Unemployment

A Better Metric to Steer Macroeconomic Policy

By Philip Cornell

Fears of inflation mean some are encouraging the Federal Reserve to act. Federal Reserve Chairman Jerome Powell acknowledged this when he said during a November 3 press briefing: “Policymakers will be watching employment levels, labor force participation, wages and quit rates, among other indicators, to make a judgment on the labor market.”

The Ludwig Institute for Shared Economic Prosperity (LISEP) has studied the metrics with which the Federal Reserve and other policymakers use to view the labor market, revealing problems with the status quo. This article explains those problems and proposes a new metric policymakers can look to in order to more accurately assesses the economy.

The most common metric used today to measure employment levels is the unemployment rate, which is simply the number of unemployed people as a percentage of the labor force. The labor force includes both those who are employed and unemployed (unemployed is defined as those who are jobless who have looked for a job in the previous four weeks). The unemployment rate, or the U-3, is also used to describe the labor market as a whole. Yet one of the shortcomings of the U-3 is the large — and important — chunk of the American population unaccounted for: those outside the labor force. The unemployment rate goes down if the unemployed get discouraged and just stop looking for a job, meaning the U-3 can look very low even while many people are jobless.

Another common metric is the labor force participation rate (LFPR), which measures the percentage of adults(1) that are currently working or actively looking for a job.(2) Like the U-3, the LFPR is problematic because it does not change at all if everyone goes from employed to unemployed; they are technically all still in the labor force. Thus, policymakers relying on the LFPR would miss key transitions of workers from unemployed to employed or vice versa.

A more useful metric is a combination of the two, the employment to population ratio, the number of people employed divided by the population; yet this metric still has limitations, which are explained further in this article.

Maximum Employment

The Fed has an obligation to Congress as part of its mandate to promote “maximum employment.”(3) Currently the unemployment rate is at 4.8%, which would indicate a good labor market. In fact, 4.8% is lower than it was at any point in the period from 1973 to 1997. But if we take into account what we see around us, from increasing homeless encampments to deteriorating towns across America, it would indicate that the labor market isn’t better than the 1980s or 1990s.

So what is behind the disconnect? The Fed is responsible for promoting maximum employment for the U.S. economy, not minimal unemployment for the U.S. labor force. Thus, maximum employment is better measured by the employment to population ratio, not the unemployment rate.

The employment to population ratio settles the discrepancy between what we see around us and what the data says. Figure 1 shows that the current employment to population ratio (marked by the red horizontal line) is lower than it has been since 1980. So the deceptively low 4.8% unemployment rate is shrouding the truth about the labor market: that many people have fallen out of the labor force entirely.

Figure 1: Employment to Population Ratio

1980–2021

Stable Prices and the Phillips Curve

The second part of the Federal Reserve’s mandate is to ensure “stable prices.” So the Fed is responsible for maximum employment while at the same time it must stave off inflation. Economists have traditionally relied on the Phillips curve relationship, which is the idea that unemployment and inflation are inversely related: low unemployment leads to high inflation and vice versa.

The mechanism underlying this relationship is that low unemployment causes employers to entice workers with higher salaries. With these higher salaries, workers purchase more goods. Thus, high demand for labor pushes up the price of labor (wages), which in turn pushes up the demand for goods. This causes prices for goods to rise, resulting in inflation. But this interconnectedness between unemployment and inflation has been under fire in recent years. Alan Blinder, former Fed vice chairman and renowned Princeton professor, published an article in the Wall Street Journal in 2018 titled Is the Phillips Curve Dead? And other Questions for the Fed.(4) The right-leaning Cato Institute answered that it was, explaining its view in a piece called The Phillips Curve is Dead.(5)

Below is a graph of inflation’s relationship to unemployment. The Phillips Curve would show a strong correlation of high inflation tied to low unemployment and low inflation tied to high unemployment (dots at the top left leading to dots at the bottom right). Cato points out the obvious: this is not a curve but a random scatterplot of points.

