Race Ethnicity and Residential Intermixing in US MSAs

Race, Ethnicity and Residential Intermixing in US MSAs Lawrence A. Brown and Madhuri Sharma Department of Geography The Ohio State University International Conference on Population Geographies IV Hong Kong

Population Growth and Increasing Diversity Total US Population 1990: 248. 7 million 2000: 281. 4 million 2006: 300. 0 million (Oct 17 th) Pct of Total US Pop Foreign Born 1950: 06. 5% 1990: 09. 5%* 2000: 13. 0% Pct Pop Foreign Born, Selected MSAs (from 49 largest) Miami: 40. 2% Los Angeles: 31. 0% New York: 24. 4% Houston: 19. 2% Chicago: 16. 0% Cincinnati 02. 6%

Over past half-century – Non-Hispanic Caucasians Fell – 83% of the Population in 1970 to 67% in 2003 Foreign-Born Rose – 6. 2% of the Population in 1980 to 12. 5% in 2005 Today – More than 11 million undocumented Foreign Born Racial/Ethnic Intermixing has Increased Notably – in Places of Consumption, Work, Residence – and Trend is Ongoing

Research Questions How Does Intermixing Vary Among MSAs? ~94% percent of foreign-born in 2000? Usually raised in context of two-group comparisons; e. g. , Caucasians vs African Americans, Asians, Hispanics Measures such as the a. Dissimilarity Index Here – Intermixing of All Groups Together, Theil Entropy Index Link Degree of Intermixing to MSA Characteristics, a particularly geographic, or spatial, approach. MSAs > 1 million pop in 2000, neither entire range nor few largest gateway cities. Strong Evidence that R/E Minorities increasingly Locate where Caucasian historically dominant, but some claim simply results in new R/E Enclaves, not heterogeneity Theil Entropy Index can address this directly

Measures: Integration/Segregation, Intermixing Single Group or Two Group Measures: Dissimilarity / Exposure Indices MSA Level - most commonly used Spatially-Modified Dissimilarity Indices MSA Level - not used Local Moran’s I, Location Quotient, Getis G Tract/Block Group Level - less used Multiple Groups Simultaneously: Theil’s Entropy Index Can be decomposed into its component parts – e. g. , MSA, Census Tract, Block Group; or races/ethnicities such as AA-A-C-H

Two Components: Diversity Score + Entropy Index D = Diversity Score for MSA/City D(i) = Diversity Score for Census Tract/Block Group i Pr(g) = proportion of R/E group g Summed over all R/E groups in MSA/Census Tract i Yields Diversity Score for the MSA/Census Tract i Diversity Scores measure R/E Diversity for the spatial unit D values differ depending on number of groups (g) proportion (Pr) of each group.

Bounds of the Diversity Score Upper Bound represents Equal Proportions of Each Group; Depends on the Number of Groups 6 R/E groups, each with Pr = 0. 167; D = 1. 799 5 R/E groups, each with Pr = 0. 200; D = 1. 600 4 R/E groups, each with Pr = 0. 250; D = 1. 386 Lower Bound D = 0 (only single group)

Entropy Index measures Intermixing at MSA level -- How evenly groups are distributed across MSA neighborhoods relative to their proportions in the MSA overall E = Entropy Index for MSA/City t(i) = Population of Census Tract/Block Group i D, D(i) = Diversity Scores for MSA, Tract/Block Group i T = Population of MSA/City E Computation -- The deviation of each unit’s (i) diversity score from MSA-wide diversity is weighted by the population of i (relative to the population of the MSA) and summed for all i E Boundaries -- From 0 = high intermixing; 1 = high clustering

Present Study 49 MSAs > 1 Million in 2000; Tract Level Data From Louisville (1. 0 million) to New York (21. 2 million); Distinct break at Philadelphia (6. 2 million ); only San Francisco (7. 0 million), Washington (7. 6 million), Chicago (9. 2 million), Los Angeles (16. 4 million), New York are larger. Focus, then, on Mid-Sized MSAs Five Racial/Ethnic Groups -- African Americans, Asians (HW-Pac Islanders), Caucasians, Hispanics, American Indians; Data Source: US Census

