Understanding the reasons for and impact of legislatively mandated benefits for selected workers.

• Author: Donohue, John J., III
• Position: Response to Christine Jolls, Stanford Law Review, vol. 53, p. 223, 2000

1. INTRODUCTION

   Judging by the long and growing list of relevant federal initiatives, the demand for legislative measures designed to benefit workers must be quite strong. Federal law facilitates the unionization of workforces, establishes minimum wages, provides for social security, promotes health insurance coverage for workers through favorable tax treatment, provides the right to take unpaid leave for certain family emergencies, and regulates pension plans and worker safety. In addition, an elaborate array of "antidiscrimination measures" has developed to serve two functions: 1) to protect workers from receiving adverse treatment because of their race, sex, religion, national origin, age if over forty, or disability (the "disparate treatment standard"); and 2) to ensure that certain classes of "protected workers" are not adversely affected by employer conduct absent a showing of a strong business justification for the challenged practice (the "disparate impact standard"). Moreover, various states have chosen to augment these substantive federal protections or the remedies available to those denied these protections. Many states have created independent prohibitions against certain forms of wrongful discharge, and all states provide benefits to workers injured on the job.

   Note that most of these governmentally mandated benefits are directed to all (or virtually all) workers--and thus are referred to as "universal mandates." Even the disparate treatment component of the major federal employment discrimination law--Title VII of the 1964 Civil Rights Act--which the public might perceive as protecting only selected groups such as women and minorities, is generally comprehensive in its protections:(1) For example, white males have often employed these laws to attack alleged unfair treatment because of race or sex. Such cases are possible because the language of the federal law prohibits discrimination on the basis of race or sex--as opposed to discrimination against racial minorities or women--and thus the disparate treatment standard has been applied symmetrically. But some measures of federal law, notably the disparate impact standard, the Age Discrimination in Employment Act (ADEA), and the Americans with Disabilities Act (ADA), seek to provide specific additional benefits to certain identifiable demographic groups: blacks and other minorities, women, those over forty, and the disabled. Thus, a neutral employment practice that disadvantages white men yet has no business justification is permissible, while the same practice would be unlawful if it were to disadvantage women or minorities.(2) Similarly, refusing to hire anyone under forty is permissible (unless it happens to conflict with Title VII's disparate impact standard by disproportionately harming women or minorities), but refusing to hire anyone over forty quickly runs afoul of the ADEA. So too, the failure to provide a disabled worker with a reasonable accommodation violates the ADA, while a nondisabled worker can lawfully be denied any reasonable accommodation (with the caveat about the disparate impact standard applying once again).

   In her richly nuanced and impressively comprehensive article, Accommodation Mandates, Christine Jolls develops the theoretical framework needed to analyze the effect of this class of laws mandating employers to provide benefits to particular, presumably "disadvantaged," groups of workers.(3) Specifically, Jolls enriches the existing economic model of universal mandates in order to analyze the impact on wages and employment of these targeted labor-market interventions.

   One can easily go on at great length about the virtues of this paper. First, it takes a broad area of disparate elements in the law and develops a unified and original theoretical framework for analyzing a variety of labor market interventions. Second, Jolls has been extremely meticulous and comprehensive in her analysis, and the piece is a model of clarity and precision. She is careful to begin with the most general analysis and then show how the predictions of theory need to be amended by additional qualifications and caveats. With very little math to tax the reader, Jolls is able to generate the type of crisp conclusions that is the mark of a good work in theoretical economics. Third, since the analysis is so rich and sophisticated, many of the conclusions will not be easily intuited through less rigorous approaches, underscoring the value added of the paper. Fourth, by tying the predictions of theory to the existing empirical literature, Jolls has provided a remarkable blend of theory and empirics that is a model of first-rate scholarship. In any paper of this degree of ambition and scope, there will frequently be points on which commentators could object, but I doubt any reader could fail to see the power and the depth of the analysis taken as a whole.

