In the debate over legislation to expand employees’ right to choose union representation in the workplace, the organized business lobby has been drumming up fears that enactment of the Employee Free Choice Act would kill jobs by forcing more employers out of business. That claim is not borne out by historical data or existing credible research, according to a new report released today by the Economic Policy Institute. In “Still Open for Business” John DiNardo, professor of economics and public policy at the University of Michigan (see full bio below), compared data on business failures among unionized and similar nonunion firms. He concluded that unionized businesses are no more likely than nonunion ones to fail.
“The conclusion that unions drive firms out of business is often discussed as if it followed inexorably from incontrovertible premises, but it doesn’t,” explained DiNardo. “ It’s ultimately an empirical question whether the simplest supply and demand story taught in introductory economics textbooks bears on the question of economic consequences of unionism on business failures – and the best evidence suggests that it does not.“
DiNardo’s paper reviews research by others which compares unionized to otherwise identical non-union workers and consistently finds that union workers are paid better. Even among identical twins, for example, the unionized twin earns more than his nonunion brother or sister. While no single mechanism has been identified that reconciles this evidence with the evidence from “natural experiments” that unionization has no effect on business failures, there are several possible ways that are consistent with the evidence. One way, for example, is that unionized firms are adapting rather than closing – often with changes that could end up being beneficial. In some cases the firm that pays a little more to its unionized employees might pay a little less to its CEO and managers, for example, resulting in less overall wage inequality in the workplace. In others the higher costs for wages might be offset by productivity gains. But DiNardo emphasizes that the best evidence shows that unionization has no causal effect on business failures.
DiNardo’s analysis explains that the theory underlying the claims about business failures, taken to its logical conclusion, would produce workplace scenarios so far divorced from reality as to be absurd: wholesale displacement of workers or business failures based on minuscule fluctuations in pay rates, for example. As he explains, the theory’s simplicity makes it seem appealing – but that simplicity is also what makes it a bad predictor of actual events. The reality, as DiNardo’s review of the data and existing research shows, is that in this case simplicity is simply wrong.
About the Author:
John DiNardo is Professor of Economics and Public Policy at the University of Michigan and research associate at the National Bureau of Economic Research. His research focuses on applied econometrics, labor economics, health economics, political science and econometrics. Recent work has included characterizing the finite sample performance of so-called semi-parametric treatment effect estimators, evaluating “accountability” standards for public schools, the labor market effects of Hawaii’s health insurance mandate, and the accuracy and reliability of pre-election polls, among other things. Some of his current projects include a chapter for the forthcoming Handbook of Labor Economics on Policy Evaluation Methods, a fifth edition of an econometric textbook, Econometric Methods, (Johnston and DiNardo), constructive proposal for attrition and non-response, and the health effects of obesity.