How do we know when to abandon a successful policy? At first blush, this might seem a contradiction in terms. Why would one ever abandon a successful policy? But most policies have costs as well as benefits, and at a certain point one might decide that the ongoing costs outweigh the ongoing benefits, even though that might not have been true earlier. Making such judgments requires a clear-eyed understanding of cause and effect.
Four current issues illustrate the problem. Consider, first, the recent five-to-four decision of the Supreme Court in Shelby County v. Holder to hold section 5 of the Voting Rights Act of 1965 unconstitutional. Writing the majority opinion, Chief Justice John Roberts reasoned that the requirement that certain jurisdictions with a long history of racial discrimination in voting must get pre-approval from the Justice Department before changing their voting laws is no longer constitutional because, whatever might have been the case in 1965, the record shows that the percentage of African-Americans who vote in those jurisdictions is now more or less on par with the percentage of White-Americans who vote. Thus, in Roberts' words, "the conditions that originally justified section 5 no longer characterize voting in the covered jurisdictions," section 5 is therefore no longer necessary, and, accordingly, it is now unconstitutional.
The illogic in this reasoning is so self-evident that it is difficult to imagine how Chief Justice Roberts could have missed it. Indeed, he couldn't miss it, because in her dissenting opinion Justice Ruth Bader Ginsburg made the point clearly. As she clearly demonstrated, the progress that African-American voters have made in these jurisdictions is due, not only to past changes, but to ongoing and active enforcement of the Voting Rights Act by the Department of Justice.
As she noted, in the twenty-five years leading up to Congress' reenactment of section 5, the Department of Justice had prohibited some 700 proposed voting laws in the covered jurisdictions on the ground that they were unconstitutionally discriminatory. In short, much of the "progress" lauded by Chief Justice Roberts was due to the ongoing and continuing effects of section 5, and as we have already seen, with section 5 out of the picture the previously covered jurisdictions are already leaping to enact discriminatory voting laws that just a few months ago would have been prohibited.
The dangers of this sort of illogic are evident in several other important and ongoing issues. A recent New York Times article on NSA surveillance programs, for example, observed that, although such programs might have been thought necessary in the "dangerous world" that existed after 9/11, "skepticism" about the necessity for such programs increases "when the world looks less dangerous." But to make sense of that contention, we need to know why the world "looks less dangerous." Is it because of changes unrelated to the NSA surveillance program or is it due to the continuing existence of the NSA surveillance program? In other words, as in the Voting Rights context, we cannot simply assume that because the world has changed for the better we no longer need to keep in place those processes and laws that got us to that better place in the first instance and that may be necessary to enable us to continue to enjoy that happier state of affairs.
A third recent example concerns affirmative action in higher education. The Supreme Court is on the verge of holding such programs unconstitutional, and the American public is skeptical of them as well. There are, of course, legitimate reasons to have reservations about affirmative action. Most obviously, by providing "preferential" treatment based on race, they can reinforce racial thinking and heighten racial antagonism. On the other hand, such programs have played an indispensable role both in promoting cross-racial interaction and understanding and in enabling traditionally disadvantaged minorities to achieve a greater degree of equality of educational opportunity.
But critics of affirmative action now suggest, among other things, that it is no longer necessary, because we have made real progress in terms of increasing the diversity of higher education. And, indeed, we have. Over the past twenty years, the percentage of Africans-Americans in college has increased from 23 percent to 35 percent. This is still well below the 44 percent of White-Americans who are currently enrolled in college, but it is dramatic progress, and it is due in no small part to the use of aggressive affirmative action programs in colleges and universities across the nation. The notion that we can now comfortably dispense with affirmative action because it is no longer necessary poses the same danger of flawed logic as the arguments about the Voting Rights Act and the NSA surveillance programs.
A final recent example of this phenomenon concerns our national policy of mass incarceration. From 1978 to 2009, the number of people incarcerated in the United States increased every year, from 307,276 in 1978 to 1,615,487 in 2009. Since 2009, however, the number of persons imprisoned has decreased (slightly) each year. Several factors seem to have contributed to this shift, including budgetary pressures -- imprisonment is not cheap. But according to a recent New York Times article, another important factor driving this change is "dropping crime rates over the last 20 years," which have "reduced public fears and diminished the interest of politicians in running tough-on-crime campaigns." Of course, it is perfectly possible, one might even say probable, that one reason for the "dropping crime rates" is that we have kept so many past offenders in prison. It is reasonable to assume, in the absence of evidence to the contrary, that if we had incarcerated "only" 307,276 individuals in the years between 1978 and 2009, the present crime rate -- and the concern about crime -- would be quite different from it is today.
My point in offering these four examples is not to suggest that there is a necessary "right" or "wrong" answer to the ultimate legal and policy questions at issue. But it is important that we at least think clearly about this issues and not blithely ignore the complexities of cause-and-effect.