By Matthew P. Lee GROWTH & RESIDENTIAL INTEGRATION TRENDS IN BULL CITY
Durham is the fourth-largest city in North Carolina and the 83rd-largest in the United States by population, with 239,358 residents as of the 2012 United States Census. Known locally as “Bull City,” Durham currently serves as the home of an expanding community bolstered by the educational, research, healthcare, and business involvements of Duke University and Research Triangle Park. Since 2000, Durham’s individual city population grew an impressive 26.7 percent, or by more than double the national average. During the same time period, the population of the greater Raleigh-Durham metropolitan statistical area increased by a monumental 47.8 percent – the highest rate of all 52 metro areas in the U.S. with over 1 million residents. This rate exceeds more than three times the 12.7 percent average growth of those 52 metro regions. (Kotkin, 2013)
Several other cities, including Austin, Texas and Las Vegas, have also witnessed comparable population explosions surpassing 40 percent. Although these surges have been attributed chiefly to domestic migration and increases in foreign-born immigrant populaces, an interesting similarity identified among the Austin, Las Vegas, and the Raleigh-Durham metropolitan areas is a characteristically low population density relative to those of other metros. None of the ten fastest growing cities from 2010 to present have had urban core densities even half of those of cities like Boston, New York, or Los Angeles. Fast-growing regions such as Raleigh-Durham lack super high-density cores and instead possess populations that expand in a more spatially dispersed fashion than those of large concentrated metros (Kotkin, 2013). This pattern is observable in Figure 1, which compares Durham’s census tract population densities per square mile between the years 2000 and 2013.
The left map of Figure 1, which depicts population densities for the year 2000, illustrates a clearly identifiable main cluster in the geographic center of the city. The majority of the census tracts in this cluster are of analogous density and are located adjacent to one another. Additionally, these tracts are more heavily populated than those in the surrounding areas. The right map of Figure 1, which portrays population densities for the year 2013, shows two additional clusters that have emerged in the southern and western quadrants of Durham. These new clusters represent growth hotspots geographically displaced from the central urban core. Their appearance corroborates the theory that populations of swiftly evolving cities like Durham expand in a spatially dispersed manner.
Coloration ranges were determined using national quantiles. (Graphics generated using Simplymap.com; data obtained from the U.S. Census Bureau)
In conjunction with expansion patterns, housing affordability is an important factor to consider in analyzing quickly developing cities like Durham. Unlike New York and Los Angeles, many of the fastest-growing places in the U.S. boast significantly lower housing prices relative to their average incomes. In metropolitan regions such as Raleigh-Durham, housing cost as a percentage of income can be less than half of those of other more tightly regulated municipalities (Kotkin, 2013). In addition to its correlation with population gains, the affordability of housing generates a number of questions regarding residential diversity in the formation of cities. Specifically for Durham, how have low housing prices and the city’s robust population growth following the turn of the century affected its degree of racial segregation? Based on the inferences of Paul Courant’s urban housing search model and the trends visible in U.S. Census data, Durham’s level of racial integration has increased.
In his paper, titled “Racial Prejudice in a Search Model of the Urban Housing Market,” Paul Courant proposes a basic framework to explain the existence of housing segregation in cities. Courant’s simplified model makes several assumptions. First, the model assumes that a housing market is composed of only two sets of people with perfect information – whites and blacks. Second, it presupposes that some but not all whites have an inherent aversion to living with blacks. Third, it assumes that searching for housing is costly and that blacks’ searching costs are systematically affected by whites’ aversion to living in non-segregated communities. Following these assumptions, Courant’s model employs a utility maximization function to represent a black individual’s preferences at any stage of his housing search throughout n neighborhoods:
Here, V represents the black searcher’s overall utility and V0 represents his utility from the best house visited up until the present. The parameter c denotes a constant search cost for both whites and blacks, while αj (assumed to vary across neighborhoods) denotes the probability that a black searcher will encounter an averse seller in a given neighborhood j. The function F(V0) is defined as the probability that a house’s utility will be below V0. Given these designations, each line of the maximization function below V0 is equal to the sum of the following:
Where V*j represents the utility from the best house examined, a black searcher will browse for houses until the equilibrium condition is met. Assuming that preferences for housing types are the same across all neighborhoods, blacks will only search where the probability of encountering an averse seller is lowest, which would entail only searching in all black neighborhoods (α=0). Even if preferences for housing types vary, blacks will still prefer to search in neighborhoods with lower values of α because white seller aversion imposes greater search costs. This inclination to reduce discrimination-related search costs diminishes demand for housing in bigoted neighborhoods and increases demand for housing in predominantly black neighborhoods. Such shifts in demand create a price differential in which black neighborhood houses are purchased by black searchers at higher prices than those of houses located in white and integrated neighborhoods. (Courant, 1978)
In summary, Courant’s model posits that anti-black prejudice perpetuates segregation by creating higher search costs for blacks and disturbing competitive equilibria in urban housing markets. Such disturbances lead to inferior buying terms for black searchers in black neighborhoods, where price has been inflated due to boosted market demand; they also lead to inferior selling terms for suppliers of housing in bigoted neighborhoods, where price has been deflated due to suppressed market demand caused by reluctance to incur search costs. Courant’s model allows for a reversal of segregation once blacks begin purchasing homes in white neighborhoods. Immigration of blacks into a given white neighborhood induces the phenomenon of “tipping” by reducing the probability that other black searchers will encounter racist sellers there. As whites who are averse to living near blacks leave the neighborhood, black search costs further decline, total demand for housing increases, prices increase and provide bigoted whites incentive to sell, and other black searchers buy houses. Through the tipping cycle, Courant’s model accounts for the possibility of integration over time. (Courant, 1978)
Although Durham has had a long history of segregation as a city in the American South, recent demographics data provided by the United State Census Bureau indicate that the area’s level of racial integration has increased since the year 2000. These data are illustrated in Figure 2, which depicts the geospatial black-white racial composition of Durham in the years 2000 and 2010. In each outlined census tract, darker shades of color express higher percentages of households occupied by the target demographic and lighter shades express lower percentages of households occupied by the target demographic. Coloration categories are divided into five equal intervals, each spanning a 20 percent range. The blue color scheme corresponds to percentages of tract households occupied by blacks and the yellow color scheme corresponds to percentages of tract households occupied by whites. For example, a region shaded with the darkest blue communicates that between 80 and 100 percent of households in that specific area are inhabited by blacks. Furthermore, a census tract shaded with the lightest yellow communicates that between zero and 20 percent of households in that specific area are inhabited by whites.
