By Matthew Kredell
USC Price research professor Adam Rose received the Best Article of 2018 Award from the Journal of Benefit-Cost Analysis for his co-authored paper “Welfare Analysis: Bridging the Partial and General Equilibrium Divide for Policy Analysis.” (Read the full paper here.)
The article evaluates the strengths, limitations and relationships between benefit-cost analysis in individual markets and across multiple, interrelated markets, and the economy as a whole.
Rose and co-author Scott Farrow, a professor at the University of Maryland-Baltimore County who works with Rose through USC’s National Center for Risk and Economic Analysis of Terrorism Events (CREATE), made progress in reconciling the partial equilibrium approach favored by government regulatory offices and the general equilibrium approach that captures a broader range of important considerations in policy evaluation.
The selection committee offered the following praise for the article: “Farrow and Rose provide a comprehensive literature-based comparison of the differences among partial equilibrium, linear input-output, and non-linear general equilibrium models that is accessible to readers.” The committee also commended the authors for their “fair-minded consideration of the arguments against their recommendation”.
“Personally, it was gratifying to receive recognition for addressing a longstanding issue in practical policy making,” Rose said. “This issue deals with what can and cannot be included in regulatory analyses of government policy, and my co-author and I are hoping we’ve torn down some barriers in utilizing this broader general equilibrium approach.”
Benefit-cost analysis compares the advantages and drawbacks of a particular action or program. Partial equilibrium analysis looks at one specific sector of the economy as functioning in isolation from all other sectors, while general equilibrium analysis is a comprehensive study of several interrelated economic variables for understanding how the economic system functions as a whole.
Currently, the U.S. Office of Management and Budget guidelines for benefit-cost analysis, which have been adopted by most federal government agencies, contain strictures against including general equilibrium effects. The authors make the case that such limitations leave out important economic impacts.
Farrow and Rose concluded that governments should relax rules against general equilibrium approaches to benefit-cost analyses.
“If we’re successful in relieving some of the strictures against including these general equilibrium effects, it will lead to more accurate cost-benefit analyses because it will include all benefits and costs that should be counted,” Rose said. “It will also take into account effects on a broader range of stakeholders and people affected by these policies than if you just look at effects on a specific industry or sector.”
The paper is the culmination of many years of research on the topic area by both authors. Rose began his research into economy-wide models with his PhD dissertation more than 45 years ago. He pioneered the application of applied general equilibrium models to evaluating the consequences of and resilience to natural disasters and terrorism, and has also done extensive work on the aggregate and income distribution general equilibrium impacts of climate mitigation policy.
“I have run into some obstacles in receptiveness of some of my results from purists in government agencies who wanted to narrowly confine impacts,” Rose said. “My co-author is from the partial equilibrium realm where cost-benefit analysis is usually employed, and to his credit questioned it along the way.”
An impetus for the article came from Rose’s time serving on an EPA Science Advisory Board panel on economy-wide modeling, in which leading experts in the field were charged with examining this issue because there are an increasing number of EPA policies that are likely to have more economy-wide repercussions, such as a nationwide carbon tax.
Director of the Center for Risk and Economic Analysis of terrorism Events (CREATE)