Diversity, Integration and Cultural Evolution: Recent Theoretical Approaches and Policy Perspectives

What diversity implies for economic outcomes and social segregation ultimately depends on various aspects of social interaction, of the socio-economic system and of institutions. People may have, to different degrees, an innate preference to interact with others of similar backgrounds, ethnic origins, habits or culture. To what extent these preferences translate into segregation of social contacts, and possibly into discrimination in socio-economic interaction, depends on the opportunities that agents face in forming social contacts. The joint effect of agents’ attitudes towards others and of their opportunities to meet others determines how selectively and intensively they socialize. Both opportunities and, to a lesser extent, preferences can be affected by “affirmative action” type of policies. For instance, admission rules can determine the shares of ethnic groups within schools, thereby affecting the patterns of students’ interaction; advertising campaigns and educational programmes can favour the flourishing of more tolerant attitudes towards other cultures. 

Preferences evolve in time as a result of many factors, with parental educational choices playing a crucial role. Such choices are usually aimed at the transmission of values and cultural traits that are, at least in part, those of the parents. These values are themselves shaped by parents’ experiences in social interaction, and may differ from society’s dominant values. The larger this difference, the higher the incentive of parents to segregate their children from society, leading to the persistence of their values and acting as a barrier to the integration of values. 

The expectation of discrimination may provide people with incentives to change their very characteristics, in order to gain acceptance within dominant groups and access to better economic opportunities. This may be more often the case when the attitudes of dominant groups are more discriminatory; as a result, the evolution of attitudes and of characteristics may end up being interdependent, and act as parts of one larger co-evolution process.

Discriminatory behaviour in economic transactions is itself affected by the segregation patterns in society and by the institutions that govern the formation of social ties. Diversity and segregation are likely to lead to economic discrimination when intergroup contacts are somehow imposed on agents rather than voluntarily chosen, and the degree of segregation itself may end up not being clearly correlated with discrimination. 

Academic research can help understand the interplay of preferences, opportunities and discrimination, and their joint evolution, by constructing systematic analytical frameworks of analysis. Economics, in particular, focuses on the use of formal models based on decision theory and on cost-benefit analysis. Central to economic analysis are agents’ incentives to form relations, to make transactions and to react to institutional constraints. Specific to the economics framework is the possibility to draw welfare conclusions based on the costs and benefits of agents’ behaviour. Below I will illustrate some of the directions taken by recent research in economics, discuss some potential advances, and relate them to policy and to the management of diversity.

Homophily: Preferences or Opportunity Bias?

A pervasive feature of social and economic networks is that contacts tend to be more frequent among similar people than among dissimilar ones. This pattern, usually referred to as "homophily", applies to social ties of different intensity (from marriage to casual friendship), and along many dimensions of similarity (education, ethnicity, gender, religion,…). For an historical account of homophily, see the seminal work of Lazarsfeld and Merton (1954) and, more recently, Moody (2001), and the comprehensive survey by McPearson, Smith-Lovin and Cook (2001). The presence of homophily has important implications for how information flows along the social network (as shown in a recent paper by Golub and Jackson (2011)) and, more generally, on how agents' characteristics impinge on social behaviour. 

The patterns of social interaction within a society are primarily affected by the prevalence of groups’ populations. In a society where people are in large majority of white ethnicity, contacts with white people will tend to be prevalent in all people’s contacts, simply because white people are easier to meet (Blau, 1977). What is striking in many types of social networks is that the prevalence of one’s own ethnicity in one’s own contact goes well beyond the relative prevalence of the ethnicity in society. This excess of homophily can be due to a bias in favour of the own ethnic group either in people’s attitudes, or of people’s meeting opportunities, or of both.

In a recent paper, Currarini, Jackson and Pin (2009) study homophily patterns in US high school friendships, getting to the conclusion that both types of bias are responsible for the observed friendship networks. Students do have a preference for interaction with members of their same ethnic group, and meet these co-members at rates higher than what we would observe as the sole outcome of biased preferences. So, students take ethnicity into account when forming social ties, and have also preferential channels of encounters with their same ethnicity. In a companion paper Currarini, Jackson and Pin (2010) show that such biases are not uniform across ethnic groups, a fact that is reflected in the realized friendship segregation (see also Boucher, 2012). These results imply that effective policies should address both students’ cultural attitudes towards diversity and their meeting constraints at school. One particular aspect of the meeting technology that seems to play a role for integration (at least in the long run) is the “meeting friends through friends” mechanism, which seems to favour integration when people reduce their discriminatory attitudes when meeting friends of friends (see Bramouillé et al. 2012)).

Cultural Evolution and the Persistence of Diversity

When it comes to studying the evolution of preferences, many channels of transmission are likely to affect people’s attitudes, including parental educational choices, social interaction with peers, government affirmative action policies, and so on. Among these, the parental choice of which cultural traits to transmit to their offspring has been investigated in some detail by a series of recent papers in economics. The evolutionary approach to cultural transmission (see Cavalli Sforza and Feldman, 1981) predicts that society would reach a cultural “melting-pot” in the long run, in which a dominant cultural trait emerges as a mix of the pre-existing ones. Bisin and Verdier (2001 and subsequent papers) have enriched this model to study parents’ investment decisions of scarce resources (time and money) for the education of children in accordance with their own culture. The main insight here is that the incentives of parents to invest depend on the degree to which their own culture is neglected (or even opposed) by society. Parents with a minoritarian culture tend to segregate their children from society in the attempt to maintain their children’s values as close as possible to theirs. By doing so, parents finally preserve their trait from the cultural “melting-pot”, and favour the persistence of diversity. These incentives help explain why so much diversity is observed and persistent in the real world, where different people and cultures interact at an ever accelerating pace. Variation of the basic model show that a melting-pot equilibrium, reflecting the values of the dominant culture, may be reached when parents choose to educate their children with values that increase their socio-economic opportunities in the future.

