Gentrification and Racial Transformation in Cincinnati, 2000-2016

Home , Residential area, Severance (land)

Gentrification and Racial Transformation in Cincinnati, 2000-2016.

A thesis submitted to the Graduate School of the University of Cincinnati in partial fulfillment of the requirements for the degree of

Master of Community Planning

In the School of Planning of the College of Design, Architecture, Art, and Planning by

Evelyn D. Ravuri July 2019

Committee Chair: L. Hollstein, Ph.D. Committee Member: C. Auffrey, Ph.D.

Abstract: Gentrification, first noticed in the largest cities of the United States in the 1960s, had diffused down the urban hierarchy by the 21st century. This new use of the urban landscape ushers in concerns over how revitalization of formerly disinvested areas of the cities and the consequent in-migration of higher-income populations has affected low-income populations. Cincinnati is one medium-sized city that has been affected by gentrification and the displacement of its low-income, largely African-American population. Using Hammel and Wyly’s (1996) gentrification criteria, it was determined that 26 census tracts in the City of Cincinnati experienced gentrification between 2000 and 2016. Fifteen of those tracts experienced displacement of the African-American population during this time. Logistic regression revealed that a gentrifying tract was 4.6 times more likely to experience displacement than a non- gentrifying tract between 2000 and 2016 indicating a possible cause and effect relationship.

By using triangulation of three data sources: U.S. census data, Google Streetview imagery, and interviews with members of Community Development Corporations in four gentrifying neighborhoods in

Cincinnati, this thesis explored possible reasons why two of these census tracts experienced displacement of African-American population and two did not. Tracts 55 (Madisonville) and 74 (Northside) were recipients of major economic investments in their business districts by public and private corporations after 2009. It is argued that substantial economic investment in these two formerly disinvested neighborhoods caused substantial increases in median housing values and median rent and led to displacement of lower-income individuals, many of whom were African-American. Google Streetview imagery examined change in the built environment in these two tracts between 2007 and 2014/16 and confirmed that these changes corresponded with gentrification. Conversely, Tract 84 (College Hill) did not experience displacement nor did its built environment change much between 2010 and 2016. Tract 84, unlike Tracts 55 and 74, had a substantial supply of high-quality housing stock at affordable prices which attracted higher-income in-migrants between 2000 and 2010. It appears that two separate housing markets operating in Tract 84 keep in-migrants from competing for the same housing opportunities as the lower-

ii income African-American population. Tract 19 (Walnut Hills) was further along in the gentrification process in 2000 than its counterparts in the other three tracts. Displacement of African-American population occurred between 2000 and 2010 as median housing values/median rents increased, but between 2010 and 2016, displacement was curtailed. Most of the post- 2010 economic investment was focused on Walnut Hill’s business district and the renovation of vacant buildings for housing and did not result in displacement between 2000 and 2010.

Interviews with CDC members revealed that the most poignant issue in their neighborhoods was the provision of affordable housing. Going forward, Cincinnati should rethink its public transportation system. All four neighborhoods were noted for accessibility to downtown and focusing on mixed-use development and densification along these nodes could lead to racially and socio-economically well- integrated neighborhoods in the City of Cincinnati and benefit all residents.

iii

Acknowledgements: The attainment of the MCP at the University of Cincinnati and the writing of this thesis could not have been accomplished without the support of my loving wife Ann Grimm. I thank her for her constant encouragement in this journey. Nor could I have endured the endless hours of work involved in this thesis without my cats: Athena and Bear (no longer with us) and Zephyr, Polaris, Pleiades, Sundae and Gelato.

I would like to thank Dr. Leah Hollstein for her invaluable guidance on the thesis. Her thoughtful criticisms allowed me to produce a better thesis. Dr. Hollstein was always helpful in ensuring that I did not go down the wrong track. I would also like to thank Dr. Chris Auffrey for his comments on this thesis.

None of this work could have been possible without Saginaw Valley State University which provided me with the year-long sabbatical to complete the course work for the MCP and provided a portion of the financial support for this.

I would also like to thank my mother and father for their constant support. Mine has been a long educational journey- starting in Fall 1983 and hopefully ending in Summer 2019, but my parents have always encouraged me. Thank you to my sisters Sandy and Lori for their invaluable sister- friendship. A special thanks goes to nephew Evan whom I love like a son. I would also like to thank Ann’s family

(Bob, Mary, Chris, Jill, Grayson, Sullivan, Terri, Chris C. and Tess) for their support through this process.

Lastly, thank you to the sixteen CDC members from College Hill, Madisonville, Northside and

Walnut Hills who provided their insight on the gentrification process in their respective neighborhoods.

Contents Chapter 1 INTRODUCTION ...... 1 Overview of the Study Area: ...... 4 Hypotheses: ...... 5 Methods: ...... 6 Justification for Study: ...... 6 Chapter 2 LITERATURE REVIEW: ...... 7 Introduction: ...... 7 Neighborhood Change: ...... 7 Theories of Gentrification: ...... 11 Demand Side:...... 12 Supply Side: ...... 13 Variations on Demand Side and Supply Side Gentrification Theories: ...... 14 Stages of Gentrification: ...... 18 Recessions and Gentrification: ...... 20 Gentrification and Displacement: ...... 24 Conclusion: ...... 30 Chapter 3 METHODOLOGY ...... 32 Introduction: ...... 32 Defining Gentrification: ...... 34 Measuring Gentrification: ...... 34 Measuring Displacement/Lack of Displacement: ...... 37 Hypotheses: ...... 38 Statistical Methodology: ...... 39 Qualitative Methodology: ...... 40 Google Maps: ...... 42 A Deeper Analysis Using Google Streetview: ...... 43 A Walk-Through the application of Hwang’s (2015) Gentrification Index: ...... 49 Conclusion: ...... 52 Chapter 4 DETERMINATION OF GENTRIFICATION AND DISPLACEMENT IN CINCINNATI, 2000-2016...... 54 Introduction: ...... 54 Median Income and Gentrification in Cincinnati: ...... 54

The Displacement of the African-American Population: ...... 69 Logistic Regression: ...... 75 Changing Distribution of the African-American Population and its Relationship to Poverty, 2000-16: ...... 79 Issues affecting Displacement Other than Gentrification: ...... 83 Conclusion: ...... 84 Chapter 5 TRACT 74- NORTHSIDE ...... 86 Introduction: ...... 86 Short History of Northside: ...... 88 Census Data: ...... 89 City and Tract 74 Comparisons - Population: ...... 89 City and Tract 74 Comparisons – Household Income, Housing and Rental Values, Educational Attainment: ...... 90 City and Tract 74 Comparisons – Housing Units, Vacancy, and Ownership: ...... 91 City and Tract 74 Comparisons – Disaggregation of Economic/Social Characteristics by Race: . 93 Comparison of African-American and White Selected Characteristics within Tract 74: ...... 95 Google Analysis: ...... 96 Examples of the Gentrification Process Using Google Streetview: ...... 101 The Built Environment in Northside as of June 2019: ...... 114 Interviews with Community Members: ...... 118 Conclusion: ...... 122 Chapter 6 TRACT 55- MADISONVILLE ...... 124 Introduction: ...... 124 A Short History of Madisonville: ...... 127 Census Data: ...... 128 A Comparison of Tract 55 with the City of Cincinnati on Select Economic, Social, and Housing Characteristics, 2000-2016: ...... 129 Select Housing Characteristics for Tract 55, 2000-2016: ...... 132 Comparison of Median Household Income for African-Americans and Whites in Tract 55 with that of City: ...... 134 Comparison of Select Characteristics of African-Americans and Whites in Tract 55: ...... 135 Google Streetview Analysis: ...... 136 Update on Built Environment in Tract 55 as of June 2019: ...... 152 Interview of Members of the Madisonville Community:...... 157

vii

Conclusion: ...... 165 Chapter 7 TRACT 19-WALNUT HILLS ...... 167 Introduction: ...... 167 Short History of Walnut Hills: ...... 169 Census Data: ...... 169 Comparison of Tract 19 and City of Cincinnati on Select Characteristics, 2000-2016: ...... 171 Select Housing Characteristics in Tract 19, 2000-2016: ...... 173 Comparison of City of Cincinnati and Tract 19 Incomes Disaggregated by Race, 2000-2016: ... 176 Select Characteristics for African-Americans and Whites in Tract 19, 2000-2016: ...... 176 Google Analysis: ...... 179 Current Revitalization Efforts in Tract 19: ...... 196 Interviews with Community Members in Walnut Hills: ...... 200 Conclusion: ...... 207 Chapter 8 COLLEGE HILL ...... 208 Introduction: ...... 208 Short History of College Hill: ...... 210 Census Data: ...... 211 A Comparison of Select Economic/Social Characteristics for Tract 84 and the City: ...... 212 Housing Market Characteristics in Tract 84: ...... 214 City and Tract 84 median household incomes disaggregated by race: ...... 217 Disaggregation of Select Characteristics in Tract 84 by Race: ...... 218 Google Analysis: ...... 222 Examples of Non-disinvested/Disinvested/Gentrified Areas in Tract 84: ...... 225 An Update on the Built Environment in Tract 84 in 2019: ...... 233 Interviews with Community Members: ...... 237 Conclusion: ...... 243 Chapter 9 A COMPARISON OF THE FOUR CENSUS TRACTS ...... 244 Introduction: ...... 244 Census Data: ...... 246 Population Change ...... 246 Comparison of Median Household Incomes by Tract for African-Americans and Whites: ...... 251 Median Housing Values and Median Rents: ...... 255 Vacancies and Homeownership: ...... 258

viii

Same Residence as Previous Year: ...... 262 Google Streetview: ...... 264 Housing: ...... 264 Stage of Gentrification: ...... 266 Summary of Interviews: ...... 269 Summary of Major Differences in Tracts Undergoing Displacement versus Those with No Displacement: ...... 273 Chapter 10 CONCLUSION: ...... 277 Displacement: ...... 279 Possible Solutions to Combat Displacement: ...... 282 Land Banks: ...... 285 Building on Transport Systems: ...... 285 Vacancies and Affordable Housing: ...... 288 A Comment on Google Streetview and Census Compatibility: ...... 289 REFERENCES: ...... 291

LIST OF TABLES 3.1 Characteristics of Selected Gentrified Tracts 40 3.2A Segment from Random Number Table 47 3.2B Number of Block Faces and Blocks Selected for Analysis 47 3.2C Stage of Gentrification by Using Three Indicator Variables 48 3.2D Hypothetical Scores for Stage of Gentrification 48 4.1 Potentially Gentrifiable Tracts in Cincinnati 73 4.2 Tracts that Underwent Displacement of African-American Population 75 4.3 Short-term and Long-term Logistic Regression Results 78 5.1 Population Change in Tract 74, 2000-2016 89 5.2 Median Household Income and Other Selected Variables, Tract 74 90 5.3 Select Housing Characteristics for Tract 74, 2000-2016 92 5.4 Comparison of Median Household Income by Race, Tract 74 94 5.5 Select Characteristics of Population by Race, Tract 74, 95 5.6 Sampled Blocks in Tract 74 and Stage of Gentrification 100 6.1 Population Change in Tract 55, 2000-2016 129 6.2 Median Household Income and Other Selected Variables, Tract 55 132 6.3 Select Housing Characteristics for Tract 55, 2000-2016 134 6.4 Comparison of Median Household Income by Race, Tract 55 135 6.5 Select Characteristics of Population by Race, Tract 55 136 6.6 Sampled Blocks in Tract 55 and Stage of Gentrification 141 7.1 Population Change in Tract 19, 2000-2016 171 7.2 Median Household Income and Other Selected Variables, Tract 19 173 7.3 Select Housing Characteristics for Tract 19, 2000-2016 175 7.4 Comparison of Median Household Income by Race, Tract 19 176 7.5 Select Characteristics of Population by Race, Tract 19 179 7.6 Sampled Blocks in Tract 19 and Stage of Gentrification 184 8.1 Population Change in Tract 84, 2000-2016 212 8.2 Median Household Income and Other Selected Variables, Tract 84 214 8.3 Select Housing Characteristics for Tract 84, 2000-2016 216 8.4 Comparison of Median Household Income by Race, Tract 84 218 8.5 Select Characteristics of Population by Race, Tract 84 221 8.6 Sampled Blocks in Tract 84 and Stage of Gentrification 225 9.1 Population of Tracts by Race, 2000-2016 247 9.2 Median Household Income by Race by Tract, 2000-2016 255 9.3 Median Housing Value and Median Rents in Four Census Tracts 258 9.4A Percentage of Vacant Dwellings in Four Census Tracts, 2000-2016 261 9.5B Homeownership Rates by Race in Four Census Tracts, 2000-2016 261 9.4C Number of Occupied Housing Units by Race 262 9.5 Percentage of Population in Same Residence as Previous Year 264 9.6 Percentage of Housing Stock Built at Different Times 266 9.7 Scores on Gentrification Index for Four Census Tracts 268

LIST OF FIGURES: 3.1 Block Face along Langland Street in Tract 74 51 4.1A Map of Cincinnati Neighborhoods 57 4.1B Map of Census Tracts in City of Cincinnati 58 4.1C Highways and Case Study Neighborhoods in Cincinnati 59 4.2A City of Cincinnati, Median Household Income by Deciles, 2000 60 4.2B City of Cincinnati, Median Household Income by Deciles, 2010 61 4.2C City of Cincinnati, Median Household Income by Deciles, 2016 62 4.3 Gentrification of Census Tracts in Cincinnati, 2000-2016 64 4.4 Percentage of Labor Force in Professions, Cincinnati, 2016 68 4.5A Displacement of African-American Population, 2000-2010 70 4.5B Displacement of African-American Population, 2010-2016 71 4.6A City of Cincinnati, Percentage of African-American by Tract, 2000 80 4.6B City of Cincinnati, Percentage of African-American by Tract, 2010 81 4.6C City of Cincinnati, Percentage of African-American by Tract, 2016 82 5.1 Northside, Tract 74, Satellite Imagery 87 5.2A Gentrified Blocks in Tract 74, Northside, 2007 98 5.2B Gentrified Blocks in Tract 74, Northside, 2014/16 99 5.3A Block 1004. Corner of Mad Anthony and Apjones, 2007 102 5.3B Block 1004. Corner of Mad Anthony and Apjones, 2014 102 5.4A Block 2019. Corner of Knowlton and Dane, 2007 104 5.4B Block 2019. Corner of Knowlton and Dane, 2014 104 5.5A Block 3010. Boarded-up Houses, 2007 106 5.5B Block 3010. Greenspace, 2014 106 5.6A Example of Late Stage Gentrification in Block 3002, 2007 108 5.6B Example of Late Stage Gentrification in Block 3002, 2014 108 5.7A Disinvested Commercial Area, Hamilton Avenue, 2207 110 5.7B Early Stage Commercial Gentrification, Hamilton Avenue, 2014 110 5.8 Construction of Knowlton Place, 2017 112 5.9A Future Site of Mixed-use Complex at Hamilton and Blue Rock, 2014 113 5.9B Mixed-use Complex at Hamilton and Blue Rock, 2016 113 5.10A Houses in Various Stages of Gentrification/Disrepair, 2014 115 5.10B Houses in Various Stages of Gentrification/Disrepair, 2019 115 5.11A Vacant Lot along Fergus Street, 2014 116 5.11B In-fill Housing along Fergus Street, 2019 116 5.12A Building in Disrepair along Knowlton Street, 2018 117 5.12B Revitalization of Building along Knowlton Street, 2019 117 6.1 Madisonville, Tract 55, Satellite Imagery 126 6.2A Gentrified Blocks in Tract 55, 2009 139 6.2B Gentrified Blocks in Tract 55, 2014/16 140 6.3A Parking Lot along Orlando Place, 2009 143 6.3B New OAR Spay/Neuter Clinic along Orlando Place, 2016 143 6.4A Disinvested Houses along Kenwood Road, 2009 145 6.4B Vacant Lot and Façade Improvement along Kenwood Road, 2016 145 6.5A Commercial Disinvestment along Whetsel Avenue, 2009 147 6.5B Boarded-up Strip of Stores along Whetsel Avenue, 2016 147 6.5C Newly Renovated Stores along Whetsel Avenue, 2018 147 6.6A Well-cared for Vacant Lot along Owasco Street, 2009 149 6.6B Newly Built Single-Family Home along Owasco Street, 2016 149 6.7A Vacant Lot along Ward Street, 2009 151

6.7B Community Garden along Ward Street, 2014 151 6.8A Vacant Lot at Whetsel and Madison, 2018 153 6.8B Mixed-use Development Located at Whetsel and Madison, 2019 153 6.9A Relatively Well-cared for House, 5805 Peabody Avenue, 2014 155 6.9B Renovated House, 5805 Peabody Avenue, 2019 155 6.10 New In-fill Houses along Ravenna Street, 2019 156 7.1 Walnut Hills, Tract 19, Satellite Imagery 168 7.2A Gentrified Blocks in Tract 19, 2007 182 7.2B Gentrified Blocks in Tract 19, 2014/16 183 7.3A House on St. James Avenue in Slight Disrepair, 2007 186 7.3B House on St. James Avenue with Slight Improvements, 2014 186 7.4A Disinvested Buildings along East McMillan, 2007 188 7.4B Revitalized Buildings along East McMillan, 2014 188 7.5A Public Space along East McMillan, 2007 190 7.5B Revitalized Public Space along East McMillan, 2014 190 7.6A Vacant Buildings along East McMillan, 2007 192 7.6B Revitalized Apartment Complex along East McMillan, 2016 192 7.7A Disinvested Business along Gilbert Avenue, 2007 194 7.7B Newly Revitalized U-Haul along Gilbert Avenue, 2014 194 7.8 Image of Windsor School being Converted to Apartments, 2018 195 7.9 Café Vivace, Up-Scale Coffee House on East McMillan, 2019 197 7.10 Renovation of Vacant Building, East McMillan, 2019 198 7.11A Dilapidated Building on East McMillan, 2018 199 7.11B In-fill along East McMillan, 2019 199 8.1 College Hill, Tract 84, Satellite Imagery 209 8.2 Gentrified Blocks in Tract 84, 207 and 2014/16 224 8.3A Spacious Ranch Houses on Bellmeadows Drive, Block 1003 227 8.3B Small Ranch Houses on Kirby Avenue, Block 1007 227 8.4 Hickory Woods Townhomes, Hamilton Avenue (Block 1000) 229 8.5A Vacant Apartment Complex on Belmont Avenue, 2011 230 8.5B Newly Remodeled Luxury Apartments on Belmont Avenue, 2016 230 8.6A Small Houses on Glenview Avenue in Disrepair, 2007 232 8.6B Rehabbed Houses on Glenview Avenue, 2014 232 8.7 Two Abandoned Buildings at Corner of Hamilton and Belmont 234 8.8 Construction Site for MSD Project located on Kirby Road 236

xii

Chapter 1 INTRODUCTION

The process of suburbanization in the U.S. after the 1950s led to population decline and economic disinvestment in central cities in the U.S. (Thomas, 1991). By the 1970s, tight housing markets and a high percentage of the population consisting of baby-boomers who were looking for affordable housing led to reinvestment in select inner-ring and downtown neighborhoods in many large cities in the U.S.

(London et. Al. 1980), a process referred to as gentrification. This process spread by the 1990s to smaller cities with relatively weak housing markets (Bereitschaft, 2014; Lees, 2006) as preferences for urban living by a subset of the American population changed to favor neighborhoods which were walkable, focused on mixed-use development, and offered an array of cultural amenities (Hammel and Wyly, 1996).

There is a major debate in neighborhood change concerning the effect of gentrification on the low-income population. Many researchers believe that low-income individuals are displaced from gentrifying neighborhoods because the cost of continuing to reside in the neighborhood becomes prohibitive (Sutton, 2018; Loukaitou-Sideris et. Al. 2017; Davidson and Lee, 2005; Kennedy and

Leonard, 2001). However, these studies rely too heavily on census data and other aggregate data pertaining to economic, demographic, and housing at the city/neighborhood level to deduce that lower- income individuals are displaced due to gentrification. Others find no evidence that gentrification displaces low-income individuals and instead suggest that these individuals benefit from additional job opportunities and amenities resulting from gentrification (Freeman, 2005; Freeman and Braconi, 2004;

Ellen and O’Regan, 2011; Crowder and South,2005). These latter studies draw their data from American

Housing Survey (AHS) data or databases designed to directly measure this effect in a specific city (for example New York City) allowing the city to track movement of individuals and their reasons for moving. The drawback of this methodology is that it is almost impossible to determine major changes occurring at the tract level. Using AHS or city-specific databases, we can only get an idea of changes at the city or national level as the number of responses from any one census tract is very small.

