Friday, August 05, 2016

Four in ten Americans decide to walk the Earth like Caine on Kung Fu


Common sense and easily accessible data suggest that the U.S.’s aging population is the main cause of the decline in workforce participation.


Anyone who’s taken a macro-economics class or listened to NPR when the government’s monthly “jobs report” comes out has been told that people who “stop looking for work” aren’t counted as part of the workforce for the purpose of determining the unemployment rate. The Bureau of Labor Statistics (BLS) defines workforce participation like this:

  • The workforce participation rate is the percentage of the population that is either employed or unemployed (that is, either working or actively seeking work).
The workforce participation rate is often used as a caveat. For example if the unemployment rate drops, a commentator might say “but we have to be careful in assuming too much from the drop, because a significant number of people stopped looking for work and are no longer being counted as part of the workforce.”
And it’s a weirdly widely accepted concept. For example, here’s Sean Hannity, Fox News’ chief bloviator and secret blockhead from Gumby blaming Obama for the labor participation rate being at its lowest point since the 70s.
And here’s 538.com’s founder and level 25 nerdist Nate Silver showing the workforce participation rate over time in a graph, and noting that a small uptick in the participation rate may indicate that “the improving job market is gradually beginning to draw more Americans off the economy’s sidelines.”  
But there are two things about this concept that have never sat well with me.
First, the idea that at some point a healthy but unemployed person may simply get too frustrated by the lack of jobs and stop looking for work doesn’t align with my understanding of the survival instinct. Are these professors and commentators suggesting that all these people just lay down to die? Decide to walk the earth like Caine on Kung Fu? That somehow frustration trumps the need for food and shelter?
Certainly it’s possible that some people may have the option to work only when the labor market is strong enough to provide copious, high-paying jobs. And certainly there are some people that would work some jobs, but when they aren’t available pursue other means of survival such as disability insurance. And some people probably walk the earth.
But in general, people need to eat and don’t want to sleep on the street, and working a job is their only option.
The second thing that gets me is that there’s an obvious problem in the calculation of the workforce participation rate: it counts all people ages 16 and older as possible workers. Workforce participation generally declines rapidly once people get to retirement age, because people that age want to retire and they have social security and/or retirement benefits to rely on for food and shelter.
So, if there was an unusually large generation say, and they all got older and started to retire, then one might reasonably expect the workforce participation rate to decline overall. It so happens that we are at precisely one of those times, as a very large generation born around 1950 started to turn 60 years old around 2010.
The first graph below shows that from 2000 to 2015 the number Americans ages 15 to 59 increased by 9 million (5%), while the number of Americans 60 years old and over increased by 29 million (49%).

The second graph show what the workforce participation rate looks like if you calculate it using the 15 to 59 population rather than the current 16+ population.* The former shows a much more realistic picture of the workforce, where about 90% of people in the age group we normally associate with work are either working or looking for work. We can also note other interesting things, like a slight dip after the economic crash in 2008, but an overall trend that is increasing.

 
*There are a few issues this comparison. First, there are certainly plenty of people that work after age 60, which would make the top line a little overstated. Second, while BLS uses 16+ for their participation rate calculation, the only data they had that was easily accessible to me (i.e. I got in less than an hour) was broken into 5-year blocks, so I had to include age 15 as well. This would make the top line a little understated. The point here is not to be as precise as possible though, it is to show large trends that illustrate the problem with the current method of measurement.

Yes it irritates me when the assorted punditry uses bad statistics to further their agendas. But perhaps even more irritating is when individuals and organizations that are generally neutral and dedicated to things like good governance use bad statistics for seemingly no reason.
Statistics are supposed to describe reality in a way that helps us understand things. Which of those two lines seems more like reality to you? Do you feel like just about everyone you know of working age is either working or looking for work? Do you feel like we have an aging population that we need to seriously consider for both the labor market and things like health care?
Or do you feel like only 6 in 10 working age people you know are working or looking for work, and that the rest are just sitting around, so frustrated they’ve decided to stop eating?