
The Texas Redistricting, Measure for Measure
Ronald Keith Gaddie
Well heaven forgive him and forgive us all,
Some rise by sin, and some by virtue fall.
– William Shakespeare
(Measure for Measure, Act ii, Scene 1)
The Texas congressional remap captured the attention of political observers as no other. This redistricting became a political inevitability due to the driving desire of the majority Texas Republicans to claim what they felt to be rightfully theirs: the majority of U.S. House seats from the state of Texas.
The
redistricting was legally unnecessary. Eighteen months before, a
federal court crafted Texas congressional maps when legislators and the
governor could not reach agreement. It was
a "least-change" map designed to remedy the population deviations among
the existing districts while also accommodating the allocation of two
additional congressional districts to the Lone Star state.1
The map perpetuated residual influences of the 1991 Democrat-drawn
congressional gerrymander, including the translation of a majority of
congressional votes for the Republicans into a majority of congressional
seats for the Democrats.
This
just would not do in the minds of Texas Republicans. Entering
the 2002 election cycle, then-U.S. House majority whip Tom DeLay
(R-Sugarland) conceived a plan with his principal political advisor, Jim
Ellis, to set up political action committees inside the state of Texas
for the purpose of winning control of the state House from the
Democrats. These committees would later become the center of the
controversy that now threatens DeLay's position as the controlling force
in the House. The plan worked. Republicans
captured 88 of 150 seats in the House and 19 of 31 in the Senate.
As
the legislature convened for its hundred-day session in early 2003,
redistricting was not a topic of discussion. Any effort to remap was
deemed just talk, hypothetical and not a subject of attention for the
press. By April it became evident that an
intellectual exercise was to become political reality, as the state
House moved forward with plans to redraw the congressional map.
Democrats found their quiver nearly bereft of arrows with which to stop
this effort.

(Source: Texas Legislative Council)
Thirty-three miles north of Texas is Ardmore,
Oklahoma. On May 12, 2003, Ardmore became
the center of the political universe, when 55 Texas Democrats arrived at
the Holiday Inn. This walkout was designed to
break the quorum in the Texas House, and had to last until the passage
of a May 15 legislative deadline, by which time House bills must be
referred to the Senate.
Each
departing member instructed the parliamentarian and clerk of the House
to lock their voting machine until they "physically return to the House
floor and advise [them] personally that [the member] wish[ed] my voting
machine to be unlocked."2 Speaker
Tom Craddick ordered the arrest of the renegades, pursuant to House
rules. Texas Rangers and state troopers searched for them, assisted by
the Department of Homeland Security (DHS), which tracked the plane of
Democrat Pete Laney. Craddick and Tom DeLay
personally requested DHS assistance. DeLay's
office also inquired about the possibility of the Justice Department
(DOJ) enforcing the legislative warrants in Oklahoma (DOJ assistance was
denied).3
The Texas legislature adjourned on June 2 without a new congressional map. A special session was immediately called by Gov. Rick Perry for the purpose of dealing with redistricting, to begin June 30. The state Senate anticipated the session by opening hearings across the state on new congressional maps.
The
Senate hearings opened on June 28, at the urging of Lt. Gov. David
Dewhurst, who sought an "independent record." These
hearings were heated, and crowds often booed the efforts of the GOP
majority. Hearings and negotiations
continued throughout the month of July. In
the midst of the hearings, the U.S. Supreme Court handed down its ruling
in Georgia v. Ashcroft,4
which further muddied the waters as Texas legislators wrangled with the
possible retrogression effects of the effort to change the maps. Eleven
state senators – ten Democrats and Republican Jeff Wentworth – prevented
the advancement of legislation with procedural delays. As
the session came to a close on July 28, ten Senate Democrats took
flight to Albuquerque.
Democrats
around the country started holding fundraisers to support the
fugitives' effort. Bush strategist Karl Rove convened with GOP Lt. Gov.
David Dewhurst to discuss tactics. The deadlock was broken on September
2, when State Sen. John Whitmire (D-Houston) was compelled to cross
lines and announce he would vote to debate the maps, stating, "I intend
to fight redistricting on the Senate floor."5 Whitmire was
evidently pressured to switch because Republicans had preserved his
senate seat in the 2001 remap as a political favor, and his marker was
called.
The
subsequent special session lacked the previous John Grisham-like drama. Still there were major issues to resolve: (1)
satisfying the Voting Rights Act, which had been rendered politically
muddy (though not legally muddy) by the Ashcroft decision; (2)
satisfying the political goals of Tom DeLay, which included the ouster
of Martin Frost, Lloyd Doggett, and numerous other Democrats; and (3)
satisfying the political goal of Speaker Craddick to have a
congressional district for Midland.
The initial two issues were disposed with when the Texas attorney general's lawyers and experts agreed that the map which ousted Reps. Frost, Doggett, and Bell while creating new minority-majority districts would pass legal muster under the Voting Rights Act. Ironically it was the last goal that held up the map until October. The new map was passed out of the House and Senate as HB-3, signed into law by Governor Perry, and then forwarded to the U.S. Department of Justice for preclearance pursuant to Section 5 of the Voting Rights Act.
Republican frustration in Texas was
understandable. In 1992, Texas Democrats polled 51% of the congressional
vote but won 70% of seats. Republicans made
steady gains in votes throughout the 1990s, commanding a clear majority
of votes but without securing a majority of seats. Democrats
continued to retain the lion's share of Texas' congressional seats
through the end of the decade. The map put
into place by the federal court continued this trend, as Republicans
again garnered a majority of the votes (54.9%) but not seats (15 of 32).
Over
one hundred potential congressional maps were crafted and reviewed. These maps boil down to two types: those that
distort rural districts and displace rural Democrats; and those that
distort rural districts and also displaced urban Democrats Chris Bell,
Lloyd Doggett, and Martin Frost. The principle differences in the maps
are the treatment of the urban Democrats' districts and the number of
minority-majority districts created in each map.
In the first type of map, changes in the major urban centers of the state were minimal. The congressional districts of Bell (Houston), Doggett (Austin), and Frost (Dallas) were left intact. Rural districts, however, were substantially reshaped to dislodge Democrats Stenholm, Edwards, Sandlin, Lampson, and Turner (and ostensibly Ralph Hall, though his subsequent party switch was not a surprise to observers), while creating an open congressional seat for the west Texas cities of Midland and Odessa. This last element was critical to winning support of Speaker Craddick. This map was the foundation map generally supported by the House and advanced by primary GOP mapmaker, Rep. Phil King.
The second type of map, often referred to as the "8-3" map in the record, accomplished the rural district goals of displacing the aforementioned six rural Democrats, while also displacing the urban, white legislators and also crafting additional minority-majority districts (increasing majority-Hispanic districts from 7 to 8 and African-American districts from 2 to 3 – hence the "8-3" map.) In Houston, Bell's 25th district (22% black, 31% Hispanic) would be reshaped into a 36% black, 30% Hispanic district (numbered the 9th) that would allow black voters to control the Democratic primary and general election. In Austin, Lloyd Doggett's compact district 10 was cut four ways, and a significant part of Doggett's Hispanic base was placed in a new Hispanic majority district, numbered the 25th, to run from southeast Austin, south 400 miles to the Rio Grande (see Figure 2).
Figure 2: Seats-Votes Responsiveness of the Balderas
Map and the HB-3 Map

This
HB-3 map shifted the balance of the congressional delegation of the
state to reflect the Republican Party's preeminence in state politics. The plan dramatically increased the number of
districts where Republicans have an electoral advantage, displaced
Democratic incumbents, either by pairing them together or with
Republican incumbents in Republican-leaning districts, and created open
seats for Republican opportunities (see Table 1). Eleven of seventeen
Democratic incumbents were displaced from over half of their previous
constituents, and the reelection prospects of nine Democratic incumbents
were imperiled.
The
abolition of Rep. Frost's district 24 was in no way accompanied by
local efforts to create a majority-minority district. The
predominantly black neighborhoods in his district were transferred to
Rep. Eddie Bernice Johnson's South Dallas district 30, returning to her
district constituents she sought in the 1996 remap that followed the Bush.
v. Vera6challenge
(see Figure 3).
Table 1: Incumbent Core Retention and Initial
Pairings, Texas 2003
|
District |
Incumbent |
Core Retention |
District |
Incumbent |
Core
Retention |
|
1 |
Sandlin (D) |
40.1 |
17 |
Edwards (D) |
7.4 |
|
2 |
Green (D) |
5.5 |
18 |
Jackson-Lee (D) |
78.8 |
|
|
Lampson (D) |
1.1 |
19 |
Neugebauer (R) |
57.5 |
|
3 |
Johnson (R) |
80.8 |
|
Stenholm (D) |
30.9 |
|
4 |
Hall (D) |
33.9 |
20 |
Gonzalez (D) |
85.2 |
|
5 |
Hensarling (R) |
62.5 |
21 |
Smith (R) |
65.8 |
|
6 |
Barton (R) |
66.4 |
22 |
DeLay (R) |
70.4 |
|
|
Turner (D) |
4.4 |
23 |
Bonilla (R) |
81.5 |
|
|
Frost (D) |
21.6 |
26 |
Burgess (R) |
50.9 |
|
7 |
Bell (D) |
18.8 |
27 |
Ortiz (D) |
86.6 |
|
|
Culberson (R) |
51.8 |
28 |
Rodriguez (D) |
46.5 |
|
8 |
Brady (R) |
38.9 |
30 |
Johnson (D) |
73.7 |
|
10 |
Doggett (D) |
40.3 |
31 |
Carter (R) |
31.0 |
|
12 |
Granger (R) |
79.8 |
32 |
Sessions (R) |
52.3 |
|
13 |
Thornberry (R) |
88.3 |
|
|
|
|
14 |
Paul (R) |
43.7 |
|
Mean |
51.76% |
|
15 |
Hinojosa (D) |
60.0 |
|
Mean Dem. |
43.22% |
|
16 |
Reyes (D) |
100.0 |
|
Mean GOP |
61.44% |
(Source: Office of the Attorney General of Texas; Texas Legislative Council; R. Keith Gaddie.)
So
how does the partisan balance and responsiveness of the Texas districts
change? The Balderas court
and experts relied on reconstituted elections within the districts,
especially on the closely contested lieutenant-governor's race, to
estimate the change in partisan balance across the districts. The number of marginal districts using the
45%-55% definition of marginal is four districts under the Balderas court map (two each Republican and Democratic
leaning), and three such districts (all Republican) in the HB-3 map. The number of safe Republican districts
increases from 16 in the Balderas court map to 19 in the HB-3 map. A net of
four seats shift out of the Democratic column and to the Republican
column.
HB-3
is not responsive to small changes in the vote, or to shifts in the
statewide majority. Responsiveness is
measured as the number of seats won by each party as the share of the
two-party vote won by the GOP changes in one-percentage point increments
above and below the actual vote. First we measure the two-party vote
share for lieutenant-governor within each congressional district, to
ascertain how many districts were carried by each party. The vote in
each district is increased/decreased at one-percent increments from the
observed vote, representing a 1 percentage point change from the
statewide average.7 This
analysis – with the same conclusions – was presented to the federal
three-judge panel that heard the challenge to the 2003 Texas map in
separate reports by John Alford of Rice University and Keith Gaddie of
The University of Oklahoma (plaintiff's and state's experts,
respectively). Responsiveness results for both maps appear in Figure 2.
In
the old map used in 2002, based on the actual vote share, 52.9%
Republican, Republicans won the majority of the vote in 18 districts,
while Democrats carried the vote in 14 districts. When
that vote share is increased by one percentage point across all
districts, Republicans pick up an additional seat, for 19, and when the
vote share is increased by another one percentage point, another seat is
gained. The districts similarly respond in
the other direction; initial reductions of the GOP vote share of minus
one-percent and minus two-percent result in corresponding seat shifts;
50.9% of the GOP vote results in half of the seats for each party.
The
HB-3 map creates a more pronounced Republican advantage in the map
design. At any value for Republican vote
share above 51.9% statewide, Republicans carry 68.75% of seats, or 22 of
32 districts. Below 51.9% of the vote
share, Republican seat share falls off with vote share, though it is
only at 45.9% of the vote for the GOP or less (54.1% Democrat or more)
that Democrats win a majority of the seats in the HB-3 map.
| Figure
3: Old South Rio Grande Valley Districts, Texas |
Figure 4: New South Rio Grande Valley Districts, Texas |
![]() |
![]() |
| Source: Texas Legislative Council |
Source: Texas Legislative Council |
.
Republican lawmakers now had to navigate the
waters of preclearance under Section 5 of the Voting Rights Act. Legal challenges loomed, including a challenge
to the legality of the mid-decade redistricting and a challenge to the
redistricting under Section 2 of the Voting Rights Act, that race was
unduly considered in crafting the maps.
A
federal court dismissed the complaint regarding the mid-decade remap. And, in the process of hearing the Section 2
challenge to the districts, the federal three-judge panel noted that it
would not hear issues of retrogression, as they pertained to Section 5
and were not a concern of the court. The
entire court challenge rested on demonstrating that the new districts
were racially motivated, or developing an evidentiary basis for a future
partisan gerrymandering challenge.
Numerous
plaintiff or intervening parties were involved in litigation against
the state, including the Texas Democratic Party; the Texas Congressional
Democrats; two different NAACP intervenors; Congresswomen Sheila
Jackson-Lee and Eddie Bernice Johnson; Hispanic intervenors LULAC and
the G.I. Forum/MALDEF; and private citizens, including Mr. Walter
Sessions who offered his name to the case (Sessions v. Perry).8 The
state's defense consisted of a legal team from the attorney
generalÕs office and private attorney Andrew Taylor of Houston. Numerous political science and voting rights
experts testified at trial.
The
general Democratic strategy had three parts: (1) get the new map denied
preclearance at the Justice Department under Section 5; failing that,
prove that the new map was either (2) an illegal partisan gerrymander;
or (3) an illegal racial gerrymander. While
the Texas trial was underway, lead counsel for the plaintiffs Paul Smith
argued before the U.S. Supreme Court for the overturn of the
Pennsylvania congressional map as an illegal partisan gerrymander. A partisan gerrymandering claim was refused by
the Texas trial court (any appeal point was subsequently mooted by the
June 2004 Supreme Court ruling in Veith v. Jubelirer9 that partisan gerrymanders are
non-justiciable). The entire legal
strategy would rely on demonstrating racial intent and effect through
either Section 5 (at DOJ) or Section 2 (in the trial court).
Preclearance
came into issue when the U.S. Supreme Court muddied the waters on
retrogression, by noting that states could comply with the Voting Rights
Act either by maintaining safe districts for minorities, or by creating
a variety of less safe "coalition" districts. Plaintiffs'
counsel and at least one plaintiffs' expert argued for the protection
of any coalition district where minorities cast votes for their
candidate of choice in the general election and that candidate
prevailed: in other words, any district electing a Democrat. This peculiar logic called on the court and
the Justice Department to accept the argument that any constituency,
regardless of the ability of a minority group to control the primary and
general election, had to be protected by the Voting Rights Act based on
general election outcomes, unless minorities supported the result. They also called on the trial court to reject
its previous logic that coalitions of minority groups be cohesive in
elections of consequence (often primaries in Texas) when determining
whether a district that was not Anglo majority but also had no ethnic or
racial minority of majority, was protected by the Voting Rights Act.
