Seats, votes, citizens, and the one person, one vote problem.
Few passages from American court cases have become as familiar as "one person, one vote." Although we know of no authoritative count of the frequency with which that phrase has appeared in print, we suspect that it may even rival the usage of such famous passages as "to be, or not to be" (1) and something is rotten in the state of Denmark." (2) References to "one person, one vote" appear not only in legal writings concerning districting practices but have become so widely used by journalists that many lay persons recognize the meaning.
One person, one vote is the yardstick against which redistricting plans have come to be compared, but the phrase did not originate in a redistricting dispute. Justice William O. Douglas first used "one man, one vote" in the 1962 challenge to Georgia's unique county-unit system, a weighting scheme roughly modeled on the Electoral College as a means for selecting Democratic Party nominees in statewide (and some congressional) contests. (3) The system, struck down in Gray v. Sanders, awarded two, four, or six unit-votes to the candidate who received the most popular votes in a county with the number of unit-votes determined by the size of the county's population. (4) Chief Justice Earl Warren appropriated the term when applying the equal protection standard to all state legislative chambers in Reynolds v. Sims, (5) and Justice Hugo Black cited the benchmark as the standard for congressional districts in Wesberry v. Sanders. (6) The concept is now applied with rigor to local, state, and congressional district plans.
Two decades after the initial usage of one person, one vote, the Supreme Court had limited the acceptable variance from absolute equality in district populations to requiring that any deviation in the populations of a state's congressional districts be explicable. (7) Courts accepted greater deviations from the ideal district population when evaluating districting practices for state legislatures and local governments. For many years, it was thought that deviations of plus-or-minus five percentage points from the ideal would pass muster--termed a safe harbor within which the legislature might act. However when Georgia devised state legislative districting plans that tested the proposition by concentrating districts at the extremes of the plus or minus five percent deviations, a trial court rejected the plans when the state failed to provide an acceptable rationale. (8) Consequently, jurisdictions now need to explain deviations in the district populations they draw for all levels of government.
In this Article, we engage the one person, one vote issue in the context of the newest debate: citizen apportionment. This approach is the basis for a strategy for shaping our understanding of the equal protection of voting rights and affecting the apportionment of legislative power. We discuss the limitations of the original one person, one vote standard, both in terms of apportioning voting power to equalize individual votes when electing representatives, and in achieving "fair" political results. We also empirically model the variation in turnout ratios by states and show how these are associated with both citizenship factors and the application of voting rights law. We then turn to the issues of citizenship and legislative apportionment and discuss the practical and legal challenges to judicial and policy approaches to citizenship apportionment.
I. CITIZENSHIP: THE NEXT ONE PERSON DEBATE
One person, one vote has become the standard for drafting districts, but the term itself is something of a misnomer. Substantial variations in the number of persons per district exist among the states due to the apportionment formula used and the unwillingness of the House of Representatives to expand its membership over the last century. (9) The Supreme Court has turned back challenges to the congressional apportionment formula, which results in deviations across the states in excess of sixty percent. (10)
Numerous congressional plans equalize populations in state districts so that deviations are restricted to no more than one person; but the actual numbers of voters vary substantially. Multiple factors affect the rate at which population translates into votes. Districts vary in the degree to which their population consists of children, non-citizens, those who have been disenfranchised for felonies, and those who do not find the issues or candidates sufficiently interesting to merit participation. In no district does the number of voters equal the number of residents; consequently, the rate at which population translates into participation vanes widely. (11)
The one person, one vote debate has moved in a new direction. For at least a decade, a little-noticed effort has sought to move the apportionment of political power away from total population toward a citizen population approach for both apportioning seats among states and equalizing voting power within states. These largely untested, somewhat novel efforts reflect an assumption that representative power and the concept of one person, one vote are vested in individual citizen rights, specifically in the right to vote. Under this approach, allocations of political power need to consider equality of the vote in the context of citizens, rather than overall population.
The argument for citizen-based apportionment is based on four assumptions: (1) the voting right is a fundamental right invested in the individual; (2) the concept of one person, one vote protects the individual vote from diminution, rather than some other concept of non-voting representative access; (3) general-population apportionment results in vote diminution on par with pre-Wesberry congressional districts due to the presence of large, non-citizen populations; (12) and (4) the original one person cases did not anticipate the presence of a large, non-citizen population concentrated in select constituencies that would lead to vote diminution.
II. THE POLITICAL RECEPTION OF ONE PERSON, ONE VOTE
Justice Douglas's coining the phrase one person, one vote may have involved a degree of public relations. As Ansolabehere and Snyder recount, the requirement that districts have equal populations outraged state legislators. (13) Linking a requirement that districts have equal populations to an equal voice in the government provided a more convincing justification for what initially proved to be a widely unpopular court decision. Subsequent to the Reynolds decision, a bipartisan coalition of U.S. senators introduced a constitutional amendment to allow for non-population based apportionment of at least one state legislative chamber. (14) The move failed by seven votes. (15) A proposal offered in the House would have removed state legislative apportionment authority from the jurisdiction of the federal courts. (16)
The pursuit of apportionment equality followed nearly two decades of litigation after the first challenge to population differences among districts went to the Supreme Court in Colegrove v. Green. (17) Decades of legislative inaction allowed silent gerrymandering to increase differences in the populations in most states' congressional districts. For example, on the eve of Wesberry v. Sanders, Georgia's largest district had three times the population of the least populous. (18) In Texas, Speaker Sam Rayburn's district had only twenty-three percent of the population of the adjacent Fifth District. Michigan's Twelfth District in the Upper Peninsula had little more than one fifth the population of that state's most heavily populated district. (19)
Post-Wesberry litigation ultimately produced a standard of strict population equality across congressional districts. (20) After the post-2000 census, seventeen states had the absolute minimum population deviations in their congressional districts while in another six states the population range between the largest and smallest districts was fewer than ten people. The one person, one vote issue seemed mooted.
III. ONE PERSON, ONE VOTE: EMPIRICAL FICTION?
Even with the advent of the one person, one vote regime, the concept is far from literally true in application. The "Redistricting Revolution" produced equally populated districts, but the revolution did not achieve districts with equal numbers of voters. Indeed some of the most under-represented districts prior to the 1960s have now become, in terms of their numbers of voters, among the most over-represented (see Table 1, next page). There are numerous examples.
In 1960, Georgia's Atlanta-based Fifth District was the second-most overpopulated district in the nation with more than 800,000 people. Its population was more than two-and-a-half times greater than Georgia's least populous district. (21) In the 2008 presidential election, Georgia's Fourth and Fifth District, which still encompass much of the 1960-vintage Fifth District, had less turnout in the presidential election than the state average and ranked tenth and eighth, respectively, among the state's thirteen districts. (22)
In New York, the Twentieth District cast 325,706 votes (23) for the two major presidential candidates in 2008 while the Sixteenth District tallied only 167,108 votes. (24) This difference far exceeds the population differences registered just prior to Wesberry when the most populous district had 471,001 people compared with 350,186 in the least populous. In California the congressional district with the highest turnout in the 2008 presidential election tallied three times the number of votes cast in the district with the fewest votes. This is a far greater disparity than the population difference in 1962, when the most populous district had less than twice the population of the least populous. In Texas, the range in votes cast in 2008 rivaled the 1960 population range with the most politically active district tallying 3.4 times as many votes as the district with the lowest turnout. Today the ballots of some voters still have several times the influence of the ballots cast in other parts of the same state.
