Does regulation kill jobs? Yes, but as editors Cary Coglianese, Adam Finkel, and Christopher Carrigan explain, arriving at that answer is far more complicated than the title of their new book. Former OIRA administrator Cass Sunstein described the matter this way: “The Republican claim that ‘job-killing regulation’ is a redundancy is as ridiculous as the left-wing view that ‘job-killing regulation’ is an oxymoron.” The truth—or what the empirical evidence reveals—is far more muddled. Regulation can eliminate jobs in affected industries, but like any government intervention, it can also “create” them in professions that must comply with the rule.

Coglianese, Finkel, and Carrigan give special attention to the human element in this debate. Regulation might result in a transfer—instead of a net loss—of jobs from one sector of the economy to another, but that transfer involves a human toll, not mere abstract data in a politically charged debate. Through a compendium of chapters by an interdisciplinary panel of experts and veterans of the “regulation versus jobs” debate, the editors bring greater clarity—through sometimes-technical discussions—to the heterogeneous world of regulation.

In the book’s first chapter, Coglianese and Carrigan note the cyclical timing of the regulation-versus-jobs debate. Not surprisingly, in periods of high unemployment, conservatives and libertarians often identify regulation as a significant impediment to job growth. Progressives, on the other hand, see regulation as another government jobs machine, capable of creating millions of jobs at little to no cost. As former Environmental Protection Agency administrator Carol Browner once argued, “[T]he EPA creates opportunities [and] creates jobs.” If it were that straightforward, it’s a wonder progressive politicians don’t run on a platform of promising more regulation in order to create more jobs. So why don’t they?

The data / The book devotes considerable attention to the current empirical evidence on the employment effects of regulation. The heterogeneity of regulation and the size of the economy make extracting statistically significant evidence on the correlation between regulation and jobs a difficult task. Even if a rule eliminates 10,000 jobs, based on an economy-wide model, the total national effects are minimal.

In his contribution to the book, Resources for the Future senior fellow Richard Morgenstern (whose work is routinely cited by the EPA in regulatory impact analyses [RIAs]) surveys the available literature and notes the methodological difficulties confronting research in this area. He estimates that $1 million in environmental spending translates into 1.5 new jobs, but he notes that estimate is “statistically insignificant.” That hasn’t stopped some regulation proponents from arguing that environmental compliance could lead to a revolution in “green jobs.” Ceres, a coalition of environmental groups, has claimed that two new air pollution rules, Utility MACT and the Cross-State Air Pollution rule (CSAPR), would generate 1.5 million new jobs—300,000 annually—from the additional $11 billion in new regulatory spending. That translates into 27 jobs per $1 million in regulatory compliance—considerably more than Morgenstern’s estimate. The EPA has noted the vast amount of uncertainty on employment effects in its analysis of the two rules, concluding that Utility MACT would either cut 15,000 jobs or create 30,000, and CSAPR would reduce employment by 1,000 or increase it by 3,000.

Morgenstern cites Massachusetts Institute of Technology economist Michael Greenstone’s estimate that Clean Air Act nonattainment counties—that is, counties with air quality below federal standards, thus making the counties subject to stricter air pollution regulation—lost 590,000 jobs relative to attainment counties. However, those figures could have been the result of job transfers to attainment counties, meaning there were no net job losses. However, even if transfers explain the county job losses, transfers also have significant negative effects on workers. For instance, job displacement can lead to a 15 to 20 percent increase in death rates in the 20 years following displacement.

So what’s the conclusion? Morgenstern notes, “There is only limited evidence that environmental regulation leads to significant job loss.”

In another of the book’s chapters, Harvard University’s Joseph Aldy and Duke University’s William Pizer examine new research on regulation and the electricity industry. They find a significant nexus between rising, regulation-induced energy prices and declines in energy-intensive manufacturing industries. For example, a 5 percent increase in electricity prices reduces employment by 1.0–1.4 percent.

Those findings have profound implications for regulations like Utility MACT and CSAPR. Aldy and Pizer examine the economic effects of CSAPR, as even the EPA admits the rule could increase energy prices by $700 million. The authors find the rule could increase prices by 2.2 percent, enough to lower employment in certain energy-intensive manufacturing industries. The implications for onerous command-and-control regulation of power plants generate even more concern.

Reform / Those who have had the pleasure of reading an agency RIA might not remember the analysis’s sober discussion of employment effects. That’s because agencies routinely exclude employment effects from their analyses. That is one problem in the current debate: an omission of data from agencies.

As Rutgers University’s Stuart Shapiro, a veteran of the Office of Information and Regulatory Analysis, details in his contribution to the book, there is a woeful lack of employment information in most analyses. Of the 56 RIAs he studied, only a minority mentioned employment, with 11 quantifying impacts. Shapiro’s solution is to create a new federal office to study the employment effects of regulations. Ideally, the office would analyze regulations retrospectively.

The University of Chicago’s Jonathan Masur and Eric Posner offer another solution for the lack of available data: simply force agencies to incorporate employment effects in RIAs. However, the authors’ research details why agencies might have avoided this practice in the past. Masur and Posner looked at an EPA regulation that the agency initially estimated would have net benefits of $159 million but, in a separate analysis, conceded it would result in 5,711 fewer jobs, which the agency didn’t incorporate into its net benefits calculation. If a central estimate of $100,000 per displaced worker is applied to the cost-benefit analysis, the net benefits turn into a net cost of $411 million. If there is continued refusal by regulatory agencies to incorporate employment effects into their analyses, OIRA will need to step in and challenge proposed rules that would eliminate jobs and impose high net costs.

Conclusion / The book is devoted to the principle that this topic needs more empirical data and a study of the human effects of job displacement. From a cost-benefit perspective, a job transfer might be an afterthought, but for those losing their job—even for a time—there are significant emotional, physical, and pecuniary effects. As book contributor Brian Mannix notes, “No corporation has ever experienced a welfare loss”; that is, businesses do not feel the effects of regulation—but people do. As the book details, the debate over regulation and jobs is more than just political posturing and abstract figures.