Should We Tax Financial Transactions?

The idea of a sort of extra-broad Tobin Tax on all financial transactions has been quite popular with the left half of the punditocracy for some time now.  Myself, I don't really see the charms.  Tiny taxes on high-volume transactions raise a lot of money, but they also cost money to record, collect, and audit, which is why few jurisdictions have 0.25% sales taxes.  And I'm not clear on what problem taxing financial transactions is supposed to solve.  It's not as if our woes were caused by legions of high-frequency traders wrecking the markets with their tiny, tiny spreads.  Nor do I think that penalizing sales would have prevented the run on the money markets, or any of the other problems of the acute phase of the crisis.  The charm seems to be entirely that it might raise some money, and it pisses off bankers.  

Of course, we have to raise money somewhere, and how better than by pissing off bankers?  It seems like this might be that most magical of policy possibilities--the tax with no downside.

Ah, but is that really so?  Ken Rogoff, who spent quite a lot of time thinking about these issues as chief economist of the IMF, points out that financial transaction taxes aren't actually pain free:

True, great thinkers like John Maynard Keynes and the late Nobel laureate James Tobin floated various ideas for taxing financial transactions as a way to reduce economic volatility. (Tobin's tax applied specifically to foreign-exchange trading.) But, since then, the idea has received close attention from many economic researchers, and, frankly, it is hard to find their research results supportive.

Such taxes surely reduce liquidity in financial markets. With fewer trades, the information content of prices is arguably reduced. But both theoretical and simulation results suggest no obvious decline in volatility. And, while raising so much revenue with so low a tax rate sounds grand, the declining volume of trades would shrink the tax base precipitously. As a result, the ultimate revenue gains are likely to prove disappointing, as Sweden discovered when it attempted to tax financial transactions two decades ago.

Worse still, over the long run, the tax burden would shift. Higher transactions taxes increase the cost of capital, ultimately lowering investment. With a lower capital stock, output would trend downward, reducing government revenues and substantially offsetting the direct gain from the tax. In the long run, wages would fall, and ordinary workers would end up bearing a significant share of the cost. More broadly, FTTs violate the general public-finance principle that it is inefficient to tax intermediate factors of production, particularly ones that are highly mobile and fluid in their response.

All of this is well known, even if prominent opinion leaders, politicians, and philanthropists prefer to ignore it. The European Commission has surely been strongly cautioned by the Fiscal Affairs Department at the International Monetary Fund, whose economists have thoroughly catalogued the pros and cons of FTTs.

So why did the Commission go forward with the idea?The most generous interpretation is that the Commission simply does not believe economists' estimates and analysis, and views an FTT as more workable than is commonly realized (a scenario that calls to mind the debate surrounding the creation of the euro). It is true that Latin American governments, particularly the Brazilian authorities, succeeded in raising more revenue from taxes on bank withdrawals (a crude version of an FTT) than most policy analysts thought possible. On the other hand, Latin America's long-term growth record is hardly an advertisement for the approach, and accounting for lost tax revenues due to lower GDP would surely yield a less impressive fiscal outcome.

Perhaps there are excellent answers to these concerns.  But I haven't seen the people supporting an FTT offering any.  I'd like to see that gap addressed before I jump on board.