IV. THE "FAIRNESS" ISSUE IN AMERICAN ELECTION LITIGATION
The enduring theme of litigation in U.S. elections is "fairness." We seek fair campaign finance laws, fair vote counts (and recounts), fair registration systems, and fair redistricting. But as a people, Americans are deeply skeptical about the ability to attain fairness. Public opinion consistently indicates skepticism about the ability to arrive at a fair and unbiased vote count, to have individual votes matter. (25)
No electoral exercise evokes more skepticism than redistricting. Electoral history is replete with examples of district boundaries manipulated to advantage parties, racial groups, incumbents, or geographic regions. (26) The gerrymander as a source of political ills is generally accepted in both the larger public and among the pundits in popular media. This skepticism persists in part because the enduring ideal of voting equality--one person, one vote--reinforces the notion that elections will reflect the majority will and the diversity of opinion in the public. (27)
One reality of the existing single-member district regime is a partisan difference in the districts won by the parties based on turnout. Political scientist James Campbell observed nearly two decades ago that Democratic members of Congress were far more likely to be elected from low-turnout districts than high-turnout districts. (28) Campbell termed this phenomenon the "cheap seats" theory, because Democratic parties have to expend fewer votes across a state to win disproportionate seats within a state. When compared to other sources of bias, estimated bias arising from turnout is substantially greater than estimated bias due to population deviations that existed before the one person, one vote era. (29) The same phenomenon is observed in state legislatures. (30)
V. ONE PERSON, ONE VOTE ... REALLY?
Now almost half a century after the launch of the "Redistricting Revolution," has one person, one vote really been achieved? Legislatures, under the watchful eye of the judiciary have spent untold numbers of hours shifting precincts and even census blocks around in an effort to achieve exact equality in the numbers of residents across a state's congressional districts. However, the Supreme Court cloaked its demands in terms not of equalizing populations, or voting-age populations, or citizen voting-age populations, but specified one person, one vote. Of course, equalizing populations is recognized as a legal fiction due to problems with achieving a complete census count, population shifts between the time of the census and drawing district boundaries, and quirks in data management.
Despite efforts to equalize populations through redistricting, vast disparities persist in participation rates among the districts of most states. In four states in 2010, the turnout in the congressional district having the greatest number of participants was at least three times as large as turnout in the district with the fewest voters. In another four states, at least twice as many voters turned out in some districts as in the district with the least interest in casting a ballot. Thus in almost one fifth of the states having more than a single district, substantial differences persist in the numbers of voters participating in a state's congressional elections. At the other end of the distribution, the district with the highest turnout in six states had less than ten percent more voters than the district with the lowest turnout. The difference in turnout between New Hampshire's two districts was less than 500. (31)
Have differences in the numbers of voters per district in a state reached the levels that triggered demands to balance populations fifty years ago? To answer that question, we compare the extent to which one person, one vote is not being achieved with the magnitude of the population differences just before Wesberry v. Sanders. Forty states had at least two districts in 1960 and 2010. As shown in Table 1, the differences in turnout in 2010 in nine of those states exceed the size of the differences in district populations as they existed following the 1960 reapportionment. For example, after the state legislature created maps to incorporate the new districts allocated to California in 1960, the most populous district had almost twice the population of the smallest district. In 2010, the California district with the highest turnout had 3.40 times as many voters as the district with the fewest voters. New York had even greater disparities than California. The most populous district drawn for the 1962 election had 1.34 times the population of the smallest district. In 2010, the New York district with the highest turnout had 3.43 times the vote cast as in the lowest turnout district. Turnout disparities are especially striking in the largest states. The states in which 2010 turnout ratios (32) exceed the population ratios of two generations ago include four of the five most populous. (33)
A possible explanation for the greater disparities in turnout in 2010 might involve population shifts during the decade. To explore the possibility that turnout differences result from the growth of some districts during the decade, we examined the turnout differences in 2002 right after the redesign of districts. The results for 2002, as shown in Table 2 below, are similar to eight years later. In eight states, six unchanged since 2010, the differences in turnout in 2010 exceeded the differences in population in the 1962 districts. In Missouri the ratios for 2010 and 1962 were identical. States in which the 2002 turnout differences exceed population differences at the eve of the "Redistricting Revolution" cannot blame decade-long population changes as the problem.
It is almost axiomatic that it should be easier to balance turnout among districts in states having fewer districts--assuming that achieving one person, one vote is indeed the objective. Less populous states are less likely to have vast disparities in ethnic concentrations, to have mega-urban areas, to have great differences in levels of partisan competition, and so forth. Or if they do, it would be easier to divide the various elements in the state's population among districts when there are fewer rather than more districts. For example, it would be impossible for each New York congressional district to have an equal share of New York City's varied population. On the other hand, it would be possible to create two very similar districts in Rhode Island or Idaho.
Let's consider this topic in the context of small states. There are nine states that had two or three congressional districts as of 2010. Six--Hawaii, Idaho, Maine, New Hampshire and Rhode island--have had two districts since 1963. New Mexico and Utah had two districts in 1963 but gained a third seat during the next five decades. Nevada had a single seat prior to 1982 when it earned a second one, and, as of 2011, it had three seats. Nebraska has maintained three seats for the last half-century, in 1962, just prior to the onset of the "Redistricting Revolution," the six states that elected their representatives from single-member districts had an average ratio of the highest turnout to the lowest turnout of 1.38. (34) At the upper extreme, turnout in one Utah district was 173% of that of the other district, while in Idaho the ratio between the two districts was 1.62. Only in Maine was turnout almost equal for the two districts with the ratio of the higher to the lower being 1.02.
At the end of the decade, after states had redrawn districts to equalize population and Hawaii and New Mexico had adopted single-member districts, the average ratio of highest to lowest turnout was down to 1.18. Idaho had the highest ratio, 1.32, while turnout was almost exactly equal in the two districts of New Hampshire (1.004) and Utah (1.002).
Redistricting to adjust for the population shifts of the 1960s resulted in an average ratio of highest- to lowest-turnout across the eight states of 1.09 in 1972. Only New Mexico at 1.32 was above 1.12. In 1982, the average for the nine states was 1.12. (35) In 1992, the mean stood at 1.09 while after the most recent redistricting the average ratio was 1.21 in 2002. The mean for 2002 is the largest for any election immediately after a reapportionment and indicates that over time, turnout has not become increasingly equal across districts within a state, even in the smallest multi-member states where equalization should be easiest if that were the goal.
