District Shapes and Interest Representation
Fair Representation in North Carolina  

Douglas J. Amy

Many of the arguments against the elongated, many-sided 12th congressional district in North Carolina are based on its shape, described by opponents as "odd" or "bizarre." Its shape does clearly deviate from what some propose as the ideal district shape : a neat, compact geographical district, preferably drawn along the lines of political subdivisions.

An aura of tradition surrounds this ideal of compact districts. Such districts are commonly thought of as "natural" and "desirable," and some believe that their creation should be one of the highest goals in districting decisions. This approach to districting often is justified by three types of assertions: (1) that compact districts are politically neutral; (2) that compact districts do a good job of representing a homogeneous set of political interests; and (3) that voters' political identities are closely tied to geographical areas.

But once we begin to examine these standard justifications for compact districts, they do not hold up to close scrutiny. Indeed, compact districts often may do a very poor job of representing the interests of voters. The ultimate purpose of districting must be to enhance the representation of interests. But there is no necessary correlation between these interests and any particular district shape. As a result, the compactness of a district is a flawed measure of the adequacy or fairness of representation. In the end, the criterion of compactness should bow to the higher value of ensuring a fair opportunity for representation of minority voters.

The Myth of "Neutral" Compact Districts

One of the main reasons that compact geographical districts are celebrated is that they appear to be politically neutral. In particular, the plaintiffs in the Shaw v. Hunt case against the North Carolina districts portray compact districts as being "race-neutral" or "color-blind." In contrast, the 12th congressional district is portrayed as race-based and designed to further certain political interests. But while the latter is true, the former is not. Compact districts are not politically or racially neutral and they also inevitably promote a certain specific set of interests.

As a matter of principle, constructing racially or politically neutral districts is an impossible goal to achieve in single-member districts -- even those drawn according to traditional criteria like compactness. All compact districts will be politically biased and work to further some interests over others.

The reason is simple: these districts are single-member districts and as such they function to represent only the interests and views of one part of the electorate: the majority of the voters who elect that district's representative. The minority of voters who support the losing candidate or candidates are thus left with no effective representation of their interests. Although winning candidates inevitably claim to represent everyone in a district, they in fact work to promote policies that many people in that district strongly oppose.

The particular minorities that go unrepresented will vary from district to district. In one district, it may be a rural minority outvoted by an urban majority; in another it may be a Republican minority outvoted by a Democratic majority or a black minority outvoted by a white majority. But in all of these cases, it is the majority interest only that is represented. In actuality, then, there are no politically "neutral" districts that do not distort representation in some way. And this is true irrespective of whether the districts are drawn to be compact or not.

Given the inherent majoritarian political bias in all districts, it is misleading to argue that an attempt to draw more compact congressional districts in North Carolina would be a "race-neutral" effort. In practice, this effort would favor the interests of the white majority and work to the disadvantage of the black minority in the state.

Indeed, it is the political and racial imbalance produced by using this approach to districting in the past that necessitated the need to draw the oddly-shaped 12th district in the first place. So the choice in North Carolina is not between race-based districts and race-neutral ones; the choice is between drawing districts in which only whites are represented (a form of race-consciousness) and drawing districts that allow some representation for blacks as well.

The illusion of "race neutral" districts persists in part because it is sometimes difficult to discern any racial intentions in the boundary lines of compact geographical districts. How could a simple, square district be biased? The problem with this belief, of course, is that it focuses on the political intentions of the line-drawers and blithely ignores the political effects of these districts.

Such districts clearly have racially biased effects, irrespective of the intentions underlying the line drawing. Indeed, it can be shown that even if compact districts were drawn with no intentions at all -- that is, were drawn by a computer in a completely random manner -- that the effect would still be to underrepresent racial and ethnic minorities.

As Bernard Grofman has commented, "blind districting is extremely unlikely to give rise to proportional results." He calculated, for instance, that with completely random districting for the U.S. Congress, "an 11% black population minority could only expect to get 1.5% of the seats on average."

It should be understood that the fact that districts are not political or racially neutral -- that all districts promote some interests over others -- is an inherent and unavoidable characteristic of the single-member district system we use to elect members of the U.S. House of Representatives.

This system is simply not designed to represent fairly all interests in a particular district. The basic purpose of elections is to produce a legislature that accurately mirrors the variety of views and interests of the electorate -- that is what democracy is all about. But single-member districts are designed to represent only part of the electorate: the majority of people in a given district.

As noted above, all of these single-member districts -- no matter how they are drawn -- will inevitably leave minorities in these districts unrepresented. Moreover, as political scientists have long known, single-member district elections tend to produce a legislature that is misrepresentative -- a body that over-represents majorities and underrepresents minorities. It is this misrepresentative nature of single-member district elections that necessitated the creation of majority-minority districts to remedy the underrepresentation of minorities in the U.S.

