Does AIG need to retain certain employees?

There are two sets of arguments about the AIG bonuses. One set of arguments concerns the abstract question of whether enacting a retrospective tax on a small number of companies is a wise, appropriate, or legal thing for Congress to do. The other set of arguments concerns the very concrete question of whether the bonuses actually serve some valuable social purpose. In this case, the supposedly valuable purpose is retention. At least, that's a "valuable purpose" if you buy the theory that particular AIG employees have a unique knowledge of the company's contracts, and are needed to unwind those positions.

Simon Johnson and James Kwak do not buy that theory. They don't even window shop that theory. And, while they've expressed their frustration with the theory before, I see they've now channeled that frustration onto the New York Times editorial page:

If A.I.G. wants to argue that complex transactions, hedging positions and counterparty relationships require employees who are intimately familiar with those trades, it should at least provide evidence that the arguments for doing so are sounder than the ones made in Indonesia in 1997, when leading bank-owning conglomerates claimed that only they understood their financing arrangements, which certainly were complex. Or the Russian bankers in 1998 who were convinced that only they and their friends could possibly close the deals that they had taken on. We heard variants of the same idea in Poland in 1990, Ukraine in 1994 (and in the Ukrainian crises subsequently), and Argentina in 2002.

Any grain of truth in these arguments must be weighed against the costs of allowing discredited insiders to manage institutions after they have blown them up. Even if the conclusion is that a few experts need to be retained, offering guaranteed bonuses to virtually the entire operation is hardly the way to achieve the desired results. We should not let people think that the best way to guarantee job security is to lose lots of money in a really complicated way.

I think this argument is pretty good, but has a couple of problems, especially when used as support for the House's tax bill. (I'm not in a position to judge how apt we should consider Kwak and Johnson's analogies to international banking crises, so I'll leave that out of it.) In general, I'm not sure why this argument shifts the burden of proof to AIG. Determining that AIG doesn't need to retain its employees is a necessary but not sufficient reason to strip away the bonuses. You also need to engage with the first question: Is a tax on bonuses a wise thing for Congress to do?

In fact, I think Johnson and Kwak's argument suggests that the House version of the tax is not the right thing to do. I agree with Johnson and Kwak's argument that blanketing the entire company with bonuses isn't the best way to retain particular employees. But this logic cuts both ways: If your argument is that certain "discredited insiders" don't need retention payments, then you aren't offering a reason to support a blanket tax on bonuses that falls across every class of employee at many different companies. You are offering a reason to fire the particular discredited employees at AIG.

Finally, I just don't share Johnson and Kwak's high skepticism of the retention argument. (I'm skeptical, but not that skeptical.) Kwak and Johnson make it sound like the individual greedy employees are making the case for their own enrichment. But since we still don't know who any of these employees are, they obviously aren't the ones arguing on behalf of their own bank accounts. As far as I can tell the strongest advocate for retaining certain employees is AIG's CEO Edward Liddy. But Edward Liddy has been CEO for less than six months and was appointed by the government. He has a salary of $1 and no bonus. And he is obviously making the bonus argument at great cost to his reputation -- which he claims is his only stake in the matter. Shouldn't that make us less skeptical of his arguments?