Some states have consistently had greater equality in turnout among their districts. The ratio for Rhode Island since 1982 has never exceeded 1.13 and has been less than 1.10 in all but two elections. During the same period, the turnout ratio in Hawaii has exceeded 1.10 only once and has been 1.05 or lower in eight elections. On the other hand, the ratio in New Mexico has fallen below 1.10 only once since it adopted single-member districts in 1968 and has been less than 1.20 in only six elections.
Even though these states with small populations often had sizable differences in turnout in congressional contests, the disparities were typically less than in larger states. Ratios in the small states never reached 2.00.
VI. MODELING CONGRESSIONAL TURNOUT DEVIATIONS FROM ONE PERSON, ONE VOTE SINCE 1962
In Table 3 (next page), we present statistical analysis of the structure of the congressional district voter turnout deviations for all states with at least two congressional districts, starting in 1962. Data are analyzed at the state level for entire apportionment periods (decades). The dependent variable in the analysis is relative turnout--measured as the ratio of ballots cast in the highest-turnout districts divided by the lowest-turnout district in the state. (36) Seven independent variables are included in the analysis. These variables were selected to allow us to test the relationship between the presence of non-citizen populations and turnout ratios, while also controlling for other historically relevant variables related to voter participation, such as the size of the higher-turnout White Anglo population, the presence of a more competitive political environment (which presumably would lead to more voter mobilization across districts), and the size of the state.
The specific independent variables included are: the proportion of the state population that is White, non-Hispanic; (37) two measures of the potentially ineligible foreign born population; (38) a nominal variable indicating whether the state is subject to Voting Rights Act section 5 preclearance; (39) a measure of the percent of the state population living in jurisdictions subject to preclearance; (40) the percent rural population in the state; (41) the number of congressional seats in the state; and a measure of partisan political competitiveness. (42) Data are subjected to linear regression analysis.
The results of the analysis indicate that, in the 1960s, 1970s, and 1980s, none of the predictor variables are significant except for foreign-born population (associated with smaller turnout ratios) and the number of congressional districts, which is significantly and positively associated with larger turnout ratios. The larger states, with more numerous and populous urban centers, also exhibited a greater propensity toward disparities in voter turnout--even after the imposition of the strict one person standard going into the 1970s and 1980s.
The behavior of the model changes substantially in the last two decades. The states with more congressional districts continue to have larger turnout ratios, as in the previous three decades. More interesting is the emergence of demographic and legal factors that are significantly related to the size of the turnout ratios. First, the presence of smaller turnout ratios--and therefore more equitable turnout across districts--is associated with a larger share of the voting-age population who are citizens. Second, the greater the share of a state population covered by the section 5 preclearance requirement, the higher the turnout ratio.
The results for the previous two decades are intriguing. The emergence of a significant relationship between the citizen share of the voting-age population (VAP) in states and the turnout ratio is coincident to the growth in the variation of citizen share of VAP across several states. In eight states, the current citizen share of the VAP has fallen by at least five points since 1992, to 89% or less of the VAP: Arizona, California, Colorado, Florida, Nevada, New Jersey, New York, and Texas. Five of these nine (California, Texas, Nevada, Arizona, and Florida) also rank among the top ten for percent non-Anglo White population, and four (all but Nevada) are also section 5 states.
The significant section 5 result indicates one of two possible explanations. The first is that it is an artifact of historic turnout issues--the section 5 coverage formula includes historic low voter participation in conjunction with the historic use of tests or devices to qualify to vote. The second possible explanation is political. The implementation of section 5 in 1991-92 was accompanied by substantial pursuit of political goals by the first Bush Administration. (43) The application of the "max Black" strategy (and accompanying "Hispanic max" in select jurisdictions) substantially increased the number of majority-minority districts in nearly all of the section 5 states--and these districts are among the lowest-turnout districts in these states. The pursuit of political goals through the DOJ resulted in the segregation of high-turnout White districts from low-turnout majority-minority districts.
Considering the arguments put forward in this paper, it is especially important to understand that the original goals of those who fought for the seemingly innocuous ideal of one person, one vote, did not consider race a salient factor. (44) So, the subsequent interplay between voting rights policy, the presence of large non-citizen voting-age populations, and the structure of turnout ratios is intriguing, and causes us to consider the role of demographic change and the institutionalization of rural representative dominion in the United States at the beginning of the twentieth century.
Historical developments borne out of the demographic effects of industrialization in the late nineteenth and early twentieth centuries contributed to overrepresentation of rural areas of American states, resulting in the dilution of a single individual's vote in urban areas and an increase in the voting power of a single individual's vote in rural areas. (45) After the court's foray into this political thicket, though, we might assume that the equalization of voting districts to conform to the one person, one vote ideal would solve this problem. But not long after the problem was solved, other voting-related legislation was passed that only problematized further the one person, one vote mantra.
The most salient was the Voting Rights Act (VRA) of 1965, passed in the wake of the Civil Rights Movement. Although the Supreme Court sustained the VRA against initial challenges, (46) the Act has had a measurable, but tenuous and problematic, effect on the ability of the one person, one vote ideal to be realized, especially after the 1982 extension and amendments that led to the creation of majority-minority districts. Some of the majority-minority districts have distinctively low turnout rates.
How should we understand these effects on the one person, one vote ideal? We suggest that recent work in political science that seeks to problematize smooth developments of politics and political institutions might be the best way to make sense of these troubling developments.
The subfield of American Political Development (APD) has grown significantly within the study of American politics over the last two decades. (47) The basic premises of APD assume that political institutions (including legislation) created at particular times for particular purposes do not arise within a political order that is devoid of preexisting institutions (and legislation) that were also created at particular times for particular purposes. Therefore, institutions (and legislation) often create friction, bumping into each other to produce a political order that is often more a jumble of conflicting aims and purposes rather than a smooth running and seamless political order that generates the results all of the existing political institutions were created to achieve. Known as "intercurrence," the idea is that researchers need to locate the historical construction of politics in the simultaneous operation of older and newer instruments of governance and in controls asserted through multiple orderings of authority, whose coordination with one another cannot be assumed and whose outward reach and impingements, including on one another, are inherently problematic. (48)
This more temporal approach to understanding the development of political phenomenon can help us understand the VRA's effects on the ideal of one person, one vote. This then allows us to see that as the ideals of equalized population were implemented initially without respect to race, the ideal soon had to operate alongside newer legislation like the VRA and its subsequent amendments and court-created standards. In this sense, one person, one vote and the VRA became layered upon one another, further preventing the attainment of equal voting power that the original ideal was supposed to realize.
VII. CITIZENSHIP AND CONGRESSIONAL SEAT APPORTIONMENT
There are observable, significant variations in voter turnout ratios and the citizen proportions in the adult population. Achieving citizen apportionment requires either a constitutional remedy via litigation or a state policy action via legislation. Any state policy effort will in all likelihood be subject to litigation. This is aside from determinations regarding the necessity or merit of this concept in the political public.
