Monday, February 3, 2014

Unemployment

Another nice graph from Torsten Slok at Deutsche Bank

35 comments:

  1. Something must be wrong with the graph. Eliminating unemployment benefits reduces aggregate demand which reduces income and reduces employment. Check your sources!

    ReplyDelete
    Replies
    1. So if the data does not square with the theory, the data must be wrong?
      (ps: the source is clearly stated: Deutsche Bank Research)

      Delete
    2. Ha ha, I am afraid not everyone got the sarcasm Russ.

      Delete
  2. Do we get a counter-metric showing the labor force participation rate in North Carolina before and after the benefit termination?

    I imagine you get interesting information either way, but it would be instructive to know if unemployment dropped because the incentive to find employment was suddenly higher, or if the incentive to pretend you were looking for employment was suddenly absent?

    ReplyDelete
    Replies
    1. The answer is "yes".

      http://blogs.wsj.com/economics/2013/12/20/what-happens-when-unemployment-benefits-are-cut-north-carolina-offers-a-clue/

      http://blog.heritage.org/2014/01/06/examining-north-carolinas-falling-unemployment-rate/

      Delete
    2. The Washington Post has a better article showing a departure between academic interpretations of the Census employment surveys and the BLS's labor force estimates: http://www.washingtonpost.com/blogs/wonkblog/wp/2014/01/24/what-happens-when-jobless-benefits-get-cut-lets-ask-north-carolina/

      Delete
    3. The WSJ article that was posted was interesting - but it points to the explanation being largely a drop in labor-force participation. People simply giving up and withdrawing from the labor force is a problem, not a good thing.

      Delete
    4. Sounds like a combination of lower participation and more employment. But either way it points to the perverse incentives of a higher unemployment benefit.

      Lower participation isn't great, but I question whether those who dropped out were really 'participating' if all it took for them to stop looking for work was a smaller benefit...

      Delete
  3. According to Evan Soltas, YoY data indicates that the drop has been more due to labour force dropouts, but the December numbers are more favourable to the view that you are (presumably) endorsing. http://esoltas.blogspot.com.au/2013/12/more-on-north-carolina.html

    Still, this graph seems like an oversimplification of a fairly complicated issue.

    ReplyDelete
  4. Interesting post John, but how about looking at Employment levels in North carolina?
    http://research.stlouisfed.org/fred2/series/NCNA

    It seems to me that North Carolina is facing what the country as a while is facing; lower labor force participation rate. If this is true, the graph you have posted is in the face of lower unemployment benefits being the impediment to higher labor force participation/job growth.

    ReplyDelete
  5. Nobody's disputing that unemployment went down - they're saying it went down because people who had been searching for jobs as part of their unemployment gave up, which means they no longer count as "unemployed" (instead, they're now simply "out of the labor force"). Why don't you post a graph of the labor force participation rate alongside it? Or would that undermine your argument?

    ReplyDelete
    Replies
    1. Basically the same thing as what I thought of -- the whole chart depends on the "modern" definition of unemployment, not on the classic, more meaningful definition (the percentage of people who are unemployed).

      This is the first time I've been disappointed in one or John Cochrane's posts, as this kind of thinking (and definition of unemployement) supports our current, extremely opressive (massive spending on military and NSA spying on us) government.

      Furthermore, it gives libertarians a bad name ("you don't care if it costs more in transportation costs to get to a four-hour per shift, sub-minimum wage waitressintg job so long as you can mark down the person as employed!").

      We should be concentrating on the abuses of big government (massive military spending, domestic spying, etc.), not going for the low-hanging fruit that isn't going to solve the problem anyway.

      To quote an old saying (which alas seems to be forgotten lately): "Get the rich off welfare and the poor won' need it".

      Alas, these days the saying seems to be "Get the poor off unemployment just so we can feel like we did something"... as the country goes down the tubes.

      Neal Reynolds

      Delete
  6. Settle down everybody. If you might look at the blog post I purposely did not make any argument, draw any conclusions or anything else. Yes, duh, unemployment goes down because people stop looking for work, stop reporting that they are looking for work, move out of state, find jobs, go on disability, go back to school, give up, etc. etc. etc. Yes, lack of employment and the employment population ratio are tragedies. The links some of you have posted tracking down which it is are great.

    But the graph is really cool. It's an interesting fact (up to data limitations) how fast reported unemployment (looking for work) went down the instant benefits ended -- no matter where it went or whether one categorizes the transitions as "good" or "bad."

    ReplyDelete
    Replies
    1. I definitely noticed your lack of commentary and figured you were just enjoying the correlation without digging into the obvious analysis.

      We're all watching too much CNN when we try to label it "Good Thing, Bad Thing".

      Delete
    2. " If you might look at the blog post I purposely did not make any argument"

      With respect, that's just disingenuous given the labeling on the chart.

      Delete
  7. North Carolina politicos did not fund its unemployment insurance properly. Of all the states in the Richmond Federal Reserve District NC was woefully under funded in comparison to only two states worse off regarding prior funding. Hence going into the Great Recession North Carolina politicos had chosen to under fund.

    NC, at the advent of the Great Recession, blew through the meager funds. They had to almost immediately apply for federal loans to replenish the exhausted fund. And they kept taking loans without repaying. Then the interest began to kick in as they did not payback in a timely manner. The problem became a big campaign issue.

    Voters booted out the Democrats and voted in the Republicans, in both state houses (first time since 1898).

    NC was billions in the hole regarding owing the Feds with interest too boot. Hence they had to have a plan to pay back the Fed’s, build a fund after the payback and still offer unemployment albeit not extended.

    And now you know the rest of the story.

    ReplyDelete
  8. Should go negative any time now. When you give trillions to corrupt bankers and Washington Plutocrats, it goes to their favorite things. Hampton's real estate is up big. Black Rock now is one of the biggest owner of people's homes. Commodities get bought up by hedge funds backed by unlimited funded crooked banks. Emerging Market were bought lock stock and barrel with Fed printed going to the corrupt investment bank rulers! Renewables are funded well...with good amount of kick backs to the politicians...Heck Obama just installed wind energy money man who faithy rewarded Obama with cash, as Ambassador to Ireland. Capitalism is dead....Cronyism Rules!

    ReplyDelete
  9. Can you post the full study if there are no copyright issues?

    ReplyDelete
  10. go to http://economics.sas.upenn.edu/~mitmanke/NC_Case_Study.pdf if you want to read Hagedorn’s paper "Case Study of Unemployment Insurance Reform
    in North Carolina" cited in the Washington Post but a bad link

    ReplyDelete
  11. I understand that John hasn't made any arguments on purpose and my guess is that the point of the graph is more on what will happen to unemployment rate regardless of what happens to employment/population rate (which is important on its own given the stated forward guidance policy of the fed is based on unemployment rate rather than employment to population ratio, hence the conclusion of the graph is important for investors who are the primary audience of that presentation)

    In any case, here is an interesting paper on this exact issue:
    http://www.nber.org/papers/w19048.pdf?new_window=1

    ReplyDelete
  12. Kevin Erdmann has done some nice work on this at his blog Idiosyncratic Whisk and his graphs incorporate the LFP.

    I believe the latest post in the series was Jan 28:

    http://idiosyncraticwhisk.blogspot.fr/2014/01/north-carolina-ftw.html

    ReplyDelete
  13. Justin Wolfers, NC vs SC:

    https://twitter.com/JustinWolfers/status/430451652563177472

    ReplyDelete
  14. My company offered a technician job to a gentleman. However he decided not to come because the salary was not much more than the unemployment benefit he was receiving. Same thing happened years ago several times. I think the benefit is too good to motivate people to work. The NC case makes clear sense to me.

    ReplyDelete
    Replies
    1. Sounds like the salary you offered was not good enough to motivate.

      Delete
    2. It doubles the state minimum wage and is good enough to support a family.

      Delete
  15. My answer to your picture-post is Germany. They have very generous unemployment benefits and their unemployment rate is 5.1%. How generous? First year 80% of the salary. Then the benefits decrease every year, but they never expire. If I remember correctly, they are about 250 euro per month at a minimum after several years of slow decrease. I hope John can explain how it is possible to have such a low unemployement and such generous, never-ending unemployment benefits. Good luck!

