Revista da EMERJ - V. 20 - N. 2 - Maio/Agosto - 2018

15  R. EMERJ, Rio de Janeiro, v. 20, n. 2, p. 8 - 53, Maio/Agosto 2018  pretations that forbid longstanding practices). 25 The Supreme Court often is called upon to resolve novel constitutional controversies, and inevitably on such occasions it creates new law. The majority’s equal protection rationale is objectionable not because it represents new law, but rather because it represents bad law–and law that the conservative Justices almost certainly would have rejected in any other setting. If it violates the Equal Protec- tion Clause to conduct a manual recount under a vague standard that might result in identical ballots being counted differently, then certainly it should be unconstitutional to use different ballot designs or different ballot-reading technologies, if these yield substantially different likelihoods of a particular vote being counted. 26 Studies have shown that undervotes were five times more likely in Florida counties using punch-card ballots than in those using more modern, optical-scan ballots. 27 Similarly, a potentially confusing “butterfly” ballot employed by Palm Beach County led to a much higher-than- -normal rate of presidential overvotes, as did ballots in other counties 25 See, e.g., United States v. Virginia, 518 U.S. 515, 568 (1996) (Scalia, J., dissenting); Rutan v. Republican Party of Illinois, 497 U.S. 62, 95 (1990) (Scalia, J., dissenting). 26 Justice Souter, who disagreed with most aspects of the majority and concurring opinions, agreed that standardless manual recounts were “wholly arbitrary,” and thus unconstitutional. Bush , slip opinion at 7 (Souter, J., dissenting). Un- like the majority, Souter did proffer an explanation for why standardless recounts were constitutionally distinct from the use of disparate voting mechanisms likely to yield different rates of nonvotes. Souter argued that local variety in the use of voting technology could be justified on the grounds of cost concerns, the desire for experimentation, and so forth. Standardless manual recounts, on the other hand, were simply arbitrary. While Souter purports to find the Florida manual recount unconstitutional on the basis of a minimum rationality standard, this is not the way that standard generally has been applied by the Court. See, e.g., Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456 (1981); Williamson v. Lee Opti- cal, 348 U.S. 483 (1955). It is not hard to conjure possible explanations for a court (or legislature) to decline to specify standards more specific than “the intent of the voter” to guide a manual recount. First, the same value in experimentation that Souter identifies with regard to voting machines seems to apply in the context of vote tabulation standards. The best way to decide upon the optimal standard might be for different counties to experiment with different standards and then compare notes. Second, “local variety” also might justify leaving it to county canvassing boards to use their discretion in defining more precise standards. It is easy to imagine the desirability of applying a different standard in evaluating ballot markings in precincts with lots of elderly voters, who might experience greater difficulty punching out chads. This sort of imagined justification is all that minimum rationality review generally requires. See, e.g., McGowan v. Maryland, 366 U.S. 420, 426 (1961) (“A statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it.”). Finally, the Florida courts ’ refusal to prescribe a more specific standard is compellingly justified by its desire not to transgress on the state legislature’s prerogatives with regard to the selection of presidential electors. For the state courts to have prescribed a standard more specific than the legislature’s “intent of the voter” would have invited Supreme Court reversal on Article II grounds. Under these circumstances, it was not “wholly arbitrary” for the Florida Supreme Court to refrain from prescribing more specific standards. 27 See Ford Fessenden, “Contesting the Vote: The Voting Machines,” New York Times, Dec. 1, 2000, A29. See also Bush , slip opinion at 4 (Breyer, J., dissenting) (noting that voters in counties with different voting machines arrived at the polls “with an uequal chance that their votes will be counted”); id. at 4 n.4 (Stevens, J., dissenting) (noting that the percentage of nonvotes–undervotes and overvotes together–in Florida counties using punch card ballots was 3.92%, while the same rate in counties using optical scan systems was only 1.43%).