The heart of the debate over citizen-based apportionment focuses on a fundamental difference in how the guarantees of the right to vote are applied. Judge Alex Kozinski, writing in partial dissent in Garza v. County of Los Angeles, articulated these differences through two theories of representation. (49) First is the "representational equality" model, which "assures that all persons living within a district--whether eligible to vote or not--have roughly equal representation in the governing body." (50) The benefits of such a model stem from equalizing population for the purpose of the representative function. These include granting residents equal access to representatives, diminishing variations in constituency service, and mitigating against unequal delivery of government services owing to district size. (51) Second, the electoral equity model" holds that:
[T]he principle of electoral equality assures that, regardless of the size of the whole body of constituents, political power, as defined by the number of those eligible to vote, is equalized as between districts holding the same number of representatives. It also assures that those eligible to vote do not suffer dilution of that important right by having their vote given less weight than that of electors in another location. Under this paradigm, the fourteenth amendment protects a right belonging to an individual elector and the key question is whether the votes of some electors are materially undercounted because of the manner in which districts are apportioned. (52)
The constitutional argument for citizen apportionment rests on a series of assumptions: First, the right to vote is determined by the states, subject to constraints from United States Constitution and its amendments, including the application of equal protection through the Fourteenth Amendment. Second, the right to vote is an individual right, and the protection of that right is invested in the guarantee of individual rights rather than group rights. Third, the right to vote is a citizen right. And fourth, apportionment of political power using districts requires that the individual vote not be diluted.
A policy remedy will necessarily rest on the previous four assumptions. It will likely be necessary to demonstrate that citizen apportionment is a general interest policy that does not violate individual voting rights in order to justify the total population deviations that will inevitably result.
The most recent litigation on this issue, Lepak v. City of Irving, (53) relied heavily on these assumptions, and placed special emphasis on an argument that when districts are created that include substantial deviations in the number of citizens or potential voters, this constitutes a geographic basis or apportioning method that dilutes the individual right to vote, and, citing to Reynolds v. Sims, that "weighting the votes of citizens differently, by any method or means. merely because of where they happen to reside, hardly seems justifiable." (54) Lepak challenged a six-member districting plan for the city council in Irving, Texas, designed to be a section 2 remedy. (55) The remedy district had approximately 40% fewer voting-age citizens and voters than the remaining five districts, even though the districts were of approximately equal total population, in ruling against the citizen-apportionment plaintiffs, the district court observed that the issue of representative quality versus voting equality has been muddled by the high court. (56)
Does Reynolds say that a districting scheme must equally distribute citizens? The muddling of the issue starts with this decision. (57) In the majority opinion, Chief Justice Warren famously observed that "[l]egislators represent people, not trees or acres. Legislators are elected by voters, not farms or cities or economic interests." (58) Chief Justice Warren's emphasis on "citizen" appears repeatedly in subsequent cases. Wesberry referred to the citizen right to an equal vote as one of the "fundamental ideas of democratic government." (59) And, in Board of Estimate v. Morris, the court observed that:
Electoral systems should strive to make each citizen's portion equal. If districts of widely unequal population elect an equal number of representatives, the voting power of each citizen in the larger constituencies is debased, and the citizens in those districts have a smaller share of representation than do those in the smaller districts. (60)
The citizen apportionment approach, when applied reactively by a court to the redistricting criteria adopted by states, requires reconsideration of the presumptive use of total district population. One person, one vote is the redistricting principle accorded highest priority. Legislatures and courts implementing maps confront this as their first, and sometimes sole, consideration. (61) This concept is consistently assessed in the context of total population, even by decisions making extensive use of the term "citizen" and "voter" when describing the dilutive nature of unconstitutional redistricting schemes. (62) Courts and legislatures have clung to the notion that precise population equality meets the one person, one vote standard, without exception or qualification.
To arrive at a determination that popularly equal districts do not satisfy equal protection of the right to vote, the Court will have to be convinced on three points. First, the Court will have to lay aside the representative model as being constitutionally equal to the equal vote model articulated by Judge Kozinski, and unambiguously endorse the concept of the right to vote as a citizen right. But herein resides confusion. Article I of the Constitution makes federal electors of all individuals enfranchised to vote for state legislature in the respective states. (63) So the right to vote is a right defined by the state, subject to the Fourteenth, Fifteenth, Nineteenth, and Twenty-Sixth Amendments. But Congress defined the right to vote in federal elections as only a citizen right in the Illegal Immigration Reform and Immigrant Responsibility Act of 1996. (64) From the standpoint of Congress, voting is a citizen right, and to exercise it without citizenship is a crime.
Second, the Court must be convinced that the concept of geography as applied in the plans overturned in Reynolds applies to districts crafted within a political subdivision or otherwise independent of the use of clearly defined concepts of geographic place. The port of entry chosen by plaintiffs pursuing citizen apportionment is geography. Drawing on the reasoning of urban Alabama voters in 1962--that where you live should not dictate the value of your vote--the new challenge assumes that including geographically concentrated non-citizens creates disparities such that the jurisdiction is, per Reynolds, "[w]eighting the votes of citizens differently." (65) This approach then depends on extending the logic of Reynolds and Wesberry to sparsely populated places--that the individual's right to vote "is unconstitutionally impaired when its weight is in a substantial fashion diluted when compared with votes of citizens living in other parts of the State." (66) The Reynolds court concluded that "the weight of a citizen's vote cannot be made to depend on where he lives"; (67) this argument requires that the Court consider whether "the weight of a citizen's vote cannot be made to depend on" whom he lives near.
Third, the Court will have to determine that citizenship apportionment is not just permissible, but mandated. A half-century ago, the Court addressed the citizenship distinction in apportionment. Critics of citizen apportionment state, correctly, that the citizenship apportionment perspective demands that the Court ignore the following from Reynolds': "We realize that it is a practical impossibility to arrange legislative districts so that each one has an identical number of residents, or citizens, or voters." (68)
The challenge presented by this dilemma is evident. In Daly v. Hunt, the Fourth Circuit reversed a North Carolina district court determination that the commission districts of Mecklenberg County violated the one person, one vote standard because of large voting-age population deviations. (69) Then, in Chenv. City of Houston, plaintiffs unsuccessfully challenged citizen voting-age population deviations that exceeded a ten-point range. The court, in rejecting Chen's challenge, noted:
The large populations in the other district will leverage the votes of the smaller number of eligible voters there. However, if a more carefully calibrated measurement is embraced in order to serve this principle of individual electoral equality, the area with the smaller number of voters will find itself relatively disadvantaged. Despite the fact that it has a larger population--and thus perhaps a greater need for government services than the other community--it will find that its political power does not adequately reflect its size. In addition, it could be argued that because the representative chosen from such a district will have a larger number of constituents, the ability of her constituents--whether or not they are potential voters--to petition and voice their opinions will be proportionately reduced .... If one accepts the principle of representational equality that representatives are chosen by a district's voters, but should represent all persons resident therein--these results may be unacceptable. The choice between these two models is stark, and because it is rarely encountered, historical and legal guidance is sparse. (70)
The Circuit Court in Chen determined that the challenge presented by these approaches to representation was anticipated by the framers of the Fourteenth Amendment. (71) According to Justice Thomas, "as long as we sustain the one person, one vote principle, we have an obligation to explain to States and localities what it actually means." (72)
To date, litigation seeking citizen-based redistricting has been unsuccessful. But, can a legislature create citizen-based districts as a policy choice?