Given that the African-American population in all cities in the U.S. has lower median incomes than the white population (U.S. Census Bureau, 2016), it is expected that displacement of African-

Americans will occur in tracts which experience gentrification as they will be unable to bear the higher- costs of residing in a gentrifying neighborhood (Owens and Candipan, 2018). In fact, some definitions of gentrification use increase in white population as a criterion for gentrification (Aka, 2010; Kennedy and

Leonard, 2001). Challenging this definition, African-American gentrification has become more frequent over the past few decades (Hyra and Rugh, 2016; McKinnish, 2010). This process usually begins with middle-class African-Americans moving into lower-income neighborhoods that are already predominantly African-American (Chronopoulos, 2016).

However, we cannot assume that African-Americans are more likely to leave gentrifying tracts than their white counterparts. While lower income African-Americans and whites may leave tracts at the same rate, racial transformation is also affected by those moving into a neighborhood. New arrivals tend to be white and they are overall wealthier than those who left (Ellen and O’Regan, 2011). This is unfortunate in that many gentrifiers (but certainly not all) are interested in an urban experience that includes racial/ethnic diversity as well as socio-economic diversity (a mixture of high, mid, and low- income households), both of which contribute to the urban vitality of neighborhoods. If individuals with less financial resources (often minorities) are unable to remain in the neighborhood because of rising costs, then this urban experience may begin to resemble the sterile suburban experience that many gentrifiers have deliberately rejected.

This thesis does not seek to answer whether gentrification displaces lower-income African-

American or white individuals but seeks to examine how gentrification has changed the aggregate racial and socio-economic composition of gentrifying tracts in Cincinnati. These changes will be examined through four cases studies of census tracts located within the City of Cincinnati. Two census tracts experienced racial and socio-economic change that identifies them as both gentrifying and having experienced displacement of the African-American population between 2010 and 2016. Conversely, two

2 tracts experienced socio-economic change which categorized them as gentrified but did not experience a significant change in racial composition. In each of the four cases, African-American median incomes will be compared for 2000 (Decennial Census) and the American Community Survey for 2010 and 2016 to determine how these individuals fared in comparison to their white counterparts in the four tracts.

The last component of this thesis specifically addresses the question of the effect that the Great

Recession has had on gentrification and racial transformation in the City of Cincinnati. According to the

U.S. Department of Labor Statistics (2012), the U.S. officially entered the Great Recession in 2008, which was attributed to low interest rates since 2001, low yields on U.S. bonds, subprime housing loans, and lax regulation on loans (Verick and Islam, 2010). Regardless of the reasons for the onset of the Great

Recession, increased unemployment and foreclosures on housing loans due to balloon loan payments caused the U.S. housing market to collapse. Many homeowners were left with mortgage debt that exceeded the value of the house (Florida, 2010). Housing markets in Florida, the East Coast, and

California underwent the greatest speculation and thus fared the worst in the housing market collapse, but the effects were widespread throughout the U.S. (Florida, 2010). Cincinnati, with a decline of 5 percent in its housing prices between 2007 and 2009, fared better than the nation which experienced a ten percent decline in housing prices (Federal Reserve Bank of Cleveland). However, Cincinnati’s unemployment rate increased from 5 percent to 10 percent between 2008 and 2009 (Federal Reserve Bank of Cleveland,

2018), and according to Kuehn (2011) both low-wage (-10.02 percent) and high-wage industries (-7.29 percent) in Cincinnati experienced a decline in the number of jobs between 2007 and 2009.

Effects of the Great Recession on gentrification in Cincinnati are unclear. It has been noted that the Great Recession had a greater impact on the minority population who experienced higher unemployment rates and were victims of unethical loan tactics which resulted in foreclosure on their homes (Hoynes et. Al, 2012), while increased unemployment may also have resulted in eviction from rental units. These conditions may have created an opportunity for those with greater financial resources to invest in neighborhoods with housing stock in good condition, but in which segments of the population

3 were economically strapped. However, this time of extreme economic uncertainty may have prevented public/private investment as well as individual gentrifiers from investing in these neighborhoods. I cannot advance a hypothesis regarding the Great Recession and its effect on gentrification/displacement on

Cincinnati at this time. The effect is likely to vary in different neighborhoods depending on myriad factors. The assumption in this paper is that gentrification leads to displacement of lower-income populations. It is assumed that lower-income individuals would be more susceptible to economic downturns than their middle-class counterparts because they have less economic resources to draw upon in times of economic crisis.

Overview of the Study Area: The City of Cincinnati’s population has declined in each decade since 1950. Between 1950 and

2010, the City lost over two-hundred thousand residents (U.S. Bureau of the Census, 2010). The greatest one decadal decline was during the 1970s and likely represents rapid suburbanization. It is also likely that this decrease reflected employment decline in industrial cities in the Northeast and Midwest and population flows to the South and West (Greenwood and Hunt, 1989). Surprisingly, the city’s population increased by about 1,500 residents between 2010 and 2016, reversing a five-decade decline. It has been noted that millennials have been more attracted to city living than their counterparts (Moos, 2016) in other age groups and this likely accounts for some of that growth.

Cincinnati has become famous in the urban planning world for its transformation of Over-the-

Rhine from a disinvested area of the city to the poster-child for gentrification (Woodward, 2016). Over- the-Rhine is not without its critics: mostly dealing with the displacement of low-income African-

American population (Brenneman, 2018). It is hoped by comparing the policies and actions that different neighborhoods in the City of Cincinnati have taken to encourage economic revitalization without displacing the incumbent residents would be invaluable for not only the City of Cincinnati, but also other cities in the U.S.

Four census tracts were chosen for this project. Tracts 19 (Walnut Hills) and 84 (College Hill) were selected for the case study because they experienced gentrification between 2010 and 2016 and did not experience displacement of the African-American population. Conversely, census Tracts 55

(Madisonville) and 74 (Northside) experienced gentrification between 2010 and 2016 and displacement of

African-Americans.

Hypotheses: The following are the hypotheses for the first part of the analysis which deals with gentrification and displacement at the city level:

1. Did tracts that began the process of gentrification between 2000 and 2010 in the City of

Cincinnati have a greater chance of displacement of the African American population than non-

gentrifying tracts by 2010?

2. Did tracts that began the process of gentrification between 2010 and 2016 in the City of

Cincinnati have a greater chance of displacement of the African American population than non-

gentrifying tracts by 2016?

3. Did tracts that began the process of gentrification between 2000 and 2010 have a greater chance

of displacement of the African-American population than non-gentrifying tracts by 2016?

The second set of hypotheses deals with the four case study tracts and focuses on why two of those tracts experienced displacement of their African-American population and two did not. These hypotheses include:

1. Disinvested neighborhoods which experienced massive economic investment over a short

timeframe (less than a decade) will experience more displacement of the African-American

population than neighborhoods in which economic investment occurred over a longer period of

time.

2. Those neighborhoods which adopted policies to deal with displacement of low-income population

will experience less displacement than those neighborhoods which did not adopt such policies.

Methods: Triangulation is used to cross-check the results from different data sources to gain a better understanding of the gentrification and displacement process in the four neighborhood case studies.

Census data from 2000, 2010 and 2016 is used to determine if displacement has occurred in these tracts.

The census data allows a detailed comparison of demographic, economic, social and housing characteristics between African-Americans and whites in the four census tracts to determine how these characteristics have changed over the timeframe of analysis. Google Streetview imagery is used to document changes to the built environment between 2007 and 2014/16 and compare these with the results from the census data. Interviews with Community Development Corporation members from the four neighborhoods was undertaken to determine the effect of the gentrification process on the policies and actions taken by the requisite associations concerning displacement of the African-American population.

Justification for Study: That gentrification is occurring in Cincinnati can hardly be disputed (Simes, 2013). Identifying neighborhoods (census tracts) where displacement of the African-American population is occurring could help understand how this process unfolds spatially and temporally. A comparison of displacement or lack of displacement in census tracts may provide insight into 1) what are the conditions under which displacement is not happening despite gentrification 2) what supports these conditions and 3) what types of policies can be enacted, conditions could be imposed, or organizations could be supported to counter the disruption of disenfranchised populations. For example, is it possible to provide subsidized housing for lower-income individuals that allow these individuals to live in the same apartment complexes as those paying market prices for housing? Could there be lower rents paid by minority business owners in gentrifying tracts to make their businesses competitive with up-scale businesses that move into the community? This insight will be used to ensure that future gentrification of these neighborhoods does not result in displacement of certain racial/ethnic groups.

Chapter 2 LITERATURE REVIEW:

Introduction: The following chapter considers five interrelated topics: neighborhood change, theories of gentrification, stages of gentrification, recessions and gentrification, and gentrification and displacement.

The neighborhood change section examines the classic studies advanced in the early to mid-twentieth century when transportation innovations and suburbanization were rapidly transforming the spatial structure of the American city. The section continues with neighborhood change studies in the last decade of the twentieth century and the first decade of the twenty-first century; a time when the economy transitioned from an industrial to a post-industrial society and changed the economic value of land close to the CBDs of U.S. cities. This changing spatial structure of the U.S. economy is expanded upon in the next section on theories of gentrification. Smith’s (1979) rent-gap-theory explores the value that can be derived by moving capital into disinvested areas of the city while Ley’s (1986) cultural theory examines changing consumer preferences for residing in cities. While these appear to be two competing theories, the literature review will demonstrate that they are both valid theories and work together to advance the understanding of gentrification. The stages of gentrification are discussed in the third section of the literature review. The effect that recessions have on the gentrification process is explored in the fourth section. Given the recency of the Great Recession, no definitive conclusions are reached concerning the impact of the Great Recession on gentrification, but it seems that the gentrification process was somewhat slowed down in many U.S. cities between 2007-2010 (Hartley and Kolliner, 2014). The last section examines gentrification and displacement focusing on the displacement of low-income population in general and then examines racial transformation in neighborhoods due to gentrification.

Neighborhood Change: Suburbanization has had a profound impact on the demographic and economic structure of inner cities in the U.S. Low cost housing loans available after WWII made it possible to purchase new homes outside

7 the central city while the mass production of automobiles and the development of highways further facilitated the process of suburban living (Long, 1988). The disamenities related to city living

(congestion, crime, pollution) could be avoided by those who relocated to the suburbs. This exodus to the suburbs was not a cross-section of the city’s population. White, middle-class families led the way, leaving a residual of poorer individuals, minorities and the elderly in a deteriorating inner city (Frey, 1979). With the exodus of population from the inner city, employment opportunities and retail soon followed their wealthy clientele to the suburbs (Grub, 1982). Lastly, tax revenues from the middle-class were transferred to the suburbs and funded schools and other services at the expense of the inner city.

For much of the twentieth century in American cities, a process of ‘filtering’ occurred as higher income households moved from aging housing stock in the central city to newer housing in the expanding suburbs (White, 1971). These higher income households left housing in relatively good condition which could then be used by households with less financial resources. This process of ‘filtering’ continued until either the housing became substandard and ceased to provide housing opportunities to even those of the lowest socioeconomic status or efforts were undertaken to revitalize the area. This residential revitalization could happen in two ways. Aesthetically pleasing, structurally sound buildings could undergo renovation or demolition of unsuitable buildings could occur and new housing built in its place.

Social scientists have a long history of studying neighborhood change in U.S. cities and it is these studies that provide a basis for the relationship between urban decline and the subsequent regeneration of cities. As early as the 1920s, Park, Burgess, and McKenzie (1925) proposed the invasion and succession theory of neighborhood change. During the early twentieth century, cities in the Northeast and Midwest were expanding rapidly and population was greater than housing supply. The newer arrivals of the twentieth century, often immigrants and rural migrants from the South, settled in the older housing of the central city which was often the only places they could afford. Wealthier residents began their retreat to newly constructed housing in the developing suburbs. Over time this process would repeat itself with newer, more expensive housing located further from the center of the city. Thus, the middle-class could

8 distance themselves from the social ills associated with the central city. Hoover and Vernon (1959) expanded on Park et. al. model and presented a neighborhood change model that is appropriate for this study. This model consisted of five stages that neighborhoods progress through as they age. In the development stage (stage 1), most housing is newly constructed single-family residential units. In stages 2 and 3, higher density housing is constructed while older units are converted to multi-family residences. In stage 4, population loss and the loss of housing units due to dilapidation occurs. In stage 5, the demolition/replacement of older housing units occurs. Unfortunately, for many cities, the lack of funds does not allow a movement to Stage 5 of the Hoover and Vernon model and many areas of a city may be stuck in Stage 4. These declines in central city populations, housing vacancy and abandonment have become major problems for many cities.

In the late 1950s, government sponsored urban renewal had resulted in the demolition of 400,000 buildings and the displacement of 1.4 million individuals from cities in the U.S. (Scott, 2017), and set the stage for what would later be known as gentrification. The government involvement in this process was largely a result of the inability of central cities to attract private investment and to renovate the urban landscape. After the 1950s, reinvestment in parts of the central city in proximity to the CBD could expect a higher premium and attracted private investments. This is a result of the rent gap (Smith, 1979) which makes it financially profitable to reinvest in the city. Decades of disinvestment in inner cities caused depreciation of the building stock which allowed both individual homeowners and investment corporations to purchase properties at a reduced rate. The difference between what is paid for the property and the potential rent that could be accrued after revitalization is referred to as the rent gap.

While retail and industry followed their clientele/workforce to the suburbs after the 1950s, this process was reversed during the 1970s as the U.S. transitioned from an industrial to a post-industrial society which focused on information-processing/banking/research at the expense of the manufacturing sector. These professional jobs were concentrated in the CBDs of U.S. cities and many employees placed a premium on proximity to the workplace as decades of urban sprawl made commutes to these

9 employment opportunties problematic. Additionally, within the past few decades, residential preferences have favored central cities and certain neighborhoods within these central cities have experienced in- migration and economic revitalization (Ley, 1986). Many individuals have become disenchanted with suburban lifestyles and have moved to the central cities to access amenities that are not available in suburban locales (Ley, 1986). By residing within the city, these residents have access to cultural, recreational, and retail amenities. These communities tend to be more pedestrian/non-automobile transport-friendly. They also tend to be located on public transport lines that allow easy access to employment opportunities.

Florida (2005) in Cities and the Creative Class introduced his creative capital approach which states that in a post-industrial society, creative individuals (artists, educators, scientists, and business professionals) are attracted to cities for not only economic motives but for life-style choices. He refers to the three T’s which include technology, talent, and tolerance. Creative cities are ones that have adapted to the latest technology which further attracts talented individuals who are looking for cities which are inclusive of a diversity of racial/ethnic and life-style alternatives (tolerant). Florida’s approach can be applied to individual cities to explain the catalysts which explain why certain neighborhoods in the city gentrify. Often these are neighborhoods in proximity to the CBD or a major university or health care complex (technology), which attracts a highly talented labor pool that desires proximity to employment, in a neighborhood that values diversity.

Whether gentrification was initiated by employment motives or residential preferences, it would likely not have occurred without the excess of housing stock in the inner cities from the previous exodus of population after 1950 (Gyourko and Glaeser, 2005). Much of this housing stock was substandard and basically unlivable, but a percentage of the housing stock was inhabitable but vacant.

This section examined the classic theories of neighborhood change which were written when the

U.S. was undergoing industrialization. Cities expanded horizontally (creating the suburbs) due to transportation innovations and the desire of the middle-class to escape the social ills of central cities in the

U.S. The U.S. entered a post-industrial phase during the late 1900s and this changed the spatial structure of cities once again. This time, there was a contraction towards the center to take advantage of an economy that became more dependent on services instead of manufacturing.

Theories of Gentrification: London et. Al (1980) summarized the four major theoretical approaches to gentrification that were advanced during the 1960s and 1970s. The approach focused on changes in population characteristics such as the aging of baby-boomers into adulthood, the decline in average household size, and the increase in dual-earning couples. These demographic changes had both economic and cultural undertones. The large number of baby-boomers entering the housing market in the 1970s resulted in the search for affordable housing opportunities in the city. The decline in household size made larger homes in the suburbs a liability for one-person households, such as singles entering the labor market or the elderly, who did not have the economic resources or the time/energy that is required for upkeep of suburban housing. In contrast, dual-earning households were less concerned with affordability in the suburbs and may have found the city convenient to their workplaces and an array of city amenities. The ecological approach incorporates the idea of sustenance organization which focuses on the diversity of different types of jobs relating to different sectors of the economy (primary, secondary, tertiary, and quaternary1). It is suggested that cities in the post-industrial era, those in the tertiary and quaternary sectors, are favorable to gentrification as these types of jobs tend to be more associated with the CBD and proximity to these jobs can be enhanced by residing in disinvested areas of the city. The socio-cultural approach focused on middle-income populations sharing space with the less affluent. It emphasized community involvement and was largely a revolt against a suburban lifestyle by the middle-income

1 The four sectors of the economy referred to in London’s study include: 1. Primary- the extraction of raw materials from the earth (e.g. agriculture, mining, lumbering); 2. Secondary- the manufacturing sector; 3. Tertiary- most services; and 4. Quaternary- includes information processing and manipulation.

11 population. The political-economic approach focused on the capitalist struggle between the classes for urban space as well as the role of government in fomenting this struggle.

Using data from 48 cities in the U.S. from 1970 to 1980, London et. Al. (1986) tested these four theories through regression analysis. They found support for each of the theories with the political- economic approach receiving the strongest support. They suggested that theories of gentrification need to integrate these four approaches and suggested a new approach, the social movement approach, be advanced; it is an amalgamation of the four other theories. The process occurs by encouraging pro- urbanism through revitalization of urban space and developing an appreciation for urban areas (socio- cultural approach). This ideology is then adopted by lending institutions and realtors who invest economic resources in revitalization (political-economic) to entice certain populations back to the city

(demographic) to inhabit a post-industrial city focused on white-collar employment (ecological).

While gentrification research during the 1960s and 1970s was guided by the five approaches discussed above, the socio-cultural and political-economic theories were further developed and dominated the literature of the 1980s. These theories of gentrification revolve around two major themes: supply side versus demand side. Supply side theories are entrenched in economics and consider gentrification to be a part of a capitalist system of moving investment from unprofitable to profitable venues (Smith, 1979).

Demand side theories claim that cultural preferences for inner city living drive the gentrification process

(Ley, 1986). This section examines these two major theories as well as rebuttals to these theories.

Demand Side: The leading advocate of the socio-cultural approach was Ley (1986) who examined the gentrification process in 22 Canadian cities between 1971 and 1981 to determine the explanatory power of the following four themes of gentrification: demographic factors, housing market dynamics, the amenity hypothesis, and the economic hypothesis. These approaches were comparable to those examined by London et. al. (1980). By using correlation analysis and principal components analysis, Ley found that the demographic characteristics (e.g. age structure and educational levels) of a city and housing

12 supply/housing prices had little to do with the causes of gentrification in the cities that he studied. Ley concluded that urban amenities in conjunction with a post-industrial economy explained most of the gentrification process in Canadian cities: that cities with a large proportion of the labor force in quaternary sector activities versus manufacturing made the inner city more attractive to middle-class individuals who could then have access to environmental and cultural amenities.

Supply Side: Smith (1979) was the leading proponent of the political-economic hypothesis and stressed that gentrification was a response to economic preferences as opposed to cultural processes. In an examination of movements of gentrifiers in the 1970s in Philadelphia, Baltimore, and Washington, D.C. he found that the majority of gentrifiers originated from other areas of the city, not the suburbs and that cultural preferences for city living cannot explain gentrification. To understand the gentrification process, one must examine the rent gap theory. The investment in suburbs after the 1940s caused many neighborhoods within the boundaries of the central city to undergo a process of disinvestment as investment flowed into the suburbs. According to the bid rent theory, rent (price) charged by a square unit of land is highest in the central business district and then is devalued as one proceeds to the outskirts of the suburbs where the lowest rent values are found (Alonso, 1967). The only way for the poor to remain on some of the most valuable land in the city was to economize on size of dwelling and/or to increase household size. Much of this was accomplished by the filtering process of housing throughout the twentieth century and the conversion of single-family residences to multi-family.

