Redistricting and Representation
Ronald M. Peters, Jr.
I have (unwisely no doubt) taken up the
game of golf recently, and among the features of contemporary golfing by
which I have been most struck is the availability of Global Positioning
System (GPS) equipment on golf carts at some courses that provides
information on exactly how far the cart is from the green (and how much
the hot dogs in the clubhouse cost). This
is but one example of how GPS technology is enhancing the quality of our
lives. Its availability in some car models
and as a foundation for internet sites that provide specific driving
instructions are further evidence of the pervasive use of global
positioning. Recently, my wife was able to
view GPS images of our neighborhood, and could easily identify our house. While it is all a bit Orwellian, it is,
nonetheless, often useful.
In reading the articles in this issue of Extensions, I have had occasion to ponder the implications of this Big Brother in the sky for our system of representative government. When I first joined the faculty of the University of Oklahoma in the mid-1970s, I was appointed to the Norman Reapportionment Commission. Norman had just recently moved from an at-large to a ward system for electing its city council members, and the city charter had been amended to create the Reapportionment Commission. The Commission was charged with annually reviewing the city's growth and making appropriate recommendations in the drawing of ward boundaries.
This was the first occasion of the Commission's meeting, and hence the first opportunity to decide what to do about ward lines in a growing municipality. The capable city planning staff presented to the commission data on current and projected growth. Since the next census was four years in the future, it was not possible to get a hard count of the residents. Instead, the city staff worked from plat maps to identify the number of new and planned dwellings. By defining the number of occupants for which the dwellings were zoned (single-family, multi-family, apartment, etc.) the Commission was able to estimate the patterns of residential and population growth. On this basis, the Commission was asked to recommend redistricting.
In the end, the Commission decided to take no action. We concluded that the wisest course of action, the charter requirements notwithstanding, was to defer redistricting until the results of the next census were available, and to recommend that, in the future, the city redraw its ward boundaries only once each decade based on the new census results. Our reasons were several. First, we concluded that redistricting ideally should be based on an actual count rather than an estimate. Second, we concluded that the best estimates that the city staff could provide were insufficiently reliable to serve as a foundation for a new system of ward boundaries. Third, we concluded that even if we had a reliable hard count of residents, it was unwise to shake up the boundaries frequently. While ten years was a sufficient period of time for a rapidly growing city to produce unequal ward populations, the disadvantages associated with redrawing the lines, thus redefining the relationships between citizens and their elected representatives, seemed to outweigh the advantages of enforcing an estimated equality in ward populations.
Of course, GPS technology does not enable one to get a hard count of residents, but it certainly does enable one to get a hard count of structures. When combined with data bases for patterns of party registration and voting participation, we now have powerful tools to assist policy makers in drawing ward or district boundaries. Geographic Information Systems (GIS) technology now drives the redistricting of local, state, and national legislative districts and has come to fundamentally shape our political system. Clark Bensen offers readers an explanation of the modern process of redistricting among the four articles on redistricting offered in this issue of Extensions. Technology has driven gerrymandering to a new level of sophistication. Geographic information is merged in sophisticated data bases with information about patterns of voting registration and behavior, allowing redistricting experts to slice polities thinly with the assurance of a skilled sushi chef.
Gerrymandering is hardly new, and the redrawing of district lines for partisan advantage is nearly as old as the republic. Nor is it new that redistricting has been used to create safe seats for members of state legislatures and the Congress. Incumbent legislators are usually complicit in designing electoral maps that secure their seats. Nor is it novel that some legislators have held safe seats over long periods of time. The influence of southern Democrats in the U.S. House of Representatives was maintained for generations by safe-seat members who built substantial seniority with little in the way of opposition at home.
But there are two kinds of safe seats. In one kind, the geographic area (say, the South) is dominated by one party and so its nominees will always win. Electoral competition only occurs when a seat is vacated, and then the competition is in the party primary, not the general election. The two parties are not competitive in that area. Call these "one-party safe seats." The nice thing about one-party safe seats is that one does not have to do an odd gerrymander to ensure control of the seat. Redistricting can then be based on natural communities of interest. In the other kind, the geographic area is competitive between the two parties. Depending upon how the district lines are drawn, the candidate of one party or the other is advantaged. These seats are contested by both parties, and the outcome is typically determined in the general election. Call these "competitive seats." Now among these, some seats are designed by redistricting to reproduce the conditions of one-party safe seats. That is, the geographic area is typically competitive, but the districts are drawn in effect to secure for each party a number of seats that will not be competitive. Call these "gerrymandered safe seats." These districts are not only or even primarily the product of a partisan effort by a majority to deny fair representation to a minority; more often, they result from collusion between majority and minority parties to protect incumbents.
Now what we see today is the creation of more and more gerrymandered safe seats. These seats are safe only because of the manner in which the districts have been drawn. Had the districts been drawn otherwise, the seats would be competitive rather than safe. Technology has facilitated the drawing of these seats, and led to the creation of districts that even Elbridge Gerry might have found odd. The most infamous case is the North Carolina 12th congressional district (later struck down by the courts) that stretched for 160 miles along I-85 in a corridor no wider than the highway right of way.