The case law, and practical use of population standards other than total population is limited. Daly v. Hunt leaves open the argument about the use of other population standards, (73) and the courts have generally conceded that some latitude exists for states to alter the districting equality baseline. The most notable case presents one example of exceptional circumstances for changing the baseline from total population. (74) In apportioning its state legislature, Hawaii chose to use registered voters because of unique circumstances. Hawaii apportioned on registered voters because of the large number of military personnel and resident tourists counted in the census but ineligible to vote in the state. (75) States are not required to use census data to remap their state legislatures. (76) More recently, the courts have upheld the authority of states to reallocate or exclude incarcerated individuals from apportionment formulas, including at least one congressional apportionment. (77)
The Hawaii case presents the most detailed treatment of an apportioning system that excluded significant numbers of persons. The state could use the alternative apportionment based on voters because not doing so would lead to a result that, according to the Court, is "substantially different from that which would have resulted from the use of a permissible population basis." (78) In advancing a citizen-based apportionment, a state must satisfy a court that such an apportionment otherwise avoids an irrational distribution of political power, and in fact reflects the "permissible population base."
VIII. ISSUES IN CITIZEN-BASED APPORTIONMENT
There are a variety of troubling consequences to citizen apportionment. Among these are problems potentially arising in the implementation of sections 2 and 5 of the Voting Rights Act of 1965; implementation problems related to data availability, validity, and reliability; and geographic consequences for the apportionment of political power among the states.
The most obvious is the implementation of the Voting Rights Act, especially sections 2 and 5. Under section 5, changes in election law cannot diminish the political access of a minority group from the status quo (the nonretrogression standard). (79) Equalizing citizen populations across districts will likely result in fewer, more heavily populated Latino-majority state legislative and congressional districts in section 5 jurisdictions of states such as Arizona, Texas, and New York. (80) In urban centers such as Atlanta, some majority-Black districts will have larger populations because of the presence of non-citizen populations. The reduction of majority-minority opportunities will necessarily constitute retrogression, unless the courts and the Department of Justice reject the previous electoral baseline in these jurisdictions for a citizenship baseline for measuring representation.
With regard to section 2 of the Voting Rights Act, problems make themselves quickly known. Under citizenship apportionment, majority-minority remedies for racially polarized voting become more difficult to implement. Under Gingles prong one, the minority population must be sufficient numerous and geographically compact to constitute a majority of the district. (81) This will necessarily drive these districts to encompass much larger total populations in some jurisdictions to simultaneously satisfy the equal citizen population criterion while also being sufficiently majority-Latino to constitute a remedy. In the representational equality context, this approach would dilute the quality of access to representation by Latinos, especially in Texas and California. (82)
Then, there are methodological challenges to implementing this proposal. Put simply, are current citizen data produced at a sufficiently accurate level to craft citizen districts? The changes in the implementation of the census-specifically, the movement away from the old "one in-six" long form--limited the ability of redistricting teams to use accurate, current, local citizenship data when drawing maps. (83) Mapmakers now make use of American Community Survey data to estimate citizen data, which has larger estimation errors than the old long form data. (84)
Citizen- or voter-based apportionment has geographic consequences for representative power. Non-citizens in the United States are sufficiently numerous to account for over thirty congressional districts. In 2000 and 2010, estimates excluding non-citizens from the allocation of seats among the states indicate that citizen-based apportionment would have allowed six states that lost seats to keep them, (85) and three other states would have gained a seat. (86) Excluding non-citizens would have removed six seats from California, and one each from mega-states New York, Florida, and Texas. (87)
Finally, there are other potential equal protection concerns regarding those who currently lack suffrage. Citizen voting-age apportionment confronts the problem that the population used to equalize apportionment will change more quickly than the general population. By the time districts are crafted and elections are conducted, we know that numerous people will have come of voting-age. Citizen apportionment, regardless of voting-age, confronts the prospect that non-citizens will become citizens. Measures other than total population introduce more moving parts in terms of population maturation and eligibility change among residents that more immediately throw equal apportionment among the districts out of balance.
The citizenship apportionment argument is initially seductive. Voter eligibility for state and federal election in the United States is universally predicated on citizenship. It is a right that is, at present, limited by citizenship and age, and can therefore only be exercised by citizens of majority. There are electoral consequences that arise from district schemes that introduce substantial variations in the citizens, citizens of majority, and voters among districts. The consequence is that some votes are worth more than other votes when selecting federal, state, and local lawmakers. And, the structure of these inequities at the congressional level are significantly associated with the presence of large, non-citizen populations and the application of section 5 of the Voting Rights Act.
But the citizenship apportionment approach has problems. The courts consistently recognize that there are other aspects of representation, and these are also subject to consideration when equalizing representative district populations. A variety of other rights and liberties require access to government, and those rights and liberties exist for all persons in the United States--not just citizen voters. Taken in combination with the practical issues of implementing citizen-based apportionment, the concept, while novel and intriguing in the abstract, becomes less attractive to courts and states.
In light of the difficulties involved in measuring fairness by number of voters or, alternatively, citizen voting-age population, it is easy to see why the courts seamlessly shifted from the rhetoric of "one person, one vote" to applying "one person, one share of a seat" when reviewing districting plans. Justice Douglas, a best-selling author, no doubt recognized the appeal of the phrase. Indeed the alacrity with which his colleagues reiterated the standard demonstrated that they saw its utility in evoking compliance from powerful political forces whose hold on authority would be undermined by disrupting apportionment schemes, many of which were generations old.
Deciding cases in the 1960s when the United States enforced restrictive immigration policies and experienced relatively little in-migration and permanent settlement by illegal immigrants, the court might reasonably have expected that eliminating gross population disparities would result in districts within each state having roughly equal numbers of citizen adults. The Justices might have also anticipated that the accelerating Civil Rights Movement would overcome or eliminate the barriers to African-American participation in the South. (88) With access to the ballot unimpeded, the Court might have expected that numbers of voters would come to be approximately equal to the voting-age population throughout a state. (89) Thus, although no one would have projected that every eligible adult would actually vote, the discount rate from person to adult to voter might have shown relatively little variation so that the short-hand, bumper sticker phrase, "one person, one vote," would not have seemed too farfetched. Considering contemporary demographics, were Justice Douglas and his colleagues applying their touchstone today, they would allow much greater deviations in population between districts than today's jurists.
(1.) WILLIAM SHAKESPEARE, THE TRAGEDY OF HAMLET, PRINCE OF DENMARK act 3, sc. 1.
