1 Introduction
2 Data and method
2.1 Diversity and segregation metrics
3 Results
3.1 Geographical perspective
Region | City | 1990 | 2000 | 2010 | 2020 | 1990-2020 | |||
---|---|---|---|---|---|---|---|---|---|
Div/Seg | Div/Seg | Change | Div/Seg | Change | Div/Seg | Change | Change | ||
1 | 3 | 1.70 | 2.07 | 22% | 2.38 | 15% | 2.86 | 20% | 68% |
0.30 | 0.26 | −12% | 0.24 | −10% | 0.20 | −16% | −34% | ||
2 | 4 | 1.92 | 2.25 | 17% | 2.47 | 10% | 2.89 | 17% | 51% |
0.35 | 0.31 | −12% | 0.27 | −11% | 0.22 | −19% | −37% | ||
3 | 6 | 2.00 | 2.33 | 17% | 2.65 | 14% | 3.08 | 16% | 54% |
0.34 | 0.28 | −12% | 0.23 | −16% | 0.19 | −19% | −44% | ||
4 | 14 | 1.88 | 2.33 | 24% | 2.66 | 14% | 3.06 | 15% | 62% |
0.31 | 0.25 | −18% | 0.21 | −16% | 0.18 | −17% | −43% | ||
5 | 10 | 1.80 | 2.11 | 18% | 2.36 | 12% | 2.75 | 16% | 53% |
0.43 | 0.35 | −18% | 0.29 | −17% | 0.15 | −19% | −44% | ||
6 | 9 | 2.43 | 2.72 | 12% | 2.90 | 7% | 3.22 | 11% | 32% |
0.25 | 0.22 | −12% | 0.19 | −14% | 0.24 | −17% | −37% | ||
7 | 2 | 1.78 | 2.05 | 15% | 2.29 | 12% | 2.67 | 17% | 50% |
0.43 | 0.35 | −19% | 0.29 | −17% | 0.22 | −23% | −49% | ||
8 | 2 | 1.81 | 2.22 | 22% | 2.47 | 11% | 2.81 | 14% | 54% |
0.15 | 0.15 | −1% | 0.14 | −7% | 0.11 | −20% | −26% | ||
9 | 9 | 2.69 | 3.17 | 18% | 3.36 | 6% | 3.60 | 7% | 34% |
0.43 | 0.35 | −7% | 0.29 | −7% | 0.24 | −13% | −25% | ||
10 | 2 | 1.72 | 2.22 | 29% | 2.59 | 17% | 3.11 | 20% | 81% |
0.14 | 0.10 | −24% | 0.09 | −11% | 30.08 | −13% | −41% |
-
Throughout each decade, diversity increased and segregation decreased in all regions. In 1990, cities, on average, consisted of two sizable racial groups, whereas in 2020, the average city had three sizable racial groups.
-
In 1990, the levels of diversity and segregation exhibited strong regional disparities among US cities. Over the course of the next 30 years, these regional differences persisted, albeit to a lesser extent. Cities in regions 9 (California) and 6 (Texas) remained the most diverse and least segregated, on average. Cities in regions 5 and 7 (Midwest) remained the least diverse and most segregated, on average. Cities in region 10 (Portland, OR, Seattle, WA) exhibited relatively low segregation in 1990 and managed to maintain this low level while increasing their diversity over the subsequent 30 years.
-
The level of diversity seems to reach a threshold at the presence of the four major racial groups. This observation may be attributed to the classification system used by the U.S. Census, which only distinguishes four significantly populous racial groups.
3.2 Temporal perspective
-
Racial diversity in US cities has exhibited a consistent upward trend from 1990 to 2020. On average, there has been an increase of approximately one significant racial group (\(N_{\mathrm{H}} \sim 1\)) in the population of an average US city over the course of 30 years.
-
Racial segregation in US cities has shown a consistent downward trend from 1990 to 2020. On average, there has been a decrease of \(H=-0.12\) in segregation over the 30-year period. The magnitude of this decrease in the index H does not have a straightforward intuitive interpretation.
