'Our Stimulus is Making the Unemployment Rate Higher'

The Senate is close to passing a new $150 billion stimulus package that would extend jobless aid for up to a year. Over the past few days I've written quite a bit about why I think extending unemployment benefits is a no-brainer. For Cato Institute employment expert Alan Reynolds, it's also a no-brainer -- that we should stop extending them.

Before we get to the chat (and it's a long chat!) let's look at the raw numbers of the worst jobs crisis in 60 years. We've lost 8.4 million jobs -- the equivalent of the entire population of Iowa and Minnesota combined -- since December 2007. About half of that number -- say, the population of West Virginia and New Mexico -- have been without a job for more than 27 weeks.

I say getting stimulus money into the hands of the cash-poor is an easy call when private demand is down and jobs aren't hiring. These folks will immediately spend the next dollars on cable, clothes, and essentials at a time when broad consumer confidence is at a year-low. Depressed economies need more churn and unemployed Americans are most likely to churn the money they receive quickly.

Of course, there is a downside to all of this poor cash. Subsidies encourage behavior. If you give money to people who are unemployed, you risk encouraging unemployment. The Congressional Budget Office acknowledges this danger, but it still calls jobless aid the most effective way to stimulate job creation. Yes,that means better than tax credits, better than tax cuts, and better than infrastructure spending.

Not everybody agrees.

"Statistically on average, we know that the intensity of search picks up when you're about to run out of benefits," Reynolds told me. The evidence is overwhelming that the stimulus added at least two percentage points to the unemployment rate by keeping Americans at home rather than pushing them to work, he told me (for more background, see his column for the New York Post).

I responded that search intensity doesn't matter as much in this recession because employers weren't hiring, anyway.

"The ratio of unemployed Americans to job openings has been at an all-time high of six-to-one for six months," I said. Why would it matter if we were discouraging Americans from looking for jobs that didn't exist? Reynolds said he hadn't seen the information I was citing. Here's what I sent him after our conversation (all data from the Bureau of Labor Statistics):

In September, unemployed Americans exceeded job openings by the largest ratio on record: 6-to-1. Three months later it was still growing. In January, the ratio fell to 5.5-to-1. That's the latest available data. The table below shows job openings and total unemployment (both listed in thousands) in January 2009 and through the last six months. The unemployed-to-openings ratio for each month is on the right:

bls unemployment data.png

I sent this chart to Reynolds. Here is the response he emailed back. My response is below:
"Economics suggests there no such thing as a fixed number of job openings. It depends on the price -- the after-tax cost to employers and the after-tax benefit to potential job seekers.[1] Taxes and transfer payments (UI benefits) can both put a wedge between employers and workers ...

Your table shows job openings up in December and early January and unemployment down since October. So, the ratio fell from 6.24 to 5.48 in a very short time.[2] More to the point, because a large number of those identified as more-or-less unemployed are either not hustling much or holding out until the last minute for a better job, then the denominator of the ratio would fall quickly if the maximum duration of extended unemployment benefits was reduced from 99 weeks to, say, 52 weeks.

Even with mediocre GDP growth of about 3% (very weak for the early stages of recovery), the unemployment rate would then drop to about 8 percent within 9-12 months, and the average duration of unemployment spells would fall dramatically too. Would those who got off benefits in one year rather than two be worse off? I quoted the 2007 OECD Employment Report saying subsidizing long spells of joblessness is typically bad for the future incomes of those we're ostensibly trying to help."[3]

[1] I agree that giving money to somebody without a job marginally reduces his incentive to look for work, but I'm not convinced that it should discourage employers from listing job openings. On the contrary, the government has cut taxes for the vast majority of workers and employers and offered numerous incentives like non-recourse loans to small businesses throughout the last year. Unless America's employers are steeped in Ricardian equivalence theory, I can't imagine why the government's stimulus policies would make them less likely to hire, especially with GDP growth strongly in the black. Yet, lo and behold, here the employers are, overwhelmingly not ready to hire.

[2] The jobless-to-openings ratio has fallen. I'm happy for that. Job openings are up 9 percent from their record low in 2009. I'm happy for that, too. But if Reynold's argument is that Americans are "not hustling" or "holding out" on jobs that are available, we should see the job openings number start to really pick up, right? Instead, it's a nose ahead of its all-time low. I see no evidence that unemployed Americans are increasingly passing up jobs to rest on the federal dole.

[3] Reynolds is right that unemployment benefits subsidize joblessness. I am willing to concede that at the margins, it might dampen some Americans' energy to seek out jobs. But the families in this Washington Post article are making $1,200 and $1,650 a month and supporting four and six children, respectively. In other words, the federal government is spending $240 per family member per month. These men have every incentive to look for more work, and they are. Moreover, their meager spending is a part of depressed private demand. I cannot imagine how canceling their benefits, which would further reduce personal consumption, would have a positive effect on demand or company revenues -- much less lower the unemployment rate by two percentage points and create 3 million net jobs in two years.

This has gone on long enough, so I'll try to sum it up my argument in a few sentences. The primary reason 15 million Americans are still out of work is that employers are not hiring. Employers are not hiring because there is no demand for their products and services. One way to juice short-term demand for their products and services is to put money in the hands of Americans most likely to spend their next dollar and get money churning throughout the economy, which can be paid back to the government in the form of taxes from those transactions, and from future transactions when we've achieved healthier employment and sustainable long-term growth.