After intervening to enforce the one-person/one-vote application of the Equal Protection Clause, the courts have been reluctant to intervene further in partisan redistricting, regarding this as a political question. At the same time, the Voting Rights Act and its progeny have led the courts to accept race as a legitimate basis of districting in some circumstances and in some states. Charles Bullock discuss the litigation in some detail and so I will not recount it here. It does appear that the only path the courts have opened to redistricting would be through the Equal Protection Clause. In Davis v. Bandemer the court only considered justiciability through the lens of Equal Protection. In the recent Vieth v Jubelirer2 the court appeared to step back from even that possibility.
Could a case be made that the courts have an obligation to attend to the health of our representative institutions, and not simply their justice? If gerrymandering districts locks up so many districts that only 10% of districts are truly competitive over time, what effect would that have on the quality of deliberative government? The Constitution imposes an obligation on the federal government to ensure to each state a "republican form of government." What does that mean? Madison calls republican government one "in which the scheme of representation takes place."3 But what is a "scheme of representation"?
This question relates not only to the nature of the districts but also to the quality of representation they produce. Congress is supposed to be a deliberative institution in which conflicts of interest are mediated by reasoned discussion. When redistricting produces monolithic legislative parties the legislative process itself is directly and deeply affected. This is one difference between what I have called "one party" and "gerrymandered" safe seats. One party safe seats tend to produce representatives who can afford to take positions in opposition to the party majority, as was often the case for southern Democrats. Gerrymandered safe seats seem to produce representatives who are beholden to the party and tend to vote the party line. Legislative process and debate is not leavened by genuine give and take among representatives reflecting diverse constituencies.
Of interest are efforts at the state level to take redistricting out of the hands of partisans. In the Arizona model, a nonpartisan commission is able to draw the lines, but it is not clear that even a nonpartisan commission can be counted upon to produce competitive districts. In Iowa, a legislative services bureau proposes maps that the legislature must then approve. This procedure is more likely to produce a nonpartisan result when the institutions of government are subject to split party control; but even then, the two parties are likely to settle upon an incumbent protection plan that diminishes electoral competitiveness. Michael McDonald provides an interesting analysis of these efforts to take partisanship out of the process.
No case has drawn more recent attention than the brawl in Austin, Texas in 2003 in which the newly empowered Republican state legislature sought to impose a new congressional district map to replace the one put in place by the courts for the 2002 congressional elections. The Republicans had a legitimate gripe. The court map was a modification of the 1990s map put in place by a Democratic state legislature that led to the Democrats electing over half of the state's congressional delegation while winning substantially less than half of the votes cast in congressional elections. The Republicans were eventually successful in implementing their own map, which of course reversed the partisan bias and promised to provide to the Republicans a number of House seats in excess of their proportion of the statewide vote. As Keith Gaddie shows, the court map was "fairer" in that it was more likely to produce a division of seats in proportion to votes than the Republican map. Nonetheless, the courts, guided by Veith, upheld the GOP gerrymander and that map controlled the 2004 election to good effect for the Republicans. One Democrat switched parties before the election; of six remaining Democrats that the redistricting plan targeted, five were defeated on November 2.
The Republican map appears to meet all requirements of the Voting Rights Act and satisfies the courts' rather exacting standards for deviation in district population. And, it must be said, the map has produced no districts as silly as the North Carolina highway district disqualified by the courts. Still, many districts are oddly drawn, and seem unlikely to reflect the communities of interest that the courts have elsewhere acknowledged as a rational foundation for drawing electoral districts. Recently the Supreme Court took redistricting observers by surprise in remanding a challenge to the Texas plan to the federal district court in light of Vieth. Since the Vieth majority had appeared to close the door on challenges to partisan gerrymanders, observers are asking what might have caused it to ask the district court to take another look at the case. Since this challenge addressed the fact that the legislature chose to redraw the lines after an election in which the Republicans took control from the Democrats, it seems likely that the court is asking whether it would always be permissible for a new legislative majority to redraw congressional district lines. If the Democrats seized power prior to 2010, could they then implement a new plan? Attention has focused on Justice Kennedy, who in Vieth had suggested that gerrymandering might yet offer justiciable issues.
An interesting comparison is to the Oklahoma map. This redistricting plan, which basically creates four Republican and one Democratic district, was proposed by the Republicans and chosen by the courts in preference to a Democratic alternative that aimed to preserve two viable Democratic districts. Yet to reach that result the Democratic plan required the drawing of strange districts that made little sense in relationship to the underlying communities they proposed to represent. The Republican plan, by contrast, offers five coherent districts that in general reflect homogenous political constituencies. Of course, that plan delivers 80% of the state's congressional seats to a party that would carry no more than 60% of the state's congressional vote in a set of truly competitive elections.
There are, then, three variables to be taken into consideration from the point of view of representation as opposed to a partisan perspective. The first addresses the proportion of competitive districts, and asks: How important to the purposes of a system of representation is it that representatives might actually have to be responsive to constituents in order to retain their seats? Are competitive safe seats desirable?
The second variable asks: How important to the purposes of a system of representation is the relationship between the representative and the fabric of the constituency he or she represents? Is it important to represent communities of interest, or can a system of representation just as well be carved in any way, shape, or form, just so long as the number of constituents in each district is within a percentage point of those in every other district?
3. James Madison, The Federalist No.10 (1787).
Alan Greenblatt, "Whatever Happened to Competitive Elections?" Governing, October 2004, 22-27.
Toobin, "The Great Election Grab," The
New Yorker, December 8, 2003, 63-80.