EIGHTH AMENDMENT MATHEMATICS (PART ONE):
How The Atkins Justices Divided When Summing

By AKHIL REED AMAR AND VIKRAM DAVID AMAR
Friday, Jun. 28, 2002

In several previous columns, we have focused on Constitution reading and Constitution writing-exploring why various clauses were written, and how these words should now be read. Today, and in our next column, we focus on the third constitutional R-'rithmetic. A series of related debates about arithmetic lay at the heart of last week's Supreme Court decision, in Atkins v. Virginia, to ban executions of mentally retarded convicts.

Atkins turned on the meaning of the Constitution's Eighth Amendment, which prohibits "cruel and unusual punishments." As explained in more detail in Akhil's column of August 8, the word "unusual" in this clause invites attention to the way standards of decency evolve over time.

Thus, when a state punishment practice is challenged as "cruel and unusual," judges often look to the practices of other states of the Union. As these practices evolve, a state policy that might have been constitutional at a given point in history can become unconstitutional-because it is much more unusual in sister states-at a later time. In short, a strong national consensus to eliminate or qualify a given punishment may render the imposition of that punishment in "outlier" States unconstitutional.

The six Justices in the Atkins majority, led by Justice Stevens, claimed that Virginia was just such an outlier. According to them, a national consensus against executing mentally retarded inmates had emerged in the thirteen years since the Court upheld this practice in Penry v. Lynaugh. Virginia, out of step with the consensus, would be imposing "cruel and unusual punishments" if it put mentally retarded inmates to death.

In a pair of dissenting opinions, Chief Justice Rehnquist and Justice Scalia, joined by Justice Thomas, sharply disagreed with the Atkins majority's claim that various numbers confirmed the existence of the requisite national consensus. What follows are a few of the "math" issues implicitly or explicitly dividing the Justices, and our views on these issues.

Question 1: Do Abolitionist States Count?

By Justice Scalia's method of counting, the possibility of a national consensus against executing mentally retarded inmates was squarely foreclosed by the fact that more than half of the states permitting capital punishment-20 of 38-allowed death for such inmates.

Though less explicit on the point, Justice Stevens's majority opinion tallied up the states very differently: A healthy majority of 30 of 50 states banned executions of mentally retarded convicts, Justice Stevens concluded-12 because they have abolished all executions, and 18 because of special rules shielding mentally retarded convicts from death.

In effect, Scalia simply refuses to count the 12 abolitionist states, while Stevens takes them into account. The result is that Scalia gets a majority one way (execution of mentally retarded inmates is okay) and Stevens gets a majority the other.

Consider the following hypothetical: If 47 states banned the death penalty altogether, yet 2 of the remaining 3 allowed executing low-IQ defendants, it would seem in ordinary English that such executions were quintessentially "unusual"-rare, uncommon-even though they would be permitted in a majority of the states (2 out of 3) allowing capital punishment. (Assume, for now, that these two states are medium-sized, rather than, say, California and Texas-we shall later discuss the issue of state size.) Under Scalia's counting method, though, there would still be no Eighth Amendment violation, for two-thirds of non-abolitionist states would support execution of the mentally retarded.

Logically, if an abolitionist state decided to allow the death penalty, but only in the most extreme cases-say, for criminal masterminds with genius IQs-Scalia would then be obliged to count the state, but not otherwise. What sense does this make? The states that are most emphatically against death, that find it truly abhorrent across the board, count the least-indeed, not at all!-on Scalia's abacus. Scalia counts them only if they compromise their moral opposition to the death penalty by allowing it in some instances.

This odd counting rule - requiring one to conduct executions to qualify for a voice as to what executions are "cruel and unusual" - seems ill calculated to determine current America's sense of decency on morally charged issues of punishment.

Still, Scalia might claim to have precedent on his side. In a footnote in a 1989 majority opinion he authored in Stanford v. Kentucky, he insisted that when analyzing the death penalty for youthful offenders, judges should exclude the abolitionist states from analysis. Arguably, the footnote represents a Court majority's imprimatur on Scalia's counting method.

