T C A Srinivasa-Raghavan: Finance ministers and bounded rationality

OKONOMOS

If the minister is arrogant, he will make mistakes

Two weeks ago, this column had reviewed a paper by Christine Jolls of the Harvard Law School and Cass R Sunstein of the Chicago Law School ("Using law to neutralise nutty behaviour", November 25), which said that sometimes it can be a good idea for governments to intervene via the law to save people from their own follies.

But even while talking of "bounded" rationality amongst individuals as the reason for such intervention, the paper assumed that governments suffered from no such debility. It, therefore, did not answer as to who would save a government from its ministers.

Indian governments are especially vulnerable. For example, this newspaper reported on Thursday that the finance minister had asked "chief commissioners of income-tax to raise as much current demand from taxpayers as possible..." and that "there was no reason why the assessment of major tax payers could not be computed in the current year itself."

Apart from indicating ministerial desperation, this means the IT department can ask taxpayers to pay whatever it thinks is needed to meet the revenue target and not what may be actually owed. The officials of the East India Company in Bengal, I might remind the finance minister, used to behave in exactly this fashion.

In a response* to the Jolls-Sunstein thesis that governments should intervene, Edward L Glaeser, also of Harvard, says that "government decision-making is likely to be particularly erroneous ... across a wide range of settings" because "governments seek to aggrandize their own authority".

Or, at any rate, ministers do.

His point has substantial appeal. He says "flaws in human cognition should make us more, not less, wary about trusting government decision-making."

This is because it is necessary, while debating paternalistic governance, to set private against public errors. It is important to recognise that psychological errors are endogenous, in that they are not forced by external circumstance.

Thus, if the minister is arrogant, he will make mistakes such as the minimum alternate tax, the cash withdrawal tax and the fringe benefit tax. The same, I would think, can be said about the latest direction to chief commissioners of income tax.

The question arises, of course, as to how we can distinguish between a ministerial direction to his subordinates and proper legislation. The latter has the force of law, while the former is no more than an order disguised as a suggestion made with humility.

The answer is recourse to the courts. But there is nothing in the law that says that a minister can't issue orders that exhibit "bounded rationality".

Even when it comes to imposing taxes that are socially inoptimal, the sovereign right to tax combined with a minister whose rationality is bounded and a legislature which doesn't care, results in some pretty odd behaviour by Indian finance ministers. Only the degree varies.

It is important to find out why otherwise sensible people, when they become finance ministers, behave so peculiarly. Glaeser discusses psychology at length and says "individuals believe, at least in part, what they hear."

He and Alberto Alesina, one of Harvard's best, had shown that 60 per cent of Americans believed that the poor were lazy, whereas, 60 per cent of Europeans believed that they were trapped in poverty.

This is crucial to understanding the phenomenon of finance ministerial caprice because he or she mostly hears only what the tax bureaucracy, corrupt to the core and a lobby in itself, tells its lord and master.

So, says, Glaeser, "one major source of cognitive errors is the supply of beliefs." In such a case, "errors will not be random but will in part reflect the costs and incentives faced by belief suppliers" in this case the CBDT.

The solution? I can do no more than quote Glaeser at length:

"Given that errors are greatly exacerbated by the suppliers of bias, situations with strongly interested parties who skew beliefs are particularly dangerous. Free entry of ideas is helpful but ... rules that prevent interventions in areas where potential providers of bias with extremely strong incentives may reduce supplier-created bias."

In short, we need laws to curb the tax bureaucracy.

**Paternalism and Psychology, NBER Working Paper No. 11789, November 2005