By the 1960s the greatest decline in rents were in the inner-city areas outside of the CBD which have been likened to a valley in the bid rent curve (Alonso, 1967). This disinvestment came about in several ways. Homeowners found that costs of repairing older dwellings could be avoided by relocating to newer housing in the suburbs. Landlords could then buy these properties and undermaintain them to extract more profit from the building. The only way to reverse this process was for housing to become

13 scarce and for reinvestment to occur in these properties (Smith, 1979). This can occur in neighborhoods with high levels of abandonment. In many cases, these buildings were inhabitable, but they had become unprofitable to retain. Gentrification occurs when the rent gap (the valley that had developed in the bid rent curve) is sufficiently wide enough for investors to rehabilitate property and invest further capital in the structure (Smith, 1979). Smith states if his theory is correct then those neighborhoods closest to the central business district would garner the first wave of gentrification while those further out could possibly experience a second wave of gentrification when prices rise too high in the inner ring (Smith,

1979).

Variations on Demand Side and Supply Side Gentrification Theories: Lees (1994) sought to integrate economic and cultural explanations of gentrification. Economic theories are based on Marxian principles (Smith) or the struggle between the bourgeoisie and the proletariat over the bidding for housing consumption while postmodernist theories examine lifestyle preferences of gentrifiers (Ley). Lees (1994) provides three examples where Marxian and cultural preferences can be interwoven to construct a more robust theory of gentrification. First, is the politicization of interest groups. This focuses on the middle-class who must simultaneously identify with the proletariat and the bourgeoise. However, postmodern society has caused the lines to blur between social classes and has caused an uncomfortable relationship between the three social classes. Lees states that gentrifiers express their allegiance to upper-class values by emulating upper-class housing consumption (economic) while at the same time retaining a social conscience which aligns them with the working-class/poor (cultural). Lees does not expand on the cultural aspect of this dichotomy, but the reader may surmise that gentrification is an adverse reaction to the sterile, homogeneous environment of the suburbs where many middle-class individuals would likely reside. Secondly, Lees argues that capital moves to disinvested areas not only because of the additional economic value that can be extracted from this movement, but to cultural preferences which may in fact guide that economic investment. Lastly,

Lees considers gentrification as a response to nihilism: defined as “an attempt to create a reality of one’s own, stolen from past realities, past cultures, in what becomes recategorization of place” (page 147). In this case, gentrification is linked with the consumption of built forms that is reminiscent of the past and pleasing to the eye (cultural) while maintaining an economic value (economics).

Hamnett (1991) argued that a theory of gentrification must treat cultural and economic perspectives as complementary to the process instead of mutually exclusive. For Hamnett there are four key questions to be answered. They include: 1) why is gentrification more prevalent in some large U.S. cities and less concentrated in older industrial cities; 2) within cities, why do some neighborhoods gentrify whereas others fail to gentrify; 3) who becomes a gentrifier and what are the reasons for this; and

4) must explain why it occurred at a certain time. Hamnett created a schema where there were three supply factors (potential gentrifiers, suitable area for gentrification, attractive central city neighborhoods) and one demand factor, the cultural preference for residing in the central city by white-collar workers. He then examined the rent gap thesis to determine under which conditions gentrification would occur. With or without the rent gap, gentrification does not occur without potential gentrifiers; nor does it occur without a demand for central city locations. Only when there is an inner-city preference for residence and a rent gap is gentrification likely to occur, illustrating that it is both economic and cultural factors that promote gentrification.

Rose (1984) found fault with both positivist and Marxist explanations of gentrification. Rose stated that positivist (consumer preference) positions do not adequately address the underlying causes of gentrification while Marxist approaches focus on commodity production which is expressed in the built landscape. Rose felt that gentrification and gentrifiers could not be aggregated into one category. One must not only examine production (Marxism) but changes in reproduction (positivism). Changes in production include economic restructuring while changes in reproduction include smaller family sizes and non-nuclear families. Rose also calls for a greater examination of the role of women in fostering gentrification. Women often have responsibilities as both economic providers and caregivers making their

15 activities difficult to coordinate in the suburbs. According to Rose, Marxist and positivist theories of gentrification do not account for ‘marginal gentrifiers’ who are not positivist consumers of amenities but must compete with other low-income individuals for affordable housing in locations which allow them access to jobs in a restructured economy. Suburban residence is not an option for many of these individuals. Rose believes that displacement that occurs in gentrification is not necessarily higher-income gentrifiers, but moderate-income gentrifiers who can marginally outbid low-income individuals for space.

Warde (1991) takes Rose’s ideas further and insists that it is gender not class that is the catalyst for gentrification. Warde states that women in a post-industrial society have new roles which break them from traditional patriarchy and put them front and center in the consumption process. Since the 1950s, women have entered the labor force in greater numbers, have attained advanced levels of education, and are now employed in higher-level positions. The investment in higher-levels of education and a greater percentage of women entering/remaining in the labor force is correlated with a percentage of women who do not marry at all or marry later and may remain childless.

Lastly, Beauregard (1990) cautions theorists who consider gentrification as a process that occurs in the same manner in all locations. He insists that gentrification can involve different actors and can impact the environment in different ways even when the process occurs in the same neighborhood.

Beauregard also admonished that once started, the gentrification process does not necessarily continue.

Capital accumulation as stated by economists is necessary, but the processes will evolve in different ways depending on those participating in the process (developers, government agencies, or even households).

By the mid-1990s, gentrification theories were being reevaluated. It had become apparent that local governments were much more involved in the gentrification process than they were previously.

While the original theories focused on capital (production) and consumption (life-style choice), a third component (government) was added to the gentrification schema. Mathema (2013) emphasized that the increased involvement of local government in the gentrification process is a result of reductions in federal and state monies appropriated at the local level. This has required alternative strategies to increase the tax

16 base of cities in the 21st century. Local governments have promoted gentrification by: 1. Inviting private corporations to invest in the city; and 2. Attracting middle-class residents back to the city who can then contribute to the tax base. Mathema’s second point can be illustrated by examining Smith’s (1996)

‘Revanchist City’ idea in which local governments focused on such activities as crime prevention and removal of homeless persons to entice the middle-class to return to the city and in effect increase the rent gap making central cities more profitable for private investment.

Freeman (2016) examined data on 21st century gentrification and found several differences from its 20th century counterpart and encourages a reworking of gentrification theory to incorporate these differences. One difference is that gentrification in the 21st century is much more prevalent than it was in the 20th century; affecting more cities and more neighborhoods within cities. Secondly, in the 1970s and

1980s, gentrifiers were focused largely on renovating dwellings in central city neighborhoods. Today, these individuals are more likely searching for newly built or already renovated high-rise condominiums with a rich array of amenities. A third difference concerns that of crime. Early gentrifiers considered crime a trade-off to access high-quality housing at low-cost in architecturally pleasing neighborhoods. For

Freeman, future theories of gentrification will need to encompass these changes.

Another perspective that needs to be considered in gentrification theory is that of transit use.

Revington (2015) regards this as an under-researched topic. New transit systems are likely to cause higher ground rent and more intensive use of land. This increases the rent gap in a neighborhood even if that neighborhood is not undergoing disinvestment (see Smith, 1979). More importantly, these transit stations may contribute to gentrification in the inner-ring suburbs and may contribute to the spatial spread of gentrification. Recall that Smith (1979) expected outlying areas to experience gentrification once investment in neighborhoods close to the CBD became too expensive.

This section examined the classic theories of gentrification which focused on the debate between consumer demand (Ley, 1986) and production (Smith, 1979) and its effect on gentrification. Other theories which attempted to integrate demand and production were also examined to better understand the

17 gentrification process (Lees, 1994; Rose; 1980; Hamnett, 1991). The section ends with an admonition that gentrification theory needs to incorporate the role of local government and the role of transit development to understand gentrification in the 21st century (Mathema, 2013; Freeman; 2016; Revinington, 2015).

Stages of Gentrification: Clay (1979) developed a model of the stages of gentrification from his analysis of this process as it unfolded in many large cities of the U.S. during the 1960s and 1970s. In Stage 1, risk-takers move into disinvested neighborhoods and use ‘sweat equity’ to rehabilitate housing stock. This sweat equity is necessary as banks are hesitant to lend in these neighborhoods. The preferred acquisition in this stage is vacant buildings that are in good condition and does not lead directly to displacement of population. In

Stage 2, small-scale investors renovate housing and the scarcity of vacant units in good condition results in these individuals outbidding the incumbents for housing and often begins the displacement process.

Developers move into the neighborhood in Stage 3 as well as individual investors who view gentrification as a money-making venture. Banks now view the area as safe for investment, which results in the arrival of additional middle-class residents with further displacement occurring. In Stage 4, land prices rise rapidly due to property held for speculation and specialized commercial districts may develop close to the business district or major institutions such as universities, hospitals, or research complexes. While Clay’s model focused largely on the economics of gentrification, Gale’s (1979) model focused on the characteristics of the gentrifiers and only included two stages. It must be noted that Gale’s study only focused on Washington, D.C., a city that may be atypical of the gentrification process due to its position as a governmental city. In Gale’s transitional stage, gentrifiers tend to be comprised of single males, childless couples, and persons under 35. They are less committed to long term residency and are willing to accept the risk of moving into a disinvested area. The second stage is the mature stage of gentrification, which occurs after substantial reinvestment from the public and private sectors and comprises more single

18 and married women. It includes households with children and older gentrifiers who are less willing to take a risk on a newly reinvesting neighborhood.

Unlike Clay and Gale who were only able to examine the very beginnings of gentrification in the

U.S., Hackworth and Smith (2001) provided a stage model of gentrification which covered the late 1950s to 2000. Their model consists of three waves of gentrification. The first wave (from the late 1950s to

1980) considers gentrification as sporadic and mostly confined to the Northeastern cities of the U.S.

Private lending institutions were risk averse to investing in these neighborhoods so what investment that occurred was funded by government entities. By the latter part of this wave developers began to invest in rehabilitation of properties that had already been devalued but had experienced further economic problems resulting from the mid-1970s oil related recession. Much of this redevelopment was stimulated by the decline of manufacturing and the rise of the service and information sectors in the central city for those working in the city, these neighborhoods would command a higher price. The second wave began in

1980 and lasted until the early 1990s. Gentrification had by then diffused to smaller cities in the U.S.

During the recession of the early 1990s, investment in gentrifying neighborhoods was curtailed as these neighborhoods were viewed as risky investments. During the third wave of gentrification (post-recession

1993) investment increased in the central city neighborhoods that had begun the gentrification process but also was noted in neighborhoods further from the central city. This era of large-scale investment was promoted by private interests and was also supported by government funds. Lees, Slater, and Wyly (2008) added a fourth wave of gentrification to Hackworth and Smith’s model. After the recession of 2001, financial agencies relaxed mortgage lending standards which stimulated speculation in the housing market and fueled the reinvestment in marginal neighborhoods. As previously stated during the first wave gentrification, banks did not risk funds in declining neighborhoods. This fourth wave process of speculation made living in gentrifying neighborhoods precarious for those with less financial resources.

Renters were unable to compete as rental prices in these neighborhoods increased substantially. It is still

19 too early to fully understand the effects of the Great Recession of 2007-09 on gentrification or the affect that it had on low-income individuals residing in gentrified neighborhoods.

Aalbers (2019) argues that a fifth wave of gentrification has occurred since the Great Recession.

During the early 2000s, capital was invested in real estate through major corporations or through individuals who purchased real estate as a form of wealth storage. This has given rise to platform capitalism which has directed investment to specific areas of a city which often have a tourist component to them or are at least cultural/art hubs for the city. One example of this new process of gentrification is that of ‘Airbnbs’ where wealthy individuals rent out their homes to capitalize on gentrification in that city.

This section examined the stages of gentrification. Early work (Clay, 1979 and Gale, 1979) emphasized that gentrification began with individuals who were willing to risk investment in marginal neighborhoods. Once that investment proved profitable, less risk-motivated individuals and banks were willing to invest in gentrifying neighborhoods. In later stages of gentrification, the process moved beyond select neighborhoods close to the CBD and attracted capital from major corporations (both national and international) as well as individual speculators who purchased property in gentrifying neighborhoods as an investment opportunity.

Recessions and Gentrification: Much of the effect of the Great Recession on gentrification needs to be gathered from studies of the effect of previous recessions on the gentrification process. One of the earliest studies concerning a recession on the gentrification process was Ley’s (1992) examination of the effect that the 1982-83 recession had on gentrification in six Canadian cities. Using the pace of gentrification during the 1970s as a comparison, Ley found that gentrification accelerated between 1981 and 1986 in these six Canadian cities which was contrary to his hypothesis that stated that gentrification would be stalled or reversed during the recession. He attributes this to the economic restructuring process that reorganized the spatial structure of cities and made inner city neighborhoods more attractive to white-collar employees than they

20 would have been in the 1970s. Bourne (1993) came to the opposite conclusion from Ley and claimed that the gentrification process of the 1960s to 1980s was a short-term process that would reverse itself with the recession of the early 1990s and that suburbanization would again become the predominant process in

North American cities. Bourne claimed that the potential supply of young gentrifiers in the 1970s would decline in the future as baby-boomers aged and found inner-city living less palatable. Bourne also predicted that a slow rate of growth in jobs in the service sector as well as the relocation of jobs to the suburbs would make inner-city living less attractive to professionals. Lastly, neighborhoods with the best potential for gentrification had already been exploited. It is obvious that these predictions did not come to fruition. Millennials, as well as aging baby-boomers, still find inner-city neighborhoods attractive while gentrification has spread from the neighborhoods adjacent to the CBD to a vast array of neighborhoods throughout the city.

Hackworth (2002) claims that the 1990 recession fundamentally changed the form of gentrification in U.S. cities using New York City as a case study. He identified four changes in this process: 1) the restructuring of the real-estate industry which has led to more involvement of corporate developers in the gentrification process at an earlier stage; 2) government (local and federal) has taken a greater role in gentrification than in previous decades; 3) opposition to the process of gentrification by disenfranchised individuals is less significant than it was in the past; and 4) gentrification has dispersed to other neighborhoods further from the Central Business District. Hackworth examined three gentrifying neighborhoods in New York City which began the gentrification process in the late 1980s to determine the effect of the recession on these neighborhoods in the very early stages of gentrification. The important point of this study is that gentrification did not decline or reverse during this recession, but that it was fundamentally different from former gentrification processes (as listed above). Lees and Bondi (1995) also examined the effect of the recession of the late 1980s/early 1990s on the gentrification process in two neighborhoods in New York City. However, their study examined neighborhoods much further along in the gentrification process than Hackworth’s study. In the Lower East Side, Manhattan, gentrification was

21 largely halted by the recession of the late 1980s/early 1990s as incumbent or potential residents were unable to afford residence in the neighborhood. Surprisingly, this is counter to how the gentrification process in this neighborhood began. During the recession of the early 1970s, the government sold assets in the city at low cost which initiated the gentrification process. The process was simply not sustainable with the economic downturn of the late 1980s. In Park Slope, Brooklyn, gentrification began in the 1950s and was aided by the investment of utility companies in the 1970s. The authors believe that the gentrification process slowed down because of saturation in the housing market, not from the economic crisis of the late 1980s/early 1990s. It is obvious from an examination of these two studies in New York

City that neighborhoods in different stages of the gentrification process are affected by recessions in different ways.

Wyly and Hammel (1999) examined whether the recession of the early 1990s negatively affected the gentrification process in eight U.S. cities (Boston, Chicago, Detroit, Milwaukee, Minneapolis-St.

Paul, Seattle and Washington, D.C.). The authors found no evidence that the recession affected gentrification. This was largely a result of housing finance methods. Banks were much more willing to lend in marginal neighborhoods or to borrowers who were higher risk after 1993 than they were prior to this time. The authors hypothesize that mortgage lending in gentrified areas should grow at least as fast as the city average. They found that between 1992 and 1997 that conventional home purchases in core gentry areas2 of the eight cities grew at 2.5 times that of the suburban rate while those in fringe gentry areas grew by 2.4 times that of suburbia. It appears that changes in the way loans were made and ideas concerning which areas of the city to invest in paved the way for further gentrification in cities in the U.S. as well as led to the economic crisis of 2008.

2 Wyly and Hammel discriminated between core and fringe gentrification in their study. They do this in three steps: 1. Examined scholarly research, planning documents, and the local press in the eight cities to obtain an inventory of gentrified tracts; 2. completed a block-by-block survey of tracts identified in step 1; 3. Used multi-variate discriminant analysis to determine the core versus fringe gentrified tracts using several socio-economic variables. Fringe gentrified tracts were identified as those where investment was modest, were spatially fragmented or were in early stages of gentrification.

There are relatively few studies that have examined the effect of the Great Recession of 2008 on the process of gentrification. This is not surprising given that it has only been a decade since this economic recession. While not specifically focused on gentrification, Solari’s recent study examined the prevalence of economic stability of census tracts in all metropolitan regions of the U.S. between 1970 and

2010 to determine how poor, middle-class, and affluent tracts fared during the Great Recession. Overall, about one-third of neighborhoods transitioned to some other type of neighborhood between 1970 and

2010. Affluent neighborhoods were a little less stable during 2000 to 2010, whereas poor neighborhoods experienced a decline during the first decade of the 21st century. Solari contributes this decline in stability to the effects of the Great Recession. Hartley and Kollinar (2014) examined the effect of the Great

Recession on gentrification in 59 of the largest cities in the U.S. by comparing income ratios between the suburbs and core areas in the city that experienced gentrification. Prior to the recession those cities that experienced the greatest growth in gentrification experienced the greatest growth in incomes relative to the suburbs. However, many of these cities experienced a decline in their incomes relative to the suburbs during the Great Recession suggesting that gentrified neighborhoods did not perform that well during this economic downturn. Lees (2009) examined the effect of the Great Recession on gentrification and found that gentrified areas in core cities weathered the economic downtown better than those in marginal neighborhoods that had just begun the gentrification process, suggesting that the momentum kept the process going through the downturn.

Coleman’s (2012) analysis of the effect that the sub-prime mortgage crisis of 2008 had on the gentrification process in 14 U.S. cities found that the recession had a minimal effect on the process of gentrification. Using 1-year ACS estimates to examine changes in major indicators of gentrification aggregated to the city level, the author concluded that gentrification in 13 of these cities was not hampered by the economic crisis. However, a process of de-gentrification occurred in Detroit due to the loss of white-collar jobs during the recession and the exodus of professionals from gentrifying neighborhoods. Unfortunately, the 1-year estimates are not available by census tract and thus the effect of

23 the economic crisis in specific neighborhoods of these 14 cities has not been uncovered and may account for the reason why Coleman came to a different conclusion concerning the effect of the Great Recession on gentrification than other researchers.

While the previously mentioned studies are hardly enough evidence to determine the effect of the

Great Recession on gentrification, it seems that this upheaval in the economy has slowed down or perhaps even reversed the gentrification process in many cities. This seems to run counter to the evidence that has been put forth for previous recessions where gentrification was not affected as much. One reason for this may be that the Great Recession was of a greater magnitude than previous recessions.

Gentrification and Displacement: Glass (1964) was the first social scientist to introduce the concept of gentrification. She noted that during the 1950s that middle/upper-income individuals were relocating to London’s East End where the housing stock (specifically Victorian architecture) was in good condition and where proximity to downtown London reduced commuting times to employment. Glass also noted that the arrival of the middle/upper income population was accompanied by a displacement of lower-income/working-class population from the East End. Several decades later, Atkinson (2000) showed that gentrification and displacement were still a concern in London’s neighborhoods. He used the Greater London Area

Longitudinal Surveys of 1981 and 1991 to determine the relationship between gentrification and displacement. The longitudinal survey tracks one percent of the population over time and gathers data on employment/education, as well as geographic location, of the tracked individual. Atkinson found that there was greater displacement of working-class/unemployed/elderly populations from gentrifying neighborhoods than there was an influx of these types of individuals into gentrifying neighborhoods suggesting that displacement was pervasive for lower-income individuals.

