close× Call Us
close×

Brief 07: Do Party Politics Undermine Ethnic Quotas?

This study uses a natural experiment and a survey experiment to examine ethnic quotas for India’s village councils.

Link to Full Study

Category: Elections, Public Service Provision

Date of Publication: Friday, February 1, 2013

EGAP Researcher: Thad Dunning

Other Authors: Janhavi Nilekani

PDF: PDF icon BRIEF-07-India-Dunning-Nilekani.pdf

Click to Download the Data

Geographical Region: Asia

Research Question:

Do ethnic quotas in village councils increase the likelihood that local politicians will represent the interests of disadvantaged groups? Can partisan politics outweigh the beneficial impact of quotas for disadvantaged ethnic groups?

Preparer: Jacob Kopas

English

Background:

In an ambitious effort to reverse historic discrimination, India’s 1993 constitutional amendment ordered village councils to periodically reserve seats for representatives of disadvantaged groups (aka scheduled castes and tribes). The assumption, supported by Chattopadhyay and Duflo’s (2004, p.1411) finding that electoral “reservation[s] affect policy choices”, was that these ethnic quotas would provide disadvantaged castes and tribes a voice in local politics, thereby incorporating their interests in village level policy-making. Research by Rohini Pande in 2003 also confirmed this finding.   Since village councils often control the distribution of state and federal welfare benefits and jobs, ethnic quotas would reduce the disparity between castes over a range of development indicators including poverty levels, literacy rates, and access to basic services.

The study employs evidence from surveys and reported distribution of welfare benefits in three states (Karnataka, Rajasthan, and Bihar) to test whether ethnic quotas promote the interests of India’s disadvantaged castes and tribes. While previous research has found some evidence that quotas increase the flow of benefits to these groups, few have brought to bear the power of randomized experiments to provide solid evidence of this effect.

In addition, the authors focus on how party politics may alter, and ultimately undermine, the beneficial impact of ethnic quotas. Although most Indian states ban displays of party affiliation in local elections, national political parties play a major role in funding campaigns for local politicians. 72% of the study’s survey respondents could readily identify the party of their village council president. Many parties also bring together multiple castes into a single coalition, which may further incentivize local politicians to favor party interests over those of fellow caste members.


Research Design:

This study employs two different research designs to test the impact of ethnic quotas. First, the researchers take advantage of a peculiarity in the quota system, which creates a so-called “natural experiment.” Villages take turns reserving seats for disadvantaged castes, and during each election cycle certain pairs of villages have a nearly random chance of being selected or not for quotas. While this “natural experiment” is not a true field experiment, the researchers present evidence that this technique comes close to an experimental randomization. The authors then surveyed villagers, council members, and local government workers to verify whether or not quotas increased access to welfare benefits and job programs. The study also uses reports of actual spending on welfare programs through a World Bank sponsored audit.

Next, the authors conducted a survey experiment to test whether party politics or ethnic identity are more influential in determining who gets welfare benefits. In this survey, the researchers read the campaign speech of a fictitious candidate for council president. In each speech, the caste and the party of the hypothetical politician were randomized. The authors then asked whether a villager would vote for the candidate and whether he or she would expect to receive a job or benefit if the candidate were to win. This allowed the researchers to measure the impact of sharing similar caste backgrounds or membership in the same political party on the villager’s response. Although this second experiment only examines villagers’ expectations for receiving benefits (and not actual benefits received), it does provide powerful evidence comparing the effects of political party and ethnic group identification.

Table 1: Estimated Average Effect of Village Council Quotas on Benefits and Welfare Programs

Survey Results (for disadv. castes) Estimated Effect Stand. Dev. Significant?
% Received job or benefit from council 1.57 (3.62) No
% Received job through MGNREGA scheme 3.39 (3.67) No
% Received benefit through any government scheme 3.28 (6.27) No
       
Outcomes in Welfare Program Expenditures (in Rupees) Estimated Effect Stand. Dev. Significant?
Ashraya Scheme -89,892.7 (-86,063.7) No
IAY Scheme -15,095.7 (87,357.1) No
Ambedkar Housing Scheme -31,681.4 (28,580.8) No
MGNREGA Scheme 39,305.7 (212,448.3) No
 
Results:

Despite high hopes from supporters of ethnic quotas, the study found little evidence that quotas had any effect on village council policies. On average, quotas for council president had no significant effect on the amount of benefits or jobs that villagers from disadvantaged castes reported receiving. Even when using data on actual expenditures, the study found no significant difference in welfare spending for programs that tend to target disadvantages castes. However, the researchers did find some evidence suggesting that sharing membership in the same political party as the council president could increase the likelihood of receiving a job or a benefit by as much as 30%.

The results of the survey experiment shed more light on the possible impact of political parties on the distribution of jobs and benefits. Despite restrictions on party mobilization, villagers were more likely to vote for the fictitious candidate if they shared party membership than if they shared caste backgrounds. Villagers were also just as likely to expect more benefits from a candidate when they shared party backgrounds as compared to ethnic backgrounds. In addition, when villagers were of different castes than the candidate, then the candidate’s political party had a much larger effect in raising expectations for receiving benefits.

Policy Implications:
  • This study shows that ethnic group quotas are not a panacea for improving representation of disadvantaged groups. Additional factors such as party politics may still affect how local bodies distribute state benefits and implement policies.
  • Political parties, particularly in areas where parties bring together multiple ethnic groups, may provide more effective means of political mobilization, and therefore may be more influential in local politics than ethnic quotas. Policy makers should keep in mind how political parties may change the impact of quotas when evaluating such policies.
  • However, it is important to note that this study does not conclude that quotas have no effect whatsoever on local politics. Quotas may still have important social or political impacts that are not directly related to the distribution of state benefits.
  • It is also important to note that while this study does imply that India’s ethnic quotas are likely not effective in reducing inequalities for disadvantaged groups, it was not able to measure the overall impact of the quota policy on India. Rather, it focused on the specific effect of village quotas during a particular election cycle, and therefore does not measure the broader impacts of the quota policy as a whole.

References

Duflo and Chattopadhyay. 2004. Women As Policy Makers: Evidence From A Randomized Policy Experiment In India. Econometrica, September. 72(5) 1409–1443. Available at: http://poverty-action.org/sites/default/files/women%20policymakers.pdf

Pande. 2004. Can Mandated Political Representation Increase PolicyInfluence for Disadvantaged Minorities? Theory and Evidence from India. BREAD Working Paper No. 024, April. Available at: http://www.hks.harvard.edu/fs/rpande/papers/bread_02412.pdf