IZA Workshop - Report


"Cultural Economics: Markets, Developments and Institutions"
(Paris, June 30-July 1, 2006)

Differences in economic outcomes across countries and individuals cannot be fully understood by considering only economic determinants. Examples include substantial differences in female labor force participation, fertility and unemployment rates. The aim of this workshop was to understand whether cultural differences are a promising candidate to understand the remaining unexplained substantial differences in economic outcomes, market structures and institutions. The workshop participants presented and discussed research using different methods and approaches to gain insights for this important question.

The first session of the workshop tried to empirically assess the influence of culture and religion on individual behavior. Anne-Célia Disdier used 47-years of panel data on birth registries in France to find that foreign media exposure influences naming patterns, mainly by introducing unfamiliar names. Media and other information was also found to be important in the paper of Tito Boeri. Using a cross-section of Italian survey data in 2004 he found that lack of information reduces the support for reforms in pension systems. Hence, information campaigns that explain reforms to citizens can play an important role in the reform process.

Luigi Guiso analyzed the role of cultural beliefs for trade and investment patterns. He argued that biased priors influence the bilateral trust between trading partners in different countries. He first established information, culture and common origin of law or common language as determinants of relative trust (controlling for constant differences in trust across countries). Interestingly these determinants mattered less for educated people suggesting that the cultural bias in priors is more important for the less educated.

Most importantly, differences in bilateral trust across countries are important to understand differences in international trade and investment. Trust is positively associated with the amount of trade, where the effect is more important for differentiated goods than for goods traded in organized exchange. Trust also has a positive effect on portfolio investment and even more so for foreign direct investment. Andrew Clark argued that religion serves as an insurance device. Using data from the European Social survey in 2002/2003 and the British Household Panel, 1991-2002, he found that catholic religion is associated with more well-being in the case of adverse events like unemployment and partnership separation. This could be due to monetary transfers from the church or family, or psychological reasons.

The second session of the workshop analyzed more the theoretical determinants of social behavior. Andrew Postlewaite analyzed social learning. He set up a model in which people consider other’s choices when making their own consumption choices. He also solved for the firms’ reaction to consumer information transmission. Assuming markets with imperfect information, experience goods and wealthy and poor buyers, he argued that the wealthy consume more frequently, thus acquire more information so that other people pay more attention to them and follow their behavior. Since this implies that the wealthy “procure” more other customers, firms have an incentive to provide better service to these wealthy people.

Giacomo Corneo laid out a theory of symbolic values. His theory could rationalize that uncertainty about self-esteem in various occupations could induce parents to teach tolerance as an insurance for their children. He also argued that parents can have an incentive to teach their children values so that the children are proud of their parents’ education. This induces persistence of the value system.

Alberto Bisin analyzed the determinants of ethic assimilation in more or less segregated neighborhoods. He derived the prediction that there is more assimilation in more segregated neighborhoods. This is because parents care more about transmitting values in more mixed and thus less segregated neighborhoods. He provided support for this prediction using the national survey of ethic minorities in the UK 1993/1994.

The third session was about the determinants of female labor force participation and fertility. Alessandra Fogli started by noting that is important to understand the big amount of variation captured by country-fixed effects in typical regressions. She argued that past aggregate outcomes in the US-American women’s country of ancestry allow to capture the values and preferences transmitted from parents to their kids. She found that the aggregate female labor force participation in 1950 (in the country of ancestry) increases the individual labor force participation of US women, controlling for education, geography and husband characteristics. An increase of the aggregate labor force participation by 1 standard deviation is associated with an increase of 0.83 hours worked in 1970.

Nezih Guner investigated the evolution of premarital sexual intercourse in the US in the last century. Solving for the steady state in a model with endogenous formation of one abstinent and one promiscuous group, he showed how technological progress reduced the cost of contraceptives and thus increased premarital sex and promiscuity.

The fourth session was concerned with understanding differences in labor market outcomes and institutions across developed countries. Yann Algan argued that stronger social and civic attitudes induce higher levels of unionization and reduce the political support for minimum wages where minimum wages are relevant for non-unionized workers. He found support for that prediction using repeated cross-sectional data on social cooperation showing that higher levels of trust are indeed positively associated with a higher union density across countries. Thus certain labor market institutions are rather resulting from weak social attitudes and more a symptom than a cause of poor labor market performance.

Thomas Philippon argued that less cooperative workers imply more concentrated firm ownership. Less cooperation of workers increases the need to monitor workers where minority shareholders free-ride on the monitoring efforts of larger shareholders. This increases the incentive for concentrated ownership if less worker cooperation induces more monitoring. Cross-country evidence indeed suggests a robust negative relationship between the concentration of firm ownership and worker cooperation. This is further supported by time-series evidence for the state of Quebec for which strike activity and ownership concentration are highly positively correlated whereas this is not the case for the rest of Canada.

Olivier Blanchard tried to explain the cross-country and time-series variation of strike activity and unemployment. He started with the observation that strike activity and subsequent unemployment are positively correlated across countries but within countries a decrease in strike activity increases unemployment. He provided a simple model of bargaining under asymmetric information within a standard matching-flow model. The model could explain the cross-country evidence by an increase in turbulence: this implies more strikes and higher unemployment in countries with low trust and has no effect in countries with high trust. In order to fit the time-series evidence, however, an additional endogenous shift towards capital-intensive technology or towards less risky and less efficient production was needed.