Rohini Pande on the Economic Implications of Identity Politics

Interviewed by Molly Lanzarotta on November 16, 2006

A basic premise of representative democracy is that those affected by policy should have a voice in its making. However, policies enacted by elected governments often fail to reflect the interests of disadvantaged minorities.

The research of Rohini Pande focuses on the economic analysis of the politics and consequences of different forms of redistribution, principally in developing countries. Pande is Mohamed Kamal Professor of Public Policy.


Q: Many countries—India, for example—are amending their political systems to set aside positions to groups, such as women and racial or religious minorities, that are perceived as being disadvantaged. What are the implications of this trend?

Pande: The first implication that is very clear in the data is that putting in place such quotas or forms of affirmative action has increased the number from these groups who are represented in legislatures. The most substantive questions about how such representation has translated into either changes in policy or welfare, for either the disadvantaged groups themselves or other groups, is still being debated.

So, for India, which is the country I've studied the most, I think we have reasonable evidence that increasing representation of a lower caste—which is the disadvantaged social group—and women, certainly increases participation by these groups in the policy process in terms of the participation in village meetings or voter turnout. There is some evidence that there is increased targeting of resources towards the policy priorities of these groups. The question of how this affects overall efficiency or public distribution is something that I think is still open for debate.

Q: Have you found that gender bias is reduced or exacerbated by political affirmative action?

Pande: At this point we don't know whether it's exacerbated or reduced, but we certainly do know that political affirmative action for women, especially in low-income countries, has been in the context of a large amount of gender bias. So, in work I've done in India, we've found that while women have different policy priorities than men, as an objective measure of performance they don't perform worse and in fact there is evidence that they seem to be less corrupt. However, when you ask villagers for their perceptions of women, systematically both male and female villagers typically rate the performance of women policy makers as being much lower than male policy makers.

So, it's clear that these are women who are functioning in an environment of gender discrimination. Now whether political affirmative action is going to change this bias or will lead to what some psychologists call a backlash is still something we don't know. In ongoing research we are investigating the extent to which perceptions of women are altered when people are exposed to a woman leader.

Q: What has your research shown are the implications of ethnic politics and political affirmative action for political corruption?

Pande: What we mean by ethnic politics is political competition where parties define their base very explicitly for an ethnic group.

Ethnic politics is something that is increasingly important in low income countries and a number of researchers have talked about how politics in many newly democratizing countries has been along ethnic lines. In research I've done in Uttar Pradesh, India's most populous state, what we find is that ethnic politics have the effect of making voters decide to vote upon multiple dimensions at the same time. So they care about political corruption, but they also care about the ethnic identity of the politicians, and what this implies is that with increasing polarization of voter preferences along ethnic lines you find political corruption increasing, especially in areas where one ethnic group is dominant.

In terms of political affirmative actions, the evidence is more mixed. When we look at women as policy makers there's evidence that women appear to be less corrupt. Now this could be because they are newcomers to the system and they don't know how to game the system yet, or it could be because they're intrinsically less corrupt. So there's certainly some evidence that newcomers into the political system seem to be less corrupt.

In terms of political affirmative action for low castes, what we've found is that in some sense, the way that political affirmative action has worked in India is by restricting candidacy to only low castes. In this case, since caste competition is reduced-everyone has to choose a low caste-what this means is that it tends to make the quality dimension more important and you actually do find over time some political competition being beneficial and corruption goes down. So I'd say overall that ethnic politics per se seems to have been associated with increasing amounts of corruption, but actually focusing in on a single dimension of individual identity-be it gender or ethnicity-as a way of choosing politicians seems to surprisingly have beneficial effects for reducing corruption.

Q: How do you see your research findings applying to the U.S. and other democracies in the developed world?

Pande: I think this idea of how to ensure that a democratic system gives representation to disadvantaged groups has universal appeal.