Early studies conducted in the U.S. overwhelmingly supported the idea that gentrification and displacement were causally linked. Sumka (1979) used the Annual Housing Survey Data for 1974-76 and

24 found that over 500,000 households were displaced each year in the U.S. due to gentrification. Sumka’s number is rather conservative in comparison to Le Gates and Hartman (1986) who reviewed 16 studies on gentrification and displacement during the late 1970s/early 1980s and estimated that at least 2.5 million individuals were displaced each year in the U.S. Marcuse (1986) used SRO (Single Resident Occupancy) data and city rent data and found that 10,000 to 40,000 households experienced gentrification related displacement each year in New York City alone. Schill and Nathan (1983) conducted a postal survey of renters in nine neighborhoods in five U.S. cities and found that 23 percent of movers were displaced due to gentrification. In contrast to the previously mentioned studies, Grier and Grier (1980) found that gentrification related displacement was only between 100-200 households each year regardless of city.

Likewise, Newman and Owen (1982) using Panel Study of Income Dynamics to examine displacement in the U.S. between 1970 and 1977 found that gentrification related displacement was relatively rare (about

1.0 percent of households being displaced). However, the low-income population (defined as those with high housing cost burdens, short occupancies, residence close to central cities of SMSAs, and those with low-education) were most susceptible to displacement in neighborhoods undergoing reinvestment.

Marcuse (1985) argued that displacement of lower-income populations through gentrification was a result of changes in the economy since the 1960s. The retraction of manufacturing employment had removed a steady source of income for less educated individuals while an increase in professional employment benefited highly educated individuals. This restructuring of the economy was accompanied by the concentration of professional employment in and around the CBD which altered the demand for land in proximity to the CBD. This change in demand for location close to the CBD and the higher incomes of professionals allowed these individuals to outbid their lower-income counterparts for properties close to the CBD. The result was displacement of lower-income population in what Marcuse referred to as ‘direct displacement’. Marcuse also introduced the idea of ‘exclusionary displacement’ which occurs when improvements in a neighborhood prevent potential new arrivals with less economic resources from moving in. Kirkland (2008) expanded upon Marcuse’s exclusionary displacement by

25 focusing foremost on race, instead of income. She found after reviewing numerous studies concerning race and approval of home loans that African-Americans were less likely to be approved for home loans than their white counterparts (controlled for economic factors) in neighborhoods that were gentrifying suggesting an added component to Marcuse’s exclusionary displacement argument; although one that is policy-based rather than economic based.

Chapple (2017) also expands upon Marcuse’s direct and exclusionary displacements with her argument that displacement in gentrifying areas does not only occur to low-income populations but also has the unintended effect of pricing middle-income populations out of the housing market. While her study focuses on gentrification related displacement in San Francisco, not representative of the housing market in the average U.S. city, she emphasizes that the restructuring of the economy has eroded the economic position of the middle-class and this is representative of changes occurring in most cities in the

U.S. One may even consider the problem more acute today than when Marcuse advanced his thesis regarding displacement in the 1980s. For Chapple, it is not necessarily affordable housing that is in short supply, but the retraction of jobs with a livable wage that is a more important factor to address.

Studies in the 1990s were relatively silent on gentrification and displacement and prompted Slater

(2006) to argue that ‘the demise of displacement as a defining feature of the process and as a research question’ (page 737) has evicted a critical perspective from gentrification research, which has resulted in detrimental effects for those displaced. By the 21st century, gentrification and displacement studies were numerous once again, but led to a major debate. Some researchers argued that low-income individuals are displaced from gentrifying neighborhoods because the cost of continuing to reside in the neighborhood becomes prohibitive (Sutton, 2018; Loukaitou-Sideris et. Al. 2017; Davidson and Lee, 2005; Kennedy and Leonard, 2001). However, these studies rely too heavily on census data and other aggregate data pertaining to economic, demographic, and housing at the city/neighborhood level to deduce that lower- income individuals are displaced due to gentrification.

One recent study that does not rely on aggregate data is Ding et. Al. (2016) who examined credit scores for 50,000 adult residents in Philadelphia between 2002-2014 to track individual moves throughout the city. The purpose was to determine if movement of low-income individuals was greater in gentrifying than in non-gentrifying tracts. Low-income residents of gentrifying tracts (10-15% moves per year) were

2-5 percentage points more likely to leave than their counterparts in non-gentrifying tracts. However, they were not significantly more likely to relocate to a lower-income tract than their counterparts in non- gentrifying tracts. Nor were low-income individuals more likely to leave gentrifying tracts during the

Great Recession than they were before the Recession. However, slightly higher mobility was observed between 2011 and 2014 and the authors attribute this to the stalling of economic progress in gentrifying tracts during the Great Recession. Once the economic crisis was over, low-income individuals were likely displaced from these tracts.

Other researchers find no evidence that gentrification displaces low-income individuals and instead suggest that these individuals benefit from additional job opportunities and amenities resulting from gentrification (Freeman, 2005; Freeman and Braconi, 2004; Ellen and O’Regan, 2011; Crowder and

South, 2005). These latter studies draw their data from American Housing Survey (AHS) data or databases designed to directly measure this effect in a specific city (for example New York City) allowing the researcher to track movement of individuals and their reasons for moving. The drawback of this methodology is that it is almost impossible to determine major changes occurring at the tract level. Using

AHS or city-specific databases, we can only get an idea of changes at the city or national level as the number of responses from any one census tract is very small.

Even researchers analyzing the same data source with different methodologies can come to different conclusions. Newman and Wyly (2006) used a mixed-methods approach as a follow-up to

Freeman and Braconi’s (2004) study on displacement in New York City using the New York City

Housing and Vacancy Survey (NYCHVS) in combination with interviews with community organizations.

Using the NYCHVS, Freeman and Braconi (2004) found that low-income individuals in seven

27 gentrifying neighborhoods in New York City had lower displacement rates than their counterparts in non- gentrifying neighborhoods. However, Newman and Wyly in their analysis of the same data set argued that the lower rates of displacement found by Freeman and Braconi was a result of their failure to count as displaced those who moved in with others (doubling-up), those who left metropolitan New York City, and those who ended up in a homeless shelter. Only by using qualitative in-depth interviews with community organizations were the authors able to document that displacement was more of an issue than Freeman and Braconi argued.

The previously mentioned studies focus more on income than race/ethnicity and thus exclude an analysis of racial/ethnic transformation in gentrifying neighborhoods. Jackson (2015) examined gentrification and racial displacement in Washington, D.C. between 1990 and 2010 and found, not surprisingly, that this displacement had led to a re-segregation of African-American population in

Washington, D.C. and less racial/ethnic diversity in both gentrifying and non-gentrifying tracts. This growing segregation and racial inequality occurred in other cities undergoing gentrification such as

Durham (De Marco and Hunt, 2018) and New York City (Chronopoulos, 2016). One solution to the problem of gentrification and displacement of low-income (mostly minority populations) might be to encourage African-American gentrifiers to invest in low-income African-American neighborhoods. Boyd

(2005) finds this solution too simplistic and not likely to yield results that differ from the influx of white middle-class into a gentrifying neighborhood. She examined Douglas/Grand Boulevard neighborhoods in

Chicago using in-depth interviews of African-American gentrifiers and the incumbent population. The interests of the middle-class (regardless of race) are often at odds to those of lower-income individuals.

Boyd feels that assuming African-American gentrifiers have no detrimental effects on a neighborhood might be too naïve and perhaps even dangerous as it gives the illusion that both African-American gentrifiers and low-income African-Americans have common interests.

Bates (2013) developed a typology of gentrification and racial displacement which she applied to

Portland, Oregon between 1990 and 2010 and can be helpful in analyzing gentrification and racial change

28 in other cities. Her typology specifically addresses at risk neighborhoods and the policies that can be developed to address displacement. The typology begins with three factors that relate to displacement: vulnerability to displacement, demographic changes in neighborhoods, and housing market appreciation.

Vulnerability to displacement examines the percentage of renters, the percentage of minorities, the educational levels, and the median incomes of a specific neighborhood. Those neighborhoods with a higher percentage of renters, a higher percentage of minorities, lower educational levels and lower median incomes than the city average are considered vulnerable to displacement. Demographic change has occurred if the increase in the white population, increase in homeownership, increase in college-educated population, and increase in median household incomes exceed those of the city average for the study period. Housing market appreciation has occurred in the neighborhood if the change in housing values over the study period exceed that of the city average. Bates identifies three stages to her typology: early, middle, and late gentrification. During early gentrification housing prices, incomes, educational levels, and whites increase but housing prices are still moderate. There is no displacement of population at this stage. Obviously, the best time to implement displacement policies is during the early stage of gentrification. By the mid-stage gentrification housing prices have increased, and demographic changes

(more whites, higher education levels, and higher incomes) have placed substantial pressure on the low- income population and has initiated the process of displacement. However, moderate priced housing is still available. Bates suggests targeted housing policies and economic development policies that support the low-income population. During late stage gentrification, housing prices are high and increasing and substantial displacement has occurred, but there is still a vulnerable population which is now in a precarious economic situation. Critical intervention is necessary at this point and the creation of affordable housing should be top priority.

One last consideration is necessary in the discussion of gentrification and displacement. Even if spatial displacement does not occur and some low-income individuals are able to remain in the gentrifying neighborhood, they may suffer from psychological/social displacement (Shaw and Hagemans,

2015; Davidson, 2009). This occurs when the political, social, and economic institutions that support low- income populations are removed from the neighborhood. The arrival of up-scale boutiques, coffee shops/expensive restaurants, and the construction of bike/walking trails are just some examples of changes that consider the gentrifiers desires ahead of the needs of the incumbent population. The political climate will also likely change as gentrifiers have the wherewithal to advance their political agendas, often at the expense of the original inhabitants of the neighborhood.

This section examined gentrification and displacement of low-income individuals. Studies in the

1970s and 1980s overwhelmingly found displacement of the low-income population from gentrification to be the norm. There was a lull in research concerning gentrification and displacement in the late

1980s/1990s. By the late 1990s, there was a major contention between those who maintained that gentrification caused displacement of low-income populations (Kennedy and Leonard, 2001) and those that maintained it did not (Freeman and Braconi, 2004; Ellen and O’Regan, 2011). The section ends with policy suggestions from Bates (2013) concerning gentrification and displacement as well as the acknowledgment that spatial displacement is not the only type of displacement that can occur in gentrifying neighborhoods. Social/psychological/political displacement may in some instances be more detrimental to populations who are able to maintain their physical location in gentrifying neighborhoods.

Conclusion: Five topics concerning gentrification were discussed in this literature review. First, neighborhood change studies focused on the suburbanization process which created new housing for middle-class individuals who sought to escape the disamenities of the city (e.g. pollution; congestion; crime). The exodus of the middle-class from the city left a substantial stock of high-quality housing which could then be inhabited by those with less economic resources. Since the 1960s, some of that same housing stock has been in high demand by the middle-class who have escaped the disamenities of the suburbs (e.g. long work commutes, lack of cultural activities, lack of pedestrian-friendly transport) and has led to

30 gentrification. The second section examined the classic theories of gentrification which focused on production and consumption processes; as well as alternatives to those theories. The third section examined the five waves (stages) of gentrification. Waves one and two (1950s-1970s) focused on using sweat-equity to rehabilitate select areas of some of the larger cities in the U.S. while wave three allowed investors to profit from revitalization begun by gentrifiers in the previous two waves. Wave four was in response to the housing crisis which relaxed lending standards and encouraged gentrification in marginal areas of the city (Lees, Wyly, and Slater; 2013). Wave five only began after the Great Recession and funnels gentrification-related investment to select areas of a city that have tourist potential or to properties which provide a high-yield investment opportunity (Aalbers, 2019). The effect of recessions on the gentrification process was examined in the fourth section. No definitive conclusions could be advanced regarding the impact of recessions on gentrification. The process appears to have different effects on different cities as well as neighborhoods at different stages of the gentrification process. Lastly, a consideration of gentrification’s effect on the displacement of low-income population is examined. This process has varying effects in different cities and may operate differently according to the stage of gentrification that a neighborhood is in at the time. Because displacement is such a disruptive force for people undergoing this process and has the potential to disrupt the functioning of communities, it will be examined in much greater detail in the following chapters of this thesis.

Chapter 3 METHODOLOGY

Introduction: This chapter examines the data sources used in the analysis as well as the methodology used to determine if gentrification and displacement has taken place at the tract level. The chapter begins with a discussion of census data and ACS data as these data provide working definitions for the gentrification and displacement process. Advantages and disadvantages of using census and ACS data are also discussed in this chapter. Using several previous studies on gentrification as a guide, a working definition of gentrification is provided using change in median household income; percentage change in population over 25 with at least a bachelor’s degree; change in median household income and change in median rent.

These variables are examined between 2000 and 2010 as well as 2010 and 2016. The city average for change in these variables serves as the benchmark to determine whether gentrification occurred during one or both time periods. Displacement is defined by using Taeuber and Taeuber’s criteria and occurs when the percentage composition of African-American population declines in a census tract over a time period as a result in declining numbers of African-Americans and increasing numbers of whites in a census tract. The statistical methodology introduces binary logistic regression as the method for determining whether displacement and gentrification are related. The second component of the analysis focuses on using Google Streetview to document changes in neighborhoods before, during, and after the

Great Recession. By using Hwang’s (2015) methodology, it is also possible to determine how far along in the process of gentrification a neighborhood was by 2016. The last part of the analysis focuses on discussions with Community Development Corporations (CDCs). These interviews provide a qualitative aspect to the previous quantitative (statistical) analysis. These discussions with community members familiar with economic and demographic change in the four selected neighborhoods attempt to uncover processes not readily accounted for quantitatively.

Data Sources:

This study will make use of the U.S. Census Bureau’s 2000 decennial census and the 2010 and

2016 American Community Surveys. The unit of analysis will be the census tract, which is the most used measurement of neighborhoods. Census tracts are designed to house between 1,000 and 8,000 residents.

When population increases beyond 8,000 residents, census tracts are subdivided whereas, when populations decrease well below 1,000 persons, several tracts are often combined. This is a serious problem for any neighborhood change researcher who needs to study changes between different time periods. Fortunately, this issue was addressed in 2000 when the Neighborhood Change Longitudinal

Database reconstructed census tracts so that researchers could have a constant boundary in which to measure neighborhood change going back to 1970. Not only did the NCLD reconfigure boundaries between census years, but it also adjusted the census data so that socio-economic characteristics were reapportioned among the census tracts that have been reconstructed. Without this matching of boundaries, measuring neighborhood change would be problematic. Brown University’s Longitudinal Tract Data Base

(LTDB) was constructed to similarly reconfigure tracts from 2000 to 2010.

A more pressing concern is the comparability of these data between 2000 and 2010. The decennial census (as of 2010) no longer collects detailed data on socio-economic variables at the census tract level. This is now accomplished through the American Community Survey. While the decennial census is a 1 in 6 sample of the American population every ten years on a specific date, the ACS collects detailed data on census tracts every year on a rolling basis. This process samples a small number of individuals each month over a five-year period to formulate estimates which are then averaged for the survey year. Unlike its counterpart the decennial census, it provides a more realistic picture of conditions as it covers a longer period of sampling and does not rely on what is occurring the day of the census. A limitation is that the sample number is much smaller than the decennial census and it has a larger margin of error than the census. These limitations are not enough to invalidate the study as the Census Bureau assures the user that these databases can be used to measure socio-economic change over time. One

33 caveat is that the ACS data should not overlap in years as it would be measuring the same variables. This is not a problem for this study in that the 2010 ACS covers one-year estimates from 2006-2010

(averaged) while the 2016 covers the average characteristics between 2012-2016.

Defining Gentrification: The term ‘gentrification’ was first introduced by Glass (1964) to describe the in-movement of higher income individuals into poor neighborhoods in London during the 1950s and early 1960s. While earlier studies seemed to conclude that gentrification resulted in displacement of poorer, often minority individuals (Lee, Spain, and Umberson, 1985; London et. Al. 1980), studies since the 1990s have been divided on the relationship between gentrification and displacement. Further confounding the issue is that there is no consensus on whether gentrification should consider racial change as part of the measurement.

For this study, racial change is not a criterion used to identify tracts that have undergone gentrification as the entire study rests on the premise that racial change is not a necessary component of gentrification.

Measuring Gentrification: The preferred method of measuring gentrification begins by identifying census tracts that are below the median household income for the reference city in the beginning of the study period (Hammel and Wyly, 1996; Vigdor, 2002). These are then classified as gentrifiable tracts. Applying this criterion for

Cincinnati yields:

Universe= 148 tracts

Not gentrifiable= these are tracts that were above the median household income for the City of

Cincinnati as of 2000. N=90

Gentrifiable= tracts below median household income for the City of Cincinnati as of 2000. N=58

Selecting a measure of gentrification is more difficult than selecting the pool of potential candidates for gentrification. Below are some examples of the measurement of gentrification.

Landis (2015)- Gentrifying tracts are tracts that increased by two deciles of income for the

reference city/metropolitan area from time 1 to time 2 (1990 to 2010).

Vigdor et. Al., (2002)- Gentrification is defined as a tract with an increase in percentage of age

25+ population with bachelor’s degree between the reference years that is at least 50% higher

than city average

Hammel and Wyly (1996)-Use 13 census variables representing socioeconomic, housing, and

total population for a baseline calculation of gentrification. The authors then identified

gentrifying tracts as ones with greater median income growth, greater median housing values,

greater growth in college educated population, and greater increase in rent than the city average.

The authors also use an extensive field study to document the gentrification process which

required a block to block study of gentrifying tracts in 23 cities in the early 1990s. Their

gentrification study required that structures in these gentrifying areas show physical evidence of

reinvestment and renovation.

Ding and Hwang’s (2016) definition of gentrification does not consist of an income variable but

is composed of a) growth in percentage of age 25+ population with bachelor’s degree or higher

that is above city median growth AND b) growth in median housing value above city median OR

c) growth in median rent above city average

I reject Landis’s definition of gentrification as it only considers income. This is problematic in that gentrifiers are often young adults who have not reached their full earning potential. Secondly, using

35 median household income discriminates against single-householders who may not make as much as dual earning households. Thirdly, educated persons such as artists often do not make as much as individuals in occupations that do not require advanced training. I do not use the Vigdor classification because, like

Landis, he only relies on one category for gentrification. While Ding and Hwang rely on two categories for classification, the absence of income would seem to eliminate some potential candidates for gentrification. I use a modified version of Hammel and Wyly’s definition that uses percentage change in income; percentage change in bachelor’s degree or higher; and/or percentage change in housing values/median rents that are above the city average for time being considered.

Justification for variables used in measuring gentrification:

1. Change in income- Individuals with higher incomes have the power to outbid those with lower

incomes and to inhabit a certain area.

2. Change in education- The transition to a post-industrial society requires higher educational levels.

A high percentage of highly-skilled jobs are found in the Central Business District or in

surrounding areas such as Universities or Hospitals (Florida, 2004). Thus, we expect

neighborhoods close to these areas to experience an influx of educated persons.

3. Change in Housing/Rent- According to Smith (1979) and Ley (1986), both proponents of

production and consumption theories of gentrification, gentrification results in increases to

median housing values and rents. For Smith, this occurs with reinvestment in disinvested areas of

the city (movement of capital to its best use) while for Ley, consumer choice results in higher

prices paid in gentrified neighborhoods.

Census tracts must have rates of growth for each of the three variables between time 1 and time 2 that exceed the city’s average for that variable in order to be considered gentrified. If a census tract experiences growth rates that are twice that of the city’s average on two variables, but falls below the city

36 average on one variable, the tract is also considered to have gentrified during that period of time. An example of this process could occur when rate of income growth is below the city average but a census tract is experiencing exceptionally high rates of growth in housing prices and college-educated individuals, but has yet to translate that into income growth.

The second part of the analysis borrows from Hammel and Wyly’s field study research; instead of canvasing neighborhoods, I use Google Earth Historical Aerial Photos to gain a general understanding of the urban landscape changes between 2000 and 2016 and then use Google Streetview to conduct an analysis of structural conditions in four tracts as of 2016 to determine their stage in the gentrification process (after Hwang, 2015; and Hammel and Wyly, 1996). This is discussed in a section below.

