Recent Underestimation of the Unemployment Rate

Tomorrow, the Department of Labor (DOL), as part of its monthly “Jobs Report”, will present the “unemployment rate” for November. At this time, it is expected to jump to 6.8%. Last month, it was 6.5%. Yesterday, I discussed some simple reasons why I think the “Jobs Number”, also part of the monthly report, is used in a misleading way. Today, I want to talk about why the “unemployment rate” can be a misleading statistic.

So what is the official unemployment rate? It is computed from a monthly survey (the Current Population Survey – CPS) of 60,000 households. Persons are classified as employed, unemployed, or not in the labor force. People are considered employed if they worked as a paid employee at any time in the previous week, worked in their own business or on their own farm, worked without pay at least 15 hours in a family business or farm, or were temporarily absent from a job due to sickness, weather, vacation, a strike, or a personal reason. People are considered unemployed if they did not meet the above criteria, were available for work, and made some effort to find work in the preceding four weeks. The unemployment rate is simply the percentage of those in the labor force who are unemployed. People who are not classified as employed or unemployed are not considered part of the labor force.

There are many potential problems with this number, and they fall into two categories. First is the problem with excluding people from the labor force. People who are not employed or are not actively seeking employment are not considered part of the labor force. In a poor economy, many people give up seeking employment. Some effort is made by the DOL to further categorize such people, calling them “discouraged workers” or “marginally attached workers” depending on how they respond to the survey. However, it is easy to imagine that such people might be less than honest about such assessments – surely no person who has been unsuccessfully looking for work for a prolonged period will feel good about him- or herself. They are more likely to say that they are not employed for some other reason. The bottom line is that, especially during a bad economy, the number of people in the labor force as calculated by the DOL is an underestimate of those who would truly like to have a job.

The second category of problems concerns those who are categorized as employed. If someone can’t find a job and resorts to “volunteering” in a family business, they are considered employed. If a person is seeking full-time employment and yet can only find a few hours employment per week, they are nonetheless considered employed. If a person cannot find work and starts a home business, they are considered employed even if the business does not produce any income.

All of these problems suggest that the unemployment rate is an underestimate of what we think of when we hear the term. Further, this discrepancy is probably greater during poor economic times.

So, is it possible to easily come up with a more reliable measure of the unemployment rate? A number of people have tried to do so by using the aforementioned numbers of “discouraged worker” and such. I suggest a simpler approach. Let us assume that at some point during the best of economic times, everyone who wants a job is either employed or looking for a job. That would mean they are all considered part of the labor force.

The DOL, as part of the CPS data series, gives historical data on the civilian non-institutional population over 16 years of age. Let’s call this TOTAL. I computed the labor force (as presented by the DOL) as a proportion of TOTAL for each month for the past several decades. Let’s call this value PROP. I assume that a “good” economy occurs at least once in a decade, and therefore PROP reaches a maximum. So, at any point in time, a better estimate of the labor force can be found by multiplying the current TOTAL by the maximum value of PROP that has been observed over the most recent 10 years. Let’s call this the Adjusted Labor Force.

A better estimate of the number of unemployed persons is the difference between the number employed given by the DOL, and the Adjusted Labor Force. We’ll call this the Adjusted Unemployed. An Adjusted Unemployment Rate is then simply the percentage that the Adjusted Unemployed is of the Adjusted Labor Force.

So, where are we now? As I mentioned, the reported unemployment rate last month was 6.5%. The Adjusted Unemployment Rate is 8.2%. This is a substantial difference. The table at the beginning of this post gives the average of the monthly values for the DOL’s unemployment rate and my Adjusted Unemployment Rate. A couple of things stand out. First, the highest average DOL unemployment rate was seen in the Reagan years. It was a full percent above the administrations both before (Carter) and after (Bush 41). The lowest unemployment was during the Clinton years.

The most interesting, and important, thing about the current Bush administration is that while the DOL unemployment rate is as low as during the Clinton years, the Adjusted Unemployment Rate is much higher. Note the final column of the table, the difference between these two measures of unemployment rate. While the difference is similar during the Carter, Reagan, Bush 41, and Clinton administrations, it is much higher in the current administration. The unemployment rate in recent years seems to be substantially underestimated. The unemployment during the current administration has been much worse than what is reported each month by the Department of Labor.