The behaviour of most finance ministers, as far as their victims are concerned, usually defies logic. But, say Roger Gordon and Wei Li in a recent paper*, the finance ministers are not to blame. A great deal, it seems, depends on whether we are talking about finance ministers in a rich or poor country.

"Tax policies differ dramatically between poorer and richer countries," they write. The latter use broad-based income and consumption taxes, while the former prefer excise taxes, tariffs, and seignorage.

"In the process, though, they collect much less revenue as a fraction of GDP than is collected in richer countries. Corruption and red tape are also far more common in poorer countries."

Gordon and Li have tried to explain why this happens. Their own model says that the real problem is tax enforcement.

When it is poor, firms shift to cash. "The government is then left relying for revenue on the remaining industries that cannot so easily shift into the cash economy to evade tax."

Eleven years ago, Gene Grossman and Elhanan Helpman had tried to explain the same thing. They had come up with the view that it was political pressures in poorer countries that mattered and that these led to some very funny tax policies.

In this paper, Gordon and Li have examined that political explanation model in detail. They conclude that while it can explain favourable tax rates for those can swing it for themselves, "it does not easily explain tariffs and seignorage. Only if sales or income tax rates cannot vary by industry to the extent desired might tariffs make sense."

They have then tried to see to which model fits the data best -- and find that the forecasts from their own model fit the data better. It appears that "a weak financial sector implies that little is lost by a firm from shifting to the cash economy as a means of evading taxes."

We know what happens. A rickety financial sector leads to a big informal economy. This results in a "perverse" tax structure to deal with the resulting pressures. The authors have documented both such relationships.

Left-wing governments should face very different pressures than right-wing governments. So their tax policies should be different as well. The authors have examined if this is true and "find little difference in tax policies across governments of different ideologies."

Their own model forecasts that the highest tax rates will be paid by capital-intensive industries because they find it hardest to shift to the cash economy.

"In the political economy model, these industries should face the lowest tax rates since they can most easily solve the internal free-rider problems and lobby the government for support." But the tests for this, they say, are not definitive.

In the political-economy model, a great deal depends on the smaller fraction of industries in poorer countries lobbying the government. Outside pressure can help in mitigating the effects of this lobbying.

But where the financial sector is weak, tax evasion is easy. "Reform efforts then need to focus on improving the quality of the financial sector." Mr Chidambaram and Dr Manmohan Singh, by allowing Congress toadies on to the public sector bank boards are doing the opposite.

Indeed, the authors also say that outside pressure to have better tax policies without simultaneously improving the financial sector, will likely cause more harm than good. VAT leakages, for example?

To my mind, the key insight the paper provides is this. "�financial reform improves not only the allocation of credit across firms but also induces a shift in government policies more broadly to ones that create fewer distortions to market allocations."

I would also say that this paper ought to provide an important intellectual underpinning for the next Budget. Mr C is likely to go for the taxpayers' throat. He may be better off improving the financial sector.