Plant Closures Cause Job Loss

David M. Driesen - Syracuse University College of Law

Posted in ,

David M. Driesen

[This is the third post in a three-part response to Jonathan S. Masur and Eric A. Posner, Against Feasibility Analysis, 77 U Chi L Rev 657 (2010).]1

Because plant closures cause job loss, any cost-benefit analysis (CBA) that counts job loss as a consequence relevant to overall well-being would have to include feasibility analysis. Hence, Masur and Posner must choose between embracing feasibility analysis and taking the indefensible position that job loss is irrelevant to overall well-being.

They seek to escape this problem by trying to disassociate plant closure from job loss. Masur and Posner point out that plant closure may not track job loss, because an employer might transfer workers when a plant shuts down.2 Their transfer argument applies to any job loss, not just to those associated with plant closure; a firm can transfer a fired worker at a still-running plant. And it is a very weak argument. Job transfers seldom occur, do not help workers who cannot or will not move, and would prove exceedingly difficult to predict. Any predicted consequence of regulation might not occur because of some deus ex machina, but, if we are to engage in any analysis of welfare effects at all, we must distinguish likely from unlikely palliatives.

Masur and Posner argue that rather than tracking shutdowns, agencies should count job loss “directly.”3 The argument for measuring jobs directly proves extremely misleading. Any agency measuring job loss “directly” would include an analysis of whether the costs imposed would lead to plant closures, as feasibility analysis demands. Agencies do this all the time with bankruptcy models and other tools, as Masur and Posner acknowledge. Messy as it may be, regulators have no more direct way of analyzing the question whether plant closures leading to unemployment might occur.

Masur and Posner stand on more solid ground when they point out that some job loss can occur outside the context of plant closures. Agencies’ tools for predicting these job losses, however, are no more direct and are more error-prone than their tools for predicting plant closures. They involve figuring out whether price increases to reflect regulatory costs would lead to consumers simply paying higher prices or instead to reduced consumption. If raising prices would reduce consumption, modelers sometimes predict that regulated firms would fire workers, but they might instead reduce wages, reduce profits, lower benefits, or lower dividends to shareholders. While predicting job loss through plant closures is not completely reliable either, it is a much safer bet that if costs bankrupt an owner or make facilities unprofitable, plants will close.

A combination of practical and theoretical considerations can justify a focus on shutdowns. First, it might make sense to focus on the most predictable job losses—those likely to occur when plants shut down. Second, plant shutdowns are much more likely to cause widespread job losses than measures that do not shut down plants. Third, plant shutdowns are much more likely to produce permanent unemployment in a significant number of cases, because increasing production at an open facility that has terminated some employees is much easier than starting a new plant once the old one has been shut down. Fourth, industry predicts job losses all the time, yet they rarely materialize. Industry lobbying on this politically sensitive point produces great potential for agency error, especially because the industry controls much of the relevant information about cost, market structure, substitute products, and so on. Confining agencies to feasibility analysis, at least for regulations governed by the feasibility principle, may reduce error.

In any event, Masur and Posner’s observation that feasibility analysis is under-inclusive with respect to job loss does not justify rejecting feasibility analysis, it only justifies supplementing it with efforts to predict job losses outside the shutdown context. In practice, agencies usually estimate both types of job loss and take them into account in promulgating technology-based regulations.

Against Feasibility Analysis’s strenuous effort to divorce plant closure from job loss fails. Any analysis that takes job loss seriously would have to estimate the number of plant closures in an industry.


David M. Driesen is a University Professor at Syracuse University College of Law.

Copyright © 2011 University of Chicago Law Review

This Legal Workshop post is based on an article forthcoming in volume 35 of the Harvard Environmental Law Review.

  1. For the previous posts, see David M. Driesen, How Cost-Benefit Analysis Incorporates and Worsens Feasibility Analysis’s Flaws (Legal Workshop, Feb 7, 2011), online at 2011/02/07/driesen-2 (visited March 1, 2011); David M. Driesen, A Modest Normative Case for Feasible Regulation (Legal Workshop, Dec 13, 2010), online at (visited March 1, 2011).
  2. Jonathan S. Masur and Eric A. Posner, Against Feasibility Analysis, 77 U Chi L Rev 657, 695 (2010).
  3. Id.

Post a Comment (all fields are required)

You must be logged in to post a comment.