YES, and it should take the form of a levy on the non-core liabilities of the banking system, such as short-term wholesale funding.

During a credit boom, when bank lending is increasing very rapidly, core funding sources such as retail deposits cannot keep pace with the increased lending so that banks tap additional, non-core funding to finance their new lending. In this way, the stage of the financial cycle is reflected in the composition of bank liabilities. The larger is the stock of non-core liabilities relative to core liabilities, the more vulnerable is the financial system to a sudden de-leveraging episode.

COMPARED to when I first began writing about financial regulation ten years ago, the subject has been transformed from the tediously arcane to the deeply political. What else could come from a situation where previously self-satisfied, highly-paid and lightly taxed bankers are bailed out by governments that are then forced to make public sector workers redundant and to shelve programmes for the less well off—hardly God’s work. It is entirely natural then, that the debates on bank taxes are driven by highly charged political sensitivities and not economic efficiency. But that is why you need a profession of dismal scientists to point this out and to say, for example in this case, that while taxes are part of the solution the proposed taxes do not solve the problem at hand.

YES, a properly designed tax on systemic risk can be an effective macroprudential instrument to reduce the probability and costliness of future crises. In addition, such a levy can be used to achieve a bail-in of the private sector during a crisis; it can serve as a tool to fund an effective cross-border resolution mechanism.

The cornerstone of any reform has to be the idea that “being systemically relevant” has to come with a cost, as it is otherwise attractive for financial institutions to choose to be systemically important and enjoy the benefits of implicit government guarantees. Funding cost advantages of too-systemic-to-fail financial institutions are substantial: the value of state guarantees embedded in ratings translates into a funding advantage of 10 to 120 basis points (depending on the individual strength of the institutions). A levy could counteract this subsidy to too-systemic-to-fail institutions.

THERE are two types of taxes: necessary evils and good taxes. “Necessary evil” taxes are sources of revenues to finance the state, which are used in spite of their negative incentive effects. For example, by taxing work the income tax induces people to remain unemployed. Good taxes, instead, have positive incentives effect. They can also be used as a source of revenue, but their primary purpose is to correct a distortion present in the marketplace. An energy producer who burns fossil fuel, for instance, does not consider the social cost of the CO2 he releases. A tax equal to the cost that the increased levels of CO2 has on the rest of society will force the producer to factor this cost in his decisions or, in economists’ jargon, to internalise the externality. Regardless of the revenues it generates, this type of tax improves social welfare. These taxes are justified, however, only when there is a “market failure”, i.e. when market prices fail to incorporate part of the costs (or benefits) an activity generates. This concept was first introduced by British economist Arthur Pigou; hence the name of Pigouvian taxes.