In an editorial last year, India Together argued in favour of reservations for lower castes. In their piece, Ashwin Mahesh and Subramaniam Vincent commented on the reasons we have reservations. Affirmative action policies they argued not only directly benefit lower castes through higher incomes, but have a larger impact on public policies when individuals from lower castes are given a voice in the decision making process. Other commentators on these pages have followed a similar line of reasoning. For instance, when advocating for reservation of Parliament seats for women, Kalpana Sharma writes that "there is a greater chance of mainstreaming women's concerns if there are more women in positions of power from where these concerns can be addressed."

"Are reservations working?" ask Mahesh and Vincent, who say that the impact of reservations on public policy would be most visible in legislatures and panchayats. Despite their arguments, none of these writers are able to provide any evidence that the legislators, once elected, actually behave in ways expected of them. The complexity of the political system means that there are a number of ways in which legislators get impeded in their work. The legal scholar Upendra Baxi argues that SC and ST legislators need to appeal both to upper-caste constituents in reserved jurisdictions and to the primarily upper-caste membership of their parties. Also, the dynamics of political parties and bargaining within legislatures can dull activism of individual legislators in favour of their communities. Or MLAs and MPs might simply concentrate on increasing their own wealth and not care about their constituents at all. Kalpana Sharma, perhaps thinking about the behaviour of Indira Gandhi, Mayawati, Rabri, and other women in power, writes that "there is no guarantee that [women’s reservation] in itself will make a difference to the status of women in the country." These are prescient words, for casual empirics do not explain what has been the actual result of political reservations in India.

When casual empirics fail, perhaps it is time for a more rigorous approach. A number of researchers in economics have started to look closely at political reservations, both for lower castes and tribes and for women in India. In an important paper published in the American Economic Review in 2003, Professor Rohini Pande of Yale University asked if reservations in state legislatures increased influence in policy-making for scheduled castes and tribes. She concluded that they did, and backed up this assertion by presented evidence of targeted redistribution policies passed by SC and ST legislators.

Legal identification of Scheduled Castes and Scheduled Tribes

Selection criteria for scheduled castes

1. Cannot be served by clean Brahmans
2. Cannot be served by the barbers, water-carriers, tailors, etc. who serve the caste Hindus
3. Pollutes a high-caste Hindu by contact or by proximity
4. Is one from whose hands a caste Hindu cannot take water
5. Is debarred from using public amenities such as roads, ferries, wells, or schools
6. Will not be treated as an equal by high-caste men of the same educational qualification in ordinary social intercourse
7. Is depressed on account of the occupation followed and, but for that, occupation would be subject to no social disability

Selection criteria for scheduled tribes

1. Tribal origin
2. Primitive ways of life and habitation in remote and less accessible areas
3. General backwardness in all respects

Source: Constitution of India

For her analysis, Pande exploits a particular feature of the Indian representative set-up. Each state legislature has seats reserved for SC and ST candidates, but the proportion of seats varies by state. There are more seats for states with higher proportion of SC and ST population, and vice versa. Also, between 1950 and 1980, the seat allocations kept changing as new data from the census became available. So these variations allow Pande to compare the policies in states with higher SC and ST representation to those with lower SC and ST reservation.

Increasing SC reservation does not have a significant impact on general spending policies such as total spending, spending on education or land reforms. However, it has a significant impact on targeted spending policies.
 •  Women's representation
 •  The merit of reservations
 •  Caste: Don't ask, don't tell
Pande's results show that reservations impact different groups differently, depending on the policy. Increasing SC reservation does not have a significant impact on general spending policies such as total spending, spending on education or land reforms. However, it has a significant impact on targeted spending policies. Increasing reservations by 1% increases job quotas for SCs by 0.6%, but does not affect spending on SC welfare schemes. This split between general and targeted policies sits well with the social structure of these groups. Compared to STs, SCs are well educated but geographically distributed, so they rely on individual specific schemes such as job quotas. An SC legislator who advocates group-specific policies cannot be sure that they will actually be used by the community that she or he wants to target.

In contrast, ST reservations have an impact on a broader range of spending policies. Increasing ST reservation by 1% decreases spending on education by 0.4%, but increases spending on tribal welfare schemes by 0.8%. Again, this matches what we know about tribal communities in India. They are remote from the major population centres yet live cohesively. So they are able to take advantage of and prefer group-specific programs over individual-specific ones.

Pande's research is one of the first threads in an emerging literature on the behaviour of elected representatives in office. In 2004, Professors Raghabendra Chattopadhyay of IIM Calcutta and Esther Duflo of MIT published their research on the impact of reservations for women in panchayats, specifically looking at Rajasthan and West Bengal. Their analysis pointed to important differences in policies enacted by panchayats headed by women and men, debunking the myth that women sarpanches are puppets controlled by men. Even in panchayats with "unassertive" women as sarpanches, the presence of a woman in a position of authority often inspired other women in the Gram Sabha to speak up, changing the dynamic of village policy making.

In another 2004 study, Professors Tim Besley, Rohini Pande, Lupin Rahman and Vijayendra Rao found that if the Sarpanch position is reserved for person from a Scheduled Caste or Tribe, then SC or ST households are 7% more likely to have access to a toilet, an electricity connection, or a private water connection via a government scheme.

Among economists, the debate on the merits of reservations is just beginning. The precise relationship between political power and policy implementation is still not clear. And there are a number of unresolved issues for everyone involved.

For political parties, there is a concern how political reservation would change the people being elected to Vidhan Sabhas, and the ideologies and policies they would back. Could this change be significant enough to change the top leadership of the party and government? If the change is perceived as minor enough, perhaps existing legislators could be convinced to vote in favour of more reservation, which includes women.

Finally, voters themselves must be concerned with reservation. How does their involvement in the political process change as a result? Does voter turnout increase or decrease when only certain kinds of candidates can stand for elections? And how can voters signal political support for or in opposition to reservation? Answering these questions is a collaborative exercise. Social scientists bring their best tools to understand what society is saying, but first the people must themselves debate and decide their preferences. Hopefully the results of reservations in the past will inform both the debate and the decision.