Using Asset Disclosures to Study Politicians’ Rents in India

A terrific paper by  Rikhil Bhavnani of University of Wisconsin-Madison. The paper was presented in World Bank’s Annual Development conference – ABCDE 2012.

He looks at corruption amidst India’s politicians using their asset disclosures before the elections. The paper is full of surprises:

Public asset disclosures, mandated of a third of the world’s national legislators, may be used to study the gains to office and corruption among politicians. I conduct such an exercise here, using data from India. Comparing the change in winners’ and losers’ self-declared family assets in the country’s two most recent state and national elections indicates that the average election winner increased his assets by 4-6% a year. 4-9% of election winners appear suspect, as their asset increases are greater than the premium that they could legitimately earn as legislators. 

Although these results are arguably modest and statistically fragile, they are driven by larger and more statistically robust estimates for members of the BJP. The suspect nature of some BJP winners’ asset changes is confirmed by an analysis of their composition. Subject to data reliability, these estimates give us reason to doubt the popular characterization of all of India’s politicians as massively corrupt. Members of the BJP, parliamentary majorities and cabinets, and politicians from constituencies reserved for minorities, are more likely to be suspect. These associations could reflect candidate selection or varying incentives for corruption. The method developed here could be used to study corruption among politicians elsewhere.

They look at elections in two periods  – 2004 and 2009. Apart from general elections they also look at state elections in the same period. In this period, Indian GDP grew by an average 9%. So the gains made by politicians (covered in the study) is much lesser than average GDP growth. Who would believe that! And the gains hardly correlate the common accusation that most politicians are corrupt in the country.

Apart from this, the difference in earnings of politicians in the two main parties is interesting. He looks at the reasons for the same:

A second result is that BJP candidates appear to enjoy higher mean asset changes, and are more likely to have suspiciously high asset changes, as compared with INC candidates. This is ironic, since the BJP claims to be the party of “governance.” Why might the estimated gains to office for INC and BJP candidates differ? One possibility is that BJP candidates are more likely to honestly declare high assets. We have no reason to believe this to be the case, however. Asset disclosure laws apply equally to all candidates, and parties are not known to monitor asset disclosures to different degrees. Further, that BJP candidates declare lower mean (and median) assets in both 2003/04 and 2008/09 than INC candidates contradicts the possibility.

A second possibility is BJP candidates’ asset increases reflect the fact that they formed the government in many states (this is, in fact, a post-treatment variable), rather than the fact that they are BJP members. This, too, appears unlikely, since the BJP formed the government in five out of the eleven states considered, while the  INC ruled both the center and four remaining states.

This leaves the possibility that the differences between INC and BJP candidates truly reflect differences in asset accrual. Such differences could be due to variation in party membership bases (the BJP is said to draw more of its membership from the middle classes, while the INC concentrates on economically disadvantaged minorities) or organizational structure (the INC is a dynastic party, while the BJP is relatively professional [Chhibber 2011]). Both these possibilities could affect candidate selection, since they might cause the parties to attract differentially corrupt candidates, or politicians’ incentives while they are in office.

Not sure whether anyone has read this study in the Parliament…:-)


The results presented here help improve our understanding of political life in India. In conjunc- tion with established facts about politicians’ careers—the long hours they work, the competitive elections they face, and the high frequency with which they drop out of politics, partly due to incumbency disadvantage (Linden 2004; Ravishankar 2007; Uppal 2005)—it may be possible to characterize Indian parties as tournaments with different payoff structures (Levitt and Venkatesh 2000). That much of the personal gains in office are concentrated among a few (in the data, 1% of politicians received 30% of the total gains to office, while 5% of politicians received 55% of the total gains to office) corroborates this notion.54 This account of politicians’ careers, if true, might help explain some general features about Indian politics, including the high candidate entry rate  (which could be caused by the large difference between the mean and median gains to office),55 the poor quality of the mean candidate (an attribute of tournaments without a primary-like screen- ing mechanism, which is absent in India) and the ubiquity of middlemen (who could be viewed as proto-candidates, working for free to get party tickets). It might also help explain systematic differences between parties.

We should be taking this research further to bring bureaucrats and judges as well:

Lastly, the method proposed in this paper could be used to study the gains to office in other contexts. As judges and bureaucrats declare their assets in India, we should be able to compute and analyze their gains to office, thereby acquiring a more comprehensive understanding of corruption in India. Similarly, the gains to office could be examined in the over one-third of countries where information on candidate assets is available to the public (Djankov et al 2010).

Apart from some crazy findings, the author takes pains to explain the methodology in the paper. It is in much details abnd gets on to you. I need to reread the paper couple of times to figure the complete methodology.

Apart from this the paper says politicians which made most gains were from states like Orissa and Karnataka. These are two states where there is mining…Then another surprise is Arunachal which gets large funds from Centre…Surprise bit is politicians from Delhi and Maharashtra don’t make much…The author says this is perhaps due to reversal in real estate prices…Not sure about that…Going by this research in Mumbai University, there should be much larger gains in Maharashtra.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: