The new highly publicized movie “On the Basis of Sex” offers a somewhat fictionalized account of the early professional life of Supreme Court Justice Ruth Bader Ginsburg. Intermingled with her life story, the film presents an idealized narrative of her early legal crusade against gender discrimination, fought in part with her late (and most devoted) husband, the eminent tax lawyer Martin Ginsburg.

Ginsburg argued or participated in several of the early influential cases on sex discrimination and went on to found the Women’s Rights Project at the American Civil Liberties Union. When she started teaching, she was one of only about 20 female law professors in the United States.  She was very much a pioneer in the women’s rights movement, motivated by her own life experiences. She had on numerous occasions been rejected from positions solely on grounds of her sex, notwithstanding her great academic distinction, and was well aware that similar obstacles fell in the path of other women who sought to make a career in the law. The film goes into these issues in depth, but I shall not dwell on them here. I am a lawyer, not a film critic, so I will comment only on Justice Ginsburg’s substantive arguments against gender discrimination

Most legal writers support Justice Ginsburg’s position that both the Due Process and the Equal Protection Clauses of the Fourteenth Amendment prohibit government discrimination on the basis of sex. I offer a split verdict on her legal efforts and of those who followed in her path. I think that she was right on the early cases that sought to get rid of senseless distinctions based on gender.  But as the law subsequently developed, she and the courts pushed the crusade too far, creating new forms of gender imbalance that the law should have resisted. Failure to understand the economics of discrimination have led courts to impose new versions of the very discrimination that the law is intended to eliminate. In general, truly competitive markets do a better job in rooting out gender discrimination than government regulation.

The antidiscrimination norm has a powerful a hold on the legal imagination. When properly applied to government action, it reduces or eliminates implicit transfers of wealth or opportunities from a disfavored to a favored group. Just that situation arose in Reed v. Reed, a unanimous 1971 Supreme Court decision handed down by Chief Justice Warren Burger. Ginsburg was one of the authors of the winning brief, but did not get to argue that case before the Court.

At issue in Reed was which of two separated adoptive parents of the minor decedent was entitled under Idaho law to administer his estate, given the absence of any designation by will. The applicable law preferred the father over the mother, solely on the basis of sex. Chief Justice Burger noted that the government normally has the power to make tax classifications, but then added that these classifications had to have some rational relationship to the ends in view, which this Idaho classification did not. The state could not justify this male preference, as it vainly tried to do before the Court, on the ground that it would save the cost of having a hearing to decide who should receive the honor. A coin flip would work every bit as well.

Reed set the table for Moritz v. Commissioner of Internal Revenue, which Ruth and Marty Ginsburg argued jointly before the Tenth Circuit in 1972. The applicable provision of the Internal Revenue Code, Section 214, provided a deduction for care of certain dependents, which was only available to “a taxpayer who is a woman or widower, or is a husband whose wife is incapacitated or is institutionalized.” Charles Moritz provided the needed care to his elderly mother but was not eligible for the deduction because he was a bachelor who never married. A unanimous Tenth Circuit relied on Reed to extend the provision to cover him on the ground that Section 214, as written, was arbitrary and capricious, and hence violated the Due Process clause. The Internal Revenue Service had offered no rational basis for letting a bachelor fall between the cracks.

Reed and Moritz are examples of the anti-discrimination norm working to remove archaic laws.  But later challenges to explicit gender-based distinctions were not so easy to defend. In the 1973 Supreme Court case of Frontiero v. Richardson, Ginsburg participated in oral argument as an amicus on behalf of the ACLU. Frontiero asked whether a female servicewoman had to prove that her husband depended on her for more than one-half of his support in order to be classified as a “dependent,” when the wife of a serviceman benefited from a conclusive presumption in favor of that position.

On the face of it, the distinction between whether a wife or husband counts as a dependent seems to reek of discrimination. But the law imposing it may have resulted in more accurate determinations for both class of dependents: it was relatively rare at the time to find women who were not dependent on their husbands but quite common to find husbands of servicewomen who were financially independent. Requiring women to prove dependency was unnecessarily costly. Not requiring husbands to prove it was open to letting in too many ineligible men. The statutory distinction was intended to minimize the sum of error and administrative costs, which is often the mark of a sensible law. Flipping a coin may have eliminated the gender-based unfairness in Reed but it would be of no use here.

