A tax on financial transactions: Good or Bad Idea?

In recent days, we have heard talk from Congress about imposing a tax on financial transactions. While there has been heated debate on the topic, there seems be more smoke than substance in most of the arguments. This morning, Paul Krugman, who seems to have made a speedy and seamless transition from economist to polemicist, has an article on why such a tax is a good idea:
http://www.nytimes.com/2009/11/27/opinion/27krugman.html?ref=opinion
As always, Krugman sees the villains here (the speculators, who else?), decides that this tax will not have much effect on the good guys (a group of long term investors, into which he puts himself and his readers) and sees potential benefits to markets from the action.

Very convenient, but not very balanced!!! I would like to provide a counter, by first examining the motives for a transactions tax and then considering the laws of unintended consequences.

Motives
As I see it, there are three motives for a transactions tax.
1. Revenue generation: As government budgets get squeezed and deficits mount, legislators are flailing around for ways to raise revenues in fragile economies. Given the sheer volume of trading volume in financial markets, even a small tax seems likely to raise huge revenues. (In a classic example of how governments compute potential revenues from taxes, the estimated tax receipts are computed by taking the existing dollar value of trades in a market and multiplying by the tax rate.... If only we lived in a static world...)

2. Punish bad behavior: As a bonus, the tax will fall most heavily on those who trade short term or in derivatives markets. If we assume, as Krugman has, that these trades are for the most part speculative, the tax punishes that "bad" behavior. (It is the same rationale that allows governments to raise taxes on tobacco and alcohol...)

3. Target the "right" entities: The perception on the part of many is that the biggest traders in derivatives markets are investment banks and hedge funds. The billions of dollars that these entities are reporting in profits, in conjunction with their absence of suitable remorse for their role in creating the banking crisis of last year, has made them easy targets. (I am quite surprised that legislators have not proposed a windfall profits tax on just the bad guys, at least as they see them... they would probably call it the Goldman tax!!)

So, what can go wrong?
1. Motives are internally inconsistent: There seems to me to be a direct contradiction between motives 1 and 2. Put another way, the only way in which this transactions cost will raise revenues is if the bad behavior in question (short term trading) continues in the future. I think legislators need to specify what their primary objective and not try to argue out of both sides of their mouths. (I know little or no chance of this happening, but no harm hoping..)

2. Speculation versus Investing: As I have argued before, I am very uncomfortable drawing the line between speculation and investing. While I might not see much benefit to short term trading, I can see how others might. To label myself as the investor and the others as speculators is self serving and wrong. Furthermore, the notion that derivatives trading is driven primarily by speculation is fantasy. I can see plenty of reasons why a long-term, value investor may use derivatives to protect and augment his returns.

3. Liquidity costs: Even if we accept the premise that short term investors create noise and pricing bubbles, long term investors benefit from the liquidity they bring to the system. In fact, the markets where long term investing is most difficult are markets where there no short term investors. (Consider the market for fine art or even real estate.... Transactions costs inflate for everyone and insiders end up dominating the market)

4. Market mobility: As trading moves of exchange floors into ether space, it is difficult to visualize how a transactions tax will work, unless it is globally coordinated. All you need is one rogue player for the system to start coming apart at the seams. Krugman argues that the clearing systems for many of these markets are centralized and that the tax can be therefore collected at these locations. While this may work in the short term, how long will it take for an offshore location (say the Cayman Islands) to set up a competitive system? (It will cost money but the potential benefits from the system will be huge.) Once that happens, any chance of regulating these markets, even in sensible ways, becomes remote.

All in all, I think this is a dumb idea that should be throttled early in the process. I am sure that you will hear variants of the concept, and they will all share a common feature. They will try to focus the tax on what they view as the markets or securities that they view as most speculative and argue that only the entities in these markets will be affected by the tax. I don't think so. Ultimately, we will all bear the cost.