A paper of mine was just published in the Review of Environmental Economics and Policy. The paper is entitled “The Social Value of Job Loss and Its Effect on the Costs of U.S. Environmental Regulations”.

The paper deals with a key issue for analyzing a wide variety of policies, which either create or destroy jobs in a local economy.  This key issue is how to evaluate the social benefits or costs of such job creation or destruction. How are these benefits or costs related to the earnings associated with the jobs? Do such benefits or costs vary with the prevailing unemployment rate in the local labor market? How much might accounting for such social benefits or costs affect evaluations of the benefits or costs of a wide variety of policies, such as the environmental regulations considered in this article?

The environmental regulatory issue is that with the Great Recession, some groups opposed to more stringent environmental regulations argued that high unemployment should lead to some cutback of regulations that might hurt job creation or destroy jobs. Current regulatory practice is to describe possible job creation or job destruction effects in benefit-cost analyses of regulations, but not to include an explicit valuation of these job effects in the formal benefit-cost analysis. The benefit-cost analysis of regulations typically focuses on the regulation’s compliance costs versus its health and other benefits.

The argument of the paper is first, that policies that affect labor demand, either positively or negatively, typically have benefits or costs that are only a fraction of the earnings associated with those jobs. Second, these benefits or costs are much greater in local economies with high-unemployment, compared to local economies with low-unemployment. Third, environmental regulations largely redistribute jobs around the U.S. rather than affecting overall U.S. jobs, and therefore such regulations only have net national costs if they redistribute jobs away from high-unemployment areas. Fourth, even if environmental regulations tended to redistribute jobs away from higher-unemployment regions of the U.S., for most real-world environmental regulations, the total social costs of any lost jobs due to regulations are only a modest percentage of the regulation’s regulatory costs, and would not tip a benefit-cost analysis.

There are large social benefits or costs associated with a person getting a job or losing a job. These social benefits or costs probably exceed the earnings involved, for several reasons: there are important psychological benefits to being gainfully employed given modern social norms; job-holding has important spillover benefits for other family members.

However, when a job is created or destroyed in some local labor market due to some labor demand change, which might be brought about by environmental regulations or economic development incentives, then only a portion of that job’s earnings actually affects unemployment. A large proportion of local job creation or job destruction leads not to changes in employment rates, but rather to changes in in-migration or out-migration. These in- or out-migrants would have fared similarly in the labor market, and so are not substantially affected by the labor demand change.

If the local labor market has high unemployment rates, then less of any job change is reflected in changes in migration rates, and more in changes in unemployment rates. Therefore, the social costs per job created or destroyed is higher in high-unemployment local labor markets.

Overall, I find that due to changes in unemployment, the social benefit or cost of job creation is about 10% of earnings in a low-unemployment local economy, and about 19% of earnings in a high-unemployment local economy.

For most environmental regulations, including possible job destruction effects does not come close to tipping the benefit-cost analysis. Even under a worst case scenario, job destruction effects, and their social costs, are typically only a small percentage of the regulation’s overall compliance costs. For the major regulations that I examine, a plausible maximum value for the social costs of job loss is less than 15% of overall regulatory costs. Because benefits are in most cases far greater than overall regulatory costs, this slight boost to regulatory costs is insufficient to make a difference to the regulatory analysis. For most regulations, more precise estimates of health benefits is more important for refining the benefit-cost analysis than including social costs of job loss.

However, in cases where a regulation particularly impacts a high-unemployment area, it may be important to consider the social costs of job loss due to a regulation.  If these social costs are high, we might want to consider ways to offset these costs. This could be accomplished by delaying the regulation’s implementation, assisting affected workers to offset the social costs, or through job creation measures in the high-unemployment area.