“The financial sector has grown too big and volatile and something must be done. A tax on financial transaction is something, therefore a tax on financial transaction must be done.”
If you can spot the logical error in that line of reasoning you just outsmarted many European policymakers.
Eleven EU countries – including France, Germany, and Spain – are still trying to sort out their differences and implement some form tax on financial transactions by as early as 2014. Such a tax is envisioned to include most financial transactions carried out by EU financial institutions, and not just on transactions carried out in Europe, but around the world. Shares and bonds would be taxed at 0.1 percent and financial derivatives at 0.01 percent.
The tax is meant to serve two goals. It would raise much-needed government funds from those institutions widely seen as having caused the current crisis and it would curb excessive speculation so as to prevent such a crisis from happening again. The European Commission has estimated the revenues to be 30-35 billion euros per year. Sounds like a win-win, so what’s the problem? Well, as the proposal is worded now, quite a few.
- Design flaws. One of the most widely used financial tools is the so-called repo (repurchase-agreement) in which, instead of lending against collateral, a financial assets is sold on Monday and bought back for slightly more on Tuesday, with the difference between the two prices being the interest rate. Each of such transactions would be taxed. A 0.2 percent daily tax would amount to a 107 percent yearly tax, completely freezing the market!
- The tax would only be on EU financial institutions. Whereas some financial assets, such as a stock of a local company are clearly local, modern financial markets are increasingly global. Imposing a tax in Europe simply means that much of the financial transactions will go elsewhere. When Sweden in the 1980s imposed a financial transaction tax not unlike the one presently being considered, they found that more than 90 percent of the market fled elsewhere, leaving actual revenues far short of those originally envisioned. (John Campbell and Kenneth Froot have a detailed analysis of the Swedish experience here). Sweden is presently strongly opposed to the current suggestion based on its past experience. The current design attempts to sidestep this problem by making a transaction taxable regardless of where it takes place as long as one EU institution is involved. Deutsche Bank selling a mortgage from Chicago to Bank of America is therefore taxable in Europe, a completely unorthodox extension of the jurisdiction of European authorities and it is unclear how this would work out in practice.
- There is no evidence that financial transaction taxes curb volatility. Careful work by the International Monetary Fund (IMF) finds very little, if any, evidence that financial transaction taxes curb volatility. And as suggested by Ken Rogoff here, they are likely to increase the general cost of credit, hurting the whole economy.
So what to do instead? The goals of curbing financial sector fragility and forcing banks to pay a share of past and future bailouts is certainly worthy. But there are better, more efficient, ways of doing that. If you want to raise more revenue, then tax bank profits or assets. If you want to make financial volatility less damaging for the rest of the economy, then raise the requirements on equity buffers from the present 8-10 percent to, say, 30 percent. This would allow the financial institutions to absorb heavier losses with less collateral damage to the rest of the economy. Both measures address directly the desired goal and are less likely to have undesired consequences.
Due to much criticism and debate, the proponents of the financial transaction tax have been forced to reconsider its implementation and design. That is a good thing. It is important that we design our financial systems based, not on what feels good or to punish perceived wrongdoers, but on what enables our economies to function the best. The proposed financial transaction tax might accomplish the former, but surely not the latter.