In spite of these serious administrative arguments for the statute, the Court struck down the law, but the Justices differed among themselves as to the correct rationale.  Four justices, led by Justice Brennan, thought distinctions on sex required the same level of scrutiny as those on race.  But four other justices favored an intermediate level of scrutiny, only to conclude that this law did not meet that lower standard. Then-Justice Rehnquist dissented. Today, no one would support this statutory distinction, given massive changes in military service and labor markets.

The misapplication of the antidiscrimination norm was much more pronounced in Craig v. Boren, a 1976 Supreme Court case in which Justice Brennan struck down an Oklahoma statute that forbade the sale of nonintoxicating” beer with the low alcohol level of 3.2% to males under the age of 21 while allowing its sale to females between the ages of 18 and 20. For regular beer, the uniform age was 21.

The basic concern was about the risks of drinking and driving. To strike down the law, Justice Brennan had to ignore the undeniable fact that men between 18 and 20 as a group are far riskier drivers than women in that same age bracket. It is, of course, never the case that all women in that age bracket are safer or more responsible drivers then men of the same age—that they’re less likely to drive dangerously—but when passing legislation to deal with uncertainties it never makes sense to ignore known probabilities. If other relevant factors might tip the scale—a record of driving violations—then these can also be taken into account as well. But the presence of additional factors does not negate the relevance of gender classifications. Every unregulated insurance company uses risk-adjusted premiums to prevent dangerous cross-subsidies of men by women. Why should the government be any different? The Equal Protection and Due Process arguments support the distinction Oklahoma made. Treating unlike cases alike is as much a form of gender discrimination as treating like cases differently.

The failure to take sex differences into account has also roiled pension markets. Women as a class outlive men by about five years, and that difference is universally reflected in private pension markets. But in Los Angeles Water & Power v. Manhart, Justice Stevens thought that antidiscrimination law overrode market preferences for employer-based plans. True, as in Craig v, Boren, sex is not a perfect predictor of life expectancy, and other factors, like smoking, influence the differences in life expectancy. But the proper response is to add that additional information into the rate classification. It is not to scrap the basic sex-related judgment as Justice Stevens did. Ignoring sex makes it certain that men will provide cross-subsidies to women in pension plans.

In normal pension and insurance markets, the amounts that any individual pays and collects are independent of all other people in the pool. But once sex differences are ignored, pool composition becomes critical: the more women in the pool, the lower the payouts for both men and women. Justice Stevens said that firms could avoid these problems by giving men and women the same lump sum which they could then use to purchase pensions (with different payouts) in the voluntary market. But his solution leaves both men and women worse off, because the dangers of adverse selection by potential insureds in the individual market will drive rates up for members of both groups.

In pension markets, these problems can be controlled. Most married couples prefer self-and-survivor annuities, and where the imbalance still exists, firms have some wiggle room to reduce the wage base for women relative to men—at the risk of facing a second round of discrimination charges. But with health care insurance, these mitigating factors are not present in the individual market, where actuarially fair rates often depend on pricing by sex and age. Now huge cross subsidies can wreck the market.

One example: it costs roughly five times as much to offer health care insurance for someone 65 years of age relative to someone at 25 years. The Affordable Care Act imposes a system of “community rating” that only allows a three-to-one ratio. The cross-subsidies are huge in contrast with market-based rates, where a person pays the same for insurance regardless of who else is in the pool. Consider the 25-year old asked to pay $100 in annual premiums. The market is in equilibrium when the younger person pays only one-sixth the amount of an older person: $16.67 versus $83.33. But the three-to-one community-rating cap means that he pays $25 of the $100 bill. Hence one-third of the total premium ($8.33/$25) is a cross-subsidy, which helps explain the constant stress that the ACA imposes on private markets.

These examples illustrate how best to apply the antidiscrimination laws. Use them to lower administrative and error costs—never to raise them. Use them to prevent cross-subsidies; never to impose them. Reed and Moritz followed these principles. Frontiero, a close case, probably violated the first of these principles. Craig, Manhart, and the ACA all violate both—leaving everyone worse off than they would be otherwise. On the Basis of Sex, like the modern antidiscrimination laws, is blind to these economic realities.

© 2019 by the Board of Trustees of Leland Stanford Junior University