But not so fast. On the very same day Stanford was decided, the Court appeared to count abolitionist states in the companion Penry decision, authored by Justice O'Connor. Stevens's majority opinion in Atkins pointedly quotes this part of Penry-so even if precedent tilted in Scalia's favor before Atkins, it no longer does after Atkins.

What's more, the reason Scalia gave in the Stanford footnote is odd, to say the least. He analogized the question of including abolitionist states to "discerning a national consensus that wagering on cockfights is inhumane by counting within that consensus that States that bar all wagering."

Huh? Those who dislike gambling often do so for reasons having nothing to do with cockfighting. But those who abhor capital punishment generally do so for reasons intimately intertwined with the reasons why similar-minded but less absolutist people would allow capital punishment in some cases, but not for the mentally retarded.

It is the roughly same moral impulse in both situations but with different degrees of intensity. Indeed, those who oppose all capital punishment would as a rule oppose even more strongly (in law talk, "a fortiori") the capital punishment of a subset of inmates, like the mentally retarded, who are especially pitiable; are at heightened risk of being erroneously convicted even if innocent (since they can aid only minimally in their own defense and may confess to things they did not do); are likely to lack a full understanding of why they are being punished; and are not necessarily subject to the same deterrence incentives that might affect others. Thus, many of the reasons one might oppose executions generally would lead one to oppose their executions in particular.

Scalia's efforts to erase abolitionist states from his tally reminds us of the Supreme Court's 1968 decision in Witherspoon v. Illinois, which erased abolitionist citizens from juries in capital cases. A person who is unalterably opposed to the death penalty can be kept off the jury for cause, according to Witherspoon. Put another way, a juror in a death case must be "death-qualified."

Even more obviously, the purpose of Eighth Amendment analysis-to canvass the moral sense of modern America-is badly compromised by dismissing altogether the abolitionist states. Even if one believes a juror must be "death-qualified" to sit, a state should not have to be "death-qualified" to have a voice in what the Eighth Amendment means.

Question 2: Does Population Matter?

Both Scalia and Stevens count states, without weighting the tally to compensate for the states' unequal underlying populations. But Stevens also counts a federal statute explicitly exempting the mentally retarded from the federal death penalty. And that statute emerged - as do all federal statutes, of course - from a lawmaking process involving greater representation of populous states in both the House and the Presidency (via the electoral college).

Scalia ignores Congress altogether; and later tartly repudiates the notion of adjusting the state tally to reflect the fact that, for example, California contains almost seventy times as many persons (and perhaps an even higher percentage of crimes and convicts) as Wyoming. Scalia labels the population-adjustment notion "quite absurd," and then confidently proclaims that what matters is "a consensus of the sovereign States that form the Union, not a nose count of Americans for and against."

Perhaps-but Scalia needs far more reasoning than he offers. (And the notion that adjusting for population is not just wrong but "absurd" is, well, absurd.)

After all, in ordinary language, the word "unusual" would not seem to have any linkage to federalism, state sovereignty, or the one state, one vote rule. If the 25 largest states, with well over 80% of America's total population and an even higher percentage of America's murders, all banned executions of the mentally retarded, such executions would indeed be statistically atypical, uncommon, rare-"unusual"-in America even if the 25 smallest states followed a different practice. Conversely, if the smallest 41 states all banned a practice, it would not be particularly unusual in America, in an ordinary language sense, if it prevailed in the 9 largest states with more than half the country's population, crime, and convicts.

The words "cruel" and "unusual" were borrowed verbatim from the English Bill of Rights of 1689 and early revolutionary state constitutions. In neither place did they incorporate a one state, one vote rule. Perhaps the same words mean something markedly different in the federal. Constitution, but the burden of proof is on Justice Scalia to demonstrate that these established law words dramatically changed their meaning when placed in the federal Constitution.

In general, Scalia has resisted this sort of changed meaning argument, and in Atkins he doesn't even seem to see the issue raised by longstanding pre-constitutional uses of the phrase. He points to the process that generated the Eighth Amendment itself, but of course, that process did weight states unequally in the House, by according larger states more representatives.

