Fed Chair Janet Yellen’s recent Humphrey Hawkins testimony excited speculation that the Federal Reserve might raise the Fed Funds rate as soon as March. The normally reticent Fed Chair seemed to turn a skosh more optimistic, noting that the economy continued to make progress and “waiting too long [to raise interest rates] would be unwise.” She did, however, hold fast to the Fed’s standard caveat that “monetary policy is not on a preset course.”1
Missing from the vast array of economic indicators the Fed monitors is the duration of unemployment. Indicators like unemployment rate, labor force participation rate (LFPR), and employment growth represent only the briefest snapshot of the job market and fail to capture the dynamics that define worker and employer experiences. The length of time that workers spend unemployed as new or reentrants to the job market and/or moving from one job to another is an important measure of just how efficient or “tight” the labor market really is.
The efficiency of the job market is an important subject of economic policy generally. However, it is an equally important qualitative measure for a Federal Reserve that aspires to support an economy that not only reaches full employment but, in Chair Yellen’s words, “maximum employment.”
The increase in the length of time unemployed Americans spend looking for work is a stark illustration of the deterioration in U.S. labor market dynamics. Compare, for example, the mean duration of unemployment (the average number of weeks a jobseeker is unemployed) to the median duration (the number of weeks that marks the midpoint where 50 percent of job seekers experience greater spells of unemployment and 50 percent less as the unemployment rate varies).
For much of the past 50 years, the relationships between mean and median duration of unemployment and the unemployment rate remained relatively stable. Gaps began to widen somewhat in the early 1990s but not dramatically.
After the financial crisis, however, even though the unemployment rate peaked at 9.9 percent vs. the 10.8 percent peak in the early 1980s, both the average and the median durations of unemployment shot up. The median unemployment rate rose to 25.2 weeks in June 2010, compared to 12.3 weeks in May 1983, and the mean rose to 40.4 in September 2011, compared to 20.8 weeks in June 1983 (Chart).
Thus, although the January 2017 unemployment rate was down to 4.8 percent, the median duration was more than 10.2 weeks and the mean, 25.1 weeks. In other words, at a time when the unemployment rate has declined to a rate considered by many to mark full employment, the duration of unemployment remains at levels comparable to peak unemployment rates in recessions past.
Behind today’s low unemployment rate is a job machine that works for some but not for all. The job market as measured by these duration numbers is improving—but very slowly. The share of unemployed who find a job in five weeks or less is up from 20 percent in the period from late 2009 through mid-2013 to just over 30 percent today. That is still well below the pre-crisis share of 38 percent. The share of long-term unemployed (15 weeks and over) still hovers at around 40 percent.2
There are many explanations, each deserving of another article. The gap between job openings and hiring is larger than usual, suggesting a lack of qualified workers. This gap could be created by skill mismatches, social factors like criminal or drug records, or geography.
However, more striking is the long-term deterioration in job creation and job destruction, and in the number and job-creating ability of new businesses. Behind the net employment number reported each month are millions of jobs that are created and destroyed.
Employment grows when job creation is greater than job destruction. However, higher rates of job creation and destruction provide opportunities for mobility that change the mix of job skills, allow workers to move up the career ladder, or find other jobs if theirs are eliminated. Since 1990, job reallocation (jobs created plus jobs destroyed) has declined from 16.3 million to 12.2 million in 2015. Job creation and job destruction have declined at similar rates, each representing 8 percent of total employment in 1990, declining to just above 6 percent for job creation and just below 6 percent for job destruction in 2015.3
High rates of job reallocation, supplemented by the creation of new companies, means there are more opportunities available to workers relative to the level of employment. The decline in dynamism, which accelerated after 2000, provides at least a partial explanation for the rise in the mean and median duration of unemployment in the post-crisis years. We went into depth on this issue back in 2011 in our interview with John Haltiwanger.
And you thought that we could get through a whole article without mentioning Mr. Trump? Wrong!
There is no way to link the loss of economic dynamism and its effects on the labor market directly to Trump voters. Still, the lingering deep structural nature of this problem suggests that President Trump’s election may reflect a more widespread and enduring economic angst than generally thought.
These are hard problems that will not be easily solved. Trump has the responsibility of addressing them. If he falls short, voter angst will likely intensify rather than just go down in the history books as a passing political mood.
1 United States Senate. Committee on Banking, Housing, and Urban Affairs. Semiannual Monetary Policy Report to the Congress. February 14, 2017. Washington, D.C. (statement of Janet L. Yellen, Chair, Board of Governors of the Federal Reserve System).
2 Bureau of Labor Statitics.
3 Haltiwanger, John. “Top Ten Signs of Declining Business Dynamism and Entrepreneurship in the U.S.” Kauffman Foundation New Entrepreneurial Growth Conference. August 2015.
Login in below to access content exclusive to clients of The GailFosler Group.
Not a client yet? For more information on the benefits of becoming a client, please contact us.