I don't agree with Ronald's analysis of "the most competitive" Texas
statewide race in 2002 as a basis for claiming a third of the Texas
congressional seats will come into play with a 5% vote swing. Checking the
Texas SoS website, The margin of victory in the 2002 LT governor's race was
5.74%, the next closest was the Commissioner of the General Land Office at
11.68% and the average of eight statewide offices, including the Chief
Justice, was 16.79%. I rather doubt that the 2002 LT governor race is
representative of the underlying partisanship of Texas.
I agree, however, that a uniform vote swing model is most likely incorrect
(anyone who reads the fine print of the academic articles by authors as
early as Tufte will see the caveats in this regard). Though, without
knowing where the qualified candidates and money are this far in advance of
an election, it is the best that we can go on at this point. Indeed, I
rather expect that the Democrats will do better in terms of their overall
vote totals simply on the fact that they are fielding more candidates in
uncompetitive races -- races that would have previously been uncontested by
a Democrat -- and Republicans are not fielding a like number in
uncompetitive Democratic districts.
I've heard it mentioned before that the "non-uniform" (for the lack of a
better word) vote swings are concentrated in competitive races. I wonder if
that is true. It could be just as likely that a larger amount of the
overall swing comes from second-tier competitive districts (because the
favored party is fielding more qualified challengers than in other
elections) or from the uncompetitive districts (because any challenger is
fielded). I suppose one could estimate the normal vote for a given election
and then observe where the largest error occurs. Anyone know of such a
study?
As for David's question, 7.5% does seem to be a large number (btw, I mean
57.5% of the two-party vote), but I haven't done a similar study for
previous elections, so I can't say with authority how unusual it is. But,
perhaps there is a paper worth writing here about where vote swings are
concentrated that would produce the statistic as a by-product.
I am skeptical that Hispanic turnout is going to be much higher in this
election. Like it or not, midterm elections are low-turnout affairs and
given the lower socio-economic status among Hispanics, they will likely vote
at the low rates as they have done in previous elections. Even in
California, where Hispanics were organized and mobilized in the 1990s by
Prop. 187, 2004 CVAP turnout among Hispanics (from the CPS) was 46.9% while
among white non-Hispanics it was 70.6% (see:
http://www.census.gov/population/socdemo/voting/cps2004/tab04a.xls, plus my
calculations).
Finally, why incumbency is collapsing into uncompetitive seats is that, as
Jacobson points out, there was a lot of sorting of mis-matched incumbents
into similar partisan-leaning districts in the last round of redistricting.
------------
Dr. Michael P. McDonald
Assistant Professor, George Mason University
Visiting Fellow, Brookings Institution
Mailing address:
(o) 703-993-4191 George Mason University
(f) 703-993-1399 Dept. of Public and International Affairs
mmcdon@gmu.edu 4400 University Drive - 3F4
http://elections.gmu.edu Fairfax, VA 22030-4444
-----Original Message-----
From: rkgaddie@ou.edu [mailto:rkgaddie@ou.edu]
Sent: Monday, May 22, 2006 5:37 PM
To: David Lublin
Cc: Michael McDonald; 'election-law'; owner-election-
law_gl@majordomo.lls.edu
Subject: Re: RE: Electionlawblog news and commentary 5/22/06
Good points made by David.
How much of that incumbency advantage is personal anymore? Oppenheimer
points out that the incumbency advantage is collapsing onto general party
safety in districts. Abramowitz has been tracking a similar collapse of
congressional partisanship into presidential partisanship. The exit poll
data in 2004 also confirm that the presidential vote and congressional
vote are now generally structured by ideology. So, if voters do go south
for Republicans, it may be that large numbers of seats have sufficient
partisan capacity to weather the storm.
But, as David also points out, these will not be uniform swings . . . I
would be concerned for Republicans in every district in California, but
also in districts where there are large Latino, citizen populations - say
the 3-10% range - that are undervoting. The current political environment
makes this undermobilized electorate potent for a general election with an
embattled GOP incumbent or a usually Republican open seat.
Even the districts in Texas showed a potential to swing dramatically,
despite the nature of the gerrymander. The graphic I attached shows the
seat swing from the 2002 LTGV in Texas (most competitive seat), based on
uniform one-point shifts in the vote statewide (Alford and I presented
similar graphics in our reports in 2003). About a third of the seats in
Texas that are currently in GOP hands could come into play based on a
five-point shift. There are structural biases, but they are insufficient
to mitigate great political tides.
_____________________________
Ronald Keith Gaddie
Professor of Political Science
The University of Oklahoma
455 West Lindsey Street, Room 222
Norman, OK 73019-2001
Phone 405-325-4989
Fax 405-325-0718
E-mail: rkgaddie@ou.edu
http://faculty-staff.ou.edu/G/Ronald.K.Gaddie-1