3.3 Groups evolution perspective
Group | Group description | Cites | 1990 | 2000 | 2010 | 2020 |
---|---|---|---|---|---|---|
Inh. | Inh. | Inh. | Inh. | |||
Sil. | Sil. | Sil. | Sil. | |||
1 (dark green) | low diversity and low segregation | Providence, Albany, Knoxville, Greenville, Minneapolis, Salt Lake City, Portland | 0.2 | 0.18 | 0.17 | 0.15 |
0.26 | 0.29 | 0.33 | 0.34 | |||
2 (red) | low diversity and high segregation | Buffalo, Pittsburgh, Philadelphia, Memphis, Louisville, Birmingham, Milwaukee, Indianapolis, Detroit, Dayton, Cleveland, Cincinnati, St. Louis | 0.19 | 0.21 | 0.21 | 0.19 |
0.39 | 0.28 | 0.17 | 0.10 | |||
3 (light green) | medium diversity and low segregation | Virginia Beach, Raleigh, Orlando,Tulsa, Oklahoma City, El Paso, Denver, Phoenix, Las Vegas, Seattle | 0.11 | 0.15 | 0.18 | 0.19 |
0.51 | 0.16 | −0.16 | −0.30 | |||
4 (orange) | high diversity and low segregation | San Antonio, Austin, Albuquerque,Tucson, San Francisco, San Diego, Sacramento, Riverside, Fresno | 0.16 | 0.15 | 0.14 | 0.13 |
0.31 | 0.22 | 0.12 | 0.08 | |||
5 (dark red) | high diversity and moderate to high segregation | New York, Washington, Chicago, New Orleans, Houston, Dallas, Los Angeles | 0.24 | 0.21 | 0.17 | 0.14 |
0.14 | 0.22 | 0.29 | 0.37 | |||
6 (blue) | low diversity and medium segregation | Hartford, Boston, Rochester, Richmond, Harrisburg,Tampa, Nashville, Miami, Jacksonville, Greensboro, Charlotte, Atlanta, Grand Rapids, Columbus, Kansas City | 0.15 | 0.19 | 0.18 | 0.16 |
0.35 | 0.10 | −0.05 | −0.14 |
-
Cities that were grouped together in 1990 based on their similarity in terms of diversity and segregation metrics continue to exhibit similarity in 2020. This indicates that the trend towards increasing diversity and decreasing segregation has impacted all cities within the 1990 groups in a similar manner. This finding is intriguing because the groups consist of cities from different regions of the US, with their only commonality in 1990 being the values of \(N_{\mathrm{H}}\) and H. However, over the course of 30 years, the evolution of racial geography has influenced their \(N_{\mathrm{H}}\) and H values in a similar fashion. One exception is group #3, where El Paso, TX, located on the US-Mexico border, has maintained its relatively low diversity due to its predominantly Hispanic population.
-
The majority of groups identified in 1990 have lost their distinctiveness by 2020. The evolution of racial geography, characterized by increased diversity and decreased segregation, has compressed the \((N_{\mathrm{H}}, H)\) space into a smaller domain compared to 1990. As a result, the groups defined in 1990 now overlap on the \((N_{\mathrm{H}}, H)\) diagram. However, groups #1 and #5 are exceptions to this trend, as they not only maintained their distinctiveness but actually increased it in relation to the other groups. However, with different groupings (\(k=5\) or \(k=7\)) such exceptions are absent.
3.4 Spatial scale limit of desegregation
Year | 〈ΔH〉 | SD(ΔH) | Max(ΔH) | Min(ΔH) |
---|---|---|---|---|
1990 | 0.146 | 0.050 | 0.35 (Greenville 0.57-0.22) | 0.08 (Los Angeles 0.38-0.30) |
2000 | 0.130 | 0.037 | 0.27 (Greenville 0.44-0.17) | 0.08 (San Diego 0.27-0.19) |
2010 | 0.120 | 0.030 | 0.21 (Greenville 0.35-0.14) | 0.07 (Los Angeles 0.35-0.28) |
2020 | 0.110 | 0.025 | 0.18 (Birmingham 0.35-0.14) | 0.06 (San Diego 0.22-0.16) |
-
During the period from 1990 to 2020, there has been a slight decrease in the gap between segregation values calculated from blocks and tracts for cities. This suggests that blocks have experienced an increase in diversity relative to tracts, thereby weakening the lower limit of desegregation scale.
-
The standard deviation of segregation gaps among the surveyed cities has significantly decreased throughout the 1990-2020 period. This indicates that the differences in segregation gaps have become more uniform across cities, suggesting a convergence in the patterns of racial segregation.
-
In 1990, cities in region 4 (Southeast) exhibited the highest values of ΔH, indicating greater disparities between block-based and tract-based segregation measures. Conversely, cities in regions 9 (California) and 5 (Midwest) had the smallest values of ΔH, indicating lower differences between the two measures. By 2020, the regional disparities in ΔH had diminished to some extent. However, cities in region 4 still displayed relatively high values of ΔH, while cities in region 9 (excluding region 5) continued to show relatively low values of ΔH.
4 Conclusions and discussion
This study | [14] | |||||
---|---|---|---|---|---|---|
Year | E | H | Year | E | H | |
Northeast | 2000 | 0.58 | 0.28 | 2000 | 0.64 | 0.47 |
2010 | 0.66 (+14%) | 0.25 (−11%) | 2010 | 0.72 (+13%) | 0.42 (−11%) | |
Midwest | 2000 | 0.55 | 0.28 | 2000 | 0.55 | 0.52 |
2010 | 0.63 (+15%) | 0.24 (−14%) | 2010 | 0.61 (+11%) | 0.45 (−13%) | |
South | 2000 | 0.68 | 0.235 | 2000 | 0.69 | 0.43 |
2010 | 0.74 (+9%) | 0.20 (−15%) | 2010 | 0.76 (+10%) | 0.37 (−14%) | |
West | 2000 | 0.75 | 0.225 | 2000 | 0.75 | 0.34 |
2010 | 0.79 (+5%) | 0.19 (−15%) | 2010 | 0.79 (+5%) | 0.30 (−12%) |