Boston University Public Health Post
June 20, 2022

COVID resulted in greater attention to the role unions can play in promoting the health of workers and their families. Blue-collar and essential workers faced the virus every day while many of their managers and most white-collar workers were able to work from home. Unions were the vanguard, advocating for the provision of masks, personal protective equipment, distancing, clean workplaces, and hazard pay. An unprecedented number of strikes occurred in the fall of 2021 resulting in what some labeled “Striketober.”  And Christian Smalls, the leader of the new Amazon Labor Union, stated that, without management’s indifference to COVID, he would never have tried to organize his co-workers.  While union membership is at a 65-year low (6% in the private sector), public support for unions is 68%, a 55 year high.

My colleague, Bozhidar Chakalov, and I recently published a literature review exploring the relationship between unions and health. We measured health not just by workplace injuries, illnesses, and absences, but by other measures involving workers and their families at home and elsewhere. These measures included mortality, hypertension, smoking, obesity, and depression among workers and low birthweight among their children.

Although we looked for studies published over a period of 41 years (1980 to 2021), we were surprised to find only 16 studies that tested for direct links between unions and workplace injuries, illnesses, and absences. And we found just five looking at the relationship between unions and other health measures. Only one study considered associations between union membership and COVID infections, reporting that union membership was connected with fewer infections.

When we expanded our search to include indirect links between unions and health we found hundreds of studies. Indirect links involve effects of unions of some work-related factor and the effects of that factor on health.  For example, economists have reported unions reduce wage discrimination. And epidemiologists, in turn, have found that discrimination likely harms health.  We identified 27 indirect links in addition to discrimination: wages, income inequality, overtime, pensions, job security, employer-provided health insurance, paid sick and family leave, Occupational Health and Safety Administration inspections, worker compensation benefits, shift work, gig jobs, piece-rate pay, job satisfaction, job strain, social support, justice at work, health-promotion programs, and self-esteem, among others.

Our review also assessed the strength of the direct and indirect links under the categories: “consensus”, “likely”, “disputed” and “unknown.” Only four links achieved “consensus”: employer-provided health insurance, OSHA inspections, dangerous working conditions, and injury deaths. Foremost among the “likely” links were wages, income inequality, and paid sick and family leave. For example, whereas economists agree that unions lift wages, there  are only likely (not  consensus) findings that higher wages benefit worker and infant health of employed parents.

One way low wages harm health is through effects on self-worth. Martin Luther King Jr. emphasized these effects when he pointed out that the only thing that makes a job “menial”  is the low wage. Economists, again, agree that unions lower wage inequality but not necessarily income inequality. This is because income includes dividends, rents, interest, and capital gains—which are highly concentrated among the wealthy—in addition to wages. Epidemiologists, in turn, do not uniformly agree that income inequality harms health. Although virtually all studies show that unions encourage paid sick and family leave, the findings pertaining to effects on worker or family health are not uniform. We nevertheless conclude that lower income inequality and paid leave likely improve health.

Finally, unions are correlated with higher non-fatal on-the-job injury rates and absenteeism. But this is not evidence that unions cause higher injury or absence rates. On the contrary, it is likely that high rates cause unionization just as the Christian Smalls/Amazon example above illustrates.

Whereas there is a sizeable amount of economic research on unions, there is little epidemiologic research. This is unfortunate for two reasons. First, public interest in unions is expanding, especially among workers under age 35. Second, as our literature review demonstrates, unions can improve health via numerous pathways for workers and their families. I predict rapid expansion in studies addressing the public health effects of labor unions. Unions are underappreciated social determinants of health.