Should we buy weapons from Iraqis?

Payments for weapons handed over in Sadr City on Monday reportedly ranged from $5 for a hand grenade to $150 for an AK-47 to $2,000 for a highly specialized mortar. It appeared that both noncombatants and Mahdi Army insurgents were taking part in the buyback.

Abdulla Abu Ghassan, a bakery owner, received $1,200 after turning in a grenade launcher, an assault rifle and ammunition, all of which he said he had kept after serving in the now-disbanded Iraqi army.

The idea, presumably, is to take weapons out of the hands of terrorists. But will this work?

Under an optimistic scenario, the weapons are gone and the Iraqis are richer. Under a pessimistic scenario, the terrorists use weapons sales to finance future operations. If the terrorists have accumulated large weapons inventories, this should not be hard to do.

The dynamic problems are formidable. By bidding up the price of weapons, we raise the return to accumulating large weapons inventories in the future. We also make the weapons market more liquid. So weapons buy-back plans become less effective over time, unless we create the expectation that price will be falling. This means we had better get it right now, if indeed we can.

In some models government does best by accumulating its own inventory. The government then precommits to selling its stock, and driving down prices, if private traders accumulate too many of the forbidden assets. Knowing this danger in advance, private traders might be reluctant to invest in large stocks. Now I am not suggesting this as a viable policy option (hey, why don't we sell off some nukes from Colorado, and lower the world price? That would show those nasty North Koreans...). Nonetheless the scenario shows just how far buying weapons can be from an ideal, incentives-driven economic policy.

Addendum: David Nishimura offers more information.