Measuring Displacement/Lack of Displacement: Displacement (from Taeuber and Taeuber, 1965) is defined as : 1) Percentage African-American in tract in beginning year (2000) is between 10 and 90 percent; 2) between beginning year (2000 for our purposes) and ending year (2010), African-American percentage of tract declines by at least 5.0 percent; and 3) the number of African-American declines and the number of whites increase between beginning

(2000) and ending year (2010). The process is then repeated between the 2010 and 2016 period. For the displacement process to have occurred, the following above conditions must all be met.

Obviously, displacement of the African-American population by whites has not occurred in all census tracts. Part of this study examines the gentrification process without displacement of the African-

American population. This will be undertaken by examining median incomes for African-American residents in 2000, 2010, and 2016 in comparison to their white counterparts in these tracts to determine if these characteristics between the two groups are comparable to the tracts where displacement occurred.

However, it is possible that displacement of poorer African-Americans by higher income African-

Americans is occurring in these tracts. I do not address this issue in this thesis, as my focus is foremost on understanding racial transition in gentrifying tracts.

Gentrification studies since the 1990s have also included a focus on Hispanic and Asian populations (Betancur, 2011). However, this thesis will not consider these populations in the models that are run to determine displacement. Cincinnati’s Asian and Hispanic populations are very small in comparison to other cities in the U.S. such as New York, Los Angeles, Chicago, and Washington, D.C.

However, racial/ethnic composition of the four tracts selected for further study on the relationship of gentrification and displacement will examine changes in the Asian and Hispanic populations if they are applicable. Cincinnati’s Asian population was only 1.5 and 1.8 percent of the City’s total population for

2000 and 2016, respectively. The Hispanic population grew from 1.3 percent of Cincinnati’s total city population to 3.2 percent between 2000 and 2010, but it still accounts for a very small percentage of the total population.

Hypotheses: Using the definitions of gentrification and displacement above, I follow Jackson’s (2016) methodology in his examination of the racial transformation of gentrifying tracts in Washington, D.C. between 1990 and 2010. He found that tracts that gentrified were likely to have undergone displacement of the African-American population in both 1990 to 2000 and 2000 to 2010. By holding poverty constant in his analysis, he was able to determine that displacement was more likely a result of gentrification than of factors related to poverty. Using Cincinnati as a case study, I replicate Jackson’s analysis by examining two periods: before the Great Recession (pre-2010) and the post-Recession period (post-2010) to 1) gauge the significance of the Great Recession on this process of gentrification and 2) to determine if the Great

Recession affected the African American population more than the white population. Several studies have noted that the Great Recession had a more negative economic impact on the African American population than the white population (Williams, Galster, and Verma, 2017; White, 2015; Hoynes et. Al, 2012).

Specific Questions relating to gentrification and racial transformation: 1. Did tracts that began the process of gentrification between 2000 and 2010 in the City of

Cincinnati have a greater chance of displacement of the African-American population than non-

gentrifying tracts by 2010?

2. Did tracts that began the process of gentrification between 2010 and 2016 in the City of

Cincinnati have a greater chance of displacement of the African-American population than non-

gentrifying tracts by 2016?

3. Did tracts that began the process of gentrification between 2000 and 2010 have a greater chance

of displacement of the black population than non-gentrifying tracts by 2016?

Statistical Methodology: Following Jackson (2016), I use logistic regression to determine if gentrifying tracts are more likely to experience displacement than non-gentrifying tracts. These regressions will give a log odds of the event (displacement) happening. Log odds above 1.0 indicate that gentrification is associated with displacement. A log odds below 1.0 indicates that displacement did not occur with the process of gentrification. I apply this procedure for the time periods 2000 to 2010, 2010 to 2016, and 2000 to 2016 to determine the relationship between gentrification and displacement.

After determining which tracts have undergone gentrification between 2000 and 2016 and which tracts have undergone displacement of the African-American population, I select four tracts for an in- depth examination of gentrification, racial change, and the effect of the Great Recession. The point of this selection is to pair two tracts that are as similar as possible in racial composition and socio-economic characteristics to isolate the effect that different policies/programs have had the displacement/lack of displacement of the African American population in those tracts. It is likely that census tracts contiguous to the Central Business District will experience a different type of gentrification than tracts gentrifying in the inner suburbs (Smith, 1979). Gentrification theory predicts that tracts closer to the Central Business

District will undergo reinvestment before tracts further from the CBD. There is a premium to residing closer to the CBD which relates to the availability of employment in white-collar occupations as well as

39 proximity to cultural amenities (London et. Al. 1980). Regarding Cincinnati, two of those tracts experienced displacement of the African-American population while two did not. To provide a fair comparison, I will examine two tracts on the outskirts of the central city and two tracts closer to the inner city. Three of those tracts had a mixed racial composition as of 2000 while one of the tracts were majority black in 2000 (defined as over 75% black). These tracts were further paired by median household income and homeownership rates as of 2000. See Table 3.1 below for further information.

Table 3.1. Characteristics of Selected Gentrified Tracts.

Median African- House- Gent. Displace Gent. Displace Own American hold 2010 2010 2016 2016 2000 2000 Income 2000 Tract 19 Walnut Hills yes yes Yes no 53.8 26,967 23.5% Tract 55 Madisonville no no Yes yes 77.7 28,424 43.4% Tract 74 Northside yes no Yes yes 45.0 25,932 37.0% Tract 84 College Hill no no Yes no 43.7 26,599 33.0% Source: American Factfinder. U.S. Census. 2017. Using Census Data (2000, 2006-2010 ACS and 2012-2016 ACS), the four tracts will be compared to assess socio-economic changes. This is a descriptive analysis of demographic/economic changes that occurred between 2000-2010 and 2010-2016. The City of Cincinnati will be used as a comparison. These indicators will include: median household income, percentage of population over 25 with a bachelor’s degree or higher, median housing prices, median rent, percentage of population living in owner-occupied housing, percentage of housing units that are vacant and population change.

Qualitative Methodology: Each neighborhood will undergo a review of planning documents to gain an understanding of the development of the neighborhood over the past several decades. These documents will be accessed through the City of Cincinnati Planning Website. These documents are published in different years and

40 likely focus on different aspects of economic development in each community, but they provide a comprehensive overview of the urban planning process.

To gain an understanding of the processes that operate in different neighborhoods and may affect gentrification and displacement, I contacted the Community Development Corporations (CDCs) for the requisite neighborhoods. These CDCs are identified through the City of Cincinnati Planning Commission website.

Questions for CDCs: 1. Can you pinpoint a year(s) when you first started to see changes in Tract x suggesting that economic reinvestment and/or revitalization was occurring? Where did the process begin within this tract? Be specific as possible. (Be sure to show map of Tract x in neighborhood y). 2. Do you know if this process of investment began in Tract x or began in a neighboring tract and diffused to tract x? Explain. 3. Can you elaborate on the how businesses, housing, and the streetscape is being revitalized in this tract? If not, please continue with generalizations concerning the entire neighborhood. 4.What effect did the Great Recession have on economic reinvestment/revitalization have on tract x?

5. Can you comment on the impact of private development agencies versus non- profit/governmental agencies in the reinvestment/revitalization process? How did the Great Recession impact the involvement of these agencies in development? 6. Has reinvestment/revitalization led to a change in the racial composition in tract x? If so, in what way? Have certain areas of the tract (or entire neighborhood) been more affected than others? 7. What do you attribute the change in racial composition (or lack of) to? Be specific. 8. Did the Great Recession differentially affect different racial/ethnic groups in your community? If so, how?

9. Can you comment on the changes in types of businesses or improvements in already established businesses in your community during the gentrification process? Can you comment on the effect of the Great Recession on the business community in your neighborhood?

The following are possible questions to be answered after data is collected. They will be answered by the author from interpretation of logistic regressions, analysis of census data, and discussions with the CDCs.

1. How was College Hill able to withstand the process of racial displacement? Is College Hill undergoing African-American gentrification? Why did College Hill only begin the gentrification process after 2010? Could this have been an adverse effect of the Great Recession? 2. Could access to employment and amenities in Downtown, the University of Cincinnati, and surrounding hospitals explain the gentrification and racial transformation in Walnut Hills between 2010 and 2016? How was Walnut Hills able to combat the displacement of the African-American population between 2010 and 2016? 3. How has economic revitalization and investment by both independent homeowners and private corporations between 2000 and 2016 affected the racial composition of Northside? 4. How has the economic revitalization of the business district in Madisonville affected the gentrification process and displacement of the African-American population?

Google Maps: After identifying the four tracts for the study I will use Google Earth to detect changes in the landscape at the tract level using a method developed by Hwang (2015). Google Earth allows the user to examine data from as early as 1984 and will allow me to track changes in the neighborhood that may not be recorded in neighborhood documents or may not be recalled by informants. The images can also be shown to the informants (from the CDCs) and information gathered regarding the gentrification process.

There is a drawback to these images. They are aerial photos and cannot be maneuvered into different profiles of the landscape. Secondly, some of the earlier images (prior to the early 2000s) are not of high quality. The major advantage is showing such large-scale transitions in the landscape such as demolition of derelict buildings and the construction of new buildings.

I seek to answer the question: How did the gentrification process unfold before, during, and after the Great Recession in each of the four tracts selected for further study (Madisonville, College Hill,

Walnut Hills, and Northside) and how can these data (landscape images) help to provide context to the process of displacement/lack of displacement in the selected communities?

Each of the four tracts will be examined for evidence of gentrification between 3 time-stamps:

1. Pre-recession= 2007

2. Recession= 2009

3. Post-Recession= 2016

A Deeper Analysis Using Google Streetview: The street audit was a popular research method used in the social sciences for examining the built environment in urban areas prior to the 21st century (Gaddis, 2018). It requires researchers to canvas an area, typically via walking or vehicle, and to document the phenomena of interest in that neighborhood.

This was accomplished with such methods as written notes, the interviewing of individuals in the neighborhood, and/or by photography. If the researcher was interested in change over time, historical data, if available, could be consulted to illustrate changes in the urban landscape. Hammel and Wyly

(1996) used the street audit approach to examine the gentrification process for 23 U.S. cities in 1995. This was a block by block survey of hundreds of tracts. Blocks are the smallest subdivisions of a census tract in which the census records data. They tend to be rectangular features that are bounded by streets or natural barriers such as streams, forests, or invisible features such as political boundaries or property lines

(U.S. Census). Hammel and Wyly then used socio-economic data from the census to construct a comparison between gentrification at a statistical level (census) and visual indicators of gentrification

(street audit). Unfortunately, many of the tracts that showed gentrification using census tracts were misclassified, illustrating that gentrification studies should include both qualitative and quantitative approaches. The drawback of Hammel and Wyly’s study is that it resulted in an exorbitant number of person hours to conduct these audits.

A new technique, known as the virtual audit using Google Streetview to examine the urban landscape has become popular in the last decade. This imagery is free to the public and can offset costs related to acquiring proprietary datasets and the time/expense of collecting data via a street audit. This imagery has been used in health studies, to aid in environmental remediation of urban landscapes, for interpreting demographic conditions of neighborhoods, and examining gentrification. While the imagery is qualitative, it can easily be converted to quantitative by creating indices of the presence/absence and

43 quantity of certain urban features that the researcher is interested in examining. The variety of research projects using Google Streetview is extensive. Google Streetview has also been used to examine landscape usage and to help in neighborhood greening projects (Li et. Al., 2017); and to determine indicators of neighborhood disorderliness which could indicate the feasibility of implementing aging-in- place strategies in certain neighborhoods according to the safety and infrastructure conditions of the neighborhood (Bader et. Al., 2017). Other researchers (Rundle et. Al. (2011) and Clark et et. Al. (2010) used the imagery to examine the urban landscape and indicators of disorder and how these can enhance or adversely affect the health of residents.

Several recent studies have expanded the Google Streetview applications to the process of gentrification which is the emphasis of this thesis. Naik, et. Al. (2017) used Google Streetview to examine changes in the urban landscape for five U.S. cities: Baltimore, Boston, Detroit, New York City, and Washington D.C. They retrieved 1.6 million street block images from the five cities and tracked urban change between 2007 and 2014. They found that blocks that were the most aesthetically pleasing/best condition were likely to experience positive change (interpreted as gentrification).

Hwang (2015) used Google Streetview to document changes to the urban landscape for tracts undergoing gentrification in Chicago between 1995 and 2014 with specific attention to how the Great

Recession affected neighborhood change. She used Hammel and Wyly’s (1996) data base for Chicago and their indicators of gentrification. The criteria to be included in the study were that the neighborhood had to have experienced disinvestment which was defined as median income for the census tract below the median for the city in 1960 (the beginning of Hammel and Wyly’s analysis). Secondly, structures within the census tract needed to show visible evidence of reinvestment. This included but was not limited to the following: structurally sound single or multi-unit buildings; reconstruction of steps, porches, windows and fences; repainting/ornamentation of structure or property; and a security system for the structure. Hwang then created a five -stage gentrification model which included disinvestment/decline; early gentrification; middle-stage gentrification; late stage gentrification; and class

44 turnover which was identified by the score on three criteria for gentrification (See table 2). These criteria consisted of: 1) a composition mix score which comprised two measures: a.) the condition of the preexisting structures which ranged from none of the structures being in adequate condition

(disinvestment) to all structures in good condition (turnover) and b) the number of new structures which ranged from none (disinvestment) to none to many in the turnover stage. The second category involved beautification which included new signage, improvements to the streetscape and landscape. The scale again ranged from none for disinvestment to many for class turnover. The last indicator was lack of disorder/decay and focused on the absence of litter, abandoned lots, and decaying structures (other than buildings). Each of the indicators was treated as a binary variable and received a 1 if present and 0 if absent. This process was carried out for each block face (one side of a block) in the census tract and was averaged to obtain a composite score which was divided by the number of block faces surveyed. Scores ranged from 1 to 0 indicating class turnover (+0.80) to disinvestment (below 0.50). These scores could then be checked to document the stage of gentrification for a specific census tract3. Quinn et. Al. (2016) used a similar methodology to Hwang (2015) to convert qualitative to quantitative data by examining indicators of physical disorder in New York City using Google Streetview. They were concerned with indicators of disorder on the mental and physical health of individuals. They assessed nine indicators of disorder by viewing 532 block faces in New York City using a yes/no response to the presence of these indictors. After creating a scale which ranked the nine indicators from weak disorder (presence of litter) to strong disorder (abandoned cars), the authors were able to determine an overall measure of disorder for each block.

Hwang and Sampson (2014) recognized that a detailed analysis using Google Streetview was not feasible for their study of 140 census tracts in Chicago to document the process of gentrification. Thus,

3 Hwang (2015) provides a detailed description of how to classify block faces by using Google Streetview in Appendix A. She provides numerous examples with photos and descriptors so that block faces can be evaluated on the three criteria for gentrification with as little subjectivity as possible. I will follow these procedures for my analysis.

45 they used a random sample of block faces. This methodology required ten block faces from at least four different blocks within each census tract. This allows a detailed analysis of a limited number of block faces from which generalizations can be drawn. My study modifies Hwang and Sampson’s methodology and uses a random number table to select blocks and block faces for analysis. Hwang’s (2015) study sampled Chicago tracts which typically had between 10 and 20 blocks that contained buildings. Blocks which primarily contained rivers, railroads, and vacant lots were excluded from Hwang’s analysis. I needed to adapt Hwang’s strategy for sampling blocks because of Chicago’s greater density of people and buildings per square mile than that of Cincinnati. The census tracts in my study range from 8 blocks

(Tract 84) to 90 blocks (Tract 55). While Tract 19 had 27 blocks, only 19 contained enough buildings to be included in the study. Much of the southern part of tract 19 is part of Eden Park. Five blocks were eliminated from tract 74 because they contained parts of streets, viaducts, riverfront or some other feature that had no buildings. I follow Hwang’s suggestion of sampling 10 block faces from at least 4 different blocks in a tract for tracts only in Tract 84. For Tract 19, I sample at least 20 block faces from 8 different blocks. Given that Tract 74 is approximately 3 times the number of blocks than that found in Hwang’s

Chicago study I include 30 block faces from at least 12 different blocks in tract 74. For Tract 55, I include at least 40 block faces from 18 different blocks to approximate the number of blocks sampled in the other three tracts. This methodology allows about twenty percent of blocks to be sampled in each of the tracts.

Tract 84 is oversampled (fifty percent of blocks) because it only has 8 blocks.

An example of the random number table is provided below. Blocks in each tract are keyed to the random number table. Numbers 94 and 89 are not usable as they do not correspond to a block segment for any of the tracts. tUsing the census 2010 block file from CAGIS, I was able to obtain the addresses of the block faces of the sampled streets (See Figures 3.2B-D for further information).

Table 3.2A. Segment from Random Number Table.

Example of Random Number Table Used for Tracts 19, 55, 74 and 84

20 17 42 01 72 33 94 55 89 65 58 60

74 49 04 27 56 49 11 63 77 79 90 31

Source: Simple-random sampling.png&exph=280&expw=424&q=random+number+table&simid=608019745423952250&select edIndex=0&ajaxhist=0

Table 3.2B. Number of Block Faces and Blocks Selected for Analysis.

Tract Sample 20 block faces selected from 8 of Tract 19 19 blocks 30 block faces selected from 12 of Tract 74 55 blocks 10 block faces selected from 4 of 8 Tract 84 blocks 42 block faces selected from 18 of Tract 55 90 blocks

Table 3.2C. Stage of Gentrification by using Three Indicator Variables (Hwang, 2015)

Disinvestment 0.0-0.49

Early/Middle Gentrification 0.50-0.64

Late Gentrification 0.65-0.80

Class Turnover 0.8 (Middle/Upper-class) Source: Hwang, 2015.

Table 3.2D. Hypothetical Scores for Stage of Gentrification in a Census Tract (Ravuri’s interpretation of Hwang, 2015) 0=not present; 1= present

Block Face Block Face Block Face Block Face A B C D 1.a. Built Environment 1 1 1 0 in Good Condition

47 b. Presence of New/Rehabilitated 0 0 1 0 Structures Average Score on 1 0.5 0.5 1 0 2. a. Efforts to 1 1 1 1 Discourage Disorder b. Personal Frontage 1 1 1 0 Beautification c. Vacant/Public Space 0 1 0 0 Beautification Average Score on 2 0.66 1 0.66 0.33 3. a. Lack of Litter 1 1 1 0 b. Lack of decaying 1 1 0 0 structures c. Lack of unkempt 0 1 1 0 vacant lots Average Score on 3 0.66 1 0.66 0 Average of 1,2, and 3 0.61 0.83 0.77 0.11 0.58 (early Average for Tract gentrificati on)

A Walk-Through the application of Hwang’s (2015) Gentrification Index:

The following describes in detail how gentrification indices were derived using Google Streetview and Hwang’s (2015) gentrification criteria for each of the block faces in the analysis. Four examples are provided, one for each of the Tracts that were examined in this analysis.

Formula for the Gentrification Index (GI): 푃1 표푟 푁1 + (퐵1 + 퐵2 + 퐵3)/3 + (퐷1 + 퐷2 + 퐷3)/3 = 퐺퐼푟푎푤 퐺퐼푟푎푤 = 퐺퐼 3 Step 1: P1: This designation determines the condition of the block face for structures that pre-date 2000. Since the study begins in 2000, it was necessary to indicate a base year for the condition of structures. This is a judgment call on the part of the interpreter. It does not require authentication of dates of buildings or other components of the built environment. To score a 1 on this indicator, the block face must appear in good condition which means that at least 75 percent of the buildings are such. For single-family residential units, the dwelling must be able to house two adults with children comfortably to score a 1 on this measure. Some indicators on this measure include: -absence of peeling paint, no structural damage, no blemishes on stone/brick/siding -porches, doors, signage (if business), walkways, should be in good condition This score is either a 1 or a 0 (present or absent). If the P1 is 0 move to N1; otherwise, move to B1.

