In a recent interview Finance Minister P Chidambaram, answering a question about the turnover tax on stock trading said, "I am concerned in the sense that the stock market has not fully absorbed the entire Budget. It is reacting to the securities transaction tax. But we can debate that. I have material to show that many countries have a similar tax at much higher levels."
What he did not say, perhaps because he has not been briefed sensibly by his advisor on such matters, is that other countries also have a lot of material to show that it is a very bad idea. For his reference, two papers are cited below.
All conclude that turnover taxes are bad. For what it worth, I have also included a paper from Australia on gambling taxes, because deep down that is how the Budget treats stock market transactions.
First, the textbook case. Conceptually, a turnover tax has the same effect, say textbooks, as a sales tax, with the difference that while sales taxes are paid as a fixed percentage of the value of a transaction, a turnover tax taxes all transactions. This makes it a killer.
For instance, if it is a shirt, a turnover tax would go into every transaction that went into making the shirt. This makes it hugely inefficient because it places a tax on the value of each transaction, thus discouraging transactions.
It also encourages vertical integration as firms try to get into intra firm trades. Therefore, it is an inefficient method for raising revenue because it discourages potentially profitable exchanges and encourages potentially inefficient mergers.
Let us now see what happened in Japan as described in an excellent paper by Hiroyuki Ono of Toyo University. He found that it "has been significantly affecting the volume of stock transactions negatively in the Japanese market, as in other countries."
He also finds that the "degree of its influence has increased over time, as stock market reforms have been underway." However -- and here is the lifeline for the finance ministry -- the negative impact is likely to be smaller than in Europe because "Japanese investors are not very sensitive to the transactions costs they face."
The moral is what every textbook says: a great deal depends on the elasticity of demand for domestic brokerage. And the truth is we don't have the foggiest about it here in India.
How could we, when Mr Chidambaram said to this paper, "Nobody can estimate what capital gains we collect, so we can only make an assumption"?
The paper by Campbell and Froot is more forthcoming. The Japanese scholar seems not to want to criticise his government very firmly and his paper is a model of Japanese politeness.
But the Americans are typically plain. They say that a great deal depends on behavioural responses but, in general, the following happens:
As a result of all this, it seems the Swedes who introduced such a tax paid a heavy price, but not the British who were cleverer. They had a system of stamp paper registration fees.
So what are lessons? There is only one, in fact: don't do it if you haven't done the homework, which I suspect, the finance ministry hasn't.
To conclude, look at gambling taxes, if only because that is how the Left would want it. The Australian experience is given here. I would recommend that everyone should read this paper.
"For some gambling activities, such as lotteries and racing, taxes are generally levied on turnover," says the government paper. Also, "gambling products are typically provided under some form of licensing arrangement for which licence fees are collected."
Some taxes are flat rate but others have progressive scale structures, which vary with the size of the operator. The one that caught my fancy is the bookmaker's turnover tax. It varies from one to two per cent. There is also interstate variation in the allocation of revenue derived from bookmaker's turnover tax.
This might be worth a study by the NIPFP.