Analyzing Regulation’s Effect on Jobs

white house aerial
Font Size:

Lisa Robinson argues that recent high unemployment highlights the need to assess employment effects of regulation.

Font Size:

U.S. newspapers’ use of the phrase “job-killing regulations” has exploded in recent years, rising by 17,550% from 2007 to 2011, according to the Institute for Policy Integrity at the New York University School of Law.  Yet are these “job-killing” claims true? Currently, little information is available on the extent to which regulations eliminate or create jobs.

In a recently published paper, Lisa Robinson, Senior Fellow at Harvard Kennedy School’s Mossavar-Rahmani Center for Business and Government, articulates a set of recommendations for developing best practices to assess regulation’s effects on jobs.  Under two executive orders, agencies are required to assess the costs and benefits of a proposed action if it is “economically significant,” having an annual economic effect of at least $100 million.  The 1993 Executive Order 12866 instructs agencies to consider employment effects as a part of this analysis; the 2011 Executive Order 13563 lists promotion of job creation as one of the U.S. regulatory system’s goals.  Yet Robinson finds that most regulatory analyses pay relatively little attention to jobs, assuming that any losses or gains are temporary transition effects given the expectation that the economy will be operating under conditions of full employment.

Recent high unemployment rates, however, have thrown the need to assess these effects into sharp relief, says Robinson.  While recognizing the complexity of such analyses, the Office of Management and Budget (OMB) has encouraged agencies to measure employment effects and has solicited public feedback on the relationship between regulation and employment.  As agencies investigate how best to incorporate job effects into regulatory analyses, Robinson suggests they consider a set of nine principles.

Robinson argues that the need for analysis will depend on the circumstances, but is particularly  important when the effect of regulation on employment is a topic of substantial public attention and political debate.  A finding that a regulation will have negligible effect on employment is as important as a finding that such impacts will be significant. Because agencies have limited resources, she proposes they conduct screening to determine the appropriate level of analysis.  To support such screening, she recommends that agencies develop a research database, that both provides access to useful studies and data sources and involves developing criteria to assess the quality and applicability of these resources.  Should screening indicate that employment effects are likely to be substantial, the agency would take further steps to estimate the number and types of jobs expected to be gained or lost.

Robinson emphasizes the importance of choosing an appropriate baseline – that is, the reference point of what employment would be without the proposed regulation.  Because the federal rulemaking process is characterized by a time lag between when the regulatory analysis is developed and when the regulation becomes fully effective, she warns against assuming that employment conditions will remain the same as currently.  In cases where uncertainty in the baseline assumptions is likely to significantly affect the analytic results, Robinson recommends that agencies compare the regulation to alternative baselines.  For example, an agency could consider one baseline that assumes a return to full employment by the time the regulation is implemented and another baseline which assumes continuing high unemployment.

Conducting such analysis requires confronting many analytic challenges, given the limited research available.  Whether regulation will increase or decrease jobs depends on industry and market conditions; production with the regulation may require more or less labor in total than production under the baseline.  Regulatory analyses should recognize the complexity of job effects, and include information on the likely timing and duration of job losses and gains, the effect on wages, and nature of the jobs and individuals affected.  Such analysis should consider the effects of regulatory benefits as well as costs; in particular, improved health and longevity will also affect productivity and individuals’ decisions regarding labor market participation. When regulation is expected to lead to both job gains and losses, agencies should report both the increase and decrease rather than solely the net effect, according to Robinson.  Failure to do so, she says, could, for example, erroneously suggest that a loss of one job offset by a gain of two jobs is no more important than 100,000 jobs lost offset by 100,001 jobs gained.

Employment gains and losses must be placed into context, Robinson argues.  A simple statement that a proposed regulation would create or eliminate “X” jobs in insufficient; the change should be compared to both employment conditions more generally and the net benefits of the rule.  Agencies should also clearly delineate what they mean by employment losses and gains.  For example, Robinson says,

[A]re they counting the number of individuals employed full-time or part-time?  Or full-time equivalent employment?  At the plant level, firm level, or industry wide?  In what industries and geographic areas and over what time period?  In the case of job losses, to what extent does the analysis include re-employment somewhere?  If new jobs are created, does the analysis consider whether they are filled by workers previously employed elsewhere or by those who were unemployed?

Finally, Robinson notes that there may be cases where certain types of employment effects should be integrated into the calculation of the net national benefits of the regulation, rather than solely included in an economic impact analysis that considers how the costs of the rule are distributed.  She indicates that labor can be conceptualized in two ways: first, as “human capital,” a resource that decreases in value if long-term unemployment causes skills to deteriorate, and second, as a source of an individual’s happiness or “utility” because it provides income used to buy goods and services.  Any such analyses should compare the value of jobs lost or gained to the value of what individuals would be doing in the absence of the regulation, including paid and non-market work and leisure, as well as job search activity and involuntary free time.

The practice of including jobs in regulatory analysis is rife with uncertainties, Robinson acknowledges.  The gaps in the available research present major challenges.  However, when concerns about high unemployment are widely-voiced, Robinson concludes that such analysis will pave the way for more informed regulatory decision-making.

Robinson’s paper appears as a chapter in the recently published book, Does Regulation Kill Jobs?, edited by Cary Coglianese, Adam Finkel and Christopher Carrigan.



This essay is part of The Regulatory Review’s six-part series, Does Regulation Kill Jobs?