Diversity Entropy (Intermixing) Most San Francisco, Los Angeles, Houston, D = 1. 21 D = 1. 20 D = 1. 18 Most Seattle, Salt Lake, Portland, E = 0. 12 E = 0. 13 Least Pittsburgh, D = 0. 41 Least Detroit, E = 0. 50 D = 0. 82 D = 0. 84 D = 0. 86 Medium Columbus, Boston, Houston, Los Angeles, Medium Jacksonville, Denver, Philadelphia. E = 0. 29 E = 0. 28 Correlation Between Diversity and Entropy = -0. 17!! Not What Might be Expected!!

Important Dimension of Entropy Index Can be Decomposed to Gauge Contribution from Components E. g. , Decompose Intermixing at MSA Level into that Portion Contributed by Block Group Compared to Census Tract Level Here -- Gauge Role of Intermixing of Caucasians vs Minorities as a Group, Compared to Role of Intermixing Among Minority Groups Alone First -- Compute D and E as in Equations Shown Earlier, symbolized as DC/M and EC/M for Caucasians vs Minorities-as-agroup & DAA/AI/A/H and EAA/AI/A/H for minorities among themselves. These combine to form E at MSA (EMSA) level as follows --

EMSA = (DC/M / DMSA) * EMSA + Pr. M * (DAA/AI/A/H / DMSA) * EAA/AI/A/H where – EMSA and DMSA are E and D in equations shown earlier Pr. M is proportion of MSA population minorities (or non. Caucasians). Hence -CM term indicates the portion of MSA intermixing attributable to Caucasians living separately from, or intermixing with, minorities-as-a-group AA/AI/A/H term indicates the portion of MSA intermixing attributable to minorities living separately from, or intermixing with, one another.

MSA Profiling Statistics Diversity Score 2000 (D) Entropy Index 2000 (E) Median 0. 82 0. 28 Minimum 0. 41 0. 12 Maximum 1. 21 0. 50 Percentiles 35 0. 73 0. 24 50 0. 82 0. 28 65 0. 91 0. 32 Hi Diversity if D = 0. 91 or Greater Med Diversity if D < 0. 91, > 0. 73 Low Diversity if D = 0. 73 or Less Nine Groups --HI-I/LO-D; MED-I/LO-D; LO-I/LO-D; Hi Intermixing if E = 0. 24 or Less Med Intermixing if E < 0. 32, > 0. 24 Low Intermixing if E = 0. 32 or Greater HI-I/MED-D; MED-I/MED-D; LO-I/MED-D; HI-I/HI-D MED-I/HI-D LO-I/HI-D

Categories of 49 MSAs based on Level of Intermixing and Diversity MSA Categ. Char. # MSAs HI/LD 3 Minneapolis--St. Paul, Portland-Salem, Salt lake Sun Belt HI/MD Hi Hispanic Hi Minority HI/HD 6 Charlotte, Denver, Norfolk, Oklahoma City, Raleigh. Durham, Seattle-Tacoma 8 Austin, Las Vegas, Orlando, Phoenix, Sacramento, San Antonio, San Diego, San Francisco, MI/LD 6 Boston, Columbus, Grand Rapids, Nashville, Providence, Rochester MI/MD 5 Greensboro, Hartford, Jacksonville, Tampa, West Palm MI/HD 3 Dallas, Houston, Los Angeles. LI/LD Largely AMB/ 8 Buffalo, Cincinnati, Cleveland, Indianapolis, Kansas City, Louisville, Pittsburgh, St. Louis LI/MD Rust Belt 4 Detroit, Memphis, Milwaukee, Philadelphia. LI/HD Gateways 6 Atlanta, Chicago, Miami, New Orleans, New York, Washington DC

