Wednesday, April 28, 2010

The IMF and Financial Sector Taxation

The IMF recently came out with some proposals for taxation of the financial sector, summarized here. There are 3 proposed schemes, which are somewhat related. These are:

1. Financial Stability Contribution (FSC)
2. Financial Activities Tax (FAT)
3. Financial Transactions Tax (FTT)

What might the role of such taxes be? First, we might argue that financial bailouts result in a redistribution toward the financial sector, and that we could compensate for this through taxation. This applies to bailouts that have already happened, and bailouts that might potentially happen. Any of these proposed taxes will result in redistribution away from the financial sector, but the IMF appears to think of the FSC in particular in terms of its role in redistribution. One could make the case, of course, that there has not been much redistribution during this recession - the Treasury will recover most or all of its TARP transfers, and the Fed appears (judging from its last financial statement) to be doing unusually well recently (though perhaps only due to its monopoly power). There is always the implicit redistribution, though, that comes from too-big-to-fail.

Second, taxes could be important in promoting efficiency. (i) We could argue that there is an externality at work here. What is it? Because of explicit deposit insurance and the implicit insurance from too-big-to-fail, financial institutions take on too much risk, and we have a hard time pricing this risk correctly in the market. There is a classic market-failure-externality problem, and we can correct it with a Pigouvian tax. We could approach this by making some calculations about how financial institutions of different sizes and functions will behave, and set the tax rates appropriately based on size so as to correct the externality. Also, we might imagine a system where we measure individual institution risk and set the tax rate as a function of risk. The IMF seems to want to think of the FAT as involving risk-adjusted tax rates. This has the same drawbacks as subordinated debt, which I discuss here. The basic problem is that we can't price the risk well, in part because financial institutions will behave in ways that make their portfolios look less risky. (ii) A second inefficiency comes from concentration in the financial industry. I have argued here that this results more from scale economies than too-big-to-fail, but in either case we could justify a tax to rake off the monopoly profits. The IMF seems to think of the FAT in this way. The FAT is essentially a value-added tax for the financial industry. The key problem here is measuring the value-added of financial activities, which is a very thorny problem. National income accountants have known for a long time that measuring the contribution of the financial sector to GDP is really hard, if not impossible. Think about it. What's an input? What's an output? Somehow assets are transformed, and we make everyone better off by redistributing risk efficiently? How do you measure the benefits of that redistribution? What happens when you have incentive problems? Yikes! The principle behind the FAT seems to have some merit, but I don't know how you implement it in a sensible way.

It's useful to think of approaches to correcting environmental externalities as an example here. With some types of pollution, market-based solutions like cap-and-trade may work, if it is relatively easy to measure the extent of the activity causing the externality. For other things, the bad activities are hard to measure, and the problem can only be corrected with direct regulation and monitoring. I tend to think that, in the case of financial regulation, that market based solutions involving corrective taxes or subordinated debt, for example, will not work to solve the basic incentive problems. What we need is better direct regulation and monitoring of risk with a systemic view in mind. Of course, there is nothing wrong with taxes on the financial sector aimed simply at correcting the negative effects of concentration and generating badly-needed revenue for the Treasury.


  1. Nice post Stephen, thanks for the review of the proposed IMF taxes. In theory, the idea of pigou taxes for the financial system seems appealing, but you did raise some important concerns.

    What I am more interested in is what happens with the revenues raised by the taxes. You mentioned the potential for generating badly-needed revenue. I'm also intrigued by the idea of a global bank bailout fund. While the name is unfortunate, I think the failure of Lehman Brothers highlights the need for an international mechanism to handle multinational financial institutions which are in distress.

    An international body empowered to either rescue failing banks or seize and liquidate them in an orderly fashion could provide additional stability to the global financial system. Funding such an institution would be tricky, but a bank tax seems to be a good starting point.

  2. It could be that this is the job that the IMF is looking for. Maybe that's why they wrote the proposal.