T C A Srinivasa-Raghavan: The economics of affirmative action

OKONOMOS/ Indian economists have shied away from examining the issue

Earlier this week, in a meeting at National Council of Applied Economic Research (NCAER), I suggested that the idea of reserving jobs for the dalits and scheduled tribes in the private sector was worth looking into. The purpose would be to ascertain just how well or badly reservations had worked.

The response of the gathered economists was lukewarm. On being pressed, one of them said it was a political issue on which the Council should refrain from commenting. Another, however, said that though it was political, it had deep economic consequences.

Bemused, I came home and searched the Web for some proper analysis of the Indian experience. There was plenty, but nothing written by economists. For them, apparently the issue does not exist (Surjit Bhalla may get on with the job).

So I turned to the National Bureau of Economic Research (NBER). It produced 15 working papers. In passing, it is worth noting that of these 15, seven were written before 1986 and eight after 1996. Why the gap?

Then I went to the World Bank site, which produced nothing. Nor did the Asian Development Bank site. This is surely quite amazing, considering how much these institutions blather on about labour markets and the need to reform them.

I did a few more searches and found that only economists employed in universities had examined the economic implications of affirmative action. Economists employed elsewhere had not, presumably because the topic is too political.

One of the NBER papers* was about whether those employed under affirmative action were less qualified than the rest. Since this fear has been widely expressed but never properly researched, I thought it would make a beginning to report the NBER paper's findings.

The authors, Harry Holzer and David Neumark, have used micro-level data on employers and employees. "Our measures of qualifications," they write, "include the educational attainment of the workers hired (both absolutely and relative to job requirements), skill requirements of the job into which they are hired, and a variety of outcome measures that are presumably related to worker performance on the job."

And what did they find after analysing a sample of over 3,200 employers in four major metropolitan areas? "Some evidence of lower educational qualifications among blacks and Hispanics hired under Affirmative Action, but not among white women." But also "our results show little evidence of substantially weaker job performance among most groups of minority and female Affirmative Action hires."

So is this true of India as well? Does the lower educational qualification get offset by performance on the job?

A more recent paper** by Pedro Carneiro, James Heckman and Dimitriv Masterov examines minority-white wage gaps. Earlier research had shown that socially disadvantaged "children and their parents may have pessimistic expectations about receiving fair rewards for their skills and so they may invest less in skill formation."

But Carneiro, Heckman and Masterov find that "gaps in ability across racial and ethnic groups open up at very early ages, long before child expectations are likely to become established." Interestingly, these gaps widen with age and schooling for some groups but not for others.

Perhaps, the most interesting finding is that the parents of socially disadvantaged children are more pessimistic when their children are young but become more confident later on. But, say the authors, "expectations data are unreliable and ambiguous".

They have also documented disparities in noncognitive traits across racial and ethnic groups and their relationship to labour market outcomes. The evidence is quite clear: preschool and family factors are critical for "explaining both cognitive and non-cognitive ability differentials by ethnicity and race. Policies that foster both types of ability are far more likely to be effective in promoting racial and ethnic equality than are additional civil rights and affirmative action policies targeted at the workplace."

Listen carefully, Dr Singh, and you will find that this is what everyone without a political axe to grind has been saying.

For reasons of space, I have mentioned only two papers here. There are 13 more that are worth downloading.

Those interested in the subject should also read the volume edited by Kenneth Arrow, which came out a few years ago and was reviewed in these columns. It is called Meritocracy and Economic Inequality. The papers in it reach very similar conclusions.

The morals are clear: one, affirmative action helps but the point of intervention is crucial; and two, Indian economists need to examine the issue.

*Are affirmative action hires less qualified?
NBER Working Paper No 5603
**Labor market discrimination and racial differences
in pre-market factors.
NBER Working Paper No 10068