Tend to be LI/LD, LI/MD

Scattered Between HI/HD, MI/MD, LI/HD

Tend to be HI/LD, HI/MD, HI/HD, MI/HD

Scattered Between HI/LD, HI/MD, MI/LD, LI/HD

Dimensions That Differentiate These Maps Era of Immigration: Pre World War II, Early 20 th Century (I); 1960 s and Later (II, III); 1980 s and Later (IV) Destination Areas of Immigrants: American Manufacturing Belt (I), Florida (II), West of the Mississippi (Great Plains + West) (III), East of the Mississippi (Mid-West + Southeast (IV) Origin Areas of Immigrants: Europe, especially South and East (I), Caribbean and South America (II), Central America, largely Mexico (III), South Central Asia, East Asia, Africa (IV) Place Characteristics Drawing Immigrants: Industrial Centers (I), Accessibility + Contiguity (II), Accessibility + Contiguity initially w/ Recruitment for Agric + Industry and “Wetbacks” (III), Refugee Resettlement Programs, Foreign Direct Investment, Universities (IV) Foreign Born Profiles of Largest MSAs (> 5 million, except Miami): Chicago, Dallas, Los Angeles, San Francisco – strong Central American, largely Mexican; New York, Miami – strong Caribbean, South American; Boston, Detroit, Philadelphia – Strong European; Washington – strong South Central Asian, East Asian, African

Can extrapolate from relationship between Brown et al MSA groups, based on immigration data, and MSA groups, based on Diversity Scores and Entropy Indices on R/E intermixing. Suggests variables/aspect/factors related to residential intermixing in urban areas. From Group I, seems that – ► ► Intermix muted in MSAs strong in Fordist mfg era Largely European foreign-born Foreign-born in US for longer time span Foreign-born relatively small share of MSA pop Group II relationships -- scattered + uninformative

Group III relationships – ► Hi intermix MSAs large share US Mex/Cent. AM pop ► Sun Belt MSAs w/ post-Ford econ activity, post-indl ► Immigration recent, from 1960 s onwards ► Urban growth since mid-20 th century substantial E. g. , Of 21 Group III MSAs, 1980 -00, 17 grew 25% or more, 11 exceeded MSA avg growth of 43%; one (Las Vegas) 237%

Group IV relationships – ► Low intermix in Cincin, Indianapolis, Washington Like Group I MSAs, in AMB, Fordist, early waves immig, largely European ► Hi/Med intermix in Charltte (61% growth), MSP (38%), Nash (45%), Raleigh (87%), CMH (27%); ~ Group III, urb growth substantial, Sun Belt MSAs, post. Ford/indust, immig 1960 s onwards ► CMH in mid AM/Rust Belt, but grew in later era than typical AMB MSA, distinctly diff econ base, “sun belt city in the rust belt” ► Hi/Med intermixing MSAs a focal point for refugee resettlement efforts + new economic activity -- technology (CMH, MSP, Raleigh), foreign auto mfg (CMH, Nash), organizational innov in finance and/or banking (Charlotte, CMH).

Correlation Values for Entropy Indices and Diversity Scores for 49 MSAs in 2000 Entropy Indices for MSA (All Groups) E-Indices (All) E-Indices (Cauc. vs. (Among Minor. ) 1 . 939(**) . 569(**) Entropy Indices (Cauc. vs. Minor. ) . 939(**) 1 0. 3 Entropy Indices (Among Minor. ) . 569(**) 0. 3 1 Diversity Scores for MSA (All Groups) -0. 172 -. 398(**) . 469(**) -. 473(**) -. 580(**) 0. 1 Diversity Scores (Among Minor. ) Diversity Overall contributes little to Intermixing Overall Diversity Among Minorities-as-a-group significantly to Intermixing Overall Diversity Among Minorities highly related to Intermixing of C N vs Minorities N In Overall Intermixing, that of C vs Minorities More Important than Among Minorities themselves Suggests that Hi Intermixing is Minorities in C Areas