Figure 2: Unemployment and Inflation

2000–2020

My view is the labor market and inflation are related; it’s just that the unemployment rate doesn’t give a good picture of the labor market. Instead, the relationship between the employment to population ratio and inflation paints a clearer picture of this relationship. Since the employment to population ratio shows employment rather than unemployment, we would expect a Phillips-type relationship to show that with high employment we would also have high inflation. This would appear as dots in the bottom left moving to dots to the top right. The graph below shows this clearly. In fact, the correlation between these two metrics is significant at the 0.0001 level (t-statistic of 8.91).

Figure 3: Employment to Population Ratio and Inflation

2000–2020

The above graph accounts for inflation, employment, and the labor force participation rate. While the Fed seeks to balance inflation with unemployment, unfortunately, in recent years low quality jobs have plagued the labor market. The mechanism for the Phillips Curve relies on the fact that workers can buy goods with their paychecks. If earnings are low, then the Phillips Curve falls apart. Since 2015, the correlation between the employment to population ratio and inflation has halved, meaning that it also has its limitations. So, is the Fed left in the dark?

At LISEP, we have pointed out that the popular definition of unemployment is misleading. We developed the True Rate of Unemployment (TRU) and True Rate of Unemployment Out of the Population, which measure the functionally unemployed out of the labor force and out of the population, respectively.(6) We did so because under the Bureau of Labor Statistics (BLS) definition, one could work for one hour at minimum wage and be counted equally employed as someone in a full-time job earning a median wage. Paying someone for one hour at a minimum-wage job will not allow that worker to demand more goods. Thus, to count workers with low-quality employment as employed and use this metric to try to predict the inflationary effect of the labor market is actually counterintuitive.

Instead, it’s LISEP’s view that we should only consider people who are functionally employed, meaning they work full time if they want to and make above a poverty wage.(7) Not only does it gives a more accurate picture of the labor market, but functional employment also provides the Fed a better predictor of the labor market’s effect on inflation. Since 2015, the correlation between LISEP’s functional employment to population ratio and the inflation rate was more than four times as strong as the BLS’s employment to population ratio, which is depicted in the graph below.

Figure 4: Functional Employment to Population Ratio and Inflation

2015-2020

As shown in Figure 4, the labor market and price relationship ­— demand for workers leading to inflation — holds. Employment and inflation are intricately tied together, but employment is not accurately measured using solely the unemployment rate (or even the traditional employment to population ratio).

Using the functional unemployment to population ratio instead, we see that when the labor market settles (which will take a little time after pandemic-related supply shocks resolve), the current functional employment to population ratio of 47% should stabilize and produce about a 2.0% inflation rate.

Conclusion

The Fed considers many different indicators of the U.S. labor market and economy in general. LISEP’s view is that the popular yet misleading unemployment rate should hold less weight. We can see empirically that the unemployment rate has not been helpful in predicting inflation. In fact, using the U-3 to try to make policy decisions could be viewed as shooting in the dark. A much more revealing indicator is LISEP’s functional employment to population ratio, which may give the Fed the flashlight it needs to steer the economy.

November 24, 2021
  1. Aged 16+ and not in an institution or military barracks.
  2. The Current Population Survey (CPS)[2] asks respondents who currently don’t have a job “Have you been doing anything to find work in the past 4 weeks?” If the respondent answers “yes” to this question, then that person is unemployed and is also counted as “in the labor force.” But if they answer “no,” then they are not in the labor force.
  3. https://www.federalreserve.gov/faqs/what-economic-goals-does-federal-reserve-seek-to-achieve-through-monetary-policy.htm
  4. https://www.wsj.com/articles/is-the-phillips-curve-dead-and-other-questions-for-the-fed-1525388237
  5. https://www.cato.org/blog/phillips-curve-dead-except-federal-reserve-cbo-modelsand left
  6. More information about LISEP’s metrics can be found on lisep.org.
  7. LISEP defines functionally employed workers as those with full-time job earning at least $20,000 2020 dollars annually.