(2.) Id. at act 1, sc. 4.
(3.) Gray v. Sanders, 372 U.S. 368 (1963).
(4.) Id. The eight most populous counties earned the popular-vote winner six unit-votes while the next thirty in size were worth four unit-votes. The remaining counties had two unit-votes each. See William G. Cornelius, The County Unit System of Georgia: Facts and Prospects, 14 W. POL. Q., 942, 942-60 (1961); Albert B. Saye0 Georgia's County Unit System of Election, 12 J. POL., 93, 93-106 (1950). Maryland also used a county unit system for many decades, though the geographic vote dilution evident in Georgia was not found in Maryland's use of the system. ROBERT S. FRIEDMAN, THE MARYLAND COUNTY UNIT SYSTEM AND URBAN-RURAL POLITICS (1958).
(5.) 377 U.S. 563 (1964).
(6.) 376 U.S. 1 (1964).
(7.) Karcher v. Daggett, 462 U.S. 725 (1983).
(8.) Larios v. Cox, 300 F. Supp. 2d 1320, 1325 (N.D. Ga. 2004) (clarifying the precision of population equality required of legislative districts and rejecting the notion of a "safe harbor" in population variation); see also Cox v. Larios, 542 U.S. 947 (2004) (affirming the reasoning in Larios v. Cox by an eight-to-one vote).
(9.) From 1840 to 1910, the U.S. House expanded its membership, reaching 435 in 1910. Since 1910, this number has remained fixed except for a brief increase to 437 with the admission of Alaska and Hawaii as states. The U.S. House uses the method of equal proportions, which is supposed to minimize the percentage differences in the population of districts across states.
(10.) See Dep't of Commerce v. Montana, 503 U.S. 442 (1992).
(11.) In contrast with the United States single-member districting arrangements, most democracies use some form of proportional representation (PR). In PR's purest form, party seats are based exclusively on the share of the votes received and have no bearing on population. Thus, in Israel and the Netherlands, where the entire nation serves as a single electoral district, any party (other than the tiniest ones) receives a share of seats almost exactly equal to its share of the votes. Only parties so small that they fail to attain the low threshold for inclusion in the parliament do not have a vote-to-seats ratio of approximately one. PR systems invariably have a minimum vote threshold for inclusion. See, e.g., AMY J. DOUGLAS, REAL CHOICES/NEW VOICES: HOW PROPORTIONAL REPRESENTATION ELECTIONS COULD REVITALIZE AMERICAN DEMOCRACY (2d ed. 2002).
(12.) Wesberry v. Sanders, 376 U.S. 1, 1 (1964).
(13.) STEPHEN ANSOLABEHERE & JAMES M. SNYDER, THE END OF EQUALITY: ONE PERSON, ONE VOTE AND THE TRANSFORMATION OF AMERICAN POLITICS 178-82 (2008).
(14.) Reynolds v. Sims, 377 U.S. 563 (1964).
(15.) Peyton McCrary, Bringing Equality to Power. How the Federal Courts Transformed the Electoral Structure of Southern Politics, 1960-1990, 5 U. PA. J. CONST. L. 665 (2003).
(17.) 328 U.S. 549 (1946).
(18.) CHARLES S. BULLOCK III, REDISTRICTING: THE MOST POLITICAL ACTIVITY IN AMERICA 35 (2010).
(19.) U.S. CENSUS BUREAU, CONGRESSIONAL DISTRICT ATLAS OF THE UNITED STATES (1964).
(20.) See Karcher v. Daggett, 462 U.S. 725 (1983) (striking down a reapportionment plan because the small population deviations among districts were not the outcome of a good-faith effort to achieve population equality).
(21.) BULLOCK, supra note 18.
(22.) See GA. SEC'Y OF STATE, VOTER REGISTRATION SYS., ACTIVE VOTERS BY RACE/GENDER W1THIN DISTRICT TYPE (2012) available at http://sos.georgia.gov/elections/ voter_registration/documentdirect%20ssvrz194.pdf.
(23.) MICHAEL BARONE & RICHARD E. COHEN, 2010 THE ALMANAC OF AMERICAN POLITICS 1081 (Jackie Koszczuk ed., 2009).
(24.) Id. at 1072.
(25.) See, e.g., Joseph Carroll, Is the Public Confident That Votes Will Be Accurately Counted on Nov. 7?, GALLUP (Oct. 27, 2006), http://www.gallup.com/poll/25189/public -confident-votes-will-accurately-counted-nov.aspx.
(26.) See BULLOCK, supra note 18; REDISTRICTING IN COMPARATIVE PERSPECTIVE (Lisa Handley & Bernard Grofman eds., 2008); JONATHAN WINBURN, REALITIES OF REDISTRICTING: FOLLOWING THE RULES AND LIMITING GERRYMANDERING IN STATE LEGISLATURES (2008).
(27.) But see Andrew Gelman & Gary King, Enhancing Democracy Through Legislative Redistricting, 88 AM. POL. SCI. REV. 541 (1994).
(28.) JAMES CAMPBELL, CHEAP SEATS: THE DEMOCRATIC PARTY'S ADVANTAGE IN U.S. HOUSE ELECTIONS (1996).
(29.) See Thomas Brunell, Partisan Bias in U.S. Congressional Elections, 1952-1996: Why the Senate Is Usually More Republican than the House of Representatives, 27 AM. POL. Q. 316 (1999) (finding that a pro-Democratic bias arises from population skews and turnout skews, while a Republican bias arises from distributional effects or gerrymandering).
(30.) See Kenneth A. Wink, 'Unwasted' Votes in the U.S. State Legislative Elections: A District-Level Approach to Operationalizing Partisan Bias (paper presented at the annual meeting of the Midwest Political Sci. Ass'n, Chi., 1993); see also Ronald E. Weber, Race-based Districting: Does It Help or Hinder Legislative Representation?, 19 POL. GEOGRAPHY 213 (2000).
(31.) Calculated by the authors from data provided by the Clerk of the U.S. House of Representatives. Election Information, OFFICE OF THE CLERK OF THE U.S. HOUSE OF REPRESENTATIVES, http://clerk.house.gov/member_info/electionInfo/index.aspx (last visited Feb. 21, 2012).
(32.) The turnout ratio is the number of votes cast in the highest-turnout district in a state, divided by the number of votes cast in the lowest turnout district in the state. This is conceptually the same as the population ratios used to describe the population disparities in Gray v. Sanders, 372 U.S. 368, 388 n.7 (1963) (Harlan, J., dissenting).
(33.) Texas, among the five most populous states as of 2010, is not included, but its omission is not due to having relatively equal turnout among its districts. In 2002, the highest turnout district had 3.79 times the turnout in the district with the fewest voters. In 2010, 3.53 times as many voters participated as in the lowest-turnout district. However, neither figure attained the 4.40 ratio for the most- to least-populous districts in 1962.
(34.) In 1962, Hawaii and New Mexico elected their two representatives at large.