Correlation Values for Entropy Index (Overall) for MSAs Socio-Economic Characteristics r-Values Per Capita Income (Total Population) -0. 008 Per Capita Income (Total Population) Logged 0. 001 Per Capita Income (African American) Logged -0. 347(**) High School or Less Education (2000, Percent) 0. 511(**) Some College or Associate Degree (2000, Percent) -0. 567(**) Bachelors Degree (2000, Percent) -0. 307(*) Graduate or Professional Degree (2000, Percent) Some College, Associate, or Bachelors Degree (2000, Percent Owner Occupied HHs in MSA (2000, Percent) 0. 023 -0. 626(**) 0. 202 African American Owner Occupied HHs ( 2000, Percent) 0. 398(**) Asian Owner Occupied HHs (2000, Percent) -0. 323(*) Hispanic Owner Occupied HHs (2000, Percent) -0. 238 African American Rented HHs (2000, Percent) 0. 550(**) Asian Rented HHs (2000, Percent) -0. 298(*) Hispanic Rented HHs (2000, Percent) -0. 291(*) Change in African American Owner Occupied HHs (90 -00, Percent) 0. 252 Change in Asian Owner Occupied HHs (90 -00, Percent) -0. 014 Change in Hispanic Owner Occupied HHs (90 -00, Percent) -0. 084 Change in African American Rented HHs (90 -00, Percent) 0. 310(*) Change in Asian Rented HHs (90 -00, Percent) 0. 212 Change in Hispanic Rented HHs (90 -00, Percent) -0. 098

NB: Significant Correlations Among Socio-Economic Characteristics Income: ► Per Capita Income of AA relates to Greater Intermixing (-0. 35) Education: ► High School or Less relates to Less Intermixing (+0. 51) ► Some College (or Assoc) or BS/BA relates to Greater Intermixing (-0. 57, -0. 31, -0. 63 when combined) Home Ownership (O) and Renting (R): ► African American O+ R relates to Less Intermixing (+0. 40, 0. 55); Inertia or Enclave Effect? ? ► Asian O+R relates to Greater Intermixing (-0. 32, -0. 30) ► Hispanic R relates to Greater Intermixing (-0. 29); O same direction, not signif Home Ownership (O) and Renting (R) Change 1990 -2000 ► African American Change, O or R, relates to Less Intermixing (+0. 25 (but not signif), +0. 31 (signif)); Inertia or Enclave effect? ?

Correlation Values for Entropy Index (Overall) for MSAs Demographic Characteristics Diversity Score for MSAs (Overall) Diversity Score (Minorities-Among-Themselves) r-Values -0. 172 -0. 473(**) Diversity Score (Caucasians vs. Minorities) -0. 081 MSA Population (2000) 0. 231 MSA Population LN (2000) 0. 239 Change in Population (1980 -2000, Percent) -0. 588(**) Change in Population (1990 -2000, Percent) -0. 392(**) African Americans Portion of MSA Population (2000) 0. 463(**) Asians Portion of MSA Population (2000) -0. 362(*) Hispanics Portion of MSA Population (2000) -0. 269 Change in Caucasian Population (1990 -2000, Percent) -0. 549(**) Change in African American Population (1990 -2000, Percent) -0. 572(**) Change in Asian Population (1990 -2000, Percent) Change in Hispanic Population (1990 -2000, Percent) -0. 211 -0. 309(*) Foreign-Born Portion of MSA Population (2000, Percent) -0. 187 Change in Foreign-Born Population, 1990 -2000 (Percent) -0. 420(**) Population who Lived in Different House in 1995 (Percent) -0. 684(**)

NB: Significant Correlations Among Demographic Characteristics MSA Diversity: ► Diversity for Minorities-among-themselves relates to Greater Intermixing (-0. 47); Diversity Overall + Caucasians vs Minorities, likewise, but not significant Population Characteristics: ► Higher proportion of Asians relates to Greater Intermixing (-0, 36); likewise Hispanics but not significant; Higher proportion of AA relates to Less Intermixing (Inertia or Enclave effect? ? ) Migration Dynamics: ► Growth Overall 80 -00 + 90 -00 relates to Greater Intermixing (-0. 59, -0. 39) ► Change in the proportion of African Americans, Hispanics, and Foreign-Born relates to Greater Intermixing (-0. 57, -0. 31, -0. 42); likewise Asians but not significant ► What does it mean that a high proportion of AA relates to Less Intermixing, while a positive change in that proportion relates to More Intermixing? ? ► Proportion of the population who Lived in a Different House in 1995 relates to Greater Intermixing (-0. 68)