In Figure 2, the top left map shows the percentage of households owned by blacks per tract for the year 2000, and the top right map shows the same data but for the year 2010.
Coloration categories are divided into five equal intervals, spanning 0% to 20%, 20% to 40%, 40% to 60%, etc.(Graphics generated using Simplymap.com; data obtained from the U.S. Census Bureau)
The top left map is marked by two solid tract clusters of the highest black percentage range and three clusters of the second highest range. Each cluster clearly contrasts with its surrounding tracts, most of which are shaded with the color for the second lowest range. In comparison, the top right map only contains one cluster of the highest black percentage range. Clusters of the second highest range still exist but they are smaller, more heterogeneous, and less contrastive with adjacent tracts. Also, a sizable swath of eastern areas have changed to the middle black percentage range. Moving downward, the bottom left map shows the percentage of households owned by whites per tract for 2000, and the bottom right map shows analogue data for 2010. The bottom left map generally reflects the inverse color gradient of the top left map, confirming the prevalence of black-white housing segregation in 2000. The bottom right map illustrates the persistence of high white percentage tracts to the west but also shows the transition of eastern tracts to the middle white percentage range, which is mirrored by a similar black transition in the top right map. In graphical representations for both racial demographics, the visibility of extreme colors decreases from 2000 to 2010, signifying that the racial integration of residential housing in Durham has improved.
Evidence of greater integration achieved through Courant’s tipping cycle is revealed by additional census data detailing changes in the number of black and white housing units per housing value segments. Although the total numbers of houses owned by blacks and whites both increased by approximately 30 percent from 2000 to 2010, rates of change segmented by housing value vary by an enormous margin. Figure 3 displays percentage changes in the number of houses owned by blacks and whites grouped by housing value brackets. Growth in the number of houses owned per value segment is highlighted in green and decay is highlighted in red. Notably, the table reports that for all housing segments valued between $70,000 and $1,000,000, percent changes in units owned by blacks are unilaterally higher than percent changes in units owned by whites. In the $400,000 bracket, the rate of growth for blacks is even more than ten times that for whites.
Relatively larger percentage increases in the quantities of high-valued homes bought by blacks in Durham reflect increasing median income for blacks (see Figure 4). Since expensive homes are likely to be built in neighborhoods with houses of comparable value and whites currently occupy a majority of Durham’s expensive homes, a portion of affluent blacks who buy nice homes probably buy them in predominantly white neighborhoods. The immigration of a few affluent blacks into white neighborhoods, as suggested by superior percent change in the number of high-value homes owned by blacks, fits well within Courant’s tipping framework and explains how Durham has become a more integrated city since the year 2000. As the median income for blacks rose and more blacks began buying pricier homes in white neighborhoods, racist whites fled, black search costs declined, other blacks moved in, neighborhoods tipped, and Durham’s degree of integration increased over ten years.
Durham’s growth over the past decades has been fierce and relentless, like a charging bull. With the city’s booming population and convenient housing costs, many people seeking a home already have much to love about Bull City. To make matters even better, residential segregation in Durham appears to be declining; using Paul Courant’s search model of urban housing markets to interpret observable trends in U.S. census data, one sees that the city’s residential communities have become increasingly integrated since the year 2000. Despite this positive trend, there are still many unanswered questions about how to best approach income and racial segregation in public environments, such as schools. Future research on such topics would be greatly beneficial to the lasting prosperity of Durham.
Courant, P. N. (1978). Racial prejudice in a search model of the urban housing market. Journal of Urban Economics, 5, 329-345.
Kotkin, J. (2013, March 18). America’s fastest-and slowest-growing cities. Forbes, Retrieved from http://www.forbes.com/sites/joelkotkin/2013/03/18/americas-fastest-and-slowest-growing-cities/
Simplymap.com; Census Data 2000 Geographies; Census Data 2010 Geographies. Retrieved from
U.S. Census Bureau; ACS_10_SF4_B25075; ACS_10_SF4_B25077; ACS_10_SF4_B992519; DEC_00_SF4_HCT066. Retrieved from