Estimates of potential employment growth

Published in

New River Investments

4 min readApr 19, 2017

Summary

Last June I wrote a response to a blog post by Stephen Williamson, of the St Louis Fed, in which I argued that an estimate of 60,000 payrolls per month was too low of an estimate for employment growth and that my expectation lied closer to 144,000/mo[1] by 2020. This post is an update to the previous estimate with revised estimates for potential labor force and employment growth over the next 3.75 years. I find capacity for new employment in the range of -2.4mm to +8.5mm for 18–64 individuals and +4.3mm to +5mm for 65+ individuals by 2020 across 3 different scenarios.

There is, in my opinion, nothing limiting increases in payroll employment below 145,000/mo as long as the economy keeps expanding and real wages keep rising. While the labor market is tight, there is still room to increase the labor force through increased participation and increase employment through lower unemployment rates.

Scenarios

Adverse non-recessionary: static labor force participation, unchanged unemployment rate
Baseline: static labor force participation, lower unemployment rate
Optimistic: higher labor force participation, lower unemployment rate

Population

The US Census Bureau, in their 2014 population projections[2], project a change in the 18–64 population of 4.3mm people and an increase of 8.6mm in the 65+ population. The changes in population by age are assumed to be linear to estimate current population levels and remaining projected growth from March 2017 to December 2020.

Labor Force Participation

Labor force size is estimated using both current labor force participation levels and higher labor force participation levels that were present as recently as the 2000s. Higher-than-present labor force participation rates are used to estimate optimistic scenario under the assumption that higher real wages could lead to increased labor force participation. I avoid using older labor force participation levels to avoid error from larger changes such as life, expectancy, prevalence of two-income households and incidence of post-secondary schooling. Because of historical low rates of change in labor force participation levels, the potential increase in labor force participation is reduced by assuming a maximum annual increase in participation equal to 1/10th of the difference between present participation rate and higher previously observed participation rate. To avoid noise, the participation rates used to estimate the higher scenario are all taken from the same period of elevated participation, February, 2000. Where current labor force participation is higher, no decline or increase was assumed.

Employment & Unemployment rates

Lower unemployment rates are estimated using levels that were present as recently as the 2000s. To avoid noise, the unemployment rates used to estimate the optimistic and baseline scenarios are all taken from the same period of depressed unemployment, February, 2001.

Results

The adverse and baseline scenarios both displayed declines in the 18–24 employment level. This is due to the “hump” of the millenial generation moving into the late-20s/early-30s. Due to static labor force participation assumptions and already-low unemployment rates, all scenarios display high employment gains in the old-age population increases.

Prime age (25–64) employment level increases were 7.76mm, 6.54mm, and 0.09mm for the optimistic, baseline and adverse scenarios, respectively. These would translate to averages of 172k, 145k, and 2k per month, respectively.

Where age-breakouts from the current population survey were more granular than the census population projections, unemployment and labor force participation rates were averaged. While the effect of this would be minimal for the 25–44 population, it could lead to under- or over-statement of 45–64 potential employment gains due to the approximately 15 percentage point difference in the participation rate between 45–54 and 55–64 individuals.

[1] https://medium.com/new-river-investments/re-dazed-confused-notes-on-labor-force-wage-growth-c63347b20408

[2] https://www.census.gov/population/projections/data/national/2014/summarytables.html