The NY Times as got it right.  I point to a 1988 Political Geography article
written by Grofman and Owen which mathematically shows that incumbent
protection and partisan gerrymanders look very similar, using an expected
seat maximization formula.  My dissertation did the same, factoring in
partisan loyalty into the model.  Cain has a similar, though not
mathematically inclined, discussion in his 1984 book, The Reapportionment
Puzzle.

There is a small difference between the optimal an incumbent protection and
a partisan gerrymander.  It is true that a partisan gerrymander will create
*slightly* less safe seats for the gerrymandering party, while in most cases
completely packing the opposition party into as few uncompetitive districts
as possible.  (Yes, there are exceptions; for example, Massachusetts where
Democrats are better off creating no Republican districts, but the general
tendency is as described.)  Certainly, the opposition's districts will not
be competitive.  The "slightly less" safe gerrymandering party's districts
are set at a level that Cain refers to as "the efficient level of partisan
strength" -- weaker than the ideal incumbent protection gerrymander, but not
so weak as to make these districts competitive.

This isn't simply an academic exercise.  In my experience in redistricting
consulting, I have observed this approach used in partisan gerrymanders.
Now that I am finished testifying on competitiveness in the Arizona trial (I
hope -- that trial has continued for over a year!) I plan to core dump
several articles on competitiveness.

=========================
Dr. Michael P. McDonald
Assistant Professor
Dept. of Public and International Affairs
George Mason University

elections.gmu.edu
mmcdon@gmu.edu
703-993-9141 (office)

-----Original Message-----
From: owner-election-law_gl@majordomo.lls.edu
[mailto:owner-election-law_gl@majordomo.lls.edu]On Behalf Of David
Epstein
Sent: Wednesday, December 10, 2003 3:16 PM
To: 'election-law'
Subject: Incoherent NY Times editorial on Vieth



I was happy to see the Times chime in with an editorial on Vieth this
morning, but then disappointed to see they clearly hadn't thought the
issue through very well.

In particular, they first decry the rise of partisan gerrymandering via
powerful computer programs, then they bemoan the lack of competitive
races. *But these work in opposite directions!* Partisan gerrymanders
mean that the majority party plans to win a lot of districts by a few
votes; i.e., these will be competitive. And if they misjudge, then they
stand to lose a lot of seats.

Contrast this with a (more typical) bipartisan gerrymander, where
everyone is safe and voters are fairly well separated into different
districts depending on their partisan affiliation. It's this latter case
that leads to lazy representatives, since they won't get significant
opposition from the other party, and to polarization within Congress,
since the electoral bases of the different representatives are getting
relatively more extreme.

Ah, well. Redistricting is tricky, and even very smart people sometimes
make very elementary mistakes when thinking about it.


David Epstein