(35.) Nevada was apportioned a second seat in 1982.
(36.) Election Information, supra note 31.
(37.) There are variations in the partisanship and turnout rates of different racial and ethnic minority groups, but Anglo Whites remain the highest-turnout racial/ethnic group. U.S. CENSUS BUREAU, http://www.census.gov (last visited Feb. 21, 2012). Anglo Whites also vote in opposition to the respective political preferences of various racial and ethnic groups more frequently than any other voting group in the United States. Id. Note that all demographic data are interpolated across the decade to account for secular change.
(38.) Id. For 1962-1970 and 1972-1980, the U.S. Census figures for foreign-born population are used to measure the potentially citizen-ineligible electorate in a state. From 1982 forward, the more precise citizen voting-age population percentage is used as a control. Id.
(39.) Section 5 of the Voting Rights Act of 1965, 42 U.S.C. [section] 1973c (2006), requires states tripping the turnout and test or device provisions of the section 4 "trigger" to have all changes in election law and procedures approved by the United States District Court for the District of Columbia or the Department of Justice before implementation. See CHARLES S. BULLOCK III& RONALD KEITH CADDIE, THE TRIUMPH OF VOTING RIGHTS IN THE SOUTH 1022 (2009).
(40.) States wholly covered by the Voting Rights Act section 5 are, since 1965, Alabama, Georgia, Louisiana, Mississippi, South Carolina, Virginia; partially covered since 1965, North Carolina; partially covered since 1970, California, New York; wholly covered since 1975, Alaska, Arizona, Texas; partially covered since 1975, Florida, New Hampshire, Michigan, South Dakota. Virginia coverage proportion is adjusted after 1997 for jurisdictions that subsequently bailed out.
(41.) U.S. CENSUS BUREAU, supra note 37.
(42.) Partisan political competitiveness is based on a "normal vote" measure of the net Republican electoral advantage in a state over the previous three national elections. INTERUNIVERSITY CONSORTIUM FOR POLITICAL AND SOCIAL RESEARCH, http://icpsr.umich.edu (last visited Feb. 21, 2012).
(43.) BULLOCK & GADDIE, supra note 39, at 358.
(44.) GENE S. GRAHAM, ONE MAN, ONE VOTE: BAKER V. CARR AND THE AMERICAN LEVELLERS (1972).
(46.) Allen v. State Bd. of Elections, 393 U.S. 544 (1969).
(47.) See, e.g., KAREN ORREN & STEPHEN SKOWRONEK, THE SEARCH FOR AMERICAN POLITICAL DEVELOPMENT (2004).
(48.) Id. at 108-18.
(49.) Garza v. Cnty. of L.A., 918 F.2d 763 (9th Cir. 1999).
(50.) Id. at 781.
(51.) Id. Critics of this perspective have referred to it as the "one resident, one equal share of access to representative" doctrine. Brief for Appellants at *27, Lepak v. City of Irving, No. 11-10194, 2011 WL 554155 (5th Cir. Dec. 14, 2011).
(52.) Garza, 918 F.2d at 782.
(53.) Lepak v. City of Irving, No. 3:10-CV-0277-P, 2011 WL 554155, at *1 (N.D. Tex. Feb. 11,2011), aff'd, No. 11-10194, 2011 WL 554155 (5th Cir. Dec. 14, 2011).
(54.) Reynolds v. Sims, 377 U.S. 563, 566 (1964); cf. Hadley v. Junior Coll. Dist., 397 U.S. 50 (1970).
(55.) BULLOCK &GADDIE, supra note 39, at 20-22.
(56.) Lepak, 2011 WL 554155, at *1.
(57.) The Reynolds decision also decided the one person, one vote issue in litigation for forty-five other states. Some notable examples of the popular malapportionment overturned by the high court on that day: in the Connecticut Assembly, the largest to smallest district had a voter ratio of 424:1; in the New Hampshire General Court, 1081:1 (one township of three people had a seat); in the Utah Legislature, 196:1; in the California Senate, where Los Angeles County had one seat, 428:1.
(58.) Reynolds, 377 U.S. at 562.
(59.) Wesberry v. Sanders, 376 U.S. 1, 8 (1964).
(60.) 489 U.S. 688, 693-94 (1989). Other decisions sometimes constrained the concept of the individual from citizen to voter. See, e.g., Hadley v. Junior Coll. Dist., 397 U.S. 50, 53 (1970) ("A qualified voter has a constitutional right to vote in elections without having his or her vote wrongfully denied, debased, or diluted."). Still other decisions even use the concepts of persons and voters and citizens interchangeably.
(61.) See BULLOCK, supra note 18, at 25-48; Nathaniel Persily, When Judges Carve Democracies: A Primer on Court-Drawn Redistricting Plans, 73 GEO. WASH. L. REV. 1131, 1139-41 (2005).
(62.) See Bd. of Estimate v. Morris, 489 U.S. 688, 694 (1989) (articulating the right of the individual citizen to an equal vote by stating that "the Court has insisted that seats in legislative bodies be apportioned to districts of substantially equal populations").
(63.) At present, all states ban non-citizen voting; to the extent that non-citizen voting occurs, it is for certain elections in select municipalities. See Gerald M. Rosberg, Aliens and Equal Protection: Why Not the Right to Vote?, 75 MICH. L. REV. 1092 (1977).
(64.) 18 U.S.C. [section]611 (2006).
(65.) Reynolds v. Sims, 377 U.S. 533,563 (1964).
(66.) Id. at 568.
(67.) Id. at 567.
(68.) Id.; see also Gaffney v. Cummings, 412 U.S. 735, 746 (1973) ("[T]otal population--even if stable and accurately taken may not actually reflect that body of voters whose votes must be counted and weighed for the purposes of reapportionment, because 'census persons' are not voters.").
(69.) 93 F.3d 1212 (4th. Cir. 1996). The court did leave room for consideration of voting-age population as a policy consideration.
(70.) 206 F.3d 502, 525 (5th Cir. 2000).
(71.) Id. at 527.
(72.) Chen v. City of Houston, 532 U.S. 1046, 1048 (2001).
(73.) 93 F.3d 1212.
(74.) Burns v. Richardson, 384 U.S. 73 (1966).
(75.) Id. at 92 ("Neither in Reynolds v. Sims nor in any other decision has this Court suggested that states are required to include aliens, transients, short-term or temporary residents, or persons denied the right to vote ... in the apportionment base by which their legislators are distributed and against which compliance with the Equal Protection Clause is to be measured. The decision to include or exclude any such group involves choices about the nature of representation with which we have been shown no constitutionally founded reason to interfere.").
(76.) Alaska, Kansas, Colorado, Illinois, Mississippi, New Jersey, Oregon, and Virginia all allow for some exclusion of either military personnel or prisoners in local or state legislative apportionment. See, for example, Groh v. Egan, 526 P.2d 863 (Alaska 1974), on Alaska's exclusion of military personnel from legislative apportionment. (Alaska, an income tax haven, has a large number of military citizens.) See also Mahan v. Howell, 410 U.S. 315 (1973); Borough of Bethel Park v. Stans, 449 F.2d 575 (3d Cir. 1971).
