I would like to offer a few thoughts on extending unemployment insurance.

Supply-side effects & the tradeoff

There are negative supply-side effects from providing unemployment insurance (UI) benefits. The best estimates I have seen suggest the current 9.5% unemployment rate is 0.5 – 1.0 percentage points higher than it would otherwise be because of previously-enacted expanded and extended UI benefits. I will start by using the bottom end of that range (0.5).

I use 5.0% to represent full employment. We are 4.5 percentage points above that. If we did not have expanded UI we would be 4.0 percentage points above full employment. That means for every 9 people out of work, one is being discouraged from taking a new job because of the expanded benefits (0.5 / 4.5). Said another way, eight people who would like a job but cannot find one are getting more generous UI benefits for each person who is getting those same benefits and choosing not to take a new job. We have to make a tradeoff between our desire to help those who want a job but cannot find one and those who would choose to stay unemployed while they have extra benefits.

The judgment call for policymakers: does an 8:1 ratio make a UI extension good policy? I say yes. If, however, the supply-side disincentive is a full percentage point, then we have a 3.5:1 ratio. That is a tougher call, but I would still say yes. Given the range of possible supply-side disincentives, I would recommend extending UI benefits when the unemployment rate is 9.5%.

I assume that most everyone would agree that at full employment it is foolish to provide more generous UI benefits. So somewhere between 5.0% and 9.5% there is a breakpoint at which the supply-side disincentive is not worth the compassion benefit of providing aid to others who want a job but cannot find one.

At a 7% rate our ratio is between 1:1 and 3:1. At an 8% rate it’s between 2:1 and 5:1. My breakpoint is around 8%. I would support a (paid for) UI extension as long as the rate is 8% or above. There is nothing magical about this judgment, and yours may differ.

Stimulus effects

The fiscal stimulus argument in favor of extending UI benefits is silly. The numbers are too small. The pending legislation would spend $33 B over the next year on UI benefits. In a $14+ trillion economy, that’s less than one-quarter of one percent of GDP. When you’re running a 10% deficit, another 0.2% is not meaningful. Sure you can argue that “every little bit helps,” and yes the unemployed are likely to spend almost all of the assistance they receive. But the reason to extend benefits is to help those people who would receive them, not because the cash will significantly accelerate economic growth and benefit everyone else. If this $33 B of spending were concentrated into a month or two it might be barely big enough to register, but it would be spread out over the next year. $3 B per month is a lot of money, but not when compared to a $1+ trillion per month economy.

For instance, the second (bolded) phrase of this statement by Speaker Pelosi is a huge stretch: “These benefits help struggling families make ends meet, boost consumer demand to spur hiring by small businesses, and strengthen our economy as a whole.” This effect is so trivially small that no one could reasonably claim to estimate or measure it.

The Republican position

While a few Congressional Republicans are arguing that UI benefits should not be extended, the overwhelming majority are for an extension as long as the deficit impact is offset. The $33 B deficit increase could be offset immediately by other spending cuts or tax increases. For those who ignored my last point (too small to matter from a macro fiscal stimulus perspective), you could easily offset the $33 B by enacting changes now that would reduce spending by $33 B beginning three or four years from now. This eliminates the “contractionary” excuse for not offsetting the spending increase.

There is overwhelming bipartisan supermajority support for a UI extension if it is offset. Speaker Pelosi and Leader Reid chose not to follow this path, presumably because it would split their side of the aisle. They instead chose a partisan route that resulted in partisan stalemate, and we are now in the midst of a traditional blame game. Do not be fooled into thinking this is a debate about whether to extend UI benefits. It is instead a debate about whether that increased spending should be offset by spending cuts or instead increase the deficit.

Reducing the supply-side disincentive

There is a simple way to reduce the labor disincentive from expanding and extending UI benefits. In 2003 President Bush proposed personal reemployment accounts as a substitute for expanded unemployment insurance benefits.

The program would have spent $3.6 B to provide 1.2 million unemployed people who are least likely to quickly find a job with a $3,000 account. That person could make regular incremental withdrawals from the account for reemployment services: job training, search services, transportation costs, child care, or even getting a suit cleaned for interviews. If the person started a new job within 13 weeks of receiving his first UI payment, he could keep any balance in the account as a cash reemployment bonus. If he did not, he could continue making regular withdrawals from the account to use for reemployment services while receiving UI benefits.

This bonus feature would reduce the incentive for workers to remain unemployed so they could keep receiving benefits. A worker who takes a job today will get part of that future stream of benefits in a lump sum payment.

Unemployed workers are not a uniform population. Some need help with their job search, while others need retraining. Still others need transportation or child care while they look for work. Rather than having government determine how to allocate these funds for the unemployed, a PRA would allow a worker to decide how best to use his account funds for any purpose related to finding a new job.

Reemployment bonuses have been tried in the past. In the 1980s experiments were conducted in Washington, New Jersey, Pennsylvania, and Illinois. Workers were provided with reemployment bonuses of between $300 and $1,000 if they found a job before their UI benefits expired. These bonuses reduced the amount of time workers remained unemployed by about a week, and the new jobs they took were comparable to those who did not get bonuses.

In the recession of 2003 Congress rejected (ignored, really) President Bush’s PRA proposal on a bipartisan basis. At the time a cynical legislative expert friend said, “Republicans hate personal reemployment accounts because they know they won’t work. Democrats hate them because they know they will.”


  • Congress should extend unemployment insurance benefits as long as the unemployment rate is above 8%. I would extend the law for six months and then reevaluate.
  • Congress should offset the deficit increase from this new spending by cutting an equal amount of spending, effective three or more years from now to remove the silly anti-stimulus argument.
  • Congress should allow States to instead use the additional funds to experiment with personal reemployment accounts for those who are hardest to reemploy.