It has become a sport among economic commentators to gauge presidential performance by the number of jobs each has created. This is questionable on several grounds: (1) differences in presidential term lengths give two term presidents an unfair advantage, (2) policies of the predecessor may influence the early period of a presidential term, (3) presidential policies may take some time to take effect, (4) policies may be targeted at different groups, and last but not least, (5) presidents may or may not have that much influence on the job market at all. The general economic situation at the start of a term, the Federal Reserve, Congress, world markets – and more – influence jobs and may have a larger effect on the “job market performance” of a president than the president itself. These factors make it difficult to compare job market outcomes under each president.[i]
Nonetheless, it is still worthwhile to compare job market outcomes for individual groups within total employment under each president. Does employment within each group keep pace with overall employment growth?[ii] If a rising tide lifts all boats, does it lift all boats equally?
In this first part of our analysis, we examine employment data by race and ethnicity from the Bureau of Labor Statistics (BLS), specifically the Labor Force Statistics from the Current Population Survey. The BLS supplies data by race for Asians, Blacks, and Whites. It also supplies data for the ethnic group of Hispanics. For simplicity, we henceforth refer to all groups as racial groups. We compare labor market outcomes for these groups using various definitions of presidential term: differentiating between one and two term presidencies and splitting presidential terms into two year sections with particular focus on later years.
We find small differences in job creation between Republican and Democratic presidents overall, but notable differences in job creation for Whites versus non-Whites: While job market outcomes for Whites again have been almost undistinguishable under Republican and Democratic presidencies, non-Whites have experienced favorable job market outcomes under Democratic presidents. These differences are even somewhat more pronounced when comparing second terms of Democratic vs. Republican presidents.
To provide background on the relative composition of each racial group in the labor market, we graph total employment as well as employment by racial groups for all post-World War II presidents. Shaded bars indicate recession periods and help gauge the economic conditions of each presidency. We focus on the last 40 years for which data by race is available. During this period, the United States saw three Democratic and four Republican presidents. Two Republican and two Democratic presidents served two terms. All presidents except for Clinton experienced at least one recession during their term, George W. Bush even two. Some presidents experienced them mid-term, others at the beginning or towards the end.
Unemployment rates were highest under Reagan and Obama. Aside from Asians (for which data only reaches back to 2003), Whites enjoyed the lowest unemployment rates throughout the last 40 years. Unemployment among Whites, who made up 72% of the U.S. population according to the 2010 US census, is most closely aligned with overall unemployment rates.
We summarize average growth of employment under Democratic and Republican presidencies. Comparing employment growth all the way back to 1954 only allows to distinguish between White and non-White employment growth. During that time period, employment growth was slightly higher under Democratic presidents not only overall, but also for both Whites and non-Whites separately.
Moving on to the time-period from 1972 onward, for which more detailed employment data by race is available, we find a similar pattern: An outright comparison of relative performance again confirms[iii] that overall job growth was slightly higher under Democratic presidents than under Republican residents. Additionally, we find that all races saw higher job growth under Democratic presidents, on average. The difference is especially large for Asians and Hispanics.
To eliminate the potential influence of predecessor policies, we compare only years three and four across all presidencies. Overall employment growth is now somewhat higher under Republican presidencies. It is also higher for Blacks and Whites under Republican Presidents. However, Asians and Hispanics experience substantially higher job growth under Democratic presidencies.
Last, we examine the long term performance of presidents to study policy persistence and purported trickle down effects. We compare growth rates during years 5-8 and during years 7-8 of two term presidents. When examining years 5-8, overall employment growth is about the same under Democratic and Republican presidents, it is slightly higher for Whites under Republican presidents, but it is higher under Democratic presidents than under Republican presidents. When examining years 7-8, total employment and all races also perform better under Democratic presidents.
The observed differences in growth rates across racial groups and presidential terms are substantial. On average, during their second term Democratic presidents created one more job than Republican presidents per 1000 employed. They created eight fewer jobs for White workers per 1000 White employees, but 14 additional jobs for Black workers per 1000 Black employees, 63 additional jobs per 1000 Hispanic employees, and 82 more jobs per 1000 Asian employees than their Republican counterparts. These differences are even larger for the last two years of two-term presidencies, in which job growth under Democratic presidents has been larger than job growth under Republican presidents overall as well as for all racial groups.
In summary, comparing the ability of presidents to create jobs is a questionable endeavor. Too many factors outside of presidential policies influence job creation. Our goal here is more modest – to simply provide facts for discussion. These facts present themselves as follows: Democratic presidents have on average about an equally good record in creating jobs, with a slight advantage for Democratic presidents. Whether this can be attributed to presidential policies are questionable[iv]. However, so far unexplained is the differential job growth across racial groups: Employment statistics reveal that Blacks, Hispanics, and especially Asians have had more favorable job-market outcomes under Democratic presidents. This effect is especially pronounced when comparing two term presidencies. Presidential policies may take time to come to fruition, so comparing second terms may be especially instructive.
In upcoming reports, we investigate these stylized facts in more depth. Our next post explores the geographic distribution of job growth within the U.S. We will shed some light on the extent to which national differences in racial employment are observed on the local level.
[i] A statistical model attempting to evaluate presidential influence on job creation would have to control for all of these factors, if not more – an endeavor that is hard to achieve given the relatively few observations that are additionally plagued by numerous statistical issues.
[ii] Analyzing differential performance across groups may eliminate some statistical issues. Moreover, certain policies are targeted at particular groups. Both aspects reduce the complexity of models required to explain presidential influence on the job market.
[iii] E.g. Blinder and Watson (2014)
[iv] E.g. Blinder and Watson (2014) find no effect of fiscal or monetary policies that can explain those differences, Campbell (2011) points towards the importance of economic conditions inherited from predecessors in explaining those differences.
Authors: Johannes Moenius, Ph.D. and Jess Chen, Ph.D.
Institute for Spatial Economic Analysis,
University of Redlands, School of Business
Contact: Johannes Moenius, Ph.D., email@example.com, 909-557-8161
Director, Institute for Spatial Economic Analysis,
University of Redlands, School of Business