The Washington PostDemocracy Dies in Darkness

Ed Lazear: This is the real unemployment rate

A job seeker fills out an application at a jobs fair in Los Angeles. (Reuters/Jonathan Alcorn/Files)

This guest post is by Edward Lazear, who was chairman of the Council of Economic Advisers from 2006-09 and is a professor at Stanford University’s Graduate School of Business.

The government announced Friday that October's unemployment rate stood at 5 percent, lower than it has been for two-thirds of the time since 2000. Many economists view rates around this level as full employment.

Despite this, wages haven't been growing. The general sense is that the labor market is far from tight, and economic growth is weak. For these reasons, the Federal Reserve has resisted raising interest rates (though it may begin doing so in December).

[Economy adds 271,000 jobs in October; unemployment rate dips to 5 percent]

Why the disconnect? As many fear, the unemployment rate may not be indicative of labor market strength. Indeed, other indicators of labor market activity, especially hiring, imply that the unemployment rate number comparable to past rates is actually around 6.3 percent.

There are two measures of the employment situation. The most widely reported is the unemployment rate, which is defined as the proportion of the labor force that is without a job. To be in the labor force, a person must have a job or be actively seeking work.

The other measure, which is generally preferred by many who study labor markets, is the employment rate, which is defined as the proportion of the working-age population (that is, 16 and above) that has a job. The employment rate is the bottom line because it measures the number actually working compared to the number who could work. But to the extent that failure to actively seek work is voluntary — for example, reflecting retirement or school attendance — taking that into account as the unemployment rate seems relevant to assessing the strength of the labor market. So which is right — the unemployment rate or the employment rate?

If credible, October's unemployment rate of 5 percent would mean that we have essentially recovered from the recession. But Friday's reported employment rate of 59.3 percent signals an economy that is well below capacity. Is the unemployment rate higher than it appears, or is the employment rate too low to reflect the appropriate strength of the labor force?

Part of the difference between the two reflects demographics, which affects the employment rate directly. An older population means that more people are retired, and more retired people yields a lower employment rate because a smaller fraction of the population is typical working age. In November 2006, 25.3 percent of the working age population was between 45 and 64 years old. Today, that number stands at around 27 percent. Additionally, the proportion over age 65 has risen from 12 percent to 13.5 percent of the working-age population over the same period. These changes can account for about half of the difference between the behavior of the employment rate and the unemployment rate during this recovery, significant but only part of the story.

To determine whether the measured unemployment or measured employment rate is too low, another indicator that is more directly related to labor market demand conditions can be used. I use the “population hiring rate,” which I create from Bureau of Labor Statistics data.  It is defined as the ratio of monthly hires (from the Job Openings and Labor Turnover Survey) to the population over 16 years of age. It is important to use the entire working-age population and not merely those employed to avoid the trap of ignoring those outside the labor force. The hiring-rate peak, annualized, was 34 percent, in November 2006 and its low was 22 percent in June 2009. The current population hiring rate is 28 percent, or about half of the way back to the peak rate.

But we must also recognize that demographics play a role in legitimately reducing the proportion of the population that would be expected to work. Consequently, my analysis uses both the hiring rate and a measure of the proportion of older individuals in the potential workforce to predict unemployment and employment rates. The approach tracks pre-recession employment and unemployment well.

Not only does the population hiring rate do a good job in predicting past unemployment and employment rates, but there is a strong logic for thinking that hires measure labor demand. A hire only occurs if an employer finds it profitable to add another worker or to replace an old one.

Using models based on pre-recovery data, it is possible to estimate what unemployment and employment rates would be if we were back in the pre-2009 era but had today’s hiring rate and demographic conditions. Using data from September, the official unemployment rate was 5.1 percent. But to be comparable to the pre-2009 period, the unemployment rate should be thought of as 6.3 percent. Labor demand is simply not strong enough to be consistent with such a low unemployment rate by historical standards.

Analogously, the peak pre-recession employment rate was 63.4 percent. Correcting for hiring and demographics makes September's rate of 59.2 equivalent to a pre-2009 rate of 61.4 percent, meaning we still have a 2 percentage point deficit when compared with the earlier peak. This amounts to about 4.8 million jobs. Put differently, the number employed has grown almost 13 million jobs since the employment trough in February 2010. But it would have to have grown by 17.5 million to make up for the recession and keep pace with growing population.

We have come back considerably from the depths of the recession that began almost eight years ago. Unfortunately, we still have a way to go. Given conditions that are consistent with an unemployment well above 6 percent -- and not a 5 percent unemployment rate -- it is no surprise that many in this economy still feel that we are not firing on all cylinders.