Recent Posts

Wednesday, January 15, 2014

Education As An Automatic Stabiliser

Chart: US Labor Force Participation Rate (20-24 years)
The fall in the Labor Force Participation Rate (LFPR) is possibly the most important analytical conundrum in this economic cycle. The falling LFPR explains how it is possible that that percentage of the population that is employed (the Employment Ratio) is not rising, while at the same time the Unemployment Rate is dropping like a rock. The subject is complicated because the LFPR is not steady over time (e.g., women entered the labour force), and varies across age cohorts. I cannot cover the whole subject in this article, rather I will look at one piece of the puzzle – the role of education in the fall of the LFPR.

There has been a lot of research published on the fall of the LFPR, but I have not found anything that I view as definitive. In fact, the two most compelling articles I have seen (both by Fed researchers) are contradictory. Although I am happy to give the pros and cons for arguments regarding nebulous topic such as whether interest rates will rise or fall this year, it seems awkward to present two papers which directly contradict each other on a narrow technical question. Once I have looked at the data in more detail, I will comment further.

Participation Rate For The Younger Cohorts

The chart at the beginning of this post shows the LFPR for the 20-24 year old civilian population in the U.S. from 1990-present. (I zoom in on this period as I am not interested in the earlier structural changes.) We see that it has fallen after the recession, and has not really recovered. Looking at these data tells me that the U.S. labour market is in an abnormal state, even if I do not have a numeric estimate of the magnitude of the problem for the aggregate labour force.

The explanation that I have seen for this is an increased number of people in this age cohort are students. (Students are not considered to be participating in the labour force; hence a higher percentage of students means a lower Participation Rate.) I have not seen estimates as to what percentage of the fall is explained by an increased percentage of students, but I want to be optimistic and assume that most of it is.

The “micro” mechanism for an increasing number of students is easily understood. When I finished my undergraduate degree in engineering, everyone applied for a number of jobs as well as grad schools. At that time (1991), the electrical engineering job market was weak, and so a number of people ended up not taking a job like they planned, but they went to grad school as a backup. I assume students are as smart now as we were then, and that quite a few have ended up in various educational institutions (including vocational training). The rising amount of student debt outstanding seems to validate this assumption.

Theoretical Dynamics

Now, let’s imagine a very small country with nice stable demographics, and we want to track the labour market status of university graduates.  During the “normal” period, 100 students graduate every year, and 20 go on to a 3-year post-graduate programme (in which we assume everyone completes in the 3 years, which is unlike any postgrad programme I had contract with). Therefore, the labour market has to absorb 100 people every year; 80 direct from undergrad, and 20 graduating from grad school.

Then, a slowdown hits the economy, and there are only 90 jobs for new graduates. This can be accommodated by having 10 extra students opting out of the job market and going to grad school.  The job market only needs to absorb 70 undergrads, and the 20 graduates of grad school. But the participation rate for the overall population we are tracking is lower, as there are more students (30 in the first-year instead of 20).

We assume that the next thing happens again, for the next two years.

The next year is where things are different. The size of the graduating class for the grad school is now 30. The job market once again has to absorb 100 graduating students (70 from undergrad, 30 from grad school) in order to avoid unemployment. Hiring would have to be even faster to return the grad school class size to a “normal” size of 20.

What happened is that the grad school acted as a buffer for students who would have been unemployed. But what happens is that buffer has a finite lifetime before those students return to the labour force. So unless the size of people entering education continues to increase, employment growth that is below population growth will eventually lead to higher unemployment.


Historically, the U.S. job market was dynamic, and had not suffered such a prolonged spell of job destruction as seen in this cycle. This meant that people who started education programmes when the economy was weak re-emerged when the economy has had jobs ready for them. What is probably happening now is that those graduates are re-emerging with the job market still in a weak state.

Since people who have just graduated are likely to undertake a job search, the odds are that they will not drop out of the labour force if they cannot immediately find a job. They thus start to put a floor under the LFPR (at least for this younger cohort).

The conclusions I draw from this thought experiment:

  1. To what extent the fall in the LFPR is due to this factor, the drop in the Unemployment Rate appears to be “legitimate” if what we are concerned about is the current spare capacity in the labour market. People in education are out of the job market, and thus do not put downward pressure on wages (to what extent supply and demand factors work in the labour market, which is not clear to me). This may help explain why inflation did not fall further.
  2. However, larger graduating classes seem to be now re-entering the market, hence the Participation rate has now bottomed.
  3. Increased job growth will draw people away from educational institutions, and so the Participation Rates in the younger cohorts will rebound upwards. Thus the low participation rate needs to be taken into account for forward-looking measures of spare capacity.

Thoughts On Education As An Automatic Stabiliser

Although not normally counted as such, it seems clear to me that education acts as an automatic stabiliser in modern economies. The students are still consuming goods and services (as marketers of RESP’s remind us Canadians…), but do not provide outputs that are sold in the market. This means that they are in a deficit position, injecting cash flow into other entities in the economy. This deficit spending is financed either via running down personal savings, or by borrowing. (To what extent the students do paid work cancels out this net stimulus.) This helps reduce the shock of a recession to the labour market as well as smoothing aggregate demand.

Education acts in a similar fashion to buffer stock employment, known as the Job Guarantee that is advocated by Modern Monetary Theory. However, in this case the government provides money in the form of a loan, instead of wages. Additionally, students are removed from the labour force for the length of the course, which lasts from a few months to a great many years for some doctoral students.

Education certainly appears to be a win-win-win solution for those who can be helped by it:

  1. The students are occupied, learn new skills, and avoid the stigma of a blank period on their curriculum vitae.
  2. Businesses have better educated, motivated and possibly better trained employees (for those who have entered vocational courses).
  3. Society as a whole benefits from the lesser amount of wasted human potential as well the hope that a better educated workforce is a more productive workforce.

However, there is little point in forcing people to train for jobs that do not in fact exist. As such, education should be a voluntary automatic stabiliser. Society still needs to have other programmes to better utilise the talents of those who have been excluded from a job, such as the Job Guarantee advocated by Modern Monetary Theory.

(c) Brian Romanchuk 2014


  1. Once again, this is a great post!

    Anecdotal but relevant: I began my undergraduate studies in the early 90s in Montreal (which was completely annihilated by the monetarist policies of the BoC during that time). Among my friends, there was no hope we would land a job after completing our degree. Many of us stayed to do graduate work. But, as you mention, we borrowed lots and spent as much...Someone I think this created the conditions for further borrowing and spending during the following decade. Just a thought.

    1. The last sentence should read "somehow I think..."


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.