Union membership in the U.S. private sector has declined considerably over time, dropping from about 35 percent in the mid-1950s to about 7 percent in the mid-2010s. Typically, the labor market impacts of this decline, which implies waning union power, have been studied in the context of wages and employment. As a result, comparatively less is known about how suppressed unionization may affect nonpecuniary working conditions. In a recent article titled “The impact of right-to-work laws on long hours and work schedules” (National Bureau of Economic Research, Working Paper 31867, November 2023), Rania Gihleb, Osea Giuntella, and Jian Qi Tan seek to fill this gap by examining the effects of shifts in union bargaining power on two factors affecting worker welfare: long working hours and nontraditional work schedules.
To estimate changes in union strength, the authors take advantage of the varying adoption of right-to-work (RTW) laws across U.S. states. These laws, now enacted by 27 states, eliminate agency-shop protections (arrangements in which workers must pay union dues as a condition of employment), thus reducing union power. Using data from the American Community Survey for the 2005–19 period, the authors compare welfare outcomes within pairs of adjacent counties located in states with differential adoption of RTW laws, while also ensuring that these outcomes trended similarly in each pair before the adoption. Long work hours are estimated with various measures, the main one being a binary variable capturing work exceeding 45 hours per week. Nonstandard work schedules are identified by examining the times at which employees arrive at work (for example, arrival between 5 p.m. and 8 a.m. would indicate a nonstandard shift).
The authors’ central finding is that the adoption of RTW laws increases the share of employees working long hours (as measured by the binary variable) by about 6 percent relative to the mean, suggesting that reduced union power weakens worker protections along this welfare dimension. This effect is stronger among male and Black workers, as well as in industry sectors with high unionization, such as construction, manufacturing, and transportation. An alternative continuous measure of working hours also suggests that the enactment of RTW laws is associated with a 0.5-percent increase in the average number of hours worked, although this effect is statistically significant only in manufacturing and transportation. Economically sizable increases on the continuous measure are also observed for blue-collar and Black workers, as well as those ages 25 to 44.
While RTW laws significantly affect the odds of people working long hours, they appear to have a weak impact on nonstandard work schedules. The authors report that, in the main sample, the rollout of RTW laws slightly increases the incidence of starting work between 5 p.m. and 8 a.m., although the strength of this effect is sensitive to data quality and method of analysis. However, some model results show notable differences by industry and worker demographics. For example, in counties adopting RTW laws, the share of people working nonstandard schedules rises sharply in the sectors of education and public administration, both of which are highly unionized. In addition, significant increases occur for Black, Hispanic, and younger workers.
The authors also offer some evidence on the effect of RTW laws on hourly wages and union coverage, topics covered in previous research. They find that, after the adoption of RTW laws, wages generally decline, especially in the manufacturing sector, as does union coverage. However, the wage estimates obtained for the main sample are imprecise and vary by period studied.
United States Department of Labor
