In Roper v. Simmons, yesterday's decision declaring the juvenile death penalty unconstitutional, the Supreme Court relied heavily on the fact that five states have gone from allowing the juvenile death penalty to banning it since the Court's 1989 decision in Stanford v. Kentucky:

  Five States that allowed the juvenile death penalty at the time of Stanford have abandoned it in the intervening 15 years — —four through legislative enactments and one through judicial decision. Streib, supra, at 5, 7; State v. Furman, 122 Wash. 2d 400, 858 P. 2d 1092 (1993) (en banc).
  Though less dramatic than the change from Penry to Atkins (“telling,” to borrow the word Atkins used to describe this difference, 536 U. S., at 315, n. 18), we still consider the change from Stanford to this case to be significant.

  As Xrlq points out, however, something important is missing from this analysis: a sense of whether the five states that changed their policy are states in which capital litigation is active or basically dormant. If the goal is to look for "evolving standards of decency" and find an "emerging national consensus," it is one thing to say that practices changed in five states; it is another to say that actual practices stayed the same, but that the law on the books in a few states changed. The difference matters because while most states allow capital punishment in theory, the actual practice of bringing capital cases and carrying out the sentences is heavily weighted in favor of a small number of states.

  I took a look at some statistics, and the stats seem to confirm my suspicion: the five states that moved to ban capital punishment for juveniles are states that have capital punishment on the books but rarely use those laws. While five states changed their laws on the books since 1989, it seems that actual practices since 1989 remained the same.

  Here are the stats I found. The juvenile death penalty has always been very rare in the United States; since the beginning of the republic, it has occurred on average about once a year. That rate has remained constant (at least in absolute terms) in the last twenty years: 22 juveniles have been executed in the last 20 years. Texas is responsible for 13 of those cases; Virginia for 3; Oklahoma 2; Georgia, South Carolina, Louisiana, Missouri each for 1. (stats on page 4 of the Streib report cited in the Roper opinion)

  The states that banned the juvenile death penalty since Stanford v. Kentucky are Washington (by the courts), and (I think, from page 7 of the Streib report). Kansas, New York, Montana, and Indiana. But Washington, Kansas, New York, Montana, and Indiana aren't states that have executed anyone for a juvenile crime at least in the last three decades — and indeed, I'm having trouble finding evidence that any of these states ever actually did so.

  More broadly, these five states are not active death penalty states even for adult crimes. Since the death penalty was reinstated in 1976, these five states have been responsible for 17 out of the 1,000 or so executions that have occurred — only about 2% of the total. 11 of the 17 executions were in just one state, Indiana. (The numbers, available here, are Washington - 4, Kansas - 0, New York - 0, Montana - 2, and Indiana - 11.) Notably, of the 4 states that abolished the juvenile death penalty by legislative decision, 2 of those don't actually have a functional death penalty for any defendant.

  If I understand the statistics correctly, the move to abolish the juvenile death penalty in five states since 1989 is essentially symbolic: none of those states have executed a juvenile in many decades, if ever, and the five states are mostly states that have capital punishment in theory but not in practice. If the "evolving standards of decency" inquiry of the Eighth Amendment focuses on actual practices, then there seems to be virtually no evidence of a changed standard since 1989.