The Melting Pot: Co-evolution of Traits and Preferences

When diversity affects personal characteristic that can, at least in part, be modified by the subject, cultural transmission and the evolution of preferences towards diversity go alongside with the evolution of characteristics. This is the case, for instance, when characteristics refer to habits on the workplace, religion, beliefs, school performances, etc. An intolerant attitude towards diversity provides minority members with strong incentives to be assimilated by the dominant group. This may prevent the persistence of minority cultures, and favour a fast homogenization of characteristics in a society, dominated by the most prevalent cultures (melting pot). Notably, this would occur with little or no decrease of intolerance in society. In contrast, societies with more tolerant attitudes provide minorities with incentives to preserve their own traits, and set conditions for an evolution of preferences towards less intolerant standards (multiculturalism). Which of these scenarios is more likely to occur crucially depends on the pace at which characteristics react to preferences and at which preferences evolve as a function of the distribution of characteristics in society. In the real world, we sometimes witness the polarization of characteristics and the contextual emergence of oppositional identities, and sometimes the complete integration of minorities in the form of their inclusion within the dominant culture. Research in this area is needed to better understand what features of a social system and of the process of coevolution of attitudes and cultural traits can lead to which of these scenarios.

Segregation, Discrimination and the Role for Policy

Segregation is particularly obnoxious when matched by the economic and social discrimination of others. While both phenomena seem to be different aspects of a generalized intolerance toward diversity, a systematic analysis of their interplay has come to interesting and unexpected conclusions (see Currarini and Mengel, 2012). Discrimination of different others has been long known in psychological and sociological disciplines as “in-group bias”, and has been shown to affect people behaviour even under the mildest definition of social groups. In the framework of laboratory experiments (see, for instance, Chen and Li, 2009), strong in-group biases emerge even when groups carry no particular meaning and are formed through a random labelling of participants. While these findings witness a fundamental inclination of people to form identities - and to favour those who share the same identity - even in response to minimal diversities, little is known about why and how such identities form in the first place. Would agents discriminate more or less in economic transactions if they were in control of their matchings? Recent results seem to suggest that the bias in favour of same group members tends to decrease when people have some control over the patterns of their social ties. In particular, while participants in a lab experiment have been observed to favour interaction with members of the same group, this has lead them to be less discriminatory towards the other group (compared to a framework where they have no control over their matches). This switch in behaviour as a reaction to a change in the matching institution (from exogenous to endogenous) has several interesting policy implications. First, and more importantly, segregation and discrimination do not seem to go together, as more intense in-group contacts (resulting from people’s voluntary choices of who to interact with) are associated with less discriminatory behaviour towards other groups. Bottom-up policies that aim at reducing discrimination by favouring social integration are more likely to be effective when based on incentives to interact across social groups rather than on rigid rules that constraint people’s patterns of interaction.          

References

• Bisin, A., Verdier, T. (2001) “The Economics of Cultural Transmission and the Dynamics of Preferences”. J. Econ. Theory 97, 298–319.
• Blau, P. M. (1977), Inequality and Heterogeneity: A Primitive Theory of Social Structure. New York: Free Press.
• Bramouillé, Y., Currarini, S., Jackson, M., Pin, P., and B. Rogers“ (2012) Homophily and Long Run Integration in Social Networks”  Journal of Economic Theory, 147(5).
• Boucher, V. (2012), “Structural Homophily”, mimeo.
• Cavalli Sforza, L.L., Feldman, M., (1981) Cultural Transmission and Evolution: A Quantitative Approach. Princeton University Press, Princeton, NJ.
• Chen, Y., and S. X. Li (2009) “Group Identity and Social Preferences," American Economic Review, 99(1), 431-457.
• Currarini, S., M.O. Jackson, and P. Pin (2009) “An Economic Model of Friendship: Homophily, Minorities and Segregation”, Econometrica 77 (4), 1003--1045.
• Currarini, S., M.O. Jackson, and P. Pin (2010) “Identifying the roles of race-based choice and chance in high school friendship network formation” Proceedings of the National Academy of Science of the U.S.A. (PNAS), 107, 4857-486
• Currarini, S. and F. Mengel (2012) “Homophily, Identity and In-Group Bias”, FEEM WP 37-2012
• Golub, B. and M. O. Jackson (2011): “How homophily affects the speed of learning and best-response dynamics"mimeo.
• Lazarsfeld, P.F. and R.K. Merton (1954): “Friendship as a social process: a substantive and methodological analysis”, in M. Berger (ed.), Freedom and Control in Modern Society, New York: Van Nostrand.
• McPherson, M., L. Smith-Lovin and J. M. Cook (2001): “Birds of a Feather: Homophily in Social Networks”, Annual Review Sociology 27, 415-44.
• Moody, J. (2001): “Race, School Integration and Friendship Segregation in America," American Journal of Sociology, 107(3), 679-716.