The basic unfairness and unrepresentativeness of our single-member district election system is much of what underlies the complaints of the plaintiffs in the Shaw case. As white citizens, they feel they will go unrepresented in a district with a black majority. And this may be the case. But it is a mistake to see this circumstance as an argument against the creation of districts to ensure minority electoral opportunity.

In reality it is an argument against the single-member district system in general. Merely changing how the single-member district lines are drawn would not necessarily guarantee representation for the plaintiffs. The plaintiffs could go unrepresented in any single-member district, even a compact one, and even in a majority white district. For example, the plaintiffs could be Democrats in a predominantly white, Republican district.

The potential to go unrepresented is present in every single-member district, irrespective of how it is drawn. It should be emphasized, however, that the danger of going unrepresented is particularly acute for blacks where, as in North Carolina, there is persistent racially polarized voting. Such voting patterns, which seriously cripple electoral opportunity for blacks wherever they are in a minority, know no geographic or geometric boundaries.

The abolition of race-based districts would not solve the problem of the injustice of district minorities going unrepresented. It would only ensure that blacks are one minority that consistently is submerged in these North Carolina congressional districts.

Logically, the most effective solution is a districting approach which eliminates the problem of minority underrepresentation in general -- an approach that ensures all groups fair representation. The only realistic way to do this is to abandon single-member districts and move to a multi-member district, proportional representation election system.

Proportional representation (PR) elections are the most common form of elections in Western democracies in large part because they guarantee fair representation for political, racial, and ethnic minorities -- and do so without having to create special minority-dominated districts. PR systems use large, multi-member districts in which seats are distributed according to the percentage of the votes received by a particular candidate or party.

Consider, for example, a district in which ten legislative seats were being elected. If the Republican party were to receive 50% of the vote in the district, it would receive five of the ten seats; 30% of the vote would win three seats.

In such a system, minorities are guaranteed fair representation no matter where they live in the district, and no matter how the district lines are drawn. Indeed, the drawing of these large district lines no longer have political implications because minorities in these districts cannot be shut out of representation. As a result, reapportionment and redistricting in PR systems become rather mundane, uncontroversial activities because they have such little impact on representation.

PR would satisfy the interests of all parties in the Shaw case. PR would ensure that blacks could have a fair opportunity to elect representatives of their choice while at the same time it would not require the creation of oddly-shaped or race-based districts in which white voters were submerged.

Thus, if the plaintiffs oppose race-based districts like the 12th congressional district, and yet also do not want to return to traditional districts that submerge the vote of blacks in white majorities, then the only other viable option for them is to support the adoption of a PR system. Clearly, PR elections would be the most simple and elegant way to ensure fair representation for everyone in North Carolina.

While such remedies currently are not authorized by Congress, there are no provisions in the U.S. Constitution that would prohibit the use of PR elections for the U.S. House of Representatives. It is an alternative remedy for minority vote dilution that deserves much more serious consideration by the courts and all parties involved in voting rights cases. But as long as alternative elections systems are not available in North Carolina, the only other way to ensure fair representation for minorities is to design districts in which they are a majority.

It is this fact that is recognized in the Voting Rights Act and its amendments. Without the ability to design majority-minority districts, racial and ethnic minorities would have little or no chance of achieving their fair share of representation in our legislative bodies. By necessity, because minority populations are sometimes not geographically concentrated, race-based districting will sometimes result in oddly-shaped, non-compact districts. Especially in large congressional districts, that may be the only way to capture enough minority votes to create a majority in the district.

While such districts may be ugly geographically, they are beautiful democratically -- because they work well to ensure that minority voters have a meaningful opportunity to elect candidates of their choice. These oddly-shaped districts may offend our sense of aesthetics, but their political effects move much closer to our core democratic ideals and make our legislatures more representative.

The Myth of Homogeneous Interests in Compact Districts

Another main justification offered for giving compactness a high priority in districting is the belief that compact districts represent natural and homogeneous communities of interest based on geographical proximity. Critics of the 12th congressional district in North Carolina claim that its constituents have little in common (except race) compared to the other more compact districts in that state. There are several problems with this assertion.
First, the assumption underlying this position is that regions themselves are somehow "regularly" shaped. This is clearly not always true, as in North Carolina's Piedmont Crescent which stretches the length of the state.

Second, it assumes that all the voters in a particular geographical region share all the same interests. But this is not necessarily true. People in a region do share certain limited interests -- such as an interest in clean air or in particular economic development projects. At the same time, however, the people in a region may be divided, sometimes deeply, over competing concerns and interests.
This, unfortunately, is particularly true in racially diverse communities and even more so when there is a history of discrimination against racial groups. Moreover, people living in the same area often do not have common opinions about pressing social, economic and political problems such as abortion, taxes, welfare, affirmative action, free trade, gun control, crime, etc. Even when a district is almost entirely rural or urban, there are still usually distinct political cleavages along the lines of class, race, gender, and ideology. When regions are full of such diversity of interests and political opinions -- as most are -- it makes little sense to insist that geographical or geometric considerations be the primary basis for representation.