N1: This is a sum of the measures for N2-N5. Note: N1 is only included in the equation if P1 receives a value of 0. However, if the pre-2000 built environment is in good shape and there is also new development on the block face, the score is taken from P1 and N1 is not considered. In other words, there is no double-counting in this index. N2: New signs or structures controlling traffic are present. This indicates public investment on the block face. Score 1 if present, 0 if absent. N3: New public courtesies present. This could include bus stops, trash cans, bike racks, etc. Score 1 if present, 0 if absent. N4: New large-scale developments on block face. Score 1 if present, 0 if absent.

N5: New Residential/commercial units are for sale or lease on the block or are under construction. Score 1 if present, 0 if absent.

Step 2: Consider each of the B and D indicators in the gentrification index. B1: Signs discouraging disorder are present along this block face. Score 1 if present, 0 if absent. B2: Beautification of personal frontage. Score 1 if present, 0 if absent. B3: Vacant area/public street frontage beautification. Score 1 if present, 0 if absent. D1: Residential blocks lacking physical disorder (garbage). Score 1 if present, 0 if absent. D2: Lacking unkempt vacant areas/public street frontage. Score 1 if present, 0 if absent. D3: Lack of boarded-up structures, abandoned or deteriorated structures. Score 1 if present, 0 if absent.

Example 1. Block Face along Langland Street (Block 3007) in Tract 74 (Northside). 0+ (0+0+0)/3 + (1+1+0)/3 = 0.66 (GIraw) 0.66/3 = 0.22 (GI)

Figure 3.1 Block Face along Langland Street in Tract 74.

Source: Google Streetview, 2007.

P1: This block face was coded as 0 indicating that its 2007 built environment was not in good condition. Note the boarded front door on the house to the left. The windows are in pretty good condition suggesting that the house may not be abandoned, but that residents enter the house through an alternate entrance. The satellite dish also indicates that there may be residents present. The house to the right has peeling paint and the front extension appears to have been sealed so that entrance cannot be obtained through those windows. The air conditioning units and the satellite dish suggest that the house is still occupied. Since P1 was 0, move to N1 analysis. N1: There were no new developments/renovations along this block face and thus it received a 0 on this trait. This refers to the condition in 2007, which requires the interpreter to infer prior conditions of the built environment from the imagery. This is not a comparison from 2007 to the 2016 end date of the study. B1: There were no signs discouraging disorder along this block face. This was coded as 0. B2: Beautification of personal frontage received a 0. The grass has been mown but a wooden structure appears in front of the house on the left. It appears to be part of the fence that has a few slats still in place. B3: Vacant area/public street frontage beautification. There was no attempt at public street frontage beautification. Note the cracked sidewalks and the litter. This score was a 0. D1: Residential blocks lacking physical disorder (garbage). While there was litter on the ground, this level of disorder did not rise to the level defined as garbage. This block face scored a 1 on this measure. D2: Lacking unkempt vacant areas/public street frontage. There were vacant lots on this block face, but this did not rise to the level of disorder. This was coded as a 1. D3: Lack of boarded-up structures, abandoned or deteriorated structures. Since I was unsure of whether the house on the left was occupied because of the boarded- up door, this block face received a 0 on this score.

Conclusion: This chapter introduced the data to be used in the analysis and the methodology for the study. The first part of the study uses census data and statistical analysis to examine the gentrification process in

Cincinnati by neighborhood to determine which census tracts gentrified between 2000 and 2016. This is followed by an analysis of whether gentrification and displacement of the African-American population occurred between 2000 and 2015. In the second part of the methodology, I turn to a qualitative analysis that examines how gentrification and displacement or lack of displacement of African-American population occurred in the four neighborhoods chosen for greater examination by interviewing individuals in the community who have knowledge of these processes. The third component turns to a mixture of

52 qualitative and quantitative methodology by using Google Streetview. While examining satellite images for evidence of gentrification is largely a subjective process, using Hwang’s methodology makes it more objective and quantifiable.

Chapter 4 DETERMINATION OF GENTRIFICATION AND DISPLACEMENT IN CINCINNATI, 2000-2016.

Introduction: This chapter focuses on the spatial evolution of the gentrification process in Cincinnati between 2000 -

2016 and its relationship to the displacement of the African-American population. The chapter begins with a discussion of the spatial distribution of median household income in the City of Cincinnati by deciles in 2000 as this is the criterion used to determine which tracts had potential for gentrification between 2000 and 2016. Secondly, the spatial distribution of tracts that underwent gentrification over the study period is discussed and linked to changes in median household income from 2000 to 2010 and 2010 to 2016. Thirdly, the displacement of the African-American population between 2000-10 and 2010-16 is mapped and discussed. The fourth component of the chapter uses logistic regression to determine if there is a statistical relationship between gentrification and the displacement of the African-American population. The chapter concludes with a discussion of the changing spatial distribution of the African-

American population between 2000 and 2016 and its causes and implications.

Median Income and Gentrification in Cincinnati: Figures 4.1A and B show maps of Cincinnati by neighborhood and tracts, respectively, for the reader’s reference. Also included is a map of the highway system in Cincinnati for the reader’s reference (Figure

4.1C). Figure 4.2A displays deciles of median household income for the City of Cincinnati as of 2000.

This decile ranking system gauges a tract’s economic performance against all other tracts in the city. The first decile includes the poorest ten percent of tracts within the city while the tenth decile includes the wealthiest ten percent. The highest median household incomes in the City of Cincinnati in 2000 were found in the eastern part of the city while the lowest incomes were found in tracts adjacent to downtown and several tracts in the west. This spatial distribution of income has been relatively stable for the latter half of the twentieth-century (City of Cincinnati, 2012). Economic disinvestment since the 1940s left a concentration of the poor in the inner-city while agglomeration of industry in the western part of the city

54 made this area less attractive to settlement by the middle-class/wealthy. Conversely, the suburbs and the eastern part of the city continued to experience economic investment as these areas had attractive housing stock, good transport connections, and a pleasant natural environment; all important components in attracting the middle-class/wealthy.

While we can expect a level of stability in the economic ranking of these tracts over time, we cannot assume that no change in ranking will occur. This is a result of three factors: the overall national/international economic system in which the city is a component; residential preferences within the city of individuals by economic status; and the investment/disinvestment decisions of private and public agencies in certain areas of the city. At the national/international economic level, economic restructuring over the past half-century has had a major effect on employment in different sectors of the economy. Specifically, manufacturing employment, a main-stay of the economies of many cities until the

1970s, has declined substantially while high-wage professional jobs have increased (Florida, 2005).

Because individuals working in different sectors of the economy inhabit different areas of the city, changes in the availability of work in different sectors of the economy can have a profound effect on the economic ranking of a tract over time, even when no in-migration or out-migration of residents from a tract occurs.

The second factor affecting economic change at the tract level is related to the first factor and concerns the residential preferences of individuals. Tracts with different amenities or disamenities may prove differentially attractive to certain economic groups and attract higher percentages of their workforce in different sectors of the economy, which would influence the ranking of those tracts within the overall system. For instance, tracts close to the CBD may benefit from the arrival of higher-income individuals, thus increasing their ranking in the system. Conversely, individuals on long-term unemployment, retirees on fixed-income, or those reliant on the welfare system are likely forced into more affordable tracts leading to a decline in the decile ranking of that tract. Lastly, public and private investment would also influence changes in median income as improvements in housing and other

55 infrastructure may attract higher income individuals to certain areas or force lower-income individuals into disinvested areas of the city.

Figure 4.1A. Map of Cincinnati Neighborhoods.

Figure 4.1B. Map of Census Tracts in City of Cincinnati.

Figure 4.1C. Highways and Case Study Neighborhoods in City of Cincinnati.

Figure 4.2A. City of Cincinnati, Median Household Income by Deciles, 2000.

Figure 4.2B. City of Cincinnati, Median Household Income by Deciles, 2010.

Figure 4.2C. City of Cincinnati, Median Household Income by Deciles, 2016.

Figure 4.3 displays tracts that have gentrified in the City of Cincinnati between 2000 and 2016.

There were 87 tracts that were non-gentrifiable, meaning that those tracts had median household incomes above the city average in 2000. Non-gentrifiable tracts were located primarily in the eastern part of the city and the inner-ring suburbs of the north and west. Of note is the paucity of tracts in the western half of the city that were potentially gentrifiable but did not undergo gentrification after 2000. Gentrification related investment focuses on cultural/recreational amenities, transportation connections, and proximity to downtown professional opportunities, none of which are available to any extent in the West.

Figure 4.3.Gentrification of Census Tracts in Cincinnati, 2000-2016.

The spatial distribution of the thirteen tracts that gentrified between 2000 and 2010 indicates that the majority were located close to downtown (Lower Price Hill, Camp Washington, Over-the-Rhine,

Corryville, and Walnut Hills). These tracts provide access to both downtown and the uptown area, including the University of Cincinnati and multiple Hospital complexes. They are densely settled tracts with walkable neighborhoods (City of Cincinnati, 2012) and located on public transport lines. The housing stock in these tracts consists of mostly attached/semi-detached rowhouses and a building inventory with local historic districts/national register historical districts which are not only desirable for their aesthetic qualities but are eligible for government funding (City of Cincinnati, 2012). Madisonville and Carthage in the northern inner-suburbs and East Price Hill in the west also experienced gentrification between 2000 and 2010. East Price Hill has a denser urban development, better transportation connections, and is closer to downtown than its counterpart neighborhoods in the west, which likely allowed it to benefit from gentrification related development (City of Cincinnati, 2012).

Between 2010 and 2016, a further 14 tracts gentrified. Several of these tracts were contiguous to already gentrified tracts surrounding the CBD. The in-fill of gentrification is apparent in Downtown,

Over-the-Rhine, Mt. Auburn, and the West End and provides a corridor of gentrification linking

Downtown to the University of Cincinnati and the Hospital Complexes. Four out-lying tracts (College

Hill, Roselawn, Madisonville, and Linwood) also experienced gentrification between 2010 and 2016. This pattern of gentrification is in line with Smith’s prediction that (1979) gentrification would begin closer to the CBD and then expand further away as more centralized areas become too expensive. This would also likely explain why Madisonville, Carthage, and East Price Hill experienced the early round of gentrification between 2000 and 2010.

Returning to the median household income maps (Figures 4.2A-C), it is obvious that a reshuffling of the economic profile of census tracts within Cincinnati has occurred between 2000 and 2016. While 15 tracts in the inner-ring suburbs of the west had median incomes above the city average in 2000, by 2016, this number declined to 9. The dividing line between east and west Cincinnati in this thesis will follow the

Interstate 74 boundary (east-west) connecting with the Mill Creek Valley/I75 corridor (north-south).

Conversely, note the pathway from downtown (tract 265) eastward along the Ohio River to Hyde Park that moved into the above average median income category by 2016. Also note the growth of higher income tracts in the Clifton/Northside/Spring Grove area by 2016. Much of this may be a result of the professional sector (University/Hospitals) that was less likely to have experienced an economic downturn during the Great Recession. It is likely that substantial economic investment during the 21st century has flowed into those tracts between Hyde Park and Downtown attracting higher-income individuals. Figure

4.4 displays the percentage of labor force by census tract employed in professions as of 2016. A simple correlation analysis between percentage of labor force employed in professions and median household income revealed that median household incomes are higher in tracts with a higher percentage of the labor force in professions. However, the correlation is not significant. A plausible explanation for this insignificant relationship may be a result of the characteristics of the professionals. Gentrification studies have indicated the preference of young professionals for residences in cities and it is likely that some of these individuals have reached their full earning power (Moos, 2016). Another national trend is preference of many professionals to live alone; thus, eliminating the potential for a two-income household and decreasing median household income. According to HUD (2012), The Cincinnati Metropolitan Area experienced a 26 percent increase in education and health service jobs and a 10 percent increase in professional and business services between 2000 and 2012. It is likely that this favored those tracts with a high percentage of labor force in professions, accelerating their performance in economic ranking while diminishing that of those tracts with lower percentage of professions. Note that the lowest median household incomes expanded in the western part of the city. This has traditionally been a middle-income to lower-income area (City of Cincinnati, 2012) and if it is representative of middle/lower income households nationally, it would have fared poorly during the Great Recession. While only three tracts in the Western half of the city were below the third income decile in 2000, this increased to nine by 2016.

However, it is unlikely that the sole cause of this spatial transformation of income was a result of economic changes of the resident population over those sixteen years. It is likely that selective internal migration occurred as those with greater economic resources settled in the east and the northern parts of

66 the City while those with lesser economic resources were forced to locate or relocate to tracts in the West.

One must also note the presence of public housing in the West (South Cumminsville, Westwood, North and South Fairmount, and East and West Price Hill) which would decrease average household incomes for the tract (City of Cincinnati, 2012).

Figure 4.4. Percentage of Labor Force in Professions, Cincinnati, 2016.

The Displacement of the African-American Population: Figures 4.5A and 4.5B display displacement of the African-American population between 2000-2010 and 2000-2016. Recall that displacement occurs when the percentage composition of African-American population of a tract declines by at least 5.0 percent with a corresponding decline in the number of

African-Americans and an increase in the number of whites. Twelve tracts experienced displacement between 2000 and 2010. These tracts were concentrated around Cincinnati’s two largest employment centers: Downton and Uptown. Tract 55 (Madisonville) was the only tract in the inner-ring suburbs that experienced displacement. By 2010-2016, fifteen tracts had experienced displacement of the African-

American population. Six of the tracts that experienced displacement between 2000 and 2010 continued to do so between 2010 and 2016 (Tracts 9,10,22,23,32, and 55). Thus, fifty percent of tracts with displacement between 2000 and 2010 experienced continued displacement from 2010 to 2016. The remaining nine tracts only experienced displacement between 2010 and 2016. While displacement was still concentrated around Downtown and Uptown, inner-ring suburbs had five tracts that experienced displacement, suggesting that the gentrification process had expanded after 2000 and may be related to the displacement of the African-American population in some of these tracts. The question that arises is whether a statistical association between gentrification and displacement of the African-American population has occurred in the short-term (2000-10 and 2010-16) and/or the long-term (2000-16). That question will be addressed in the next section.

Figure 4.5A. Displacement of African-American Population in Cincinnati by Tract, 2000-2010.

Figure 4.5B. Displacement of African-American Population in Cincinnati by Tract, 2010-2016.

Table 4.1 shows potentially gentrifiable tracts in the City of Cincinnati and whether gentrification and/or displacement occurred between 2000-10 and 2010-16. Table 4.2 displays the change in African-

American population for gentrifying and non-gentrifying tracts for 2000-10 and 2010-16. Three tracts experienced mass displacement by 2016, defined as a percentage reduction of African-American population by half of its 2000 population:. Tracts 9 (Over-the-Rhine), 32 (Corryville), and 74 (Northside).

Tract 9 went from 71.6 percent African-American in 2000 to only 30.3 percent in 2016. Tract 32 went from 52.7 percent African-American to 23.6 percent between 2000 and 2016; while Tract 74’s African-

American population declined from 45.0 percent in 2000 to 21.3 percent by 2016. It is important to note that gentrification for each of these tracts had begun in 2000 and it took several years for the process of displacement to occur. This is evident in that most of the reduction in African-American population occurred between 2010 and 2016 for tracts 9 and 74. The African-American population may have been able to withstand the early pressures of gentrification prior to the Great Recession, but the combined effects of gentrification and the Great Recession probably adversely affected the economic situation of

African-Americans in these census tracts and led to displacement. There is some data to support this hypothesis concerning the gentrification, recessions, and displacement. Lees (2009) found in her study of gentrification and the Great Recession in New York City that neighborhoods further along in the gentrification process fared better than neighborhoods that had just begun the gentrification process.

Thus, pressure from gentrifiers in newly gentrifying neighborhoods of New York did not lead to as much displacement of the incumbent population. The rapid displacement of African-American population in tract 32 is likely a result of investments made by the adjacent University of Cincinnati for student housing and business revitalization around the University’s main campus (City of Cincinnati, 2012). Most of the displacement in tract 32 occurred between 2000 and 2010 (going from 52.7 to 28.7 percent African-

American).

Table 4.1. Potentially Gentrifiable Tracts in Cincinnati, Gentrification and Displacement 2000-2016

Gentrified 2000- Gentrified 2010- Gentrified Displacement Displacement Tract 10 16 2000-16 2000-10 2010-16

2 Yes No n/a No No 7 Yes Yes n/a No No 9 Yes Yes n/a Yes Yes 10 Yes Yes n/a Yes Yes 11 No n/a Yes Yes No 16 No n/a No No Yes 17 Yes Yes n/a No Yes 19 Yes No n/a Yes No 22 No n/a No Yes Yes 23 No n/a Yes Yes Yes 25 No n/a No No No 26 No n/a No Yes No 28 Yes Yes n/a No No 29 No n/a No No No 30 No n/a No Yes No 32 Yes Yes n/a Yes Yes 33 No n/a Yes No Yes 36 No n/a No No No 37 No n/a Yes Yes No 38 No n/a No No No 39 No n/a No No No 47.02 No n/a Yes No No 55 No n/a Yes Yes Yes 61 Yes No n/a No No 63 No n/a No No No 66 No n/a Yes No No 68 No n/a No No No 69 No n/a Yes No No 72 No n/a No No No 74 Yes Yes n/a No Yes 77 No n/a No No No 78 No n/a No No No 80 No n/a No No No 82.02 No n/a No No No 84 No n/a Yes No No 85.01 No n/a No No No 85.02 No n/a No Yes No 86.01 No n/a No No No

88 No n/a No No No 92 No n/a No No No 93 No n/a No No No 94 Yes No n/a No No 95 No n/a No No Yes 97 No n/a No No No 100.02 No n/a No No No 108 Yes No n/a No Yes 110 No n/a Yes No No 254.01 No n/a No No No 257 No n/a Yes No No 263 Yes No n/a No No 264 No n/a Yes No No 265 No n/a Yes No Yes 267 No n/a No No No 269 No n/a Yes No Yes 270 No n/a No No No 271 No n/a No No Yes 272 No n/a No No No 274 No n/a No No No *n/a refers to not applicable

Source: U.S. Census 2000 and ACS 2010 and 2016.

Table 4.2. Tracts that Underwent Displacement of African-American Population and Gentrification between 2000 and 2016. Total Tract Number and Percent Percent Percent Percent Neighborhood AA 2000 AA 2010 AA 2016 Change 2000-16 9 (OTR) 71.6 64.1 30.3 -41.3 10 (OTR) 68.9 55 35.6 -33.3 11 (Pendleton) 83.8 75.4 69.1 -14.7 16 (OTR) 84.2 85.4 69 -15.2 17 (OTR) 82.2 83.4 72.2 -10 22 (Mt. Auburn) 85.6 74.2 66 -19.6 23 (Mt. Auburn) 82.2 74.5 64.3 -17.9 32 (Corryville) 52.7 28.7 23.6 -29.1 33 (Corryville) 47.7 43.1 28.2 -19.5 55 (Madisonville) 78.8 68.3 53 -25.8 74 (Northside) 45 40.4 21.3 -23.7 108 (Madisonville) 57.8 53.3 37.4 -20.4 265 (Downtown) 46.6 41 32.5 -14.1 269 (West End) 89.8 87.6 79.3 -10.5 271 (West End) 75.7 84.4 76.6 +0.9*

* The West End tracts are over 75.0 percent African-American and are some of the poorest tracts in the city. Gentrification in the surrounding tracts of OTR and Downtown have likely caused some displacement of the population. Note the decline in the African-American population between 2000-10 for tract 271, the decline in tract 269 between 2010-16, and the increase for tract 271 between 2010-16. It may be that poor African-Americans in these two tracts went through a series of displacements during this time but largely remained within the boundaries of these two tracts.

Source: U.S. Census.

Logistic Regression: Table 4.3 shows the results from the logistic regressions. Recall that logistic regression measures the relationship between two variables. By using displacement as the dependent variable in logistic regression, we can predict the effect that gentrification (independent variable) will have on displacement.

We cannot discern that gentrification does or does not cause displacement (causation), but we can begin to study the relationship between these two events (correlation) and attempt to uncover the causative factors of displacement. As mentioned previously, thirteen tracts gentrified between 2000 and 2010

(Figure 4.3) while twelve tracts experienced displacement (Figure 4.5A). However, only 4 tracts experienced both gentrification and displacement during this time period. Thus, 30.8 percent (4/13 tracts) of the gentrifying tracts experienced displacement while 17.7 percent (8/45) of potentially gentrifiable tracts (those that met the conditions for gentrification but did not undergo it) experienced displacement in this time period.