This
strategy was suspected by the state's attorneys and speculated about by
lawyers for the Legislative Research Council, or "Leg Council." Plaintiff's expert Allan Lichtman and attorney
Jerry Hebert both framed such expansive opinions regarding Section 5 to
the Senate redistricting panel in July. Lichtman, in his initial
assessment of the Balderas map
post-Ashcroft, argued that any district with as little as 5% minority
population, where the minority voters elected a candidate of choice in
the general election, was protected from retrogression under Ashcroft. In
other words, any district anywhere in Texas where minorities voted and
where Democrats were elected could not be altered against the will of
the minority group. The plaintiffs would
narrow their strategy in the period of discovery leading up to trial,
arguing mainly for the preservation of districts 24 and 25 (Frost and
Bell) and for the restoration of the design of the South Rio Grande
Valley districts to the Balderas court design. These districts held out
plausible hope for a Section 5 challenge, in that they were not
districts with Anglo majorities and they were districts where minority
voters made substantial contributions to forming winning political
coalitions
The
decision of the Justice Department was controversial. There
is a strong indication that professional staff were poised to challenge
the Texas map and deny Section 5 preclearance, but that they were
overruled by the political appointees in the Justice Department. During preclearance investigation, the
professional staff were especially
interested in the potential minority performance and control in the old
districts of Chris Bell and Martin Frost, and the prospects for
non-performance of the new Hispanic-majority districts of the South Rio
Grande Valley.
The
trial court, meanwhile, rejected the argument that the new districts
violated Section 2 of the Voting Rights Act. State
demographers presented evidence consistent with the Cromartie decision
that the new minority districts were a product of partisan motives
rather than racial motives. The plaintiffs'
experts provided sufficient support for the state's case in support of
its new map under cross-examination that a state attorney observed of
more than one expert, "he doesn't realize it, but he's working for us."10
The effect of the new map on minority
representation is already evident. In three
minority-majority districts – 9, 25, and 28 – competitive primaries were
held. In the new district 9, freshman
Chris Bell was denied re-nomination for another term by African-American
Justice of the Peace Al Green. The decisive
35-point result defied the argument of critics of the map, in that Green
was most likely the candidate of choice for black voters in this very
Democratic district.
In
the new, heavily Hispanic district 25, the results are mixed in terms
of effect. Anglo incumbent Lloyd Doggett
defeated South Valley judge Leticia Hinojosa by 28 points. This
district, like new district 9, is safely Democratic and should return
Doggett to Congress, though Republicans recruited a formidable
challenger, Hispanic Texas Utilities Commission Chair Becky Klein. Next
door in District 28, incumbent Congressman Ciro Rodriguez (D) narrowly
lost re-nomination to former Secretary of State Henry Cuellar. This contest was settled by an extended
recount and lawsuit, which overturned the winner from the initial count
(Rodriguez). The results of the new
minority districts have increased descriptive representation by electing
one new minority congressman, Al Green in district 9.
The
partisan political consequences of the remap are far-reaching. The balance of control in the U.S. House could
very well be determined by what happens in Texas. In
district 4, incumbent Ralph Hall switched parties. In district 10,
Lloyd Doggett abandoned the unwinnable district to the Republicans, and
there is no Democrat running. Max Sandlin
and Nick Lampson face districts that are on average five points more
Republican than before redistricting. Nearly 99% of Lampson's
constituents are new to him. Charlie Stenholm (TX-19) and Martin Frost
(TX-24) face the task of unseating incumbent Republicans Randy
Neugebauer and Pete Sessions, though both Democrats are well funded and
bring ample seniority and experience to bear against the partisanship of
the new districts. Frost faces the daunting
task of convincing some of the oldest, most Republican suburbs of Dallas
to take him on as their representative. Frost's website notes the huge
minority population in new 32 (36% Hispanic, 8% African-American), which
is a similar combined-minority population to his old haunts. But only
about 13% of registered voters are Hispanic, and the black vote is
one-third that of old district 24. The
electorate is 18 points more Republican than Frost's old haunts.
If
we consider the "hit list" – the ten Anglo-white Democrats in the Texas
delegation – most of them could be gone in the next Congress. Ralph Hall changed parties. Jim Turner retired
(minus two). Of Max Sandlin, Nick Lampson, and Chet Edwards, two will
likely not be reelected. Stenholm and Frost
must beat Republican incumbents to return, though the odds favor
Stenholm relative to Frost. Bell was
defeated for renomination. Gene Green (D-29) is cruising to reelection.