   Section II of this comment will describe the outlines of Jolls' approach and highlight her major theoretical conclusions. Section III attempts to illustrate the value of her theoretical framework by applying it to a few areas that Jolls did not explore in this paper--the law of sex harassment and disparate impact.(4) Section IV raises some broad questions about why accommodation mandates would be adopted if they don't entirely advance the interests of their intended beneficiaries as the Jollsian framework suggests. Section V explores some of the limits of partial equilibrium analysis of labor market interventions in light of work examining the distributional impact of minimum wage laws and recent controversial empirical work questioning the theoretical predictions of such analyses. Section VI concludes with some general comments on whether the adoption of accommodation mandates is sensible public policy.

2. OVERVIEW OF JOLLS' PAPER

   Jolls begins her paper with the established partial equilibrium analysis of universal mandates, which she then modifies to assess the impact of the targeted mandates that she refers to as "accommodation mandates."(5) This established framework is based on the simple price-theoretic model in which the intersection of the supply and demand curves for labor identifies the equilibrium wage and quantity of labor hired. How will this equilibrium change if employers are legally required to add a benefit for every worker? The simplest, perhaps trivial, case is where the law tells every employer to give an annual bonus of, say, $1000 to every worker. The basic economic model says that the demand curve will shift down reflecting the fact that every worker is suddenly less attractive by $1000 than they were previously, and the supply curve will also shift down by the same amount because each job is now suddenly $1000 more attractive to workers. The net effect is that: 1) the wage will fall leaving the worker with the same total compensation (wage plus bonus) that existed prior to the passage of the law; and 2) the quantity of labor hired will remain unchanged.(6)

   Of course, it is easy to see why this Coase-like invariance prediction would not occur in the short run. Some workers would be governed by existing collective bargaining agreements that would prevent the employer from lowering the wage until the end of the contract was reached. Even those workers who were not covered by a union contract might have explicit or implicit contracts that would prevent employers from immediate downward wage adjustments. Nonetheless, over time employers would likely have greater flexibility to respond to the mandated bonus, and one would suspect that they would then lower the wage or reduce future pay increases to offset the mandated bonus. As a result, one might expect that in the long run such a universal mandate would have little or no effect on the total compensation and employment of workers as the simple model predicts.

   From this familiar foundation, Jolls launches her comprehensive assessment of accommodation mandates versus mandates directed at workers as a whole. To make this advance, she notes that the accommodation mandate will only directly influence the supply and demand of a portion of the labor market--that governing the intended beneficiaries. Moreover, these intended beneficiaries are generally (and, in the examples that Jolls discusses, uniformly) defined by employment discrimination law to be members of protected classes--for example, the disabled and women of childbearing age. As a result, federal law imposes legal restrictions on lowering wages and reducing employment for these protected workers, which implies that employers are legally constrained in how they respond to targeted mandates in a way that does not exist when they face universal mandates.

   Jolls shows that, because of the failure to consider the impact of antidiscrimination law, naive extrapolations from the analysis of universal mandates to the targeted mandates that are the focus of her interest can lead to an array of misleading predictions. Jolls analyzes three cases that depend on the extent to which the legal constraints actually impede the employer responses: 1) the wage and employment restrictions of employment discrimination law are both fully binding; 2) the wage restrictions alone are binding; and 3) neither restriction is binding.(7) For each of these three cases, Jolls carefully details how the impact of the targeted mandate is influenced by whether the value of the mandated benefit to the workers is greater or less than the cost to the employer of supplying it. Jolls shows that, unlike the case of the comprehensive mandate discussed above, accommodation mandates have a greater scope for benefiting their intended beneficiaries (at the expense of the so-called "nondisadvantaged workers") when both the wage and employment restrictions of antidiscrimination law are fully effective. Indeed, this finding can hold even when the value of the benefit is less than its cost, which Jolls notes is...