Correlation Values for Entropy Index (Overall) for MSAs Built-Environment Characteristics r-Values Foreign-Born Entered Before 1965 as Percent Total F-B 0. 475(**) Foreign-Born Entered 1980 -2000 as Percent Total F-B -0. 412(**) Median Value of Owner Occupied HH (2000) -0. 227 Median Year Housing Structure Built -0. 644(**) Houses Built Before 1970 as Percent of 2000 Stock 0. 651(**) Houses Built Between 1970 -1990 as Percent of 2000 Stock -0. 641(**) Houses Built Between 1990 -2000 as Percent of 2000 Stock -0. 547(**) Manufacturing Employment 1980 (Percent Total Employment) 0. 347(*) Managerial/Professional Employment 1980 (Percent Employment) -0. 310(*) Manufacturing Employment 1990 (Percent Total Employment) 0. 382(**) Managerial/Professional Employment 1990 (Percent Employment) Manufacturing Employment 2000 (Percent Total Employment) Managerial/Professional Employment 2000 (Percent Employment) -0. 050 0. 346(*) -0. 095

NB: Significant Correlations Among Built-Environment Characteristics ERA of Growth; Fordist (F) vs Post-Fordist (P-F) ► Foreign-Born Entering Before 1965 (F) relates to Less Intermixing; Foreign-Born Entering 1980 -2000 (P-F) relates to Greater Intermixing (+0. 48, -0. 41) ► Higher Employment in Manufacturing (F) 1980, 1990, 2000 relates to Less Intermixing (+0. 35, +0. 38, +0. 35) ► Higher Employment in Managerial/Professional (P-F) 1980 relates to Greater Intermixing (-0. 31); also 1990 and 2000 but not significant Housing Characteristics: ►Median Year Housing Built Relates Inversely to Intermixing (-0. 64) (i. e. , higher the year, higher the intermixing) ►More Housing Built 1970 -1990 and 1990 -2000 relates to Greater Intermixing (-0. 64, -0. 55) ► More Housing Built Before 1970 relates to Less Intermixing (+0. 65)

Backward Step-Wise Regression 14 Variables initially; Chosen from original 51 on basis of -(i) Strength of zero-order r; (ii) Intercorrelations among candidate variables (threshold of 0. 60) (iii) Relevance in terms of Explanatory Frameworks and Research Criteria: F value of 0. 05 for Entry; 0. 06 for Removal Correlation among Independent Variables Diversity Score (Minorities-Among-Themselves) and African American Portion of MSA Population (2000) Change in Population (1980 -2000, Percent) and Population who Lived in Different House in 1995 (Percent) Houses Built Between 1990 -2000 as Percent of 2000 Stock Change in Foreign-Born Population, 1990 -2000 (Percent) and Foreign-Born Entered 1980 -2000 as Percent Total F-B Houses Built Between 1990 -2000 as Percent of 2000 Stock Population who Lived in Different House in 1995 (Percent) and Houses Built Between 1990 -2000 as Percent of 2000 Stock Foreign-Born Entered 1980 -2000 as Percent Total F-B and Houses Built Between 1990 -2000 as Percent of 2000 Stock = -0. 61 = +0. 68 = +0. 61 = +0. 80 = +0. 84 = +0. 60 = +0. 62

In Short – Great Deal of Interrelationship between Diversity Score (Minorities-Among-Themselves) African American Portion of MSA Population (2000) Change in Population (1980 -2000, Percent) Population who Lived in Different House in 1995 (Percent) Houses Built Between 1990 -2000 as Percent of 2000 Stock Change in Foreign-Born Population, 1990 -2000 (Percent) Foreign-Born Entered 1980 -2000 as Percent Total F-B Hence, Population Change Overall and Dynamics related to the Foreign Born appear to be a significant Driver of Racial/Ethnic Intermixing