    ReplyDelete
    Replies
    1. "First year 80% of the salary. Then the benefits decrease every year, but they never expire."

      Untrue. It is 60% of your average monthly salary of the last 6 months if you have no child (max 67% with child) for a maximum of 24 months if you are 58 years old at least (less if you are younger). The key element is: each time you decline a job offer, your benefits are suspended for 12 weeks.

      Delete
    2. You are right. It is only 60% and 67% with a child (€ 2964 max). But after 12 months, if a person is not eligible to receive full unemployment benefits, one can apply for Hartz IV benefits. They are open-ended and are € 382 (2013) + the cost of adequate housing (including heating) and health care. (Source: Wikipedia) Compare this with South Carolina!

      Delete
  16. This is an interesting graph. Before people start reading too much into this graph concerning the effects of the termination of the Emergency Unemployment Compensation (EUC) program in NC last July, they should probably consider two things.

    First, BLS labor force statistics at the state and county level are constructed differently than the national labor force statistics from the Current Population Survey (CPS). The CPS is a sample that includes about 110 thousand observations for the US. Once you extract state and local subsamples, with a few exceptions, you end up with small samples and a lot of sampling uncertainty. For this reason the BLS treats statistics from local subsamples of the CPS as noisy signals of the ‘true’ labor market statistic. Essentially, the BLS solves a signal extraction problem, assuming that the true statistic is related in a stable manner to another statistic that is more reliably observed. (BLS Handbook of Methods, Ch 4, http://www.bls.gov/opub/hom/) In the case of unemployment that variable is the number of people receiving unemployment benefits. In June 2013 the number of regular unemployment claims paid was about the same as the number of EUC claims paid. In July the number of EUC claims paid was only one tenth of the June EUC numbers. So unless the BLS introduced a structural break in July of last year in its statistical model relating true unemployment to UI benefit recipients in NC, the decline in the unemployment rate featured in the graph just reflects the EUC recipients being phased out in NC.

    Second, sampling uncertainty for the CPS state subsample is quite large, even for a state of the size of NC. One of the previous comments mentions the note of Mitman et al (http://economics.sas.upenn.edu/~mitmanke/NC_Case_Study.pdf ) who calculate the NC unemployment rate based on the NC subsample of the CPS up to December 2013. In the most recent version of their note, Mitman et al now provide the standard deviation for the sampling uncertainty of the NC subsample of the CPS, about 0.6%. That means that for the period July to December 2013, only the December unemployment rate is outside a two standard error band of the June value. On the other hand, the national unemployment rate has been two standard deviations below its June value since July. So I am not sure how informative the NC CPS subsample is about the effects of the EUC expiration in NC.

    ReplyDelete
  17. "First year 80% of the salary. Then the benefits decrease every year, but they never expire." Really! It cannot be true unless Germany is wonderland. Where did you find it? What is the other side of the coin?

    "If I remember correctly, they are about 250 euro per month at a minimum after several years of slow decrease". Can you live with only 250 euro per month?

    ReplyDelete
  18. Kevin Erdmann is at it again with a great graph re: North Carolina (and the rest of the US)

    http://idiosyncraticwhisk.blogspot.fr/2014/02/the-end-of-eui-in-north-carolina-and.html

    ReplyDelete
  19. "But the graph is really cool. It's an interesting fact (up to data limitations) how fast reported unemployment (looking for work) went down the instant benefits ended -- no matter where it went or whether one categorizes the transitions as "good" or "bad". "

    Turn it the way you want - but the insinuation "less benefits => more (of these lazy fellows got to) work" is not only bad style, but bad economics!

    Si tacuisses - philosophus mansisses!

    ReplyDelete
  20. An unemployed person has to certify that he is looking for work in order to collect unemployment checks. As soon as the checks stops, he stops lying, and is no longer counted as being in the labor force. The size of the labor force drops, and the unemployment rate falls because the number of people working represents a larger share of the reduced labor force, even if nobody gets a job.

    ReplyDelete
    Replies
    1. Now that's the best explanation of the lot!

      Delete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.