In the United States the debate has mainly been around how to incorporate not just low income groups but specifically African Americans into the political arena. The U.S. has a long history of gerrymandering- until the early 1990s there was allowance for racial gerrymandering which now has been ended, so you can only have partisan gerrymandering. But there was explicit recognition of the fact that perhaps by creating majority minority districts you could actually increase the representation these groups get. In the U.S. there's been some disappointment with the belief that increasing gerrymandering didn't in the end increase African American presence in the legislatures in very large numbers, and I think it's something that's still debated.

Another country that's interesting is France. In France, as in a number of European countries, there's been a lot of emphasis on female affirmative action politics. So in France in 2001 for the first time they introduced a requirement that 50 per cent of the party lists have to be female candidates. Recent studies suggest that this has essentially been used to ensure that male incumbents stay in power. So they basically field women candidates in areas where either there was large gender bias in the population or male incumbents were very strong. So they fulfill their quotas, but male incumbents still get elected. So that's a case where simply putting women on party lists seems to have had very little effect and some people suggest that's partly because of an existing gender bias in the population, which I think will resonate with anyone working in a low-income setting.


Answers to questions submitted via e-mail:

Q: What do you think about the long term implications of caste and religion based quotas being enacted in India? In specific, do you think that it will be more divisive, rather than uniting, for the country in the long run? Would it lead to structural divides between the lower castes and the higher casts, and between the Muslims and the 'others' (as the government has now proposed religion based quotas for Muslims)? Can it lead to further partitions of the country in the time to come, like what happened to Yugoslavia, etc?

-Ashesh S.
Mumbai, India

Pande: I think there are two questions here-whether one should target resources according to religion/caste, and whether one should create religion or caste-based quotas in institutions and government.

In the first case, it might be argued that allocating resources to less economically developed minorities on the basis of identity creates a sense of injustice among the poorer members of the majority that can easily be exploited by those seeking power-and that improvements in their welfare could be more effectively and fairly achieved just by targeting poverty. However this fails to account for institutional bias against these groups-the recent Sachar committee report for instance provides suggestive evidence that Muslims are significantly less likely to get large loans.

In the second case, again, one would expect institutions to be mostly populated by, and biased in favor of the majority, and to recruit from among the majority. Quotas, though potentially unfair to some in the majority in the short term (it's impossible to know exactly the effect of discrimination on recruitment, or to balance it exactly), can reduce unfairness and discrimination in the long term by allowing some of the minority to positions of power, and reducing the bias of the institution. It is unfortunately true that minorities who are discriminated against have an unappealing choice of suffering covert discrimination in silence, or taking part in targeted programs that improve their lot, but may also provide a pretext for the majority to make the discrimination less covert. Perhaps the answer to this is to ban identity-based political parties, so politicians find it harder to make electoral gains from inciting religious and caste-based conflicts-but this would clearly not be easy in the India we have now.

Q: What are the similarities between the American caste system that was structured between white Americans and African American slaves and the Indian caste system that excludes the "untouchables?" Are the "untouchables" visibly different and detectable from other Indians?

How do the exclusions of women and untouchables compare? Which is more absolute? In Indian culture, how do degrees of difficulty compare for achieving inclusion and opportunity for women versus "untouchables?" For other groups? Are the Brahmins and other upper classes more accepting of women than untouchables? White women versus African American women in America?

How do you compare the acceptance of African Americans in executive elite positions in America with the acceptance of Indian immigrants in such positions? Are there different degrees of difficulty? Are Indians more or less accepted/resisted?

How do Indian immigrants view accepting African Americans versus white Americans as their interchangeable human equals?

-Edward J.
South Orange, NJ

Pande: These are very interesting questions, but unfortunately I have to restrict myself to a brief answer, and I think a comprehensive answer might extend to a thesis, or two!

That said, you may be interested in looking at the following:

For information on caste in India: The Untouchables: Subordination, Poverty and the State in Modern India (Oliver Mendelsohn, Marika Vicziany, Cambridge University Press, 2006).

For information on women in India: Women in Modern India (Geraldine Forbes, Cambridge University Press, 2006).

For information on the Indian immigrant experience in the US: Ethnic Routes to Becoming American: Indian Immigrants and the Cultures of Citizenship (Sharmila Rudrappa, Rutgers University Press, 2004).

Print print | Email email