I wish this editorial wasn’t at CATO

“Conservatives” hate the unemployed so the facts on display in this editorial are obviously ideologically tainted. (h/t Wilkinson)

Look at the JOLTS data yourself: unemployment is being driven by a slow hire rate (fire rates are at about their normal level). Hiring rates could be depressed because of low demand for workers or for low supply of job hunters. GDP is rising which suggests demand for workers should be increasing. Unemployment benefits, on the other hand, have risen substantially so there are incentives to stay unemployed. Finally, we know the disincentive effect is large because of research by conservative stalwarts such as Larry Summers and Alan Krueger ((both somehow sneaked their way into high posts in the Obama administration)).

21 thoughts on “I wish this editorial wasn’t at CATO”

  1. Can’t productivity improvements displace demand for workers as well? Maybe that’s more of a large scale question that doesn’t specifically apply to the current recession.

    Glancing at some of the JOLTS summaries, the only areas where I see even a small rise in openings since 9/2008 are in finance and real estate. Everything else shows a drop.

    Conceptually, I can see how extending unemployment benefits can extend unemployment; I even know a few anecdotes to support this. If there are fewer jobs to go around, though, to stretch a metaphor: how does playing shorter music in a game of musical chairs result in fewer losers?

  2. Reynolds: “The unemployment rate has not been rising because of growing numbers of newly jobless people. Indeed, initial claims for unemployment benefits are way down.”

    Huh? They are over 50% higher than they were a year and a half ago.
    http://modeledbehavior.com/2009/11/12/new-claims-day-10/

    I echo swong; job openings are way down in the JOLTS data you linked to, and layoffs/discharges are up too – aren’t these the relevant driver for the search model that drives unemployment equilibrium, whether or not you can draw a job during a time period?

    I know you like focusing on the ‘individual decision making under uncertainty’ of individuals in these models, but why is the firm’s hire rate not endogenous? If there’s a mass of unemployed people, why not take some time to search out the most productive – why does it have to be a function of individuals? It could be suppressed based on the high level of job hunters. I think there’s some anecdotal evidence of this in places like the WSJ:
    http://online.wsj.com/article/SB10001424052970203872404574257983795638374.html

  3. “Can’t productivity improvements displace demand for workers as well?” This is a question about employment, not unemployment. (But the answer is I don’t know. It hasn’t happened in the last couple of centuries when their was dramatic technological change.)

    The problem with your musical chairs analogy is that the number of chairs adjusts to the number of people playing. Employers consider the number and quality of job candidates a job opening will produce when they decide whether or not to create a job opening.

    I’m glad you see the logic of unemployment benefits increasing unemployment. Because this story is based on sound logic, it has to be true. The real question is how important are the effects implied by the story. The answer, by every empirical study I’ve seen, is that UI benefits have substantial effects on unemployment.

  4. Right – you’ve blogged about this before. Layoff rates aren’t really higher than they have been, but hiring rates have dropped (like, according to the JOLTS data you linked to). The whole “clogged pool drain” thing.

  5. According to that site, job opening levels and rates are the lowest ever since 2000. Sounds like low demand for workers to me.

  6. ss, openings are endogenous. Look at the hire rate and the fire rates instead. If demand was low, you’d see more firing. Also, its hard to square increasing wages and productivity with a reduction in demand.

  7. Do you have data on rising wages handy? Ditto for productivity? I’m mostly curious.

    I’m pretty sure per-worker productivity has been rising pretty fast for the past decade at least. I’ve read at least a few articles saying that wages have been relatively flat for a little while, but data > mass media writeups…

  8. Also this one from the AP.

    Anecdotally – I have many more recruiters bugging me for part time and short term contract work than in the past. There are a number of plausible explanations for that that don’t have to do with the economy, though. The layoff survivors I know tend to have a higher workload now, without a wage increase to go with it.

    So why aren’t businesses capitalizing on this flood of cheap, willing labor?

  9. swong, bls.gov for data… look at the right hand side. gabe, CEO surveys and a dollar buy you a Coke. Besides, we know hire rates are low. The question is why are they low.

    Great queation swong. Why aren’t there more hires?

  10. Mike, the timing matters. Layoffs are down since these benefit increases went in. Similarly, claims have been declining for most of the year. I’m too lazy to check, but it would be nice to know when the UI extensions were made. I know there was a big one in February and one recently, but before that?

    Regarding search models: congestion is a strange (IMHO) feature of some models. I’ll throw my own anecdote out there… when I was involved in hiring we had a set date for the hire and we wouldn’t move it if there were more than the expected number of applicants. More applicants actually would push the process faster. I haven’t written down an explicit model, but I imagine UI extensions decrease the average quality of applicants (b/c applicant’s outside option is higher so they apply to more jobs they’re underqualified for). This reduces the number of openings because firms find them less profitable.

    BTW, just to show I don’t hate unemployed people: my preferred alternative policy would be to give people a lump sum when they lose their jobs. UI subsidizes being unemployed. If we think that number is too big, then we shouldn’t subsidize the production of it.

  11. I’m having a hard time finding solid numbers on it, but as of april, only 5.8 million out of 13.2 million unemployed

    http://www.usatoday.com/money/economy/employment/2009-04-09-unemployed-but-no-benefits_N.htm

    That’s less than half, a point I don’t see anywhere in the CATO article. The eligibility requirements are fairly strict for collecting UI.

    The lump sum transfer could work, but how would you prevent gaming of the system? I.e. firms fire, workers collect, workers get hired again? It’s a similar problem to a hiring tax credit. Also, any lump-sum reasonable for normal times would be insufficient for the current crisis.

  12. gabe, we’re talking about marginal effects here. If the article is correct, half of the unemployed would be eligible for benefits whether or not we were giving incentives to stay unemployed! You can argue that this goes to effect size, but your stat only makes Krueger’s result look that much more impressive.

    BTW, doesn’t gaming happen anyway? Up where I grew up, logging companies would “layoff” workers in the rainy season so they could collect UI. Anyway, don’t we have police to investigate fraud?

    I don’t get your last comment. We have poor relief don’t we? Unemployment insurance is insurance, not poor relief.

    Look, I’m all about helicopter drops… hand out money to people who say they need it, fraud be damned. Just don’t make the hand-out contingent on finding a job.

  13. I agree its about marginal effects. I was unaware that so few people qualified for UI, and it seems like an important point.

    Yes gaming happens anyway, but a lump-sum would make it worse. This would be higher than 100% of wages for some period of work (daily, weekly, hourly), so it would be much more prone to gaming than UI. Marginal Effects, right? We do have olice to investigate fraud, that doesn’t mean that a system that encourages layoffs and rehires can’t be criticized.

    We do have poor relief, but what I meant is that any lump-sum appropriate for times of full employment would be too small currently. If average unemployment spells are 4-week, than a lump-sum payment scaled to that duration would be too small for durations of 32 weeks. (Just making up numbers). I was just trying to get more details of this lump sum plan. (Would it vary over the business cycle or not), etc.

  14. California unemployment claims open for about a year, combined with a maximum disbursement cap – something like a set dollar amount in a two year period. To continue receiving money, you have to mail in a list of jobs you’ve applied to every week, including company and contact information. You also have to report any income you make from part time work, which is subtracted from your benefits.

    I was picked for a phone screen when I applied a few months ago – if I understood correctly, you are officially expected to look for work “full time” while you are collecting benefits, and you officially must offer an explanation if you turn down any offers (this is probably a little looser in practice).

    So in this state at least, it’s contingent on looking for a job, not finding one.

  15. The policy under consideration was extension of UI. Take the money allocated to that extension and distribute it among the unemployed (or those unemployed at some date before the passage of the bill) equally. I’m not making claims about optimal policy in “normal” times. I’m saying a lump-sum transfer is better, because of the incentive effects, than these UI extensions.

    BTW, the work requirement for UI eligibility is meant to get around the problem with gaming you talk about (e.g. firing and then rehiring).

    Optimal policy in normal times would be lump-sum-ish poor relief like the EITC, food stamps and housing vouchers (the last two ARE lump sum if they’re sent to everyone). These would also act as automatic stabilizers.

  16. swong, btw, I’m happy everything worked out for you in the end.

    I’d say, though, that UI benefits are contingent on looking like you’re looking for a job. If you’re looking for a job, then you look like you are, but there’s other ways to look like your looking for a job. For example, you could apply for jobs that you’re underqualified for…

    I should make the usual qualifiers about marginal analysis here: it just takes a few people to act in this way for the analysis to go through. Not everyone has to act on incentives for incentives to matter.

  17. Well, my total disbursement over a four week period amounted to a single check for exactly $60. My case may have been an outlier.

    I agree on the “appearance of job hunting” vs “job hunting” method of gaming the system.

    Application spam happens even in the best of times – it seems there’s a positive feedback loop between companies inflating job requirements and candidates inflating resumes. See also: degree inflation, positions requiring longer experience with a piece of software than that software has existed, etc. Total honesty offers no incentive whatsoever in either a job opening or an application.

    In other words, the market has evolved some really bizarre mating rituals. I should bring some shiny pebbles to my next job interview.

    It may certainly be the case that UI exacerbates this problem.

  18. Doug, I’m not sure what explains the increase in mass layoffs but a relatively constant fire rate. Unions may be part of the story since nearly half of those mass layoffs were in manufacturing. More likely is a trade story. There was this paper on voxeu a couple weeks ago trying to explain the drop in trade… maybe those authors know something about this mechanism.

    In any case, the underlying mechanism has to explain only a small uptick in layoff rates, but much higher mass layoffs. It might have something to do with the low voluntary separation rate right now. In normal times lower quality workers would separate themselves and try to find a better match. They don’t want to brave the labor market right now. Firms are having more mass layoffs events to get rid of these workers.

    I don’t think the “skewed” incentives of the increase in UI would explain this apparent contradiction in the data, though. My guess is that those incentive affects have a much bigger impact on the unemployed decisions about finding work then they would about firm/worker’s decisions about separation.

Comments are closed.