicago: Chicago University Press.

    Book  Google Scholar 

  • Verba, S., Nie, N., & Kim, J.-o. (1978). Participation and political equality: a seven-nation comparison. New York: Cambridge University Press.

    Google Scholar 

  • Verba, S., Schlozman, K., Brady, H., & Nie, N. (1993). Citizen activity: who participates? What do they say? American Political Science Review, 87(2), 303–318.

    Article  Google Scholar 

  • Welzel, C., Inglehart, R., & Deutsch, F. (2005). Social capital, voluntary associations and collective action: which aspects of social capital have the greatest ‘civic’ payoff? Journal of Civil Society, 1, 121–146.

    Article  Google Scholar 

  • Wilson, J. (2000). Volunteering. Annual Review of Sociology, 26, 215–240.

    Article  Google Scholar 

  • Zukin, C., Keeter, S., Andolina, M., Jenkins, K., & Delli Carpini, M. (2006). A new civic engagement? Political participation, civic life, and the changing American citizen. Oxford: Oxford University Press.

    Google Scholar 

Download references


Thanks for advice and data analysis assistance to Rita Costa, Cátia Nunes and Jonah Kushner, and Scenes Project participants at the University of Chicago. Direct correspondence to Terry Clark, University of Chicago, 1126 E. 59th St. #322, Chicago, IL 60637, USA; Filipe Carreira da Silva, ICS-UL, Av. Aníbal de Bettencourt 9, 1600-189 Lisboa, Portugal.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Filipe Carreira da Silva.



Fig. 1
figure 1

Rising membership of cultural activity groups. Note: Percentage of individuals in each country sample who belong to education, arts, music or cultural activities organizations in each WVS wave and total variation (1981–1984; 1989–1993; 1994–1999)

Fig. 2
figure 2

Path diagram of core model. Note: The paths show direct effects from the two boxes of variables on the left hand side. The bottom boxes with their paths leading to circled arrows indicate interaction or mediated effects, which shift the paths from the left hand side. How and why these mediated effects differ is the main focus of subsequent sections of the paper

Fig. 3
figure 3

Membership in voluntary organizations, 1981–2004. Note: Comparison of levels of membership in voluntary organizations (question: “Please look carefully at the following list of voluntary organizations and activities and say … which if any do you belong to?”). WVS data (1981–1984 and 1999–2004 wave)

Fig. 4
figure 4

Arts participation predicting democratic politics, by context (multilevel regression analysis)

Table 1 Multilevel regression analysis: predictors of democratic politics (new political culture context)
Table 2 Multilevel regression analysis: predictors of democratic politics (class politics context)
Table 3 Social participation predicting representative democracy component (OLS regression)
Table 4 Impact of cultural membership on democratic politics (OLS regression)

Rights and permissions

Reprints and permissions

About this article

Cite this article

da Silva, F.C., Clark, T.N. & Cabaço, S. Culture on the Rise: How and Why Cultural Membership Promotes Democratic Politics. Int J Polit Cult Soc 27, 343–366 (2014).

Download citation

  • Published:

  • Issue Date:

  • DOI: