Wednesday, March 05, 2008

What the Unemployment Rate Misses

In his Economic Scene column in today's New York Times, David Leonhardt discusses the challenges of measuring unemployment and using the unemployment rate to assess the state of the labor market. In a nutshell, we have a fairly low official unemployment rate and yet many people not working. In this excerpt, he focuses on a distinction that his colleague Paul Krugman once glossed over (to much fanfare in my first month of blogging):

There are only two possible explanations for this bizarre combination of a falling employment rate and a falling unemployment rate. The first is that there has been a big increase in the number of people not working purely by their own choice. You can think of them as the self-unemployed. They include retirees, as well as stay-at-home parents, people caring for aging parents and others doing unpaid work.

If growth in this group were the reason for the confusing statistics, we wouldn’t need to worry. It would be perfectly fair to say that unemployment was historically low.

The second possible explanation — a jump in the number of people who aren’t working, who aren’t actively looking but who would, in fact, like to find a good job — is less comforting. It also appears to be the more accurate explanation.
As we discussed briefly then, the BLS does collect measures of unemployment that progressively relax the condition that individuals have to be actively looking for work. Leonhardt characterizes them as "broader but not especially useful." I don't think they should be dismissed so readily. They are found in Table A-12 of the monthly employment report. You can get the historical data here. Let's go to the picture.

The 4 curves are as follows:

  • Blue: The conventionally measured unemployment rate, currently at 5 percent and low by historical standards. This is the number unemployed divided by the total number in the labor force (employed plus unemployed).

  • Red: Add people classified as discouraged workers--those who have given a job-market related reason for not currently looking for a job--to the unemployed. The increase is very slight--historically between 0.1 and 0.4 percentage points.

  • Yellow: Add people classified as marginally attached (beyond being discouraged)--those who currently are not looking for work but indicate that they want and are available for a job and have looked for work sometime in the recent past. This currently adds 0.8 percentage points to the unemployment rate, which is typical of the full 14 year time period.

  • Green: Add people classified as employed part time for economic reasons--those who want and are available for full-time work but have had to settle for a part-time schedule. This number is currently 9 percentage points of the labor force (augmented to include those marginally attached or employed part time for economic reasons).
The last measure seems to be a pretty good measure of labor underutilization. What does it tell us about what the conventional unemployment rate misses? In April 2006, the gap between the two shrank to 3.4 percentage points, compared to 4.1 percent today. The latter figure is about the size of the gap that prevailed around the recent peak in the unemployment rate in 2003. The gap was greater than 4.1 percent in most of the months shown prior to 1997. The gap was as narrow as 3 percentage points as the unemployment rate reached its lowest values in 2000.

The more comprehensive measures of labor underutilization are available and are consistent with the story being told in the article, though you have to get to "employed part time for economic reasons" to get much of a gap. I think they would be more "useful" to journalists if journalists chose to report them.

Barry Ritholtz also comments on the story and refers back to a measure of the "augmented unemployment rate," which doesn't include the economic part timers but also doesn't require that those who want a job have actually looked for one. (This information can be calculated from Table A-1 of the monthly employment report.) At present, there are about 5 million who "want a job" among the roughly 80 million who are not in the labor force, or about 6.25 percent. This proportion has stayed around 6 percent for several years.


Anonymous said...

Women are dropping out of the paid workforce in large numbers.

spencer said...

Time after time we see press reports of hundreds of people lining up to apply for jobs at a newly opened Wal-Mart. Some people like to see all these jobs applicants as a measure of how great Wal-Mart is. Maybe, but rather i view them as a measure of the concealed or hidden unemployment and underemployment in the overall economy.hundreds of people apply for a mediocre job at Wal-Mart is a signal of how our system is not functioning very well.

Anonymous said...

The obvious correlate to look at is wages. Stagnant wages in a growing economy say something.


Anonymous said...

There are 1.5 million real estate agents who are seriously underemployed, not counted in these numbers because they are officially employed. Also my friend the independent mortgage broker, I notice he's no longer showing up at Starbucks in the morning for a latte.

Other self-employed types are taking a huge whack as well . . . those in construction, my dentist (people put off going when the going gets tough), veterinarian (ditto, Buddy isn't getting his annual tooth cleaning), etc. This is a large class of (now underemployed) people.

Ivan Kitov said...

The figure you presented is a good illustration of the political flavor of your statements. You just want the number to be as large as possible. On the other hand, all curves in the figure look like similar. In that sense, having one of the definitions of unemployment is enouth to obtain all others - these definitions just represent different, but fixed, portions of some "true" unemployment. Theoretically, one can use any definition for economic study since they are equivalent.
Politically, in the year of elections, Democrats like the upper exstreem, and Republicans - the smallest possible value. If Dems will win, one caould expect the next election will bring opposite situation in unemployment definition used.

Real problem is - what does drive unemployment changes?

Ivan Kitov said...

Results of some short study.

Figure 1. Ratio of part time emloyees and unemployed (BLS definition). Theis ratio has been increasing since 2003 - mopre and more people got some paid job. UE rate has been decreasing since 2003. So, now more people likely tolose part time job and move to unemployed. Not sure that this is better.

Figure 2. Imagine that people react to the opportunity to get paid job somehow. Then increasing labor force would encourage some of them to join labor force when they were not in labor force. With some obvious lag. Decreasing change rate of labor force would apparently discourage these people and some more who lost jobs. When the labor force change rate suddenly turns from some elevated growth rate to lower one, i.e. the labor market starts to grow not at the rate to accomodate all who want job as a reaction to previous strong market - unemployment increases. The opposite turn results in decreasing unemployment, with some lag behind the change in labor force growth rate. This effect with lag time of 4.5 yeras is observed in reality as Figure 2 demonstrates:

Therefore, labor force changes lead unemployment changes by 4.5 years. No R or D - just economy, where labor force has not been growing evenly.
The next short-term move in the UE is likely downward, but then - growth will be observed

Figure 3 provides a smoothed view on noisy monthly estimates of labor force and unemployment rate.

Anonymous said...

Jobs are difficult to find and that too well paying jobs. The unemployment rate is low but people are unhappy with their employment prospects, so either they have taken up low paying jobs to get by, or have stopped looking for work and hence the unemployment rate stays low despite problems in the labor market. A better barometer to judge job prospects is Scorelogix Job Security Score (see job security entry in Wikipedia and you will see why economy is the number one driver of jobs and unemployment, and we know that economy is not doing well. Scorelogix Job Security Index is a good measure to gauge job security for the country and it has been awful for some time now and things don't look too good in the near future. Everytime economy slows down or goes into recession I find the number of unemployed people I know spike whether or not unemployment rate truly reflects the extent of the problems people have in finding jobs.