Lump Sum Unemployment Insurance

I’m not quite sure what Mitt Romney had in mind last night when he suggested replacing Unemployment Insurance with “personal accounts,” but it put me in mind of the system they use in Austria as described by David Card, Raj Chetty, and Andrea Weber in “Cash-on-Hand and Competing Models of Intertemporal Behavior: New Evidence from the Labor Market.”

The basic conservative take on Unemployment Insurance is that it increases unemployment by, in effect, paying people to not take jobs. Progressives counter that Unemployment Insurance is a high-multiplier form of Keynesian stimulus since unemployed people have a high propensity to spend their checks. What’s more, there are obviously humanitarian reasons you might want to hand out UI. One way of squaring the circle would be to make UI a bit more insurance-like and say that there’s a lump sum payment you get associated with job loss. Then you can do whatever you want with the money. Get a new job next week, and you have a windfall. Stretch it out and keep looking for three months. Whatever. This ought to have the claimed benefits for UI without the claimed harms. In Austria, they have a system like this. Except it only applies to workers who’ve amassed at least three years of experience at their previous employer. That discontinuity allows for a helpful empirical study of the impact of lump sums on worker behavior as follows:

Whoops! You actually get the reverse of the result predicted by the conservative theory. Even though the lump sum payments have no employment-disincentive effect, recipients of lump sum payments take longer to find jobs. It seems that the real issue here isn’t the incentives at all, but the impact on liquidity (full argument here). It seems to me that switching to a lump sum system might be a good for separate reasons. But the evidence from existing lump sum systems seems to mostly be that UI doesn’t have the incentive effect that Romney seemed to be worried about.