Friday, October 15, 2004

Reviewing the Comments on Unemployment

We have had a few more comments on the Krugman posts below. They are interesting questions and merit more discussion. A comment on the original post early this morning asked:

Are you sure you understand the precise definitions used in the household survey because you seem confused in the comment above?

My original post listed the definitions of discouraged and marginally attached workers that form the basis of the U-4 and U-5 measures of the unemployment rate. I wouldn't say that I was confused about the definitions, since I included them in the original post. The issue is whether Ryan's (an earlier commenter) situation would be classified by the CPS as being discouraged. I thought that it would. Here is how Ryan described his situation:

I was let go and looked for work for 6 months before deciding that what I really needed to do to get a decent paying and more stabile career path was to go back to school. If I had been able to find a job, I'd be working.

The original commenter commented again on yesterday's post and stated:

If I give up looking for work to go back to school because the labor market sucks then I am not counted as discouraged, marginally attached, or part-time due to economic reasons and yet I have most assuredly "stopped actively looking for work and therefore dropped out of the unemployment statistics."

As I noted, that's far from clear to me. Someone who goes back to school is out of the labor force. My conjecture, however, is that this does not mean that he is not also discouraged or marginally attached if he did so, as in Ryan's case, only because he couldn't find a job. If he is discouraged, then he adds to both the numerator (unemployed plus discouraged) and the denominator (labor force plus discouraged) of U-4. His situation would raise that measure of the unemployment rate, and the fact that this measure of the unemployment rate has fallen by the same amount as the standard measure (counting discouraged workers as out of the labor force) would invalidate Krugman's point. Since this is a question of interpretation, I e-mailed and asked for guidance. I'll post with any answer that I receive.

Both new comments have mentioned the employment-population ratio, with the original commenter kindly sending me the link to this post from January 6, 2004, at Brad DeLong's website. I acknowledged in my original post that Krugman was correct in stating that the employment-population ratio was unchanged since June 2003, and Brad's post graphs this ratio since 1950 and notes how much it has fallen from its historical high since the start of the recession.

However, the employment-population ratio can fall not only because some people may become discouraged but because some people voluntarily choose to leave the labor force of their own accord. This is why the fact that it has not increased is not enough to support Krugman's point. My wife was working full time in 2000, working part time for non-economic reasons in early 2003, and is now voluntarily out of the labor force. Someone else is doing the job she used to do, and that ultimately allowed another person to become employed in her place (i.e., someone was hired to replace her, then someone was hired to fill her replacement's old job, etc., until someone who was unemployed or not in the labor force became employed.)

There remains no piece of evidence to support Krugman's statement that "...unemployment declined only because some of those without jobs stopped actively looking for work, and therefore dropped out of the unemployment statistics." He cannot use the word "only"--that language is falsified by the presence of even a single person with my wife's job history whose departure made room for a new person to become employed. He says "stopped actively looking for work." The BLS allows for alternative definitions of the unemployment rate that accommodate the presence of people who are not "actively" looking for work and are thus not in U-3 but are in U-4, U-5, or U-6. That all of these measures have declined since U-3 peaked in June 2003 casts doubt on his statement more generally.

As I have laid it out here, Krugman's argument (cutting him a break and substituting "primarily" for "only") can only be right if there is evidence to show that 1) there are people who have dropped out of the labor force involuntarily in such a way that they would not be counted as discouraged, marginally attached, or working part-time for economic reasons, and that 2) the number of these workers substantially exceeds the number of people who left voluntarily for reasons unrelated to the job market.

There is no empirical evidence for either piece. Brad provides a good discussion, raising some possibilities, and asks the question, "Why? What's happened to change the relationship between changes in employment and changes in the labor force? And what does it mean?" He concludes with, "It is a mystery to me." Paul Krugman, in a highly partisan op-ed, asserts a particular answer that is in direct conflict with measures of unemployment that are designed to address the question.

A fact-checking procedure at the New York Times would have required him to cite evidence for his assertions before putting them in print. He would have had to prove his point, rather than (at best) making a harsh sounding statement with enough ambiguity so that, ex post, one could find a particular interpretation and some circumstances under which it may be true. This sort of fact-checking is one of the reasons why publications from, say, The Brookings Institution, are so highly regarded. In articles or book chapters that I have written for them, they put their authors through that test, and that makes the publications better.

That the New York Times does not see fit to do so is one of the reasons why the mainstream media is losing credibility, and alternative media like blogs--where people like me and people who obviously start with different perspectives from mine can figure things out together and check their arguments--are growing in importance.

Thanks again for your comments.


rustbelt tom said...

While economist argue over the household versus payroll survey and other statistical mishmash, there is severe slack in the labor force, and especially in the "rustbelt."

People cannot eat statistics, a point apparently lost on the Bush administration.

John Snow gave a speech this week in northwest Ohio about our terrific economy, while northwest Ohio continues to set records in personal bankruptcies and foreclosures. The Dems are pouncing on that.

Is our economy healthy if one major region is, for all intents, still in a recession (conceding the rustbelt may not fit the economist's definition of a recession)?

Anonymous said...

Ok, the time for being polite is done. You're wrong about Krugman and you are showing yourself to be ignorant. You owe him an apology. Also, the tightness available in an academic publication of the Brookings Institution is not available int he NYT op-ed. You have spent many more words than you would be allowed in the NYT already defending a proposition and you are still wrong.

1. You listed the original definitions, but you failed to understand them. Someone going to school full time will not be listed in any of those definitions. No matter what you wish those definitions to mean or what you think they mean, they don't mean that. Check with you colleagues at Dartmouth or any friends you have at the BLS about how the Household Survey is conducted. Again, You're wrong and your insistence that this may be a matter of opinion is wierd.

2. You use one piece of anecdotal information to try and refute Krugman, but you misconstrue what he said. He did NOT say that evey person who dropped out of the labor force did so because they stopped looking for work. What he said was the drop in the unemployment statistics was only due to a set of people dropping out of the statistics. This can be true even if your wife dropped out of the labor force as you described. The person wh took her job may not have been unemployed. That person may have been reentering the labor force so even though the job did not "disappear" no net employment gain happened.

3. The mystery is easy to solve. The survey is not designed to reveal fine distinctions about why people move into and out of the labor force. The methodology is designed so that it is repeatable and so that comparisons can be made between time periods. The level of detail that would be required to make determinations as to whether someone dropped out of the labor force due to economic reasons is not possible. It's self-reported and is for a specific tiem period. If you decided to go back to school then you are not going to seek out jobs right then even if the decision was based on your previous inability to find a job. Popular accounts of Prescott's Nobel prize suggest to me that he theorized a similar mechanism.

Anonymous said...

Hi Andrew,
Just read your post and appreciate the intelligent perspective you are bringing to gaining some insight into the (rather complex) unemployment statistics.

Furthermore, with your obvious willingness to carefully consider your readers comments, and modify your position or correct your facts should something be incorrect, you wonderfully illustrate the point in your final paragraphs about the rise of blogs and the (evolving) reduction of influence of the mainstream media.

Kirk from Colorado