A Fresh—and Data-Driven—Look at the Effects of Unionization

A Fresh—and Data-Driven—Look at the Effects of Unionization

Wednesday, November 16, 2016

Unions have long been a lightning rod in American society. On one hand, they’re blamed for everything from the auto industry’s struggles to the current public employee pension crisis. But proponents will point to how they’ve ensured worker safety and provided livable wages and benefits, among other advances.

But for all of this controversy, there have been relatively few empirical analyses on the true impact of unions. Assistant Professor Aaron Sojourner has made a significant contribution to that with his groundbreaking look at nursing home industry unionization.

As Sojourner notes, previous studies have often used too wide of a lens. “They’ve compared unionized companies such as Ford with, for example, Walmart or Toyota, which are not unionized. But there are a lot of other differences between those organizations. It’s hard to control for them.”

Sojourner focused on a single industry, nursing homes, and made a very narrow, well-chosen comparison. He focused only on nursing homes where workers voted in a union election and the vote was close. To approximate a true experiment, he compared outcomes in nursing homes where the union just won an election to outcomes where unions just lost an election. Unionization activity in the industry has been on the rise in recent decades, so there were many hundreds of homes with close elections to examine. And the U.S. government’s close monitoring of the industry provided a rich trove of data to examine.

Sojourner’s research did uncover that unionization cut staff hours—not an unusual result of unions bargaining for higher wages and benefits. “We saw an overall 16 percent decline in hours per resident per day with nurse’s aides,” he notes. “And registered nurse hours fell by about half.”

But here’s the surprising thing: The data didn’t reveal a corresponding decline in the quantity or the quality of care. “Unionized nursing homes seem to be able to attract better nurse’s aides, perhaps because they’re paying more. Or maybe aides stay longer and build more skills,” he says. “We also didn’t see an impact on the survival of unionized nursing homes—they’ve stayed open as long as nonunionized ones. “Put together the decline of staff hours and the lack of the change in care quality,” he adds, “and it appears labor productivity actually goes up with unionization, at least when it comes to nursing homes.”

The competitive context may matter, however. The positive productivity effects appeared especially in local areas with fewer nursing homes and less competition between them. Unionization in less competitive product markets did not show evidence of productivity increases.

Watch Sojourner further describe his findings:

“Impacts of Unionization on Quality and Productivity: Regression Discontinuity Evidence from Nursing Homes”
Sojourner, Aaron J., Frandsen, Brigham R., Town, Robert J., Grabowski, David C., Chen, Min M, ILR Review (2015)