My first thought upon reading that Goldman is restructuring bonuses for its top 30 executives was that this is designed to preempt much more extreme measures, like an American version of the British bonus tax. My second thought was, "Great, we should do this for every bonus on Wall Street, every year." (At least every bonus over $50,000 or whatever.) According to the Times, the bonuses will be paid out in special "shares at risk," which is basically stock that can't be sold for five years and can be clawed back if the recipient loses Goldman a bunch of money. But only this year. Next year, back to cash bonuses, apparently. 

And my third thought? That thought number two may trump thought number one. That is, Goldman may have intended this as a way of relieving pressure for more sweeping bonus reform. But the idea is so eminently reasonable it's hard to believe Goldman hasn't just massively increased pressure on itself (and every other major bank) to adopt this approach permanently, for all executives.

After all, I can see a principled argument for only leveling a 50 percent bonus tax this year (a la the British), since the banks have benefited from enormous government support of late, but presumably won't benefit from similar levels in the future. But what's the principled argument against permanently restructuring bonuses so as to better align the interests of executives with those of shareholders and taxpayers, or doing so for all executives? Why should those interests be aligned for this year's bonus, but not next year's? Why should they be aligned for executive number 30 but not 31? Once you've signed onto the idea that bonuses shouldn't create perverse incentives, as Goldman has, it's hard to see how you draw a line across time and people.