Union workers in the United States are now more satisfied with their jobs than nonunion counterparts, according to a new study, offering a counterpoint to seminal scholarship dating back to the 1970s that linked union membership to a decrease in job satisfaction.
The working paper, published April 26 by the National Bureau of Economic Research, also found that union membership boosts reported life satisfaction irrespective of demographic differences such as wealth, gender, age and educational background. Researchers hypothesize this may be because unions protect members against underemployment, a growing concern in the United States.
More than 40 years ago, eminent scholars including Richard Freeman in 1978 and George Borjas in 1979 first proposed a negative relationship between unionization and individuals' job satisfaction in the U.S., a causal relationship that many empirical studies have since supported, including the work of Cornell University's Tove Hammer and Ariel Avgar in 2005. Multiple explanations for this phenomenon have been proposed: less content workers might opt for union protected jobs, and union positions might be more numerous in less desirable or more demanding industries, which could impact satisfaction despite better pay and benefits under union protection. Finally, union members might quit positions less frequently and remain in jobs for a longer duration than their nonunionized peers.
From the perspective of Benjamin Artz, co-author of the NBER paper and a professor of economics at the College of Business at the University of Wisconsin Oshkosh, the hypothesized relationship seemed odd, since union workers are often paid substantially more in wages and receive better fringe benefits than nonunionized workers. The landscape of organized labor had also begun to change, with the industries and occupations inhabited by union workers transforming, and he and his colleagues hoped to dig deeper than conventional explanations.
"[W]e've seen an interesting increase in union density recently, implying that unions have now done a good job of preventing as much job loss among its members than among nonunion workers," Artz told The Academic Times. "These changing circumstances generated interest in revisiting the relationship between job satisfaction and unions."
Using data on two cohorts of older and younger workers from the National Longitudinal Surveys of Youth in 1979 and 1997, researchers found that overall job satisfaction among private-sector union members has improved since the Great Recession. Among those surveyed in both cohorts, union workers had a lower mean job satisfaction than their nonunionized counterparts — except in the period since 2008, when unionized workers are significantly more satisfied with their jobs.
For the younger cohort surveyed in 1997, the mean job satisfaction of nonunionized workers was higher than that of union members until the onset of the Great Recession. After 2008, the expressed satisfaction gap between union-affiliated and nonaffiliated workers narrowed until approximately 2011, when the overall job satisfaction of union members grew higher than for nonmembers.
The positive association between membership and expressed satisfaction was stronger in the study's fixed effects model, suggesting that the emergence of positive union effects on job satisfaction was not caused by workers joining or leaving labor unions. In other words, the sorting of workers by union status does not appear to be behind the shift in expressed satisfaction. This offers a counter-perspective to previous findings by Alex Bryson, also a co-author of this paper and professor at University College London's Social Research Institute. In 2004, Bryson found that a selection effect, rather than a causal relationship, accounted for differences in satisfaction among unionized and nonunionized workers.
Artz and his fellow researchers did not find that rising job satisfaction can be explained by changes in the union wage premium, unions' wage bargaining tactics or the rate at which employed union members quit their jobs compared to the nonunionized. A decrease in unions' ability to lower quit rates was observed among older workers, but not for the younger cohort. Instead, the answer might lie in unions' capacity to shield their members from underemployment, a recurring problem faced by nonunion workers.
According to the study, Artz and his co-authors were the first to consider the interaction between unionization, underemployment and employee well-being in their estimation. Their results indicate that the observed positive shift in union membership's effect on job satisfaction following the Great Recession is partially a result of lower underemployment rates among union members, in addition to unions' ability to minimize the detrimental impact of underemployment on members' overall job satisfaction.
This insight could prove particularly important in the wake of the COVID-19 recession, with data from the U.S. Census Bureau and Labor Department pointing to persistently high rates of hardship and unemployment that are particularly pronounced for workers in low-wage industries.
Artz and his co-authors are not the first to challenge longstanding perspectives that advanced the theory of union membership exerting a negative impact on job satisfaction. In another NBER working paper from 2020, David Blanchflower, economics professor at Dartmouth College, and Bryson also found that the partial correlation between union membership and employee job satisfaction is positive and statistically significant. A 2002 study conducted in Canada examined differences in reported job satisfaction between union members and nonunion workers, with results suggesting that the negative relationship between union participation and job satisfaction disappears when controlling for disparate working conditions. Still earlier, in a 1990 study, economists discovered that unionization has a significant positive effect on satisfaction when job attributes are controlled.
Building on prior literature, as well as their own NBER paper, Artz says he and his colleagues would next like to investigate why union membership's effect on job satisfaction appears to have shifted following the Great Recession. Perhaps a particular aspect of union jobs has improved compared to nonunion positions, with more enjoyable employment undertaken by union members; or maybe a decrease in job satisfaction among nonunion workers is to blame. Regardless, Artz believes that organized labor will play a vital role in determining the quality of life experienced by many Americans, especially in less competitive labor markets.
"The U.S. is a country built on capitalism and competition. When labor markets are sufficiently competitive, unions are less necessary and may in fact be problematic," Artz said. "However, the costs and uncertainty associated with workers switching jobs and some employers' substantial market power provide for a less competitive landscape. Poor working conditions and income inequality result."
In response to the problems of American capitalism, the U.S. has typically turned to labor unions as a temporary fix, although they can be a divisive topic among Americans today. Critics commonly express concern about the potential threat posed by organized labor to business operations and economic prosperity. According to Artz, there is no reason to presume that workers rights under union membership must be detrimental to companies' success and the economy's stability over time.
"Prior to the 1980s, union workers and the firms that employed them enjoyed spectacular economic gains and security," Artz said. "The economy has indeed changed since then, but the problems workers face have not. Reestablishing unions' place in the labor market may be one solution."
The study, "Unions increase job satisfaction in the United States," published as a working paper by the National Bureau of Economic Research on April 26, was authored by Benjamin Artz, University of Wisconsin Oshkosh; Alex Bryson, University College London; and David G. Blanchflower, Dartmouth College, University of Glasgow, and NBER.