About the mixedmetro project
Welcome! The mixedmetro project is a cooperative venture to explore the racial and ethnic diversity of US metropolitan neighborhoods. It involves researchers in the Departments of Geography at three US academic institutions: the University of Georgia, the University of Washington, and Dartmouth College. See the Credits page for details about the research team, funding sources, and recommended citations for this research.
The parts of this website reporting on household-scale diversity stem from research and analysis undertaken while the authors were conducting research approved by the Center for Economic Studies at the U.S. Census Bureau. It has undergone a Census Bureau review more limited in scope than that given to official Census Bureau publications. Research results and conclusions expressed are those of the authors and do not necessarily indicate concurrence by the Census Bureau. It has been screened to insure that no confidential information is revealed.
Please browse through the maps and explore how they depict the patterns of diversity and segregation in these cities, metropolitan regions, and states. If you have questions, comments, or suggestions for improvements to this website, please contact us via the Comments page - we'd like to hear from you.
Using this site
We currently offer four sets of maps, as well as two sets of transition matrices (see below). All the maps classify census tracts based on racial composition and diversity.
You can begin by selecting the name of a city, metro area, or state from the drop-down menus. On the individual map pages, you can chose to view census tracts - a neighborhood-like spatial unit delimited by the Census Bureau - using data from the 1990, 2000, or 2010 Census. You can also zoom in and out and display background features using the sliding scale.
- One set of maps allows users to examine census tracts with high concentrations of immigrants for 16 selected cities.
- The second set of maps provides windows on the neighborhood racial composition of the largest 53 metropolitan areas in the US (those with populations over 1 million).
- The third set of maps shows the neighborhood racial composition of all 50 US states, as well as the District of Columbia.
- The fourth of maps allows users to simultaneously examine concentrations of households headed by three different classifications of mixed-race couples for 16 selected cities.
Neighborhood racial segregation and diversity jointly considered
Our intent is to create a classification system that will allow readers to easily comprehend (i.e., "see") a broad range of neighborhoods on the basis of racial composition with an explicit eye towards the notion of diversity. The classification system emerged from detailed explorations of the impact of many configurations of census tract racial composition on a common measure of diversity - scaled entropy.
We scaled the standard entropy diversity measure so that Ej ranges between a minimum of 0 and a maximum of 1. For our computations, we calculated Ej based on individuals in 5 racial groups created from the 2000 and 2010 Censuses (white, black, American Indian, Asian and Pacific Islander, "Other Race"), plus Latino/as. The Asian and Pacific Islander category is a combination of two categories on the 2000/2010 Censuses which conforms to the 1990 single aggregate category of Asian and Pacific Islander. We opted for this aggregation to ensure compatibility between this analysis based on 2000/2010 data and future analyses based on comparisons of 1990 and 2000/2010 data. After experimentation, we arrived at the following organization:
These categories of entropy capture the level of diversity in neighborhoods but they do not convey which racial group is dominant at low and moderate levels of diversity. Accordingly, we further distinguish "low diversity" (darker shading) and "moderate diversity" (lighter shading) tracts by the racial group with the largest share. In so doing, we present a new mapping of race that describes neighborhoods as "low diversity, white-dominant"; or "moderately diverse, Asian dominant"; and so on. We use orange for predominantly white neighborhoods, pink for Asian locales, green for black census tracts, yellow for American Indian communities, and blue for Latino residential quarters. Brown areas represent "highly diverse" neighborhoods. In these neighborhoods no group is dominant; no group has more than 45 percent of the population, meaning by definition that these places have at least 10 percent of their population made up by a third group (or combination of remaining groups). These are places constituted not only by roughly equal numbers of African Americans and whites; they include other racialized groups such as Latinos, Asians, or American Indians.
- Low Diversity: Ej is less than 0.37 (or when one group is over 80 percent of tract population)
- Moderate Diversity: Ej is between 0.37 and 0.74
- High Diversity: Ej is greater than 0.74 (and no group is over 45 percent of tract population)
Summary statistics for the MixedMetro tract classification scheme are presented and discussed here.
About the concentrations of immigrants, and of mixed-race households
Additional overlay options identify tracts which have concentrations of immigrants, or of different types of mixed-race households. These are superimposed on the patterns of neighborhood racial diversity. We deploy the familiar location quotient to measure and depict concentrations of immigrants or mixed-race households. For example:
LQj = (Pij / Pj) / (Pim/Pm)
LQj is the location quotient for black-white headed households in census tract j; Pij is the count of such households in tract j; Pj is the total population of tract j; Pim is the count of all black-white headed households in metropolitan region m; and Pm is the total population of metropolitan area m. A similar approach is used to calculate location quotients for high concentrations of immigrants, in any of the six largest immigrant populations for a given metro area.
This ratio equals one when the proportion of immigrants, or mixed-race households in a neighborhood is the same as the proportion in the metropolitan region. Census tracts with high concentrations of the variable in question are outlined using two different line widths. The spaces outlined with thick lines reference "super-concentrations" of immigrants or mixed-race households — neighborhoods identified with location quotients over 2.58 standard deviations above the metropolitan mean — while thinner lines show those with location quotients from 1.96 to 2.58 standard deviations above the mean. The data for these overlays come from confidential files from Census 2000 (for mixed race households) and for both Census 2000 and 2010 (immigration).
Transition matrices, which tally the census tracts that changed classification during the decade and also count the number that did not, allow users another perspective on neighborhood change taking place over the two decades from 1990 to 2010. These tabulations are available for each state and metropolitan area and can be accessed on the front page, or from the individual map pages.