Doggett won a safe, Hispanic seat. Three of
the Anglo Democrats will definitely not return to Congress from Texas;
two will definitely return but will represent Hispanic-majority
districts. Of the remaining five Anglo
Democrats, it appears that no more than two will make it back for the
next Congress.
The
real political consequences of the Texas remap are best summed up by
the presiding judge, Patrick Higginbotham. The Texas trial was
unaffected by the looming Veith partisan-gerrymandering decision, but the
Texas panel was not mute on the topic. Higginbotham observed that:
We decide only the legality of Plan 1374C,
not its wisdom. Whether the Texas
Legislature has acted in the best interest of Texas is a judgment that
belongs to the people who elected the officials whose act is challenged
in this case. Nor does the reality that
this is a reprise of the act of 1991 State Legislature weigh with the
court's decision beyond its marker of the impact of the computer-drawn
map. This extraordinary change in the ability to slice thin the lines
brings welcome assistance, but comes with a high cost of creating much
greater potential for abuse. Congress can
assist by banning mid-decade redistricting . . . the only check on those
grasps of power lie with the voter. But,
perversely, these seizures entail political moves that too often dance
close to avoiding the recall of the disagreeing voter. We
know it is rough and tumble politics, and we are ever mindful that the
judiciary must call the fouls without participating in the game. We must
nonetheless express concern that in the age of technology this is a very
different game.11
Coda
November 3, 2004: The Texas redistricting
accomplished its political goals. On November
2, four Anglo, Democratic congressmen – Max Sandlin (37.7%), Nick
Lampson (43.0%), Charles Stenholm (40.1%), and Martin Frost (44.0%) –
were defeated for reelection, while Chet Edwards (51.2%) survived by a
narrow margin. Texas Republicans held 20 of 32 seats in the state
congressional delegation. The future of the
Texas map is still not certain. The U.S. Supreme Court has directed the
district court in Sessions v. Perry to revisit their assessment of the Texas map
in light of the decision by the high court in Veith v. Jubelirer.
1. Balderas v. Perry, 536 U.S. 919 (2002).
2. Associated Press, "Text of letter sent by missing Democrats to House officials to lock voting machines," May 12, 2003.
3. In October
2004, DeLay was reprimanded by the House Ethics Committee for this
action.
4. Georgia v.
Ashcroft, 123 S.Ct. 2498
(2003).
5. R. G.
Ratcliffe, "Whitmire ends speculation, will stay in Austin," Houston
Chronicle, September 5, 2003.
6. Bush v.
Vera, 531 U.S. 98 (2000).
7. Janet C.
Campagna, "Bias and Responsiveness in the Seats-Votes Relationship," Legislative
Studies Quarterly 16:81-89.
8. Sessions
v. Perry, Civil Action No.
2:03-CV-00354 (E.D. Tx., Marshall Division, 2003).
9. Vieth v. Jubelirer, 2004 W L 894316 (U.S. Pa.).
10. Comment to
author by anonymous source at the Sessions v. Perry trial.
11. Sessions
v. Perry, page 2.
R. Keith Gaddie
(Ph.D., University of Georgia) is Professor of Political Science at the
University of Oklahoma. His most recent book is Born to Run: Origins
of the Political Career (Rowman and Littlefield, 2004), and he has
also written articles for The Journal of Politics, Social Science
Quarterly, Political Research Quarterly, Legislative Studies Quarterly,
Cato Policy Analysis, and State and Local Government Review, as well
as other journals. In 2003 Professor Gaddie served as a consultant
and testifying expert on redistricting to Attorney General Glenn Abbott
of Texas. He has also been involved in redistricting litigation in
Oklahoma, New Mexico, South Dakota, Wisconsin, Illinois, Virginia, and
Georgia. His email address is rkgaddie@ou.edu.