New evidence shows that unions played a major role in reducing income inequality in the United States in the decades when organized labor was strong.
But it also demonstrates that the decline in union power since the 1960s — which may be exacerbated as a result of a recent Supreme Court decision — has contributed to the widening gap between rich and poor.
The new insights come from a working paper, “Unions and Inequality Over the Twentieth Century: New Evidence from Survey Data,” by four economists: Henry Farber, Daniel Herbst and Ilyana Kuziemko of Princeton, and Suresh Naidu of Columbia. They establish that unions have constrained income inequality far beyond their own membership ranks.
While the scholars can’t pinpoint the precise mechanism at work, they speculate that unions have indirectly increased pay at firms nervous that their own employees might organize. Unions have also lobbied for higher minimum wages and pushed to hold down executive salaries. They have also advocated for broader access to health care, countering a key channel through which income inequality can harm all of society.
The findings are particularly relevant in light of the Supreme Court’s June 27 decision in the case of Janus v. American Federation of State, County and Municipal Employees. The court ruled that states can no longer require public employees who are represented by a union — but have chosen not to formally become members — to contribute to the costs of collective bargaining. That will certainly hurt unions financially, and it may lessen their already diminished power.
Income inequality began its steep rise in the 1970s. Economists have been arguing about the origins of this trend since, with the primary explanations falling into two camps.
The dominant narrative, described in “The Race Between Education and Technology” by Professors Claudia Goldin and Lawrence Katz of Harvard, is that scientific progress has given the most educated workers the upper hand through “skill-biased technological change.” The theory goes that companies have bid up wages for workers who are best able to adopt new technologies, while demand for other workers has stagnated. This narrative is bolstered by rising levels of earnings for college-educated workers.
But another explanation for increasing income inequality centers on the erosion of the minimum wage and the decline of unions. Economists in this camp emphasize changing norms, institutions, and politics — not just market forces — as important drivers of the widening gulf between rich and poor.
The debate has real-world consequences.
If market forces are primarily responsible for the growing inequality, then the best we can do, from an economic standpoint, is to try to buffer their negative effects on low-skilled workers through “post-market” policies like taxes and social welfare programs. But if institutions matter more, then we can reduce inequality with market-oriented policies that, for example, bolster the minimum wage or ease the formation of unions.
Until now, the study of unions’ effect on inequality has essentially started with the ’70s, because good data has been hard to come by for any time before then. But it was hard to tell a complete story about how the rise and fall of unions affected economic inequality because the data is confined to a time during which unions were already in decline.
In the new study, the four scholars have mined newly available Gallup Organization data going back to the 1930s, based on surveys of American households that include questions about political beliefs as well as union membership, education, and income. A rich trove of these older surveys is now publicly available at the Roper Center at Cornell University.
The four economists painstakingly cleaned and coded hundreds of these surveys spanning nearly 90 years. The data encompass the growth of unions during the 1930s and ’40s, their heyday in the ’50s and ’60s, and their slow decline to the present.
Union workers now earn about 20 percent more than nonunion workers in similar jobs. Remarkably, this union premium has held steady since the 1930s.
Throughout this period, the biggest boost from union membership has gone to the least educated workers, who have, in turn, driven the rise and fall of union membership. The decades following World War II, when unskilled workers formed the union movement’s backbone, marked the most rapid decreases in income inequality. Wages for nonwhite workers were particularly strong then.
But increasing wages for low-skilled union members is just one channel through which unions can reduce income inequality. Unions can also affect the earnings of nonunion workers.
To capture such effects, the researchers broadened their lens to include the entire distribution of workers and their wages beyond those who are in typically unionized jobs and industries. They found that, going back to the 1930s, more unions meant more income equality. During years and in states where workers were more likely to be unionized, income inequality was lower.
I will admit freely that I’m predisposed toward unions. I’ve seen their effects in my own life. My father was a high school dropout, but as a unionized mechanic at the United States Postal Service he earned a solid wage. His union paycheck (along with my mother’s low paid, nonunion job), financed a house, a car, and Catholic school educations for three daughters.
Before I trained as an economist, I spent six years as an organizer, forming unions among secretaries, library workers, and lab assistants at Harvard and the University of Minnesota. I saw firsthand the increased economic security that unionization brought these predominantly female workers, in the form of higher wages, more generous pensions, and paid maternity leave.
Thanks to the new research, evidence going back nearly a century now shows that unions have formed a critical counterweight to the power of companies. They increase the earnings of the lowest skilled and sharply reduce inequality.
But the Supreme Court’s decision will curtail the capacity of unions to organize and represent workers. The court ruled that unions can no longer collect “agency fees” from those government workers whom they represent but who have chosen not to join. These fees have helped pay for contract negotiations as well as prevent the free-rider problem that arises when only some pay for benefits enjoyed by everyone.
Incomes in the United States are now as unequal as they were in the 1920s. The gulf between rich and poor will widen if, as I fear, unions are weakened further.
Susan Dynarski is a professor of education, public policy and economics at the University of Michigan. Follow her on Twitter: @dynarski.