One reason the unemployment rate has come down from 9.3% at the start of 2011 to 7.4% last month is the decline in the Labor Force Participation Rate (LFPR). The LFPR is the number of people with jobs (employed) or looking for jobs (unemployed) divided by the civilian, noninstitutional population (see here for exact definitions). At the start of 2011 the LFPR was 64.3% and the latest value is 63.4%. That may not seem like a big drop, but it represents about 2.4 million people, and if the LFPR had held steady at 64.3% over the last two-and-a-half years, the unemployment rate would currently be 8.7% instead of 7.4%.
If the LFPR is decreasing, that means there are more adults not looking for work. This could either be a positive or negative sign for the economy. If the decline is due to discouraged job hunters giving up their search for a job, that would be a negative sign. If the decline is due to a natural aging of the population because of demographic trends (more older folks retiring), that would be a positive (or at least neutral) sign.
The Bureau of Labor Statistics publishes a slew of data on this issue from its Current Population Survey (also known as the Household Survey). Let’s take a look at some of this data and see if it can shed any light on why the LFPR is decreasing.
Population by Age Group
Conventional wisdom is that the country’s population is skewing older as the baby boomer generation enters their later years. The chart below shows the breakdown in the civilian noninstitutional population for each July since 2013.
In the last two years, the fraction of the population over the age of 65 increased by 1.2%, and the fraction between 55-64 increased by 0.4%. In contrast, the percentage between 45-54 decreased by 0.7%, between 35-44 decreased by 0.4%, and between 25-34 by 0.4%. Given this trend of fewer adults (percentage-wise) in their prime working years and more adults in their later years, we would expect the labor force participation rate to decline over this period, and in fact it does:
Labor Force Participation Rate By Age
The chart below shows the LFPR by age group since 2003. Note that these are not seasonally adjusted; both the 16-19 (dark blue line) and 20-24 (red line) cohorts show strong seasonal patterns as high school and college students look for summer employment. There is a strong upward trend in the LFPR for 65+ workers (light blue line at the bottom of the chart), from 14% in 2003 to a peak over 19% in 2013. Given the persistence of this trend in both good (2004-6) and bad (2008-10) economic environments, one could conclude this is not purely a reaction to the economic climate, but rather due to noneconomic factors, such as an increase in health and ability to work for those over 65, or change in cultural expectations about the appropriate age to retire.
The chart below shows a blow-up of the 65+ cohort from the chart above. There has been a clear linear trend over the last 10 years for more older workers. Note the downtick the last two months: In May the participation rate for those 65+ was 19.2%, and this dropped to 18.8% in June and 18.3% in July. Two months isn’t nearly enough to declare a new trend, and I would expect the rate to mean-revert back to the trendline as it has the previous times it’s dipped, but it’s something to keep an eye on the rest of this year.
If there is evidence of LFPR decline due to discouraged workers, it would have to be in those under the age of 55, since the participation rate has been increasing above age 55. The chart below shows the three cohorts between 25-54. Each series has declined between 1.5-2.0% since the great recession.
This decline in LFPR since 2008 for people between 25-54 is similar for men and women, though the decline for men is slightly greater. In rough terms, the men’s LFPR fell 2% and the women’s LFPR fell 1.5%. (Therefore, the often-heard argument that the overall decline in LFPR is due mostly to more women choosing to stay home with their children doesn’t hold water.)
This puts a negative slant on the falling unemployment rate; the falling LFPR for people in their prime working years indicates an increasing number of discouraged workers. The LFPR falling due to an ageing population, however, is a natural demographic occurrence, and doesn’t have negative economic connotations. How much of the recent fall in the unemployment rate was caused by each of these two items, and what can we expect going forward?
Unemployment Rate Calculation
From July 2009 to July 2013, the LFPR fell 2.2% from 66.2% to 64.0%. Isolating the effect of the ageing population, we find that it accounts for just about half the fall in LFPR, or 1.1% (I calculated this by keeping the participation rates by age constant at 2009 levels). Isolating the effect of the change in participation rates (younger workers participating less), we find that 1.2% of the decline in LFPR is due to people opting out. Thus about half the decline in LFPR was due to the ageing population, and half was due to younger-aged cohorts participating less.
What does this mean for the future of the unemployment rate? Suppose the participation rates hadn’t changed over the period. As mentioned before, this would mean the LFPR would be 1.2% higher than it is today, or about 2.9MM people greater. Suppose also that all of these people started looking for work tomorrow. Then both the labor force and the ranks of the unemployed would increase by 2.9MM people, which would increase the unemployment rate by 1.7%. Obviously this isn’t going to happen, but you can see the headwind for the unemployment rate if people start opting back in.
Countervailing this is the continued ageing of the population. If the population trends shown in the stacked bar chart above continue, and LFPR by age group holds constant, the overall LFPR would fall by about 0.2% per year just to population shifts, and the unemployment rate would fall about 0.3%.
To recap, the ageing population provides a small tailwind for the unemployment rate, to the tune of 0.3% per year. If participation rates by age group stay the same, this will be the key demographic driver of the unemployment rate. However, I would expect participation levels to increase toward levels seen before the great recession which will mean a significant headwind to further reductions in the unemployment rate, as a large group of people reenter the workforce.