Following up on Mark Scarberry’s comments, Bryan Wildenthal agrees:

I think Mark’s conclusion is persuasive.
I (unlike Mark) personally have no doubt that a state’s voters could (with
proper timing) act as the “Legislature” to change how a state’s electors are
chosen. I.e., I disagree with the Article II theory set forth in Bush v Gore
by Rehnquist, Scalia, and Thomas, which (as I recall from the Dec. 2000
discussions on ConLawProf, and as I gather from Mark’s present posting) Mark
supports (or at least leans toward).
I also tend to think a state’s voters could, in theory (at least as far as
the US Constitution and federal law are concerned) make a change in the
state’s method of elector selection contemporaneously and simultaneously
with the voters’ choice of presidential slate — i.e., both decisions could
be resolved together on Nov. 2 without being improperly retroactive or
“un-chusing” the state’s electors. For example, this might work if the
Colorado initiative, by its own terms, became effective immediately on Nov.
2, if it wins a majority.
But I agree, based on what Mark has presented, that the Colorado initiative
simply will not do that. By the time it would take effect, apparently on
Nov. 3, the electors will already have been chosen (on Nov. 2, as federal
law requires) under the pre-existing method set by Colorado law. That cannot
be undone later, neither a day later, nor (as the Florida Legislature
brazenly and improperly came close to doing in Dec. 2000 by purporting to
appoint electors itself) several weeks later, nor ever. (I think I recall
Mark and others arguing that Florida voters may have failed to make a choice
in Nov. 2000, thus allowing what I view as the Florida Leg’s attempt at a
post hoc hijack, but I guess we don’t need to rehash all that again.)
Anyway, I kind of like the fact that Mark (who I assume supports President
Bush) and I (a declared Kerry partisan) agree that (apparently) Colorado
cannot change from the winner-take-all system in time to affect this
election. Because we now stand ex ante the election, we have no way of
knowing whether this stance will be crucially favorable to the political
fortunes of either Bush or Kerry (possibly neither). I personally am glad
winner-take-all should (apparently) apply this year regardless of the vote
on the Colorado referendum, because I hold out hope that Kerry will win a
majority in Colorado, and thus all 9 electoral votes. I am willing to take
the risk that Bush might win them all.
How about other con law profs similarly taking a stand ex ante, before
knowing which position will benefit the academic’s preferred political
choice?
Bryan Wildenthal
Thomas Jefferson School of Law

But others are not so sure. Richard Winger notes here that this reading might be inconsistent with the Florida legislature’s (aborted) attempt to choose a different slate of electors during 2000. And Dan Lowenstein notes here that it is “plausible to say that the ‘choice’ occurs on the day that the voters vote, and not on the effective date. After all, the choice of electors at the November 2 election is not certified until days or weeks after. That does not prevent us from saying the choice was made on November 2. So the choice how to allocate the electors to the candidates is in the same sense made when the voters vote, regardless of when the initiative becomes effective.”
Dan also makes a point I’ve been making for a while: whatever legal challenges to an election that can be brought before an election should be brought. And if they are not brought before the election, the legal challenge should be barred by the doctrine of “laches” after the election. Otherwise, people have an “option:” if the election goes the way I like, I don’t raise the legal issue; if it goes the other way, I do.
Apparently a suit was just filed in federal court (see here, but I don’t have any details). My theory on why no suit has been filed before is as follows: Republicans are more likely than Democrats to be hurt by the Amendment, because Bush is more likely than Kerry to win in Colorado. A pre-election suit is most likely to be taken in state court rather than federal court (because of Article III case or controversy requirements in federal court), but the state Supreme Court recently sided with Democrats, rather than Republicans, on the important question of the permissibility of Colorado’s re-redistricting. Republicans would prefer to litigate in federal court, especially because the Tenth Circuit is viewed as more conservative than the Colorado Supreme Court. If the Tenth Circuit, for example, held that Amendment 36 violates Article II of the U.S. Constititution, the Supreme Court would likely not get involved. But if the Colorado Supreme Court held there was no Article II violation, the Supreme Court would be put in the position of potentially deciding the outcome of two presidential elections in a row.