While there was a reshuffling of the racial composition of Cincinnati’s census tracts between

2000 and 2010, there is not significant evidence to suggest that gentrification and displacement are associated at the tract level. The logistic regression results displayed a 1.8 odds ratio, meaning that there was an 80% greater likelihood of displacement occurring in a gentrifying tract between 2000 and 2010 versus a non-gentrifying tract. While this seems like it should be significant, it fails to meet statistical significance at the 0.1 level. Between 2010 and 2016, a further fourteen tracts gentrified. The results from this logistic regression (also 1.8) show there is still no significant relationship between gentrification and displacement. During this time, six tracts experienced both gentrification and displacement (42.9%), while 9 tracts experienced displacement (20.5%), but no gentrification4.

It is clear from these short-term results that there are processes other than gentrification that are related to the displacement of the African-American population, as non-gentrifying tracts also experienced some displacement of the African-American population. Recall again that displacement in these tracts has resulted in the numerical growth of the white population and the numerical decline of the

African-American population resulting in a 5.0 percent or greater decline in African-American population. Thus, whites were displacing African-Americans in several of these tracts but the factors that

4 Poverty is used as a control variable in the logistic regressions and follows Jackson’s (2015) methodology for gentrification and African-American displacement in Washington, D.C. There is no relationship between poverty and displacement in this analysis for Cincinnati (evidenced by the 1.0 odds between poverty and displacement). Research as documented that high poverty tracts are often likely less attractive to potential gentrifiers than tracts with moderate levels of poverty (Kinder Institute of Urban Research, 2018; Landis, 2015) making displacement less likely to occur. Neither are tracts with low percentages of poverty likely to experience displacement as the African- American population which has settled in these higher-income tracts would be better equipped with economic resources to remain in the tract. Conversely, tracts with moderate poverty rates may have the highest rates of displacement and thus it may be concluded that the relationship between poverty and displacement is non-linear.

76 lead to gentrification have not occurred: increased educational levels, increased median household incomes, and increased median house value/rent that exceed the city average. One possibility is that lower-income whites experienced displacement from gentrifying tracts and moved into poorer tracts thus displacing some of the African-American population. Much of the middle-class/low-income white population in Cincinnati is of Appalachian origin (Maloney and Auffrey, 2013) and is concentrated in the western half of the city (for example East Price Hill, Sedamsville, and Camp Washington). The white

Appalachian population has traditionally had lower levels of educational attainment than the non-

Appalachian white population and has largely worked in manufacturing, making their socio-economic status more akin to the African-American population than non-Appalachian whites. These similar socio- economic characteristics between white Appalachians and African-Americans may result in competition for affordable housing in certain neighborhoods. Furthermore, it is likely that the Great Recession negatively affected low-income African-Americans and white Appalachians to a greater degree than their wealthier white counterparts (Nakajima, 2013). Unfortunately, this is supposition as this project does not test for this scenario.

Before concluding that gentrification and displacement are unrelated it is necessary to statistically test this relationship over the long-term (2000-2016). The relationship is significant at the 0.05 level with the odds of African-American displacement 4.6 times higher for tracts that underwent the gentrification process anytime between 2000 and 2016 than tracts that did not undergo gentrification. Eleven out of 26

(42.3%) gentrifying tracts experienced displacement while only 4 out of 32 (12.5%) potentially gentrifiable tracts experienced displacement. These results show that displacement and gentrification are linked, but it does not indicate that every tract has experienced displacement. In the next chapter, an examination of tracts that underwent gentrification and displacement, as well as those that experienced gentrification and no displacement, will be analyzed to determine possible reasons for these differences.

These results indicate that African-American displacement is more likely to occur as the process of gentrification intensifies (Gale, 1979). Table 2 shows tracts that experienced gentrification and displacement of African-American population between 2000 and 2016. Six tracts (9,10,17,28, 32, and 74)

77 began the gentrification process and experienced displacement during the 2010 to 2016 timeframe. Only tract 32 experienced greater displacement during 2000-2010 than 2010-2016. As previously stated, this is likely a result of the investment of the University of Cincinnati in this tract. It is likely that early gentrifiers use ‘sweat equity’ to improve their dwellings and their neighborhoods. These tend to be individuals who are risk takers: the young, singles, artists, or gay individuals. They are likely to find vacant housing in good condition at affordable prices. During this stage of gentrification, banks and developers are less likely to invest in the neighborhood because the value of the neighborhood has not yet been established. As the gentrification process continues, middle and upper-income individuals will be attracted to the neighborhood. These individuals are good risks for banks and investment will flow into the neighborhood. Developers will likely expand on this and provide additional housing and commercial enterprises. Local governments will seek to appease middle/higher income individuals and provide other types of services and infrastructure as these individuals will add revenue to the city tax base. These improvements will increase the cost of living while affordable housing will likely become less available.

By this stage of the gentrification process, most of the vacant housing in good condition will have been utilized. It is no surprise that the cumulative effect of the gentrification process should lead to the displacement of low-income population by 2016.

Table 4.3. Short-Term and Long-Term Logistic Regression Results for Gentrification and Displacement of African-American Population in Cincinnati, 2000-2016. Exp B S.E. Wald Significance (B) Gentrify 2000-10 0.604 0.734 0.678 0.41 1.83 Poverty 2000 0.028 0.022 1.592 0.207 1.029 Constant -2.525 0.916 7.598 0.006 0.08 Nagelkerke= 0.068

Gentrify 2010-16 0.603 0.667 0.818 0.366 1.828 Poverty 2010 0.001 0.019 0.003 0.959 1.001 Constant -1.233 0.835 2.182 0.14 0.291 Nagelkerke=0.021

Gentrify 2000-16 1.538 0.664 5.36 0.021 4.655 Poverty 2010 0.005 0.019 0.074 0.786 1.005

Constant -2.119 0.948 4.996 0.025 0.12

Changing Distribution of the African-American Population and its Relationship to Poverty, 2000- 16: Figures 4.6A-C display the percentage of the population that is African-America by census tract in

2000 for the City of Cincinnati. Tracts with at least 75 percent African-American were concentrated in the central basin and extending eastward to Roselawn. Three tracts in the west and an additional 8 tracts located in the inner-ring suburbs of the north and east had at least 75 percent African-American population. Conversely, a concentration of tracts in the east and nine tracts at the interface of the western suburbs had less than 10 percent African-American population. Tract 72 (Clifton) and Tract 265 (Mt.

Adams) existed as islands of very low African-American population (less than 10 percent).

By 2016, the West End and Mt. Auburn neighborhoods had lost a substantial percentage of

African-American population. It is likely that the gentrification process had spread from Over-the-Rhine and Downtown into the neighboring tracts of the West End and Mt. Auburn inducing further displacement. Conversely, in the western half of the city, 5 contiguous tracts reached African-American populations exceeding 75 percent. This is the location of several public housing projects. When these 5 tracts are added to 10 tracts with at least 50 percent African-American population, there is a swathe of contiguous tracts in the West with greater than 50 percent African-American population. Unfortunately, these correspond with the decline in median household income between 2000 and 2016. The movement of the African-American population to the inner-ring suburbs may be heralded by some as a move upward in economic status, but the median household income maps show that this is not the case. In the past several decades, poverty in U.S. cities has been associated with the inner-ring suburbs as investment opportunities have favored locations closer to the CBD (Murphy and Allard, 2015; Cooke, 2010). Public housing has also moved to the inner-ring suburbs (particularly evident in Winton Hills and Madisonville).

Thus, we may conclude that the dispersal of the African-American population to the inner-ring suburbs may be re-concentrating poverty in a different area of the city.

Figure 4.6A. City of Cincinnati, Percentage African-American by Tract, 2000.

Figure 4.6B. City of Cincinnati, Percentage African-American by Tract, 2010.

Figure 4.6C. City of Cincinnati, Percentage African-American by Tract, 2016.

Issues affecting Displacement Other than Gentrification: While gentrification is likely driving some displacement of the African-American population,

Cincinnati’s status as a shrinking city and the Great Recession certainly confound these processes. The

City of Cincinnati’s population declined from 500,000 in 1950 to about 300,000 in 2016, a 40 percent decline (U.S. Census). However, the housing stock declined at a much slower rate. Between 1950 and

2016, the City only lost 1.9 percent (going from 162,591 units in 1950 to 161,162 units in 2016) of its housing stock (U.S. Census). While urban renewal in the 1950s to 1970s led to a reduction in housing stock in areas closer to Cincinnati’s CBD (Hampton, 2014), housing units were added to the inner-ring suburbs, most of which had at least fifty percent of their housing stock built after 1950 (City of

Cincinnati, 2012).

If so much housing stock is available to a declining city population, then why isn’t everyone enjoying an abundance of high-quality, low-cost housing? One reason is that demographic trends changed significantly since the 1950s; specifically, average household size has declined. There has been an increase in single-parent head of households, an increase in the number of single professionals who seek to live alone, and an increase in the elderly population, many who also live alone (Mather and Jacobsen,

2012). The large percentage of the baby-boomer population in the post-childbearing years by the early

2000s would also contribute to a decline in household size. In fact, 44.3 percent of the City of

Cincinnati’s households consisted of one person in 2016 (U.S. Census, 2016). Secondly, vacancy accounted for 16.5 percent of the housing stock in Cincinnati in 2016 whereas this was only 1.1 percent in

1950. While many of these structures need to be demolished, others could be renovated (City of

Cincinnati, 2012).

Once declining household size and the increase in housing vacancy are accounted for between

1950 and 2016, much of the excess housing stock that one would assume exists by 2016 (due to declining total city population) would have been reclaimed. In 1950, the average household in the City of

Cincinnati consisted of 2.7 persons (U.S. Bureau of the Census, 1953) whereas that declined to 2.1 by

2016 (U.S. Census, 2016). However, there is still plenty of high-quality housing in rustbelt cities such as

Cincinnati (Mallach and Brachman, 2010), but low-income populations are unlikely to be able to afford this option. Thus, we may assume that there is a spatial mismatch between population and housing and that this may fuel the displacement process. This mismatch is not dependent upon gentrification.

The Great Recession also affected housing options and displacement of the African-American population. During the early 2000s, easy access to housing loans allowed many African-Americans to purchase homes and often relocate to another neighborhood. Rents were lower in 2000 than they were by

2016, which may have allowed African-Americans to resist displacement in their chosen neighborhoods.

With the Great Recession, African-Americans who lost their jobs and/or experienced housing foreclosure, may have sought refuge in neighborhoods with lower housing costs and/or public housing options. A recent report entitled ‘The City of Cincinnati: 2015-2019 Consolidated Plan and 2015 Annual Action’ noted that one of the biggest problems that Cincinnati faces is the availability of affordable housing (City of Cincinnati, 2015). This makes an analysis of the displacement of low-income populations, (whether white or non-white), that more pertinent to the future functioning of the city.

Conclusion: This chapter examined the change in median household incomes, the gentrification process, and the racial transformation of population between 2000 and 2016 in the City of Cincinnati. The spatial distribution of the ranking of median household income shifted during the study period. The east and the urban core increased their ranking in the economic hierarchy while the western half of the city declined in economic ranking. This change in the spatial distribution of median household incomes was accompanied by a change in the spatial distribution of the African-American population: higher concentrations of

African-Americans seem to be associated with a decline in economic ranking. Tracts undergoing gentrification were examined to determine if there was a relationship between this process and the displacement of African-American population. A spatial examination showed that tracts that underwent gentrification experienced increases in income ranking and declines in African-American population,

84 strong evidence that the gentrification process is changing the economic and racial profile of Cincinnati’s neighborhoods. An additional logistic regression was performed to determine whether tracts that gentrified between 2000-2016 had 1) a greater decline in percentage composition of the tract that was

African-American between 2000-16 and 2) whether or not the tract experienced an increase in its decile ranking of median income compared to the city between 2000-2016. Results from the logistic regression found that a one percent decline in African-American population was associated with a 4.6 percent greater likelihood that the tract had gentrified between 2000 and 2016 (significant at the 0.05 level). Likewise, tracts that gentrified between 2000 and 2016 were three times more likely to experience an increase in their decile ranking of median household income between 2000 and 2016 (significant at the 0.1 level).

Logistic regression was applied to determine this relationship statistically. In the short-term (2000-10 or

2010-16) gentrification and displacement were not statistically significant. However, in the long-term

(2000-16), gentrification and the displacement of the African-American population were related. Thus, it is imperative that the City of Cincinnati examine policies concerning how gentrification can contribute positive aspects to not only the middle-class gentrifiers, businesses, and local government but also to marginalized low-income populations. Otherwise, low-income (largely minority) populations will continue to be concentrated in the bypassed urban core and in the inner-suburbs which have now become some of the most disinvested areas in U.S. cities, effectively removing this segment of the population from making valuable contributions to the city.

Chapter 5 TRACT 74- NORTHSIDE Introduction: This chapter is the first of four chapters that will examine the gentrification process in the four neighborhoods that were selected for analysis. As explained in Chapter 3, Tract 74 in Northside was selected for further analysis because it experienced a substantial decline in its African-American population between 2000 and 2016 that was likely due to displacement associated with the gentrification process. The first part of the chapter draws upon census data to determine changes in the economic, demographic, and social characteristics of the tract’s population that have occurred between 2000 and

2016 and which may be linked to gentrification. The second part of the chapter uses Google Streetview to examine changes in the built environment at the block level between 2007 and 2014/165. Although not its intended purpose, Google Streetview allows the user to document the process of gentrification over time.

Used in conjunction with Hwang’s (2015) gentrification inventory it is possible to determine the stage of gentrification that a block face (one side of a street in a block) has progressed over time. The strategy relies on a random sample of blocks from a gentrified tract as a total census of the blocks in a tract is not feasible due to time constraints. The original intention of this research was to use Google Streetview to document the effect of the Great Recession on the gentrification process in the four neighborhoods; I have since determined that that is not possible due to the availability of imagery. While I have shown that the gentrification process advanced in Northside between 2007 and 2014/16, I cannot determine in what way the Great Recession affected this process. The third part of the chapter includes interviews with informants from NEST (Northsiders Engaged in Sustainable Transformation).

5 For most block faces, 2014 is the latest image date by the end of the study period (2000-2016). If 2016 images are available, they are used.

Figure 5.1.Northside, Tract 74, Satellite Imagery.

Short History of Northside: Northside (the northern portion of the former Cumminsville village/neighborhood) was settled in the late 1700s at the crossroads of two Native American trading routes. Today these routes are known as

Spring Grove and Hamilton Avenues (Cumminsville History, 1792-1914). By the early 1800s,

Cumminsville consisted of a few log cabins, a tavern, and a tannery. Population in Cumminsville grew as the Miami-Erie Canal and Cincinnati-Hamilton-Dayton Railroad allowed the populace to move out of the

Cincinnati basin by the mid-1800s and in 1874, with a population around 4,500, the community was annexed by the city of Cincinnati (Cumminsville History, 1792-1911). In the early 1900s, Knowlton’s

Corner, located at the intersection of Spring Grove and Hamilton Avenues, was one of the busiest retail areas in Cincinnati (www.Northside.net). Cumminsville’s economic and population decline began with the arrival of the automobile and suburbanization, as population, industry, and retail relocated to the suburbs.

Further disruption to Cumminsville occurred when Interstate 74 connections via Colerain Avenue were completed in 1974. This connection resulted in a dissolution of Cumminsville into the two distinct neighborhoods of Northside and South Cumminsville. While the impact of a new highway exit on

Northside might be viewed as positive for its social and economic evolution, this dissection severed South

Cumminsville into an isolated area no longer connected to Northside. Fortunately, Northside’s location on

Hamilton Avenue provided access to the University of Cincinnati and Downtown to the southeast and the suburbs to the north. With the severance of Northside and South Cumminsville, Northside inherited the business district and it is this area that has seen revitalization in the past few decades. According to

Maloney and Auffrey (2013), in 1970, the socio-economic status of the four census tracts in Northside was classified in category 2 (out of 4) and were considered poor. However, between 1970 and 2010, three of the four census tracts in Northside were re-categorized as moving up to category 3. It appears that

Northside has embarked on a process of gentrification.

Census Data: City and Tract 74 Comparisons - Population: Table 5.1 displays demographic/economic characteristics for Tract 74. Between 2000 and 2010,

Tract 74’s total population decreased by 38.4 percent and then rebounded by 22.7 percent between 2010 and 2016 (U.S. Census). The decline in population between 2000 and 2010 was comparable between the

African-American population (-33.7 percent) and the non-Hispanic white population (-30.7 percent).

However, the African-American population declined by another 44.9 percent between 2010 and 2016. In

2000, 45 percent of the tract’s population was African-American while by 2016 only 17.4 percent of the tract was African-American. In contrast, the non-Hispanic white population grew by 36.3 percent

(essentially recovering its loss during the first decade of the 21st century) between 2010 and 2016 to 66.8 percent of the total population, up from 53 percent in 2000. The Hispanic population increased from 18 individuals to 163 between 2000 and 2016, comprising 10 percent of the tract population by 20166. It is evident that this tract has undergone rapid racial/ethnic transition between 2000 and 2016.

Table 5.1. Population Change in Tract 74, 2000-2016

2000 – 2010 – 2016 – 2000 2010 Percent 2016 Number of Number of Number of Percent of of Tract Percent of Individuals in Individuals Individuals Tract Total Total Tract Total Tract in Tract in Tract Total 2,210 1,598 1,670

African- 975 646 356 44.1 40.4 21.3 American Non-Hispanic 1,182 861 1,115 53.8 53.9 66.8 White

Hispanic/Other 53 91 199 2.1 5.7 11.9 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

6 The Hispanic population will not be examined in this chapter as the focus is on displacement of racial/ethnic minorities and not the growth of those populations.

City and Tract 74 Comparisons – Household Income, Housing and Rental Values, Educational Attainment: One indicator of gentrification is an increase in median household incomes that exceed the city’s average (Hammel and Wyly, 1996). The median household income for Tract 74 increased from $25,932 to $44,429 between 2000 and 2016, a 41.6 percent increase (see Table 5.2). To put this in perspective, we can look at the growth of median household income for the City of Cincinnati. Its median household income increased from $29,493 to $34,629 over the same time period; a 17.4 percent increase. More importantly, Tract 74’s median household income was below the city average in 2000 and 2010, but by

2016 was above the city average. Another indicator of gentrification is an increase in the percentage of college graduates residing in a neighborhood (Hammel and Wyly, 1996). In 2000, only 11.7 percent of

Tract 74’s population over 25 had a bachelor’s degree. By 2016, this percentage increased to 44.3 percent suggesting that this tract is undergoing a major socio-economic transformation. To compare, the City of

Cincinnati had only 33.8 percent of its population over 25 population with college degrees in 2016.

Table 5.2. Median Household Income, Percentage Population with at Least Bachelor’s Degree, Median Housing Values and Median Rent for City of Cincinnati and Tract 74, 2000-2016: Tract City City City Tract Tract 74 Cinti. Cinti. Cinti. 74 2010 74 2016 2000 2000 2010 2016 Median $25,932 $29,493 $31,976 $33,681 $44,429 $34,629 Income Educ Bach + 11.70% 26.60% 29.00% 31.70% 44.30% 33.80% Median Housing $56,000 $93,000 105,500 $129,700 $115,500 $120,300 Value Median $494 $444 $606 $593 $810 $662 Rent Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

A third indicator of gentrification according to Hammel and Wyly (1996) are median housing values and median rents that outpace the growth of the city. Median housing values in Tract 74 increased from $56,000 in 2000 to $115,500 in 2016, a 106 percent increase. Most of this increase occurred between 2000 and 2010, a time when housing speculation throughout the nation was occurring. Housing values for the City of Cincinnati are included for reference. Note the increase in the City of Cincinnati’s

90 median housing values from $93,000 to $129,700 between 2000 and 2010, a 39.5 percent increase and then the 7.2 percent decline between 2010 and 2016. It is likely that this decline was a result of the housing debacle that occurred with the Great Recession. Conversely, median housing values in Tract 74 increased between 2010 and 2016, although the growth in housing values between 2010 and 2016 was much less than that of the growth between 2000 and 2010. Median housing values were much more comparable between the City of Cincinnati and Tract 74 in 2016 than they were in 2000. In 2000, Tract

74 was only 60.9 percent of the median housing value of Cincinnati whereas this increased to 96.0 percent by 2016. As can be imagined, a growth of 104.1 percent in median housing values in Tract 74 in less than two decades would likely make it difficult for those with less financial resources to purchase housing in Tract 74. Median rent increased by 34.0 percent for the City between 2000 and 2016 compared to an increase of 64.0 percent for Tract 74. By 2016, renters in Tract 74 were paying an average of $148 dollars more than their counterparts in the City. For those with limited financial resources, there were cheaper housing options in other Cincinnati neighborhoods than in Northside and it may be that this increase in median rent led to displacement of lower-income individuals.