There are reasons to believe that geographic representation may not be as important today as it was in the past. In colonial times, many citizens were tightly bound to their geographical regions. Indeed, it would not be unusual for a person to grow up, raise a family, and die within 20 miles of where he or she was born. But that is not the case today.

Increasingly Americans are less tied to and less identified with a particular geographic locality. According to the U.S. Census Bureau, almost 20% of Americans move every year. 50% of the U.S. population changed addresses between 1985 and 1989. As a result, when people identify themselves politically, it often has less to do with their place of residence than with their particular ideological or party preferences.

As argued by journalist Hendrik Hertzberg, "The ties that bind are no longer just the ties of propinquity. The communities that count are communities of interest or belief -- communities that are at least as likely to be national as local. Ideology, profession, class, even racial and sexual identity -- those are the soil our roots grow in now, as much as or more than where we live."

In other words, for the purpose of representation, it may make little sense to group together people by geography. Instead, they should be grouped together by the political interests by which they define themselves. In the Supreme Court's decision in Shaw v. Reno, Justice O'Connor questioned the legitimacy of districts that link together people who "may have little in common with one another but the color of their skin." But a better question to ask is whether it makes sense to have districts that link together people who may have little in common with one another but the place in which they live?

Again, this is not to suggest that there are not some valid regional interests that deserve representation. The point is that these are not the only interests or the most important interests deserving representation; and that the primacy given to compact geographical districting only ensures that some other interests are not fully represented.

If the accurate and fair representation of competing interests is the goal of elections, then these interests -- not just traditional geographical considerations -- need to be taken into account in the districting process. Indeed, it is this logic that underlies the creation of majority-minority districts to ensure that minorities have a fair chance to be represented in our legislatures.

To underscore this point, it is useful to note that other Western democracies do not put the primary emphasis on geographical representation that we do here in the United States. Most of these other countries use different election systems (usually PR elections systems) that allow for the representation of a variety of interests: geographical, ideological, ethnic, etc. This is done by creating large, multi-member districts, in which several members of the legislature are elected in proportion to the vote received by particular parties.
PR systems recognize that geography is but one basis for representation and that citizens should be allowed to group themselves together in other ways to win representation. In the preference voting form of PR (a form already used in the U.S.), citizens can vote for a particular candidate because of his or her place of residence, or ideology, or gender, or race, etc. In this way, citizens form "voluntary constituencies" that come together on the basis of a variety of political criteria to elect particular candidates.

In contrast, our current single-member district system forces a certain kind of representation on voters -- it presumes that geography is what matters most -- while PR allows voters to be represented in the ways they feel most appropriate. This is another reason why PR has become the most common and popular election system among Western democracies, and another reason why it deserves more careful consideration here in the United States.

Conclusion: Fairness Over Compactness

The ideal of compact districts persists largely as a matter of political habit and tradition, not because it actually enhances the process of representation. The usual justifications for emphasizing compactness in districting do not stand up to critical scrutiny. As shown above, such districts are not "race-neutral" and they actually do a poor job representing all the voters in a district. Moreover, supporters of compact districts assume that voters in such districts have homogeneous interests and that geographical considerations play a central role in people's political lives -- neither of which is usually true.

As the standard justifications for the importance of compact geographical districts are not persuasive, the creation of such districts should not be given the highest priority in the districting process. When necessary, the criterion of compactness must yield to the goal of enhancing the process of representation -- in this case, the specific goal of ensuring a fair opportunity for representation of racial and ethnic minorities traditionally shut out of the political process in North Carolina congressional elections.

The notion of sacrificing compactness to higher democratic goals is in fact not a new one. For example, by insisting on the goal of equal populations in districts, the courts have given this a higher priority over maintaining compactness. The result has been the creation of districts that are not particularly compact and that often violate the traditional lines of political subdivisions.

But the courts have found those districts to be acceptable because they further the more important goal of establishing districts of equal population. The logic in the Shaw case is similar: the traditional criterion of compactness should yield to the higher good of enabling racial and ethnic minorities to have a fair chance to elect officials to represent them.
Douglas Amy is a professor of politics at Mount Holyoke College. He wrote Real Choices, New Voices: The Case for Proportional Representation Elections in the United States (Columbia University Press, 1993).
Recent Articles
October 19th 2009
A better election system
Lowell Sun

Election expert Doug Amy explains how choice voting can "inject new blood" into the elections of Lowell (MA), and give voters a greater incentive to participate.

October 16th 2009
Haven't Detroit voters spoken enough?
Livingston Daily

In Detroit, there have been three mayors in the past two years and the current one has come under scrutiny. Perhaps a system like instant runoff voting will help bring political stability to motor city.

August 21st 2009
Black candidate for Euclid school board to test new voting system
Cleveland Plain Dealer

Limited voting, a form of proportional voting, will be used in Euclid (OH), in the hopes of allowing better representation of minorities.

July 2nd 2009
Reforming Albany
New York Times

FairVote's Rob Richie responds in a letter to the editor making the case for proportional voting systems to bring substantive reform to New York's legislature.