Note also that Scalia's approach would seem to leave Americans in the territories and the District of Columbia wholly out of the equation.

Moreover, the 12 abolitionist states are likewise on average smaller than the others; indeed many of these 12 rank among America's smallest. In order of ascending population, the 12 are Vermont (49th in population), Alaska (48th), North Dakota (47th), Rhode Island (43rd), Hawaii (42nd), Maine (40th), West Virginia (37th), Iowa (30th), Minnesota (21st), Wisconsin (18th), Massachusetts (13th) and Michigan (8th).

Just as Scalia's exclusion of abolitionist states calls to mind Witherspoon, so his tart dismissal of a "nose count of Americans" reminds us of the 2000 election, where Bush supporters (among whom Justice Scalia surely ranks high) often sneeringly dismissed the fact that Gore won the national nose count as utterly irrelevant.

Question 3: Do Victory Margins Matter?

Justice Stevens not only counts the number of states banning executions of the mentally retarded, but also notes the "overwhelming" legislative majorities by which many of these states acted. Scalia finds this, too, "quite absurd."

But the very process by which the Eighth Amendment was adopted-the process Scalia himself invokes-did focus on the significance of legislative supermajorities, by requiring the Amendment to win by a two-thirds vote in each House of Congress. More generally, the actual vote counts in state legislatures would seem relevant evidence of the depth and breadth (or lack thereof) of modern America's concern about executing the mentally retarded.

On the other hand, some lawmakers might cast insincere votes if they know that a bill already has enough votes to pass. Thus, judges should treat vote counts with special caution.

Still, a clear pattern of overwhelming support in many state legislatures, suggests that lawmakers (with their fingers on the public pulse) sense strong sentiment among their constituents - a fact also confirmed by various poll data cited by the Atkins majority. (Using this poll data raises special concerns, however, as we shall explain in our next column.)

Question 4: How much is enough?

Should a mere 30 of 50 states-reflecting less than 60% of America's population and crime-be enough to invalidate the practice of other states as "unusual"? Here, the dissenters raise a powerful point.

Scalia responds to this argument by suggesting that perhaps such executions are not specially unusual or infrequent; the explanation is simply that there aren't that many mentally retarded convicts. Perhaps, but if in fact it could be shown that mentally retarded convicts are statistically far less likely than others to be executed even in those states that permit their execution, and if the reason for that lower likelihood is that society views their execution with special reservation or revulsion, this actual practice on the ground would indeed seem relevant to proper Eighth Amendment analysis.

In an ordinary language sense, the word "unusual" would seem to focus not merely on the law on the books but also on the law as actually applied. Laws exist allowing jaywalkers to be jailed; but being jailed for jaywalking in America surely is "unusual." (Whether it is also "cruel" is, of course, another question.)

Moreover, examining law as actually applied properly brings constitutional institutions other than the legislature into the frame of Eighth Amendment analysis. Criminal laws are often written in overbroad ways precisely because it is understood-and in some respects constitutionally required-that such laws will be softened in practice by merciful discretion exercised by prosecutors, grand juries, criminal trial juries, trial judges, governors, and parole boards. Each of these institutions represents the public, too-and helps define what modern America really does believe and practice when it comes to punishment.

The four issues canvassed today do not exhaust the interesting mathematical questions raised by Atkins. In our next column, we will discuss whether and how the Court should count recent trendlines, popular opinion poll data, European opinion, and the views of "widely diverse religious communities."


Akhil Reed Amar and Vikram David Amar are brothers who write about law. Akhil graduated from Yale College and Yale Law School, clerked for then-judge Stephen Breyer, and teaches at Yale Law School. Vikram graduated from U.C. Berkeley and Yale Law School, clerked for Judge William Norris and Justice Harry Blackmun, and teaches at U.C. Hastings College of Law. Their "brothers in law" column appears regularly in Writ, and they are also occasional contributors to publications such as the New York Times, the Los Angeles Times, and the Washington Post. Jointly and separately, they have published over one hundred law review articles and five books.

Ads by FindLaw