In his opening statement for his press conference following the latest meeting of the Federal Reserve, Chairman Ben Bernanke once again called out the influence of lower participation rates on unemployment (all quotes come from the transcript of the press conference):
"…we have seen ongoing decline in labor force participation, which likely reflects discouragement on the part of many potential workers, as well as longer-term influences such as the aging of the population."
Without separating out the specific contributions of each factor, Bernanke proceeded to state that the decline in the participation rate of 0.3 percentage points over the past year was an unfavorable development in labor markets:
"Not all labor market developments over the past year were positive, however; notably, the labor force participation rate fell by about 0.3 percentage points over the past year, and real wages remained about flat."
During the question and answer period, Pedro da Costa from Reuters asked "…how high do you think the jobless rate would be if it were not for the decline in participation? I've heard estimates as high as 10 to 11 percent. And could you put the labor market in that context?" In response, Bernanke separated out the cyclical from the structural movements in the participation rate but did not offer an adjusted figure for the unemployment rate (emphasis mine):
"I think there is a cyclical component to participation and in that respect, the unemployment rate understates the amount of sort of true unemployment if you will in the economy. But on the other hand, there is also a downward trend in participation in our economy which is arising from factors that have been going on for some time including an aging population, lower participation by prime-age males, fewer women in the labor force, other factors which aren't really related to this recession.
Over the last year, the unemployment rate has dropped by eight-tenths of a percentage point. The participation rate is dropped by three-tenths of a percentage point, which is pretty close to the trend. So in other words, I think it would be fair to say that most of the improvement in the unemployment rate, not all, but most of it in the last year is due to job creation rather than lower participation. I would also note that if you look at the broader measures of unemployment that the BLS publishes including part-time work, including discouraged workers and so on, you'll see that those rates have fallen about the same amount as the overall standard civilian unemployment rate. So, I think that there has been progress and it's obscure to some extent by the downward trend in participation."
So is the decline in participation in the past year an unfavorable development or not? In the above response, Bernanke did an about face by contradicting his apparent lament in the opening statement that this 0.3 percentage decline is "not a positive development" in the labor force. Perhaps he did not mean to imply that this development was bad (not positive does not have to equate to bad), but given the context, I have to assume he meant it was an undesirable development.
This topic caught my interest particularly because I have seen many pundits, bloggers, and the like decry any improvement in employment because of the decline in the participation rate. They rarely acknowledge the structural component of this decline that has little to nothing to do with the recession; this decline will likely continue even once the labor market firms up. The commentary during this press conference highlighted for me the nuances of the participation rate and how it can be misapplied in understanding labor force dynamics. A demographic breakdown of the participation rate well demonstrates the strong influences of structural factors. These distinctions are important since without them it is easy to overstate the case on the ominous implications of the decline in participation and thus miss the positive developments in the labor force when they happen.
It is well-recognized now that the labor force participation rate did not start declining after the Great Recession. It has dropped since reaching a peak in 2000. It is possible that the 2000 peak was an unsustainable one in a "normal" economy (whatever that looks like now!). By breaking up the data simply by gender, we immediately observe that the rapid increase in female participation was compensating for an ongoing, multi-decade decline in male participation. In 2000, the boost from female participation finally ended. The decline in the female participation rate is the singular change since the Great Recession - from 59.5% in January 2007 to 57.3% in August 2013. The gender-based data below are not seasonally adjusted (all data and charts are from the St. Louis Federal Reserve).
U.S. Civilian Labor Force Participation Rate with Gender Breakdown
At opposite ends of the age spectrum, important dynamics are at work: young adults are increasingly entering college, especially women. Since the 1970s, the percentage of women aged 18-24 enrolled in college has steadily increased from 20% to 45% now. Many of these young people are delaying participation in the labor force. Older adults are leaving the workforce in greater numbers as baby boomers retire. So, if we narrow the scope to adults aged 25-54, we find a much less dramatic decline in the participation rate than for the overall adult population, especially for women. In the graph below, I had to use "activity rate" for the gender breakdown, which is essentially the same thing as the participation rate but (I believe) it includes the entire adult population, not just civilians. The seasonally adjusted numbers for women seems only available through 2012, and the series for men seems ONLY to have seasonally adjusted numbers. So, I am admittedly patching things together here…
U.S. Civilian Labor Force Participation Rate with Gender and Age (25-54) Breakdown (Activity Rate)
Interestingly, the participation rate for adults 65 and over has actually increased slightly since the recession, even for senior citizens with disabilities. (Data only available to 2008).
Civilian Labor Force Participation Rate for Adults 65 and Over, With and Without Disability
Of course the negative spin on these data is that (at the margins) senior citizens have been forced by poor savings to remain in the labor force past typical retirement age. However, we also know that with people living longer and healthier, more people are able to work longer into life and perhaps appreciate the opportunity to supplement their savings.
Regardless, any way the data are sliced, the structural issues are evident in the labor force participation rate. Still, some are, strangely enough, worried that the unemployment rate will increase once the labor market shows substantial improvement. This concern is directly related to the knock that the unemployment rate would be much higher if the participation rate were higher. Staff economist Andreas Hornstein at the Federal Reserve Bank of Richmond is working on some important insights on this question that could potentially turn conventional thinking on its ear. In a first draft of a working paper called "The Cyclicality of the Labor Force Participation Rate" (May, 2013), Hornstein uses labor flows to suggest that the unemployment rate will NOT go up as a result of non-participants returning to a strengthening labor market. Here is the key finding from the paper (LFP is the acronym for Labor Force Participation) (emphasis mine):
"Many observers of the U.S. labor market perceive the LFP rate to be below its long run trend and the unemployment rate to be above its long run trend. In fact, the low cyclical LFP rate is seen as keeping the cyclical unemployment rate from being even higher, because poor employment prospects have induced discouraged unemployed workers to leave the labor force and have prevented marginally attached inactive participants from a return to job search. In this paper I have documented that direct observations on transition rates between unemployment and out of the labor force are inconsistent with this perception. It turns out that at times of high unemployment, unemployed workers are less likely to exit the labor force and inactive workers are more likely to return to the labor force. This pattern holds not only for the period since the Great Recession, but more generally for most of the post WWII period, and it would have introduced a positive correlation between cyclical movements of the unemployment rate and the LFP rate. Yet we have observed a negative correlation between the two rates. I have shown that the negative comovement is actually induced by movements in the unemployment rate itself, and by a pro-cyclical transition rate from inactivity to employment without an intervening unemployment spell. To summarize, a low cyclical LFP rate simply seems to reflect a high current unemployment rate, rather than to indicate an elevated future unemployment rate."
Hornstein's results and calculations are certainly surprising given the conventional wisdom. Hornstein's analysis of the labor flow data, isolating cyclical and structural (trend) features, show that a high unemployment rate induces a higher likelihood that participants will enter the labor force as unemployed and a lower likelihood that they exit the labor force from an unemployed state. Thus, the unemployment rate and the participation rate increase together. The observed negative correlation between these rates is entirely explained by the decrease in likelihood that a participant directly transitions from inactivity to employment.
Time will tell as always, but once/if we see the unemployment rate continue to improve even as the participation rate increases (on a cyclical basis back to trend), we will know why. In the meantime, it seems that the constant hand-wringing about the aggregate participation rate's dampening impact on the unemployment rate is misplaced angst. The focus should remain on the health of the labor market for those who need to or who have chosen to find work.