According to a survey carried out by Handley , “19 of the 60 countries that delimit constituencies indicated that respect for communities of interest was a criterion considered by the boundary authority” . That selection of countries includes such disparate nations as Australia, Germany, Hungary, Pakistan, Nepal, and Papua New Guinea, all defining the term in slightly different ways. Such definitions range from the very vague “coherent area” and “homogeneity of community” to the more precise “community of economic, social, and regional interests” and “concentration of minority/indigenous populations.” So while there is a vague idea throughout the world of such a thing as a COI, different countries hold different ideas of what exactly that looks like. Unlike the vast majority of the countries surveyed by Handley , the United States has a very decentralized redistricting system, with each state running its own affairs. This means that the degree to which COIs factor into the process varies with the state; some do not consider them at all while others emphasize them a great deal. According to Levitt , 22 states take COIs into account when drawing state legislative districts, and 13 do the same for congressional districts. This practice is most common among states in the Western United States, because they tend to use independent commissions to draw district lines, which use “fair districting criteria” like respect for COIs . Even though a fair number of states have provisions for respecting COIs, many have had difficulty enforcing those provisions since they lack a precise definition for the notion . Some states have tried to buck this trend by laying out a more specific definition, again concentrated in the West. The state of Colorado references “ethnic, cultural, economic, trade area, geographic,danish trolley and demographic factors.” Alaska, in turn, defined communities by interviewing hundreds of residents on the interests they shared .

Among those countries and states that do include respect for COIs as a redistricting criterion, there is clearly a wide disparity in how they define those entities. Despite this lack of agreement, certain common threads appear across various definitions. One is that there is a geographic or spatial element to the concept. Morrill called the COI “the most geographic [of the redistricting criteria], in the sense that a major concern of geography is to identify the regional structure of a society…the territories with which citizens strongly identify, and whose integrity they want to maintain” . Stephanopoulos concurred, arguing that people who live nearby tend to have common interests and values and also feel more connected to each other . From the very beginning of California’s use of the criterion for redistricting, the state very clearly defined it as a territorial concept—a particular area with certain interests . This remains the case today, as the California Constitution defines COIs as “contiguous populations” . Even Monmonier , though holding that COIs were becoming large and fragmented due to advances in transportation and communication, recognized that geography continued to be relevant today . There is thus good reason to think of COIs as geographic regions of some type. This fact calls for a brief excursus into the topic of regions in geography. In this essay, he defined geographic regions as pieces of the earth’s surface with certain shared properties. Regions can be thought of as roughly two-dimensional features, usually contiguous and compact, with a common theme or character. Montello surveyed textbooks in human and physical geography and concluded that regions could be categorized into four types: administrative, thematic, functional, and cognitive.

Administrative regions are those formed by fiat and defined by precise boundaries and uniform membership functions. I propose that this type of region may be further classified into jurisdictions and their divisions. Jurisdictions are regions under the control of a distinct administration, and include countries, states, counties, cities, towns, school districts, and many other special purpose districts. Divisions, on the other hand, are sub-regions of a jurisdiction created for an administrative use, yet without administrations of their own. Examples of such entities include the Census tracts and blocks of the United States; the electoral districts of a state, county, or city; the voting precincts of a county; the service areas of a police or fire authority; the neighborhoods of a city ; and the parcels of a city. Thematic regions are defined by measurable themes and attributes, such as demographics and economics. Functional regions are determined by various forms of interaction among places, like commuting patterns or trade between cities. “Finally, cognitive regions are produced by people’s informal perceptions and conceptions,” for example, “my neighborhood” or “Southern California” . The latter three types of regions lack the potentially precise boundaries and uniform membership functions that characterize administrative regions, meaning that they cannot escape a level of “fuzziness” about their extent, and most seem to have a “core” and “periphery” to them. With these distinctions in mind, one can assess which type of region a COI might be. If the first common thread among definitions for the COI is its geographic nature, then the second is the objective or thematic aspect of the concept. The basic idea is of “a group of people that share common social and economic interests” . More specific thematic attributes are given in the redistricting law of certain states, like Colorado and Alaska . California refers to urban, rural, industrial, or agricultural areas, as well as “areas in which the people share similar living standards, use the same transportation facilities, have similar work opportunities, or have access to the same media of communication” . Morrill noted that “communities are revealed through patterns of work; of residence; of shopping; and of social, religious, and political participation” .

