![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
The role of a person's group identity and sense of integration into society as a determinant of the person's productivity and capability has been vastly underestimated in the literature. We talk of policies to subsidize the poor and give direct support to alleviate poverty. These are important but, in the long run, it is critical that we instill in people a sense of belonging and having certain basic rights as citizens. This paper tries to advance this perspective by building a new model where a person's community identity matters, ex post, in determining whether he or she will be poor, even though all persons are identical ex ante. The paper also draws on data collected from a non-governmental organization-run school in Kolkata to illustrate the role of a school child's sense of ‘belonging’ in determining how the child performs academically. The theory and the empirical work are inputs into the larger and more general idea that when people feel marginalized in a society they tend to ‘give up’. A substantial part of the paper is devoted to the policy implications of these analytical ideas and empirical results in the context of policy-making.
Acknowledgements
This paper is based on my keynote address to the HDCA International Conference in Jakarta on 6 September 2012. The paper had been work in progress for quite some time before the address and continued to be written and revised after it. As a consequence, I have accumulated a large amount of debt to various scholars who have commented on it. I am grateful, in particular, to Talia Bar, David Crocker, Supriyo De, Rohini Kohli, Hyejin Ku, Ortrud Lessman, Eleonora Mavroeidi, Priya Mukherjee, Siddiq Osmani, Nupur Ray, Ranjan Ray, Frances Stewart, Ashutosh Varshney, Maria Monica Wihardja and, especially, Sudarno Sumarto who, as a discussant for my lecture, commented extensively on it.
Notes
These figures are from the Labour Force Survey 2003, conducted by Statistics South Africa.
Although I pose this paper as a critique of methodological individualism, some may consider a broader definition of methodological individualism, which allows group identity to matter in the description of what a person can do and the well-being a person has. For instance, in Amartya Sen's work on famines, the so-called entitlement mapping takes us from an individual's endowment to what the person is entitled to (see Osmani, Citation1995). I would argue that this mapping must also allow for the group identity in determining the person's entitlement. If someone's definition of methodological individualism allows for such group influences, with such a person my difference will be simply a semantic one.
Which castes are the favored ones can vary from region to region. Also, while caste hierarchies are fairly stable, it is not as if they do not change. Social processes, like sanskritization and the garnering of political power, can cause caste groups to gain or lose over time.
I am using the more bombastic term ‘supermodularity’ instead of the (in this context) equally good term ‘convexity’, to clarify that I could have worked with a more general model where each contract may have a different effect on output. Such a model would use a production function, g, as follows: y i = g(e i , x 1, … x n ), where x j is an indicator variable, which takes a value of one if the jth contract is signed and zero if it is not signed. The general assumption I want to use says that if, for some j, x j is changed from zero to one, the increase in output that occurs with this is greater if the value of (x 1 + … + x j − 1 + x j + 1 + …+ x n ) is higher, with e i being held constant.
A similar set of experiments recently conducted by Field and Nolen Citation(2005) with South African children—Blacks, Whites and Coloreds—finds similar results, especially with boys. Of course, race, unlike caste, is visible. So an announcement of race is not as revelatory as the announcement of caste. What Field and Nolen do, therefore, is to consider situations where no mention is made of race and situations where the atmosphere is ‘charged’ by giving questionnaires on race.
Between these two, education seems to be overwhelmingly the more important cause (Glick and Sahn, Citation2006).
The full test question is reproduced in the Appendix to Chapter 2 in Basu Citation(2010).
The connection between a person's status and poverty, on the one hand, and his or her cognitive abilities and intelligence, on the other, is being increasingly observed in empirical studies. These are discussed persuasively in Esther Duflo's Tanner Lectures earlier this year (Duflo, Citation2012).
It remains a bit of a puzzle why this does not happen for children who live with their guardians, instead of the parents. It is possible that when asked whether their parents talk to each other, since their parents do not live with them, they gave erratic answers to the question.
For a discussion, see Bhargava Citation(1993), Arrow Citation(1994) and Basu Citation(2008).
I am personally in favor of using affirmative action, including job quotas, even when there may be no efficiency gain to be made, purely to build up role-models within communities that have faced discrimination over long periods of time. (This has to be used in limited measures keeping in mind that this may involve efficiency tradeoffs.) This is a purely normative stance of mine that does not have anything to do with the argument developed in this paper. What is different about the argument being presented here is that it points to why there may be a case for affirmative action even if one had no normative commitment to correct intertemporal inequities.
