Raymond Fisman, Florian Schulz and Vikrant Vig have a fascinating new working paper: Private returns to public office, which gives us new insights into Indian politics.
We know that elections in India are typically rather close. There is something almost capricious about who wins and who doesn't. The random outcome of an election can, then, be interpreted as randomised allocation into control and treatment. One candidate gets elected, another candidate is very much like him but doesn't get elected.
We can then ask the question: what happened to the wealth of the bloke who got elected? The authors say that the rate of growth of the assets of the person who won grows by 300 to 600 bps per year when compared with the person who lost.
The strongest effects are observed for a person who makes it to being a minister: his asset returns are 1300 to 2900 bps higher than the person who did not win the election.
The paper reminds us of the conditions under which the most fruitful economic research happens today. It's got to be a live and interesting question. A high quality dataset has to be the engine; without good quality data, research is just garbage-in-garbage-out. It's got to persuade us that that the claimed answer is correct. Too often, we in the research profession are failing on these three tests.
We know that elections in India are typically rather close. There is something almost capricious about who wins and who doesn't. The random outcome of an election can, then, be interpreted as randomised allocation into control and treatment. One candidate gets elected, another candidate is very much like him but doesn't get elected.
We can then ask the question: what happened to the wealth of the bloke who got elected? The authors say that the rate of growth of the assets of the person who won grows by 300 to 600 bps per year when compared with the person who lost.
The strongest effects are observed for a person who makes it to being a minister: his asset returns are 1300 to 2900 bps higher than the person who did not win the election.
The paper reminds us of the conditions under which the most fruitful economic research happens today. It's got to be a live and interesting question. A high quality dataset has to be the engine; without good quality data, research is just garbage-in-garbage-out. It's got to persuade us that that the claimed answer is correct. Too often, we in the research profession are failing on these three tests.
If you want do one better pick up the data on Bureaucrats in IAS. Their returns from a job on services compared to those who does a job with similar salary structure in a different field. The conclusions are likely to confirm elite capture, Politician-Bureaucrat nexus and such other maladies.
ReplyDeleteI stopped reading after the first line:
ReplyDeleteWe study the wealth accumulation of Indian parliamentarians using "public disclosures"
The EC data is as poor as any other in India. However, here are two redeeming features:
ReplyDelete1. Yes the Legislators may have much more black money. However, the will be some consistency in the ratio between black and white. An increase in white may imply a greater increase in black. However, this keeps the data relevant.
2. The data is after all competitively screened. Opponents are always digging at the data to find some underreporting to disqualify the candidate. Therefore this data has some pressure which other govt. data does not.
I don't agree with your second point. If anything, we have politicians going down a negative spiral of competitive corruption. Due to complete absence of rule of law, any information against an opponent is merely used as a tactic for under-the-table negotiations. In the current scenario, all candidates are equally culpable and vulnerable, so the criticism from an opponent is limited by his own vulnerabilities (just look at how limited the BJP is, in attacking the government on corruption, since their own laundry is equally dirty). On the contrary, a clean person cannot enter the environment at all, and has to tow the line - just ask young bureaucrats. And, the people don't care.
DeleteA better analysis would start from the public money and trace the flow to various entities. Its always easier to just follow the money, instead of starting backwards from dubious assumptions to infer corruption.
Suppose a politician has two pools of money - white and black. The EC disclosure does a good job of showing the white component. As Shubho emphasises, there is competitive screening, so we can pretty much trust the data that is submitted. That is, the politician is unlikely to visibly own land or shares or a business, which is not reported to the EC.
ReplyDeleteThe politician thus owns one pool of highly visible assets and another pool of highly concealed assets.
There are two distinct rates of return on the two pools. It can well be that the rate of return on the not-observed pool is much higher or much lower. In addition, there can be movement between pools. It is relatively hard to move out from the white pool to the black pool. It is possible to move from the black pool to the white pool. To some extent, if politicians are trying to go legit (owing to various pressures) then there would be a flow of money from the black pool to the white pool, thus exaggerating the rate of return on the white pool.
In short, to the extent that being an elected MLA or a minister increases the pressure on managing the black pool, if politicians tend to switch their assets from the black pool to the white pool, this would generate excessive returns on the white pool, which would tend to undermine the main result of the paper.
I will highlight certain portions of the paper which are slightly confusing from a legal point of view.The authors claim at page 2, in the Introduction, that the data was 'gathered via India's Right to Information (RTI) Act, which required all candidates standing for public office at all levels to disclose the value and composition of their assets.' It needs to be clarified that the obligation to disclose assets arises from rules framed under section 75A(3) of the Representation of the People Act, 1951 (RoPA) and not from RTI Act. RTI Act is a general act which does not specifically impose any obligation on any politicians to provide such information. It is only by way of an application that certain information may be demanded from public functionaries under RTI but that is not the same as saying that the statute imposes on them a duty to disclose such information while standing for public office.
ReplyDeleteHowever, irrespective of whether the data was obtained under the RTI or the RoPA, it is genuine data in the eyes of law. If it is under RoPA, then according to section 81 of the Evidence Act, 1872, there is a presumption that such document containing data of the assets [Eg. Form I of Members of Lok Sabha (Declaration of Assets and Liabilities) Rules] is genuine. If it is under RTI, then under section 79 of the Indian Evidence Act, 1872 it will again presumed to be genuine. Therefore as to the genuineness of the data there can be no doubt at least legally speaking (assuming they were obtained under either of the statutes).
The Indian politics is a big business in the world, The EC expression has a best work of see the white Ingredient.
ReplyDeleteone call that's all
Interesting related paper on "legal corruption" in Mumbai, with a similar comparison between gains made by winners vs losers of elections:
ReplyDeleteAnatomy of legal corruption in India: Honest graft by politicians