Yet the Senate went home for the holiday weekend without extending [unemployment insurance] benefits. How was that possible?I prefer to be charitable. Ignorance (I guess that is the clueless and confused part) is a good explanation for a lot of behavior, and is easily cured with education.
The answer is that we’re facing a coalition of the heartless, the clueless and the confused.
Why is unemployment insurance (UI) provided by the government, and not by the private sector? To my knowledge, and perhaps surprisingly, economists who analyze UI systems typically don't answer this question. I think the answer is closely related to why (in my opinion) a well-designed government-run health insurance program works well. In the UI business, standard problems of moral hazard (behavior of the insured affects the chances of suffering an insured loss) and adverse selection (the insurer has a hard time differentiating bad insurance risks from good ones) are severe. The insurer of unemployment cannot observe how much effort the unemployed put into searching for work, and cannot determine whether the unemployed are excessively choosy in terms of the jobs they will accept. As well, as with private health insurance, a market in private unemployment insurance would likely degenerate, by way of the adverse selection problem, to a state of affairs where only high-risk (for unemployment) workers buy insurance policies, and the high prices of the insurance keep the low-risk workers out of the insurance pool. This argument is far from airtight, though, as we now have to ask why the private market seems to work in terms of auto insurance and homeowners' insurance, but not for health and unemployment. However, let's go on.
If we accept that the government should be providing unemployment insurance, we also have to accept that there are some frictions in loan markets that prevent unemployed people from smoothing their consumption over time in the face of unemployment shocks. Credit market frictions, or what we think of as credit market constraints, can make government tax policy matter. For example, a government transfer - unemployment insurance benefits - can increase economic welfare, in spite of the fact that the government will have to finance the transfer by increasing someone else's tax burden, either now or in the future.
Thus, we think that governments can improve welfare by providing unemployment insurance, and there can be a cyclical role for generous unemployment insurance when aggregate economic activity is low and credit market frictions are more severe. Unemployment insurance is targeted toward people who are credit constrained, and therefore arguably a more efficient policy tool than broad tax cuts. However, the government is still faced with the severe moral hazard and adverse selection problems inherent in all UI programs, and needs to design these programs so as to provide the correct incentives. It is well-known that more generous unemployment insurance increases the average unemployment rate - this is well-documented in data from Europe and Canada relative to the US, for example, and is consistent with standard search theory.
The theory of dynamic incentives, as applied to UI, in work by Shavell and Weiss, Hopenhayn and Nicolini, for example (see this paper for the references) tells us that the UI system in the US is likely far too stingy. Generally, the idea is that replacement rates (the ratio of the UI benefit to income when working) should be higher, and benefits should continue for much longer, declining over time. In normal times, UI benefits in the US conform to a replacement rate of about 50% for 26 weeks, then go to zero. In most developed countries, UI is much more generous. For example, in Canada, UI benefits continue for longer, are geared to local labor market conditions (probably a bad idea, actually), and include maternity leaves, parental leaves, compassionate leaves, and illness.
What's the conclusion for current policy in the United States? Krugman is right, in that extending UI benefits under the current conditions would be a good idea. Incentive problems matter, but theory tells us that we should cut the unemployed some slack during a downturn. For the long term, reform of UI in the United States would be useful, but as usual, that would be complicated. UI is in part governed by a framework of Federal law, but the details and funding are worked out by the States. Good luck with that.