City and Tract 74 Comparisons – Housing Units, Vacancy, and Ownership: While not directly an indicator of gentrification, it is helpful to examine changes in the number of housing units, housing vacancy rates, and percentage of housing that is owner-occupied (see Table 5.3).

The number of housing units increased slightly between 2000 and 2016, from 998 to 1,0427. It is likely that these additional dwellings were either newly constructed, older buildings converted from commercial/industrial use to housing/apartment units, or larger housing units subdivided into smaller units. We can assume that this number of additional dwellings is at least 44 (the difference between the

7 American Can Lofts located on Spring Grove Avenue is an industrial building converted to 110 apartments. The 2012-2016 ACS data likely undercounts residents in these apartments due to the recent conversion. The Gantry Development located at Hamilton Avenue and Blue Rock with 130 apartments was completed by 2016 but was not yet occupied.

91 total number of dwellings in the two years); it could be more as we do not have an accurate count of the dwelling units that were demolished during this time. For 2000 and 2010, vacancy was around 10 percent, but by 2016 it jumped to 19.7 percent. This increase in vacancy may be due to the recession which caused residents to abandon houses/apartments that became too expensive and seek housing elsewhere. Literature has demonstrated that minorities and low-income non-Hispanic whites were able to get housing loans during the 1990s and early 2000s which became difficult to repay after the collapse of the housing market

(Hoynes, et al.). The combined processes of the recession and gentrification may have forced residents out of the neighborhood. Given the increase in housing values and rent from 2000 to 2016, the cost of living may have become prohibitive for low-income individuals who had previously considered making

Northside their home. Homeownership rates in Tract 74 increased from 37 to 51 percent between 2000 and 2010. This is not surprising given that loans during most of that period were relatively easy to acquire. The decrease in homeownership to 44.8 percent by 2016 is likely a result of defaults on housing loans, as discussed above.

Table 5.3. Select Housing Characteristics for Tract 74, 2000-2016:

2000 2010 2016 Percent Owner- 37 51 44.8 occupied Housing Percent Vacant 11.1 10.1 19.7 Number of Housing 998 1,034 1,042 Units Same Residence One X[1] 68 79.2 Year Ago Source: American Factfinder. 2000 Census ; 2006-2010 ACS; and 2012-2016 ACS.

Residential stability (measured as percentage of population residing in the same residence as the previous year) increased from 68 percent in 2010 to 79.2 percent in 2016 (U.S. Census Bureau). Several reasons may account for the increase in residential stability between the two years. In 2010, the combined

92 effects of gentrification (an increase in housing costs), the recession (a loss of jobs), and bank lending practices could have forced residents out of their homes/apartments. Homeowners who borrowed loans at high interest rates may have been unable to pay their mortgages and were forced into foreclosure.

According to A Decade of Foreclosures in Hamilton County, Northside experienced the sixth highest number of foreclosures between 2006 and 2012 of the neighborhoods in Cincinnati. During the 1990s and early 2000s many housing loans were made to high-risk borrowers (Verick and Islam, 2011). These were applicants with limited monetary resources for down payments and poor credit history. One type of loan was ARM (Adjusted Rate Mortgage) which led to large payments due at one time. This situation of high interest loans was likely made much worse if the homeowner was unemployed during the Great

Recession. Renters may also have been forced out of their homes as rents increased during the gentrification process and loss of employment forced renters to seek housing in neighborhoods with lower costs of living. By 2016, the recession was over and who were able to survive the combined economic recession and increase in costs of living caused by gentrification were likely higher income individuals.

This will be discussed in greater detail in the next section which disaggregates African-American and non-Hispanic white demographic and economic characteristics.

City and Tract 74 Comparisons – Disaggregation of Economic/Social Characteristics by Race: Table 5.4 compares African-American and white median household incomes for Tract 74 with those of their counterparts in the City of Cincinnati. Note in all three years that African-American incomes in Tract 74 were higher than the city average for African-Americans, with 2010 showing a 60 percent higher median income. Between 2000 and 2010, it was likely that poorer African-Americans were displaced as median household incomes of the African-American population in Tract 74 increased8. It is

8 It is possible that African-Americans with higher incomes moved into Tract 74 while there was a massive exodus of low-income African-Americans between 2000 and 2010. The high mobility rates of African-Americans in Tract 74 for 2010 (in table 5) indicates that this was possible. This would amount to affluent African-Americans participating in the gentrification process. I find this unlikely as this time corresponds with the Great Recession which had a greater negative economic effect on African-Americans than whites (Hoynes et. Al., 2012). It would also be difficult to explain why these African-American gentrifiers lost interest in Tract 74 by 2016.

93 my hypothesis that some of the African-American population was able to remain in the neighborhood, perhaps spending down their cash reserves; these reserves may have been depleted by 2016, initiating displacement.

The average median household incomes in Cincinnati for African-Americans give a perspective on the toll that the Great Recession had on African-Americans. During the 2000 – 2016 period, African-

American median household incomes were stagnant. Non-Hispanic white median household income in

Tract 74 grew faster than that of their non-Hispanic white counterparts in the City. While substantially below city median household incomes in 2000 and 2010, Tract 74 closed the gap by 2016. It is likely that lower-income whites left the tract between 2000 and 2010; we know that there was a decline of approximately one-third of the tract’s white population during this time while median incomes increased.

The increase in median household income between 2010 and 2016 was likely a result of higher-income whites moving into the tract. We can assume this as the white population increased during this time period.

Table 5.4. Comparison of African-American and White Median Household Incomes for Tract 74 and City of Cincinnati, 2000-2016.

2000 2010 2016

City-Wide, $20,984 $22,816 $21,814 African-American Tract 74, African- $22,857 $36,250 $22,143 American City-Wide, Non- $36,467 $47,067 $50,937 Hispanic White Tract 74, Non- $27,333 $32,361 $49,286 Hispanic White Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Comparison of African-American and White Selected Characteristics within Tract 74: Table 5.5 shows educational attainment, median household income, and residential mobility data9 for the African-American and non-Hispanic white population in tract 74 for 2000, 2010 and 2016. One should be extremely careful in interpreting these data as they are 2006-10 and 2012-16 American

Community Survey estimates which have high margins of error. Note the decline in educational attainment between 2010 and 2016 for African-Americans. Also note that there was almost no mobility of the African-American population in the year prior to the 2012-2016 ACS.

Table 5.5. Select Characteristics of Population from Tract 74 by Race, 2000-2014.

Percent of Population Over Same House 1 year Median Household 25 with at Least a ago Income Bachelor's Degree African-American 13.30% 97.20% $22,14 2016 African-American 22.00% 45.30% $36,250 2010 African-American 5.90% X $22,857 2000 White 2016 0.486 0.727 49286 White 2010 30.20% 81.20% $32,361 White 2000 15.90% X $27,333 Source: ACS 2006-10 and 2012-16.

Conversely, less than half of African-Americans were in the same house in the year prior to the

2006-10 ACS survey. This suggests that the recession and the process of gentrification caused movement from the tract and it may be that African-Americans with higher incomes and educational levels left the tract as both educational levels and median incomes for African-American declined between 2010 and

2016. The substantial decline in total African-Americans between 2000 and 2016 suggests that African-

Americans with lower-income/lower educational levels did not move into tract 74 in large numbers. It is likely that many African-Americans were priced out of the housing market in Tract 74 between 2010 and

2016 due to gentrification. Conversely, note the increase in educational attainment for non-Hispanic

9 The ACS does not disaggregate median housing price or rent by race and thus these data are not included in the table.

95 whites between 2010 and 2016 (along with increases in median household income) and the increase in residential mobility (decrease in percentage of persons reporting living in same house as previous year).

Young adults are more likely to be involved in gentrification and are more likely to be mobile than their older counterparts (Speare, 1975). In 2010, the white population’s lower rates of residential mobility may have been a result of the recession. Spatial lock-in is a term that is used to describe homeowners who owed more on their houses than what they were worth (Chan, 2001). With declining housing prices, many would-be sellers may have delayed putting their houses on the market. In addition, high interest payments may not have allowed the homeowner to move without taking a financial loss.

Google Analysis: Figures 5.2A and B show the blocks that were selected for analysis from Tract 74 and their gentrification status as of 2007 and 2014/16. Table 5.6 displays the scores from the Google Streetview analysis for Tract 74. Twelve blocks were selected using a random number table and each of the three/four sides of the block were examined10. This resulted in 31 block faces for analysis. Fourteen block faces were excluded as they were alleys and it was not possible to examine them with Google

Streetview11. In 2007, 24 block faces were characterized as in some form of disinvestment (scores below

0.50). Figure 5.2A displays the spatial distribution of gentrification for the sampled blocks in Tract 74 as of 2007. Note that gentrification was concentrated in the Northwest Quadrant of Tract 74. These results are not surprising as gentrification tends to concentrate in select spots which have such features such as interesting architecture and/or access to employment/recreational/cultural opportunities (Hammel and

Wyly, 1996). This sample shows that investment is located along Hamilton Avenue, a major artery between downtown/University of Cincinnati and the suburbs. The location along Pullan Avenue in the

North is the boundary line for Tract 78, a tract that was not a candidate for gentrification in 2000 as it

10 Please see chapter 2 on methodology to see a full explanation of this process. 11 The best the user can do in Google Streetview is to position the image on a cross street and look down the alley. I excluded all alleys because this was insufficient to carry out the detailed analysis using Hwang’s criteria.

96 already had a median income above the city’s average and contained stylish housing stock in good condition. The average score for all of Tract 74 on the gentrification stage was 0.41 in 2007, characterizing the tract as an area of disinvestment. Evidence of this disinvestment was apparent in several blocks which contained boarded up buildings or structures that were in such disrepair that they needed to be demolished. However, four of the twelve blocks were characterized as gentrified as of 2007.

Blocks 1010, 3013, and 3003 were characterized as being in the early stage of gentrification while block

3002 was characterized as being in the middle stage of gentrification.

Figure 5.2B displays gentrified blocks in 2014/16. Note the expansion of gentrification in the

Northwest Quadrant as well as the gentrification process that is found along Mad Anthony Street and surrounding blocks. A total of 11 of the 12 blocks were gentrified by 2014/16. However, most of these blocks were in the very early stages of gentrification. Eight block faces were not gentrified (below 0.50).

This resulted in a tract score of 0.56 on the stage of gentrification scale for 2014/16. This score indicates early gentrification. Recall that Tract 74 moved from a candidate for gentrification as indicated by its median household income and other characteristics of population and housing stock in 2000 to gentrified in 2010. It also continued that process from 2010 to 2016 (see Chapter 3). An examination of the built environment using Google Streetview supports the census data used in Chapter 3 to designate Tract 74 as gentrified.

Figure 5.2A. Gentrified Blocks in Tract 74, Northside, 2007.

Figure 5.2B. Gentrified Blocks in Tract 74, Northside, 2014/16.

Table 5.6. Sampled Blocks in Tract 74 and Stage of Gentrification for 2007 and 2014/16

Key: Disinvestment below 0.50 Early between 0.50-0.57 Middle between 0.58-0.64 Late between 0.65-0.80 Turnover over 0.80

2007 2014/16 Difference between 2007 and 2014/16 Block Score Score and Stage of Gentrification 1004 0.22 0.57 Went from disinvestment to early stage 1009 0.36 0.63 Went from disinvestment to middle stage 1010 0.55 0.66 Went from early stage to late stage 2011 0.39 0.5 Went from disinvestment to early stage 2019 0.36 0.6 Went from disinvestment to middle stage 2022 0.36 0.39 Very little change; considered disinvested 2029 0.44 0.52 Went from disinvestment to early stage 3002 0.67 0.72 Remained in late stage 3003 0.55 0.55 No change; at early stage 3007 0.22 0.5 Went from disinvestment to early stage 3010 0.11 0.5 Went from disinvestment to early stage 3013 0.55 0.55 No change; at early stage Tract 0.41 0.56 Average Source: Google Streetview and Hwang (2015).

The most frequent change between 2007 and 2014/16 was a transition from disinvestment to early stage gentrification. This occurred in five of the twelve blocks. Only in two blocks (1009 and 2019) did gentrification occur extremely rapidly (defined as going from disinvestment to middle-stage gentrification). Block 1009 is located adjacent to Block 1010 which was already gentrified as of 2007 while block 2019 is in proximity to the newly constructed (as of 2016) mixed-use development complex at Hamilton/Blue Rock. Block 2022 was the only block in the sample that did not gentrify by 2016. Its

100 location close to the Mill Creek and its history as an industrial/commercial strip on the periphery of Tract

74 may have made reinvestment less likely to occur.

Examples of the Gentrification Process Using Google Streetview: Below are several examples of the stage of gentrification for representative blocks and how I arrived at that designation. Keep in mind that Google Streetview does not allow us to view images prior to 2007. It is thus impossible using this tool to discern whether a block had gentrified prior to 2007 or if it had ever been through a process of disinvestment12. Regardless of these limitations, much information concerning the gentrification process can be gained by examining these images between 2007 and

2014/16.

Figure 5.3A is a Google Streetview image of a building at the corner of Mad Anthony and

Apjones in 2007 (Block 1004). Note the disrepair of the building in the 2007 image, particularly the boarded-up windows and doors. The same building in 2014 had received substantial reconstruction. This included new windows, window boxes and shutters, and painting of the exterior. Note also the beautification of the yard with shrubbery. This block face13 received a score of 0.36 in 2007 but in 2014 it improved to 0.5914.

12 Most gentrification studies focus on census tracts as the unit of analysis as blocks are not large enough in population or territory to indicate whether a neighborhood has undergone gentrification. 13 These block face scores are averaged to create the block score. They will not match the score in Table 6. 14 Recall that buildings are not coded for stage of gentrification. The entire block face was examined using the coding criteria. This example at Mad Anthony and Apjones was the most impressive as it displayed a total rehabilitation of the structure. The rest of the block face had experienced gentrification, but not to the scale that this building had.

101

Figures 5.3A and B. Block 1004. Corner of Mad Anthony and Apjones. 2007 and 2014.

Source: Google Streetview, 2007.

Source: Google Streetview, 2014.

102

Another example of gentrification between 2007 and 2014/16 is found at the corner of Knowlton and Dane in Block 2019 (see Figures 5.4 A and B). The 2007 Google image shows a building with peeling paint and a streetscape that is free of litter but somewhat unkempt. One of the doors was boarded- up to prevent access. The lack of boarding of windows suggests that unauthorized entry was not a major issue at this location. By 2014, note the rehabilitation of the property. The building was painted, a newly constructed porch extends from the rear of the structure, and a wooden fence offers privacy. The grounds had been beautified. This block face received a score of 0.11 in 2007 and a score of 0.73 by 201415.

15 Unlike the building at Mad Anthony and Apjones in Block 1004, this building is representative of the gentrification process on this block face.

103

Figures 5.4A and B. Block 2019. Corner of Knowlton and Dane. 2007 and 2014.

Source: Google Streetview, 2007.

Source: Google Streetview, 2014.

104

On Lingo Street in Block 3010 there were several small houses in disrepair as of 2007 (See

Figures 5.5A and B). Note the boarding of the lower windows and doors to prevent unauthorized access.

It is likely that these buildings were not candidates for gentrification. The construction of these houses from aluminum siding/wood would make them not as durable as brick or stone. The lack of disrepair in the front of the houses suggests that this street was relatively well cared for in 2007. By 2014, these houses were demolished and replaced with greenspace. This block face was coded as a 0.11 in 2007 but improved to a 0.50 by 2014. The 0.50 was given to the block face because of the removal of the structures in disrepair and the clearing for future gentrification.

105

Figures 5.5A and B. Block 3010. Boarded up Houses and Greenspace, 2007 and 2014. Lingo St.

Source: Google Streetview, 2007.

Source: Google Streetview, 2014.

106

Block 3002 is an example of gentrification in both 2007 and 2014 (Figures 5.6A and B). This segment is along Williamson Place, a street perpendicular to Pullan Avenue. Note the upkeep of the houses in 2007. This block face scored 0.78 in 2007 indicating it was in the late stage of gentrification.

By 2014, the block face received a 0.89 indicating that it had attained turnover status. This was determined by the rehabilitation of the house on the right and the substantial investment in beautification of the yards.

107

Figures 5.6A and B. Example of Late Stage Gentrification in Block 3002.

Source: Google Streetview, 2007.

Source: Google Streetview, 2016.

108

Block 2011 is located along Hamilton Avenue and is an example of commercial gentrification

(Figures 5.7A and B). Investments along the sidewalk, particularly the lamp posts, indicate that this segment of the street was primed for gentrification as of 2007. The block face received a score of 0.44 in

2007, largely a result of the failure to modernize the buildings. Note that by 2016 that the building on the right had been rehabbed with aluminum siding to improve appearance. Also note the improvement in the center building with the bright colored awnings while the building to the left added a new business sign overhead. These investments resulted in a score of 0.55 for the block face in 2016.

109

Figures 5.7A and B. Example of Early-Stage Gentrification of a commercial area along Hamilton Avenue.

Source: Google Streetview, 2007.

Source: Google Streetview, 2016.

110

While only twenty percent of Tract 74’s blocks were sampled, I feel that this in an accurate representation of the gentrification process in Tract 74. Unfortunately, the random sample did not include any of the blocks which had experienced large-scale corporate investment. Knowlton Place (Figure 5.8), a senior living facility newly constructed in 2017; The American Can Lofts Building, an old industrial building rehabbed around 2014; and the Gantry Development Project built at the intersection of Blue

Rock and Hamilton opened in 2016 (Figures 5.9A and B) are additions to the built environment that are too recent to have had much of an effect on the gentrification process between 2000 and 2016. These are large-scale projects that have attracted a lot of attention. They tend to take attention away from small- scale revitalization projects. It is these small-scale projects that likely led the way to bigger investments in

Tract 74. This would be in keeping with the gentrification literature which indicates that early stages of gentrification occur as risk takers settle in neighborhoods and make incremental improvements to their properties which are then expanded upon by larger investors at a later stage (Clay, 1979; Gale, 1979).

111

Figure 5.8. Construction of Knowlton Place in 2017.

Source: Google Streetview, 2017.

112

Figures 5.9 A and B. Site of Mixed-Use complex at Hamilton and Blue Rock in 2014 and 2016.

Source: Google Streetview, 2014.

Source: Google Streetview, 2016.

113

The Built Environment in Northside as of June 2019: On June 25, 2019, I canvassed the sampled blocks in Tract 74 that were a part of this study to document changes in the built environment from 2016 to present-day. Most of the large-scale developments were in place by 2016, but gentrification at a smaller scale has continued in Tract 74.

Single-family homeowners continue to renovate their homes, some in-fill of new housing was in evidence and businesses have invested in new facades. While much progress in revitalizing Tract 74’s built environment has occurred in the past decade, there is much more that could be accomplished. Many streets have dwellings that are in various stages of rehabilitation, or completely new, along with dilapidated dwellings that need to be demolished.

Figures 5.10A and B display images along Mad Anthony Street. It shows the beginning stages of gentrification. Figure 5.10A shows houses in disrepair as late as 2014. In the 2019 image, note that the houses to the left have new windows, new roofing and new embellishments in the landscape while the two houses to the right are in disrepair. However, note that new windows have been installed in the top floor of house on the far right suggesting that this house was being renovated by 2019.

114

Figures 5.10A and B. Houses in Various Stages of Gentrification/Disrepair along Mad Anthony.

Source: Google Streetview, 2014.

Source