(77.) Fletcher v. Lamone, No. RWT-1 lcv3220, 2011 U.S. Dist. WL 6740169 (D. Md. Dec. 23, 2011) (upholding use of prisoner-adjusted census data for congressional districts that reallocated felons within the state to their previous residence, including over 1,000 persons to residences out of the state).
(78.) Richardson, 384 U.S. at 93.
(80.) California and Florida would also confront the problem of expanded Latino districts, but these constituencies are not located in section 5-covered jurisdictions.
(81.) Thornburg v. Gingles, 478 U.S. 30, 50 (1986).
(82.) Brief of Respondent-Intervenors at *35, Lepak v. City of Irving, No. 11-10194 2011 WL 554155 (5th Cir. May 16, 2011).
(83.) This issue also came up in the Lepak litigation.
(84.) AMERICAN COMMUNITY SURVEY, http://www.census.gov/acs/www (last visited Feb. 22, 2012). The American Community Survey uses a five-year rolling survey of households to generate information previously included in the old long form.
(85.) Indiana, Michigan, Mississippi, Oklahoma, Pennsylvania, and Wisconsin.
(86.) Montana, Kentucky and Utah.
(87.) In Oklahoma, U.S. Rep. Ernest Istook would have avoided the discomfort of nearly losing his sea--only averted due to the retirement of other members of his delegation--if non-citizens had been excluded when apportioning seats among states, as Oklahoma would have retained six seats instead of dropping to five.
(88.) After a decade of glacial growth in Black registration in the South, African-American registration experienced an increase of almost fifty percent between 1962 and 1964. HARRELL R. RODGERS, JR., & CHARLES S. BULLOCK III, LAW AND SOCIAL CHANGE: CIVIL RIGHTS LAWS AND THEIR CONSEQUENCES 25-26 (1972).
(89.) Survey work done in the early 1960s found disfranchised southern African Americans eager to exercise the franchise, investing it with greater symbolism than Whites did. DONALD R. MATTHEWS & JAMES W. PROTHRO, NEGROES AND THE NEW SOUTHERN POLIYTICS (1966).
Ronald Keith Gaddie, Professor of Political Science, the University of Oklahoma; Editor, Social Science Quarterly.
Justin J. Wert, Assistant Professor of Political Science, the University of Oklahoma.
Charles S. Bullock, III, Richard B. Russell Professor of Political Science and Josiah Meigs Distinguished Teaching Professor, the University of Georgia.
Table 1: Low, Median, and High Turnout Ratios for US Representatives in American States Since 1962 Year Low Median High N * 1962 1.017 2.39 18.81 42 1964 1.015 1.92 17.48 43 1966 1.004 1.56 4.08 44 1968 1.031 1.428 4.799 45 1970 1.002 1.604 5.931 44 1972 1.003 1.358 4.09 43 1974 1.041 1.727 8.947 44 1976 1.007 1.49 5.137 44 1978 1.023 1.574 7.535 43 1980 1.037 1.524 6.482 42 1982 1.001 1.37 4.597 44 1984 1.001 1.397 3.533 44 1986 1.011 1.522 3.695 44 1988 1.018 1.538 3.224 44 1990 1.029 1.73 4.686 44 1992 1.019 1.354 5.746 43 1994 1.004 1.473 5.706 42 1996 1.018 1.357 5.251 43 1998 1.033 1.496 4.997 43 2000 1.049 1.399 4.351 43 2002 1.002 1.427 3.79 43 2004 1.015 1.366 4.104 43 2006 1.05 1.388 4.53 43 2008 1.022 1.349 4.671 43 2010 1.002 1.374 3.533 43 * All states with two or more congressional districts. The 'N' of states floats for two reasons: (1) states did not report ballots for uncontested elections, preventing the computation of a ratio; (2) until 1968, New Mexico was excluded because the congressional delegation was elected at-large rather than from districts. Table 2: Population and Turnout Ratios in States in which Recent Turnout Ratios Exceed the 1962 Population Ratio (Computed By Authors) State 1962 Population 2010 Turnout 2002 Turnout Ratio Ratio Ratio California 1.95 3.4 3.47 Florida 2.14 2.45 2.23 Illinois 1.98 2.91 2.89 Maine 1.09 1.12 1.12 Massachusetts 1.27 2.06 1.75 Minnesota 1.29 1.33 1.16 Missouri 1.34 1.44 1.34 New Jersey 2.29 2.6 2.16 New York 1.35 3.43 3.02 Virginia 1.72 1.44 1.89 West Virginia 1.39 1.24 1.47 Table 3: Modeling Congressional Turnout Deviations from One Person, One Vote Since 1962 1962-70 1972-80 1982-90 b (s.e. (b)) b (s.e. (b)) b (s.e. (b)) Intercept 5.206 1.157 -5.934 White Non-Hispanic -.031 -.004 -.001 Pop. (.012) (.005) (.010) Foreign Born -.109 .040 Population (.059) * (.025) n/a % Citizen of VAP n/a n/a .075 (.071) VRA Section 5 .997 .214 -.299 coverage (1.644) (.199) (.429) VRA Section 5 -.021 .002 .006 % coverage (.017) (.003) (.005) Percent Rural -.008 .003 -.007 Population (.013) (.006) (.006) Congressional Seats .054 .070 .082 (.017) ** (.007) ** (.016) ** Competitiveness -.359 -.342 .075 (.843) (.324) (.784) Adjusted [R.sup.2] .120 .550 .143 N 218 216 220 1992-00 2002-10 b (s.e. (b)) b (s.e. (b)) Intercept 10.308 6.506 White Non-Hispanic .004 Pop. (.003) * 005 (.003) * Foreign Born Population n/a n/a % Citizen of VAP -.104 -.060 (.016) ** (.015) ** VRA Section 5 -.038 coverage (.108) -.127 (109) VRA Section 5 .003 .006 % coverage (.001) ** (.001) ** Percent Rural .012 Population (.003) ** -.002 (.003) Congressional Seats .054 .043 (.004) ** (.004) ** Competitiveness .057 (.207) -.240 (.152) Adjusted [R.sup.2] .771 .685 N 214 215 * The dependent variable is the ratio of ballots cast in the highest-turnout districts divided by the lowest-turnout district in the state. * p > .10, two-tailed (.05 one tailed); ** p > .05, two-tailed.
|Printer friendly Cite/link Email Feedback|
|Author:||Gaddie, Ronald Keith; Wert, Justin J.; Bullock, Charles S., III|
|Publication:||Stanford Law & Policy Review|
|Date:||Mar 22, 2012|
|Previous Article:||Democrats at DOJ: why partisan use of the Voting Rights Act might not be so bad after all.|
|Next Article:||Drafting proper short bill titles: do states have the answer?|