Regression Analyses with Intermixing (Entropy Index) Multiple Regression Analysis R-squared = 0. 77 Zero-Order Correlations Beta Values t-test Sign. 0. 506 Constant b-Values 3. 664 0. 001 DEMOGRAPHIC CHARACTERISTICS Diversity Score (Minorities-Among-Themselves) -0. 473(**) African American Portion of MSA Population (2000) 0. 463(**) 0. 307 0. 289 2. 976 0. 005 Asian Portion of MSA Population (2000) -0. 362(**) -0. 442 -0. 166 -1. 673 0. 103 Hispanic Portion of MSA Population (2000) -0. 269(**) Change, Population (1980 -2000, Percent) -0. 588(**) Change, African American Population (1990 -00, Percent) -0. 572(**) -0. 077 -0. 178 -1. 501 0. 142 Change in Foreign-Born Population (1990 -00, Percent) -0. 42(**) -0. 064 -0. 605 -2. 602 0. 013 Population, Lived in Different House, 1995 (Percent) -0. 684(**) -0. 544 -0. 369 -1. 711 0. 095 -0. 412(**) 0. 246 0. 255 1. 537 0. 133 BUILT ENVIRONMENT CHARACTERISTICS Foreign-Born Entered 1980 -00 as Percent Total F-Born Managerial Employment 1990 (Percent Total Employment) -0. 05 Manufacturing Employment 1990 (Percent Total Employment) 0. 382(**) 0. 486 0. 283 2. 593 0. 014 Houses Built Between 1990 -00 as Percent of 2000 Stock -0. 547(**) 0. 387 0. 337 1. 21 0. 234 Some College, Associate or Bachelors Degree (2000, Percent) -0. 626(**) -0. 513 -0. 238 -2. 134 0. 04 Per capita Income (African American) Logged -0. 366(**) SOCIO-ECONOMIC CHARACTERISTICS

Final Model Variables: Education and Income: ► Some College or Greater Ed relates to More Intermixing ► African American Income relates to More Intermixing (but didn’t enter; r=+0. 45 w/ Education) Diversity: ► Higher Asian Element relates to More Intermixing ► Greater Change in the African American or Foreign-Born Elements relates to More Intermixing ► Higher African American Element relates to Less Intermixing ► Higher Hispanic Element and Increasing Population relate to More Intermixing (but didn’t enter; r=high w/ other diversity variables) Built Environment Characteristis ► More Manufacturing in 1990 relates to Less Intermixing ► More Mobility (Changing Houses) relates to More Intermixing ► More Recently Built Housing relates to More Intermixing (but wrong sign; r= high with diversity variables) ► More Recent Entry of Foreign-Born relates to More Intermixing; i. e. , MSA is a F-B magnet (but wrong sign; r=high with diversity variables)

Generalizations I: Empirical Overall, Housing Market for R/E Minorities Opened Up (M-LP) everywhere. But place-to-place differences are found; e. g. , Distinct difference between Rust/Sun Belt MSAs; Why? ? What are some Common Denominators? ? MSA Growth: Above avg growth Last Quarter 20 th Century = > Intermixing Post-Fordist Economic Activity More open landscapes (Older MSAs often enclosed (choked) with strong, long-standing municipalities; Institutionalized barriers? ? ) Expansion of Housing Stock E. g. 17 MSAs with High Intermixing, within 29 Higest Growth 1980 -00, 30% or greater MSAs in Flux = Greater Intermixing High levels of Housing Turnover Minorities gaining greater share of MSA population; inflow

Generalizations II: Empirical (Continued) Population Composition MSAs w/ long-standing Europe component = < Intermixing Older MSAs? More Institutionalized? ? Fordist Swoon? ? Substantial In-Migration of Others Less Effect? ? MSAs w/ long-stand non-Europe diversity = > Intermixing? Substantial in-migration = > Intermixing (e. g. , Hispanics) Higher Education and Minority Incomes = > Intermixing Housing Tenure: Owning vs Renting doesn’t impact Intermixing; R/E Presence seems more important? ? Diversity: Alone doesn’t relate to Intermixing Diversity Among Minorities = > Intermixing Seems to be intermixing w/ Caucasians; not new enclaves Population Change/Flux + Dynamics related to Foreign Born appear to be a significant Driver of Racial/Ethnic Intermixing