Mac Donald and Cain described it as a “clustering of some measurable social or economic characteristic,” though stressing that there is a crucial subjective component as well . Finally, topographic barriers such as mountain ranges and rivers, as well as important transportation links, are often used to delineate COIs . Such references to various socioeconomic attributes suggest that COIs qualify at least as thematic regions, if not also functional regions if transportation and communication links factor in. A third important thread is the subjective or cognitive element . Geographers and others have long recognized that people tend to identify with the place in which they live . Gardner and others have argued that this sense of identity is disrupted when district boundaries are constantly redrawn and/or disregard salient communities. When describing the principles many believed should guide boundary reorganization of local government areas in England, Prescott referenced one stating that “the boundary should be drawn to cater for local sentiment and regional patriotism” . Morrill contended that districts should be meaningful entities with which constituents can identify. Grofman introduced an idea that he called the “cognizability principle,” vertical aeroponic tower garden which refers to the ability of residents to cognize their district by being aware of the general configuration of the boundaries, thereby facilitating their “identification of and with the district” . Mac Donald and Cain maintained that residents of a COI “have to perceive and acknowledge that a social, cultural, or economic interest is politically relevant” . Perceptions of such interests do not necessarily correlate with demographic attributes, but may instead reflect environmental and cultural concerns, or even things like where parks and fire services are located . It seems then that the beliefs of individuals can be used to identify COIs as cognitive regions.

In sum, a COI can be defined as a geographic region with both objective and subjective elements to it. One may then see it as either a thematic, functional, or cognitive region, with the electoral district itself being an administrative region. Because thematic, functional, and cognitive regions often correlate and the first two are easier to identify than the third because their attributes are more directly measurable, experts tend to just measure the objective and assume the subjective will closely adhere . However, situations may exist where a thematically and functionally homogeneous area is too large or small to constitute one district, in which case cognitive distinctions should be used to ensure that people end up in districts where they feel the strongest sense of attachment to and belonging toward, and are the better represented for it. My research will therefore focus on the nature and scale of these more neglected cognitive COIs, and explore how closely those cognitive regions coincide with the administrative regions. If a COI can take the form of a cognitive region, one may wonder about the scale of such regions. To be clear, by scale, I mean absolute size, not cartographic scale. So when I refer to larger scales, I am describing larger areas, not smaller areas. Clearly COIs are almost never the exact size of an electoral district; they are not equipopulous, as districts are required to be . Therefore attempts to respect COIs when drawing district boundaries can never fully succeed; the hope is merely to do the best job possible. Granting the fact that such communities may differ in size from electoral districts, the question arises as to how they compare in size to districts at various scales. Each state in the U.S. is divided into congressional, state upper house, and state lower house districts . Furthermore, each of these three levels features a wide variation in size. At all levels, districts can range from large, sparse, and rural to small, dense, and urban. This is not to mention the many types of local districts that exist, such as those for county boards of supervisors, city councils, schools, and special purposes. One might wonder whether COIs are relevant at each of these levels, and whether a district may be too large or small for the COI criterion to meaningfully apply. It is informative to ask whether the criterion to preserve COIs is indeed being applied by authorities at multiple district levels. A quick survey of several jurisdictions in the United States reveals that many do in fact utilize this criterion at different scale levels. As mentioned above, as of 2011, 13 states considered COIs for both congressional and state legislative districts, and an additional 9 states considered it for just state legislative districts . California’s redistricting commission is tasked with drawing districts for Congress, both houses of the State Legislature, and the state’s Board of Equalization , and it uses the same criteria for all four bodies, including respect for COIs . This is especially notable considering how large and geographically diverse that state is, meaning this criterion is applied for a district spanning huge stretches of the Mojave Desert just as much as it is for one made up of a few densely-packed neighborhoods in innercity Los Angeles. Furthermore, even local jurisdictions apply the criterion for their relatively small districts, although they often use neighborhoods as a substitute for COIs. For example, San Francisco and San Diego have employed citizen commissions for drawing city council districts that tried to carefully consider which neighborhoods should land in which districts . Even the relatively small city of Santa Barbara sought to respect the small COIs within it when determining the boundaries of its city council districts . These findings demonstrate that the criterion is used for many different types of districts of various sizes. Given the application of the criterion at multiple district levels, scholars seem to diverge into two camps regarding whether that application is actually meaningful at all of those levels. In other words, they differ on how large or small a COI can truly be said to be. One camp has asserted that COIs do exist at multiple scales, but are not the same thing at one scale as they are at another. For example, Cain and Hopkins said that they “tend to be identified with neighborhoods in local government more than at the state or federal level. Statewide, COIs are likely to be defined as agricultural regions, coastal areas, and the like” . Morrill outlined three types distinguished by their scale: at the broadest scale, urban, suburban, and rural communities; at the regional scale, small metropolitan areas and media markets; and at the local level, city districts and large neighborhoods .