The one person, one vote standard in redistricting: the uses and abuses of population deviations in legislative redistricting.

• Author: Brunell, Thomas L.
• Position: Law Review Symposium 2011: Baker v. Carr After 50 Years: Appraising the Reapportionment Revolution

ABSTRACT

Since the Redistricting Revolution of the 1960s, which began with a series of Supreme Court decisions that forced states to draw equally populated districts for the United States House and state legislative chambers, the standards for congressional and legislative districts have evolved differently over time. Today's standards call for virtually no population deviations for congressional districts, though legislative districts can typically deviate up to 10 percent. In practice this means that districts can vary by as much as 5 percent above and below the ideal population within a state. Based on a review of data from forty-seven states after the 2000 redistricting cycle, this Article demonstrates that these population deviations are a simple tool for those that redraw electoral boundaries to create a partisan gerrymander. If one party controls the redistricting process, districts from the opposite party are typically overpopulated and districts favoring the party in control are usually underpopulated. Based on this phenomenon, this Article argues that courts ought to abolish the "10 percent" rule.

INTRODUCTION

Following the 2010 Census, each state used data from the results of the decennial headcount to redraw state legislative and, in those states with more than one district, congressional electoral boundaries. This process is time consuming, expensive, susceptible to litigation, and vitally important for many politicians. Most voters are unaware that redistricting is taking place and they are even less likely to know if and how it affected the voting districts where they reside. Remapping electoral boundaries provides politicians an opportunity to affect the distribution of seats within their own state and to improve individual members' chances of being reelected. This Article's focus is one important aspect of redistricting: the extent to which those in charge of redistricting use population deviations across districts in state legislative redistricting.

Modern redistricting revolves around a handful of accepted criteria or principles, most of which are not consistently enforced. Compactness, or the shape of districts, is one principle that is not heavily enforced. (1) Protecting communities of interest is a rather amorphous criterion that can mean virtually anything when it comes time to litigate a map, but at its most basic level, it means preserving other political boundaries, such as county and municipal lines. (2) Contiguity, which requires that all parts of a district be connected, however, is strictly required and enforced. (3) Another criterion that courts routinely enforce is that of one person, one vote ("OPOV"), which mandates that districts within a state be nearly equal in population. (4)

In this Article, I review the relevant case law pertaining to the OPOV standards and explain the differences between congressional and state legislative standards for OPOV. Then, I examine how the more relaxed approach to state legislative districts is used primarily for political purposes. And, lastly, I argue that courts ought to require all representative districts to abide by the very strictest of standards: equal population. (5)

1. THE LEGAL HISTORY AND LANDSCAPE (6)

   In Baker v. Carr, (7) the Supreme Court finally stepped into the "political thicket" of redistricting and addressed legislative malapportionment, or the creation of districts with dramatic population variances within the same state. In a series of famous cases in the 1960s, the Court decided that the Equal Protection Clause of the Constitution served as grounds to strike down legislative redistricting maps that had egregious population deviations with a single state. (8) In the early 1960s, for example, the average state Senate district in California was comprised of just under 400,000 people. (9) The standard deviation for state Senate districts at the same time was over 900,000 people, indicating a large variance in the population of these electoral districts. (10) The variance stemmed, in large part, because Los Angeles County, the most heavily populated county in the state, had just one seat prior to the Supreme Court's insistence on population equality. (11) After the OPOV decisions, however, Los Angeles County was represented by 14.5 people in the state Senate. (12)

   The idea behind the Baker v. Cart and related decisions is straightforward: If one voter lives in a district with 50,000 people and another voter in the same state lives in a district that only had 5,000 people, then the voters in the second district cast a far more powerful and important vote than their fellow citizens in the first district. Legislatures represent people, the Court recognized, not land, trees, or square mileage. (13)

   Baker concerned the redistricting process in Tennessee. In the 1960s, the Tennessee constitution required the state legislature to redraw legislative districts after each decennial census. (14) Despite this requirement, the state's legislative district lines remained static between 1901 and 1960. (15) Because the population growth during this time was, by and large, in the state's urban areas, these stagnant district lines led to vastly underrepresented urban areas. (16) The plaintiffs for the case included citizens from Memphis, Knoxville, and Nashville--the major urban areas of the state. In a lengthy opinion, Justice Brennan, writing for the majority, carefully considered justiciability, jurisdiction, and standing. (17) In the process of finding no bar to the plaintiff's claim, he disposed of Colegrove v. Green, (18) a case in which the court refused to intercede in redistricting related matters. (19)

   Other related and equally important redistricting decisions followed shortly after Baker. In 1963, the Court heard Gray v. Sanders, (20) which involved a Georgia law that required votes in primary elections for statewide offices to be tallied using a county-unit system. (21) The Supreme Court ruled that this system violated the Equal Protection Clause because the weight of a vote was not equal, but rather depended on the size of the county in which a voter resided. (22)

   At this time, many states mimicked the federal legislature by apportioning the two state legislative chambers on different basis. For instance, in 1962 the Alabama legislature proposed an amendment to the state constitution which would have apportioned the state senate by county--each of the sixty-seven counties would get one senator. (23) Though this is analogous to the United States Senate, the Court held that "State[s] [must] make an honest and good faith effort to construct districts, in both houses of its legislature, as nearly of equal population as is practicable." (24) Thus, the court killed the "federal analogy" for having even one chamber of the state legislature being apportioned according to something other than equal population.

   While a majority of the Justices clearly saw a need to equalize the population across districts within a state, they were also clear that they did not think that perfect population equality was required:

   So long as the divergences from a strict population standard are based on legitimate considerations incident to the effectuation of a rational state policy, some deviations from the equal-population principle are constitutionally permissible with respect to the apportionment of seats in either or both of the two houses of a bicameral state legislature. (25) This is the language that led us down the road of permissible deviations from strict population equality for state legislative districts.

   A year later, in Wesberry v. Sanders, (26) the Court ruled that states must draw districts for the U.S. House of Representatives so that the population in each district is nearly as equal as possible because each person's vote in a state ought to be weighted equally. (27) Unlike those cases involving state apportionment procedures, the Court based congressional OPOV standards on Article I, section 2 of the Constitution, which requires that states receive seats in the House "according to their respective Numbers." (28) Over time, the federal judiciary has interpreted this as requiring nearly absolute equality. (29) But since the justification for legislative district equality comes from the Equal Protection Clause, "judicial deference to state interests and practical necessity" is required. (30) Thus, there is a distinction between the level of equality required based on the office for which the state legislature is drawing the lines.

   Later, the Court in Karcher v. Daggett (31) established a two-step test with respect to population equality for congressional districts. Under the Karcher test, a court must first assess whether the state could have reduced or eliminated the population differences with a good faith effort. (32) Then, if the challengers "can establish that the population differences were not the result of a good-faith effort to achieve equality, the State ... bear[s] the burden of proving that each significant variance between districts was necessary to achieve some legitimate goal." (33)

   Courts have found that small population deviations across congressional districts pass muster when the state demonstrates a compelling governmental interest to justify these deviations from OPOV. (34) Courts sometimes, however, find that very small deviations are sufficient to force a change in the map. For instance, the 2000 congressional redistricting plan in Pennsylvania had a deviation of just nineteen people, but the people that drew the plan had no compelling explanation for why these deviations existed. (35) They simply decided that the nineteen-person deviation was close enough, and the district court found that this explanation failed to meet the second part of the Karcher test. (36)

   The court's decision, at first blush, seems ridiculous. After all, census data are not perfect. (37) And, even with perfect data, when this...