I wrote about the issue of *which* Congress chooses back in 2001:
https://groups.google.com/d/msg/soc.history.what-if/yMtjG1ksWOQ/cBAe_70cvWcJ
***
Putting the 1948 presidential election into the House of Representatives
requires no great effort. Just (a) have Dewey make a marginally better
campaign in Ohio (a change of less than 3,600 votes there will do), and (b)
get Henry Wallace on the ballot in Illinois, and assume he will do about as
well there as in Michigan or Wisconsin (about 2 percent--even if half of it
is from people who didn't vote or voted for minor left-wing candidates, as
long as the other half is from Truman voters, Dewey carries the state). The
result is that the Electoral College stands at 250 votes for Truman, 242 for
Dewey, and 39 for Thurmond...
(Alternatively, you can keep Wallace off the ballot in Illinois and keep
the state in the Truman camp--but change 9,000 votes in California in
addition to the 3,600 in Ohio.)
Previously, I did not think this would make much difference, because I
assumed that under the Twentieth Amendment, the new Congress decides on the
President and Vice-President in the event there is no majority in the
Electoral College. The new, 81st Congress was heavily Democratic (though I
don't have a state-by-state delegation breakdown for the House) and would
therefore presumably choose Truman. (I don't think the Dixiecrats controlled
enough state delegations to matter--though again I will have to check.)
Certainly that is in accord with the *purpose* of the Amendment--the whole
point of pushing back the opening date of the new Congress to January 3 (17
days before the presidential and vice-presidential terms start) was to
prevent a lame duck Congress from making the choice. And yet...there is
nothing in the text of the amendment which specifically states that only the
new Congress can make the choice. In a 1980 Atlantic Monthly article
entitled "Deadlock: What Happens if Nobody Wins", Laurence M. Tribe and
Thomas M. Rollins argue that "The outgoing Republican Eightieth Congress...
could have responded by moving up the date for picking among Democrat Truman,
Republican Thomas Dewey, and States' Rights candidate Strom Thurmond. In any
year, this tactic would surely stir popular protest, but a partisan Congress
could decide to take the heat: the re-elected members are likely to be from
safe districts; the lame duck members have little or nothing left to lose."
http://www.theatlantic.com/issues/80oct/deadlock2.htm Tribe and Rollins
argue that this is so contrary to the history and principle of the 20th
Amendment that "the silence of the Constitution's text becomes almost
irrelevant. The document should be interpreted to forbid lame duck
manipulation of the presidency, although a partisan Congress might decide
differently -- and it is anyone's guess how far the courts would go to halt
the lame ducks as they tramp across the spirit of the document for their own
narrow ends." Indeed, I could easily see the Supreme Court of 1948 refusing
to decide this, holding it was a "political question." (In analyzing the
plausibility of the Court so holding, please let's not get into another
debate on Supreme Court decisions of 2000. We are talking about the court of
the 1940's, which expressed a marked reluctance to get into what Justice
Frankfurter in *Colegrove v. Green*, 328 US 549 [1946] described as the "political
thicket.")
(The Tribe-Rollins article also has some interesting scenarios on the 1980
election,
http://www.theatlantic.com/issues/80oct/deadlock.htm but they
depend on Anderson doing much better than in OTL and the Reagan-Carter race
being much closer.)