Generalizations III: Empirical (Continued) Enclave Effect? ? AA Presence (Proportn; point in time) relates to < Intermixing AA Change, also A + H, relate to > intermixing Suggests (i) Inertia Effect; (ii) Traditional Urban Enclaves continue; but (iii) New Enclaves not being formed. What about Ethnoburbs; Is their omission a scale effect? ? Frameworks: Consistent w/ Assimilation (salt + pepper effect) Consistent w/ Market-Led Pluralism (R/E intermixing, spread) Not Consistent with Stratification (R/E intermixing, spread) Not Consistent with Resurgent Ethnicity (but could be result of scale and/or broad scope of analysis) Suggests Class- rather than Culture-Process

And So It Goes -At the Frontier!! Thanks for Your Attention

Map of 49 Metropolitan Areas Regional Variations Northeast and Midwest with low to medium diversity and low intermixing. South and West with high to medium diversity and high and medium intermixing. Chicago and New York – Low Intermixing yet High Diversity What is happening in large MSAs? Does Population size decrease intermixing ? Why are West and South doing better in terms of racial residential intermixing? ? ? Fig. 4: Diversity and Entropy Indices

Major Patterns of Residential Intermixing Frey (2005) maps the Metropolitan areas loosing population and gaining population (1990~2000) • Rustbelt with mostly low-medium diversity and low intermixing. Fig. 5 : Regional Variations for Spatial Dispersion Source: Frey, W. H. , Brookings Institution Living Cities Census Series, September 2005 • These areas have slow economic growth and slow or losing population. • South and Southwest with High-medium diversity and Intermixing • South & Southwest have fast growing population and Economy

Diversity & Intermixing - View of Regional Spatial Variation Fast growing Areas HIHD, MIMD, MIHD Slow Growing Areas MILD, LIMD, LILD Fig. 6 a, b : Regional Variations for Spatial Dispersion Source: Frey, W. H. , Brookings Institution Living Cities Census Series, September 2005

Interesting Patterns Inertia of Rustbelt Slow Growing Economy Slow Growing population Low Diversity Fast Economic Growth Population Growth New Construction Medium to High Diversity and Intermixing Fig. 8: Regional Variation Characteristics

Micro-level Analyses – 3 MSAs in Ohio MSA Size Diversity Theil’s Entropy Score Index (Et) based (D) on tract level data Theil’s Entropy Index (Eb) based on block-group data Index Ratio = Eb/Et Cincinnati 1979202 0. 553 0. 345 0. 375 1. 008 Cleveland 2945831 0. 715 0. 393 0. 444 1. 128 Columbus 1540157 0. 687 0. 260 0. 289 1. 111 Fig. 12: Micro-level Clustering • All the 3 MSAs are more clustered at levels of block-groups than at tracts, Index Ratio greater than 1. • Block-level clustering-highest for Cleveland, lowest for Cincinnati. • Columbus – MSA shows more mixing (0. 289) in comparison to other two, yet block-level clustering in Columbus > Cincinnati. • It shows how different each MSA is in terms of their overall representation of mixing patterns at different scales.

Variables and Entropy Indices Why do patterns differ ? ? ? Rust Belt – Slow Economy Growing Population, New Constructions, Economic Growth and Medium to High Diversity and Intermixing Fast Growing Population New Economic Development Increasing Minorities What such pattern?

Interesting Patterns Inertia of Rustbelt Slow Growing Economy Slow Growing opulation Low Diversity Fast Economic Growth Population Growth New Construction Medium to High Diversity and Intermixing Fig. 8: Regional Variation Characteristics

Categories Counts Metropolises 1: LDHI 2 Minneapolis and Salt Lake 2: MDHI Seattle, Raleigh-Durham, Portland, Orlando, Oklahoma, 6 Norfolk 3: HDHI 4 Las-Vegas, Austin, San-Diego and Sacramento 4: LDMI Rochester, Columbus, Grand-Rapids, Indianapolis, 6 Louisville, Providence 5: MDMI West Palm Beach, , Tampa, San Antonio, Phoenix, Nashville, Kansas-City, Jacksonville, Hartford, Greensboro, Denver, Boston, Atlanta, Charlotte-SC 13 6: HDMI 6 Dallas, Houston, LA, Miami, San Francisco, W. DC. 7: LDLI 4 St. Louis, Pittsburgh, Cincinnati, Buffalo, 8: MDLI Memphis, Milwaukee, New Orleans, Philadelphia, Detroit, 6 Cleveland, 9: HDLI 2 Chicago, New-York

Correlation Values for Diversity Score (Among Minorities) for MSAs Socio-Economic Characteristics r-Values Per Capita Income (Total Population) . 348(*) Per Capita Income (African American) 0. 141 High School or Less Education (2000, Percent) Some College or Associate Degree (2000, Percent) -. 431(**) 0. 23 Bachelors Degree (2000, Percent) . 335(*) Graduate or Professional Degree (2000, Percent) 0. 255 Owner Occupied HHs in MSA (2000, Percent) -0. 214 African American Owner Occupied HHs ( 2000, Percent) -. 618(**) Asian Owner Occupied HHs (2000, Percent) . 381(**) Hispanic Owner Occupied HHs (2000, Percent) -0. 008 African American Rented HHs (2000, Percent) -. 638(**) Asian Rented HHs (2000, Percent) . 465(**) Hispanic Rented HHs (2000, Percent) Non-Caucasian HHs in MSA (2000, Percent) Change in Non-Caucasian HHs in MSA (90 -00, Percent) 0. 171 -. 338(*) 0. 027 Change in African American Owner Occupied HHs (90 -00, Percent) -. 556(**) Change in Asian Owner Occupied HHs (90 -00, Percent) . 382(**) Change in Hispanic Owner Occupied HHs (90 -00, Percent) -. 485(**) Change in African American Rented HHs (90 -00, Percent) -. 593(**) Change in Asian Rented HHs (90 -00, Percent) . 499(**) Change in All Other Races Rented HHs (90 -00, Percent) -. 478(**)

Correlation Values for Diversity Score (Among Minorities) for MSAs Demographic Characteristics Diversity Score for MSAs (Overall) Diversity Score (Minorities-Among-Themselves) r-Values 0. 116 1 Diversity Score (Caucasians vs. Minorities) -0. 138 MSA Population (2000) 0. 186 MSA Population LN (2000) 0. 201 Change in Population (1980 -2000, Percent) 0. 176 African Americans Portion of MSA Population (2000) -. 614(**) Asians Portion of MSA Population (2000) . 445(**) Hispanics Portion of MSA Population (2000) Change in Caucasian Population (1980 -2000, Percent) 0. 079 -. 395(**) Change in African American Population (1980 -2000, Percent) -0. 278 Change in Hispanic Population (1980 -2000, Percent) . 353(*) Change in Asian Population (1980 -2000, Percent) . 525(**) Foreign-Born Portion of MSA Population (2000, Percent) . 327(*) Change in Foreign-Born (2000 -90, Percent) . 356(*) Population (>5 yrs) Lived in a Different House in 1995 (Percent) 0. 152

Correlation Values for Diversity Score (Among Minorities) for MSAs Built-Environment Characteristics r-Values Foreign-Born Entered Before 1965 as Percent Total F-B -0. 127 Foreign-Born Entered 1980 -2000 as Percent Total F-B 0. 098 Median Value of Owner Occupied HH (2000) . 424(**) Median Year Housing Structure Built -0. 026 Houses Built Before 1970 as Percent of 2000 Stock -0. 013 Houses Built Before 1970 -1990 as Percent of 2000 Stock 0. 045 Houses Built Before 1990 -2000 as Percent of 2000 Stock -0. 038 Manufacturing Employment 1980 (Percent Total Employment) 0. 036 Managerial/Professional Employment 1980 (Percent Employment) . 357(*) Manufacturing Employment 1990 (Percent Total Employment) -0. 041 Managerial/Professional Employment 1990 (Percent Employment) 0. 095 Manufacturing Employment 2000 (Percent Total Employment) 0. 01 Managerial/Professional Employment 2000 (Percent Employment) 0. 271 Change in Manufacturing Employment (1990 -2000, Percent) -0. 129 Change in Managerial/Professional Employment (1990 -2000, Percent) -0. 087 Change in Manufacturing Employment (1980 -2000, Percent) 0. 034 Change in Managerial/Professional Employment (1980 -2000, Percent) -0. 083