The article is a statistical regression analysis of the relative vote-shares of different social groups in reserved and non-reserved constituencies.

Affirmative action comprises an important element of public policies in enhancing welfares of disadvantaged populations. In India, affirmative action in the form of reserving seats in electoral constituencies has been in effect since 1950. The objective of this reservation policy is to guarantee the political representation of specific groups at all levels. This article shows empirically that, as far as voters' turnout rates in Lok Sabha elections are concerned, the system of constituency reservation does indeed help to serve this objective.

Political reservation in India initially began for historically disadvantaged groups designated in the constitution as scheduled castes (SCs) and scheduled tribes (STs) and has been recently extended to women and other sections in local elections (such reservation is beyond the scope of this article). In reserved constituencies, as is well known, only candidates belonging to reserved groups can stand for elections. How does this electoral system affect the voting behaviour of disadvantaged groups and other groups? In constituencies reserved for SC candidates (SC constituencies for short), are SC voters encouraged to vote by the system? Does reservation “discourage” non-SC voters, who quietly boycott the election? Since non-SC voters are usually the majority in SC constituencies, SC candidates need to appeal to non-SC voters to win elections so that non-SC voters may have an incentive to vote in SC constituencies. In order to throw light on such questions, we undertook a statistical regression analysis of the relative vote-shares of different social groups in reserved and non-reserved constituencies. We used micro-data on voters collected as part of the National Election Study 2004 (NES04).

NES04 was conducted by the Centre for the Study of Developing Societies (CSDS). It offers the largest and most comprehensive election database of its kind in India. We used NES04 micro-data on 27,000 voters in 420 randomly-selected parliament constituencies. The sample voters were interviewed after the 2004 Lok Sabha election on their voting behaviour, political opinion, and background.

In the main empirical analysis, a regression model was estimated with each voter's voting (dummy variable) as the dependent variable, separately for each group of voters (either SC, ST, OBC, Other Hindu or Others). The explanatory variables include the status dummies of his/her constituency with respect to reservation (either SC, ST or general constituency) and various socio-economic characteristics of the constituency. The regression coefficient on SC or ST constituency dummies is our estimate for the impact of political reservation on voter turnout.

The results are summarised in the Table (see above). With regard to ST constituencies, we were not able to obtain meaningful results, possibly due to the small number of ST constituencies relative to SC constituencies.

The findings

Three interesting findings emerge from the Table.

First, the turnout of SC voters in SC constituencies is significantly higher in statistical terms than the turnout of SC voters in general constituencies. As the Table shows, the turnout of SC voters is 4.5 percentage points higher in SC constituencies than their turnout in general constituencies. The difference is not only statistically significant but also politically significant — 4.5 percentage points compared with the national turnout rate of 58.1 per cent in the Lok Sabha election of 2004. To put the point differently, SC voters are thus significantly more likely to vote when they reside in SC constituencies.

Second, non-SC voters are slightly less likely to vote in SC constituencies than they do in non-SC constituencies, but the difference is small (-0.7 percentage point) and statistically insignificant (column 2).

Third, the statistical insignificance remains intact even if we further divide non-SC voters into ST, OBC, Other Hindu, and Others (columns 3 and 4). Our results thus show that OBC and Other Hindu voters in SC constituencies are as likely to cast their votes as they are when they belong to general constituencies.

Summary and implications

Let us conclude this article with the summary of our findings and their implications. First, political reservation increases the turnout of SC voters in Lok Sabha constituencies reserved for SC candidates. This implies that reservation not only guarantees parliamentary representation but also promotes the mass participation of disadvantaged classes in the electoral process. Second, non-SC voters including relatively “upper” caste voters are not “discouraged” by the system of reservation to vote in SC constituencies. This implies that there is a general acceptance of political reservation in the Indian electoral system. These findings thus clarify how the electoral reservation in India contributes to the achievement of affirmative action. To promote the overall welfare of disadvantaged populations, however, other aspects of political reservations as well as economic reservations are important, an issue on which further research is called for.

(Yuko Mori is Research Fellow of the Japan Society for Promotion of Science and Takashi Kurosaki is Professor at the Institute of Economic Research, both affiliated to Hitotsubashi University, Tokyo. The authors are very grateful to Professor V. K. Ramachandran for comments. They also acknowledge the CSDS for allowing them the use of NES04 microdata. A full paper version is available on request.)

More In: Comment | Opinion