Tuesday, June 6, 2017

LFPR and the New Macroeconomics

The civilian labor force participation rate (LFPR) tells the percentage of the population that wants to work. That is, it counts those with jobs and adds those who are looking for jobs and relates that number to the size of the population. Not everyone wants to be in the labor force -- some are too young or too old. Some are busy getting education. Some are sick. Some don't want to work for a variety of reasons. So LFPR is never close to 100%. 

US LFPR generally increased after World War II until early 2001 after it reached a little more than 67% of the population. Since then it has been falling and was recorded as 62.7% in May 2017. This roughly 4% decline is meaningful -- 4% of the US population of 230 million people is about 9 million people who no longer participate in the labor force. To put that number of 9 million in perspective – that’s about how many people work in manufacturing. That’s like everyone in New Jersey deciding they would no longer take or look for a job. No New Jersey jokes please. 

This new 16-year trend is important. I am going to argue that it is very important and may constitute the beginning of a new phase of macroeconomics and policy. As I said last week, macro is becoming obsolete. Monetary and fiscal policy are out of bullets. Supply-side policy has political downsides. So what’s left?

The answer might reside in the LFPR. Today’s experts repeat over and over that the lackluster economic growth predicted for the future is caused by lack of business spending on capital and a reluctance of people to join the labor force. One could go further and say that the former is related to the latter – firms are pessimistic and won’t invest more because they see LFPR as a major problem and do not see a government that is doing anything about current economic challenges.

Future macroeconomic theory and policy, therefore, should be focused on LFPR. I have mused in this blog in the past that if labor is not forthcoming and if the labor that does come is not prepared for the jobs of the future, then maybe we should focus on that mismatch. In macro we usually take that mismatch as secondary and hope it will be solved by national economic growth induced from traditional monetary and fiscal policies. But that puts the cart before the horse. Maybe today we need to focus on labor mismatch and if we solve that then maybe economic growth will improve in the process.

This post today is a humble beginning in this direction, and my only goal is to shed some light on the data. Today I look at some of the data as it relates to the LFPR. I got the data from the FRED service at the St. Louis Federal Reserve Bank. I look at data from 2002 to 2017. The goal is to better understand or break down the above-mentioned roughly 4% decline in labor participation in the USA.

Consider first, men versus women. The table below shows that LFPR for both men and women fell between 2002 and 2017 – but it fell more for men – falling almost twice as much.
                        Women    Men   Gender Gap
2002                 59.6          73.9    14.3
2017                 57.0          69.0    12.0
Change             -2.6          -4.9    

Next, look at age. In 2002 almost 84% of those in the prime work ages (25-54) looked for and/or found work. Younger people worked too – 76% was the LFPR for those aged 16-24. Those 55 years or older had a much lower rate at 34%. The changes in the next 25 years are interesting. For the regular working ages the LFPR went down by only 2%. Those at the younger end found participation rates falling by at least twice as much as their seniors. As for the older folks, they are participating dramatically more – an increase of almost 6% in their LFPR!

                        25-54    55+  16-19 20-24
2002                83.7      34.2    76.7   75.4
2017                81.7      39.9    71.9   70.4
Change            -2.0      +5.7    -4.8    -5.0

Finally I look at education. The first column looks at high school graduates 25 years and older; the second is college graduates 25 years and older. The impacts of college education on LFPR are dramatic. While college-educated people did participate somewhat less in 2017, the change for high school grads was much larger – almost five times as large.
                        High School              College
2002                          64.4                75.4
2017                          58.0                74.0
Change                      -6.4                 -1.4

This excursion through some data is meant to be a first step in looking deeper into a major macroeconomic challenge. Surely this is not enough data to form solid conclusions. Curious minds would wonder about other and finer breakdowns as they relate to education, training, age, race, location, industry, and more.

What is going on in the last 16 years? This data suggests that the largest groups to explain a slowdown in labor participation are young males with less education. Surprisingly, older people who should be enjoying time on Alaskan cruises sipping JD seem to be increasing their participation.

A scientific friend of mine said that most good science starts with data and ends with understanding. Labor force participation data needs to be better understood. Then perhaps we will know WHY participation is flagging and perhaps what we can do about it. Let's get back to work!


  1. A very interesting subject. Some ideas and or postulations:
    1. Technology has changed the face of the labor market in that many sectors of manual or hands on labor have been replaced. Where did the people go who were replaced? Most likely to lesser paying service jobs or retired. ( Would explain part of the stagnant wage rate for the same period of time.)
    2. High School only educated youth have fewer choices where climbing up the economic ladder is a possibility ( ergo see technology).

    Let me provide an example: In my business we buy x-ray parts from Asian manufacturers. We assemble them, pack them and sell them to American dentists. These digital based parts and equipment replace at least two hygienists in the single dental office with 3 operatories. The existing employees all are trained to use the equipment but his training does not carry over to any non dentist job in case they get fired or laid off.(over-reliance on specialization) One such technologies is a intraoral CT scanner that builds a 3D picture of the structure of the teeth. A tooth replacement or implant can be done in 10 less hours than 5 years ago by sending the digital scan to a lab that uses a computer driven 3D printer to create the tooth from Zircon. This whole process takes 1 person and 2 hours vs 3 people and 8 hours. There are 150K dentists in the US and if 1/4 of them did orthodontics then the total displacement for the 3,000 labs who build the tooth the old way would be 12,000 technicians. The story is the same and the advent of digitizing is leading us through a tranformation change similar to the beginning of the industrial age.

    The answer is less people...which is happening due to less births ...with exception of illegal immigration...where jobs are more menial....but the real answer is a different kind of K-12 education system that is also adaptable to colleges. Not much can be done with the 40 to 60 age group. Uber is probably a better example of this change.

  2. I went to SSRN and did a search for papers post in the last 3 years with "Labor Force Participation Rate" in the title, abstract or full text. I got 826 papers. Any worth reading?

    1. I will start working on that right now! :-)
      The problem, of course, is not what academics are saying but more about how policymakers approach our problems. A direct approach to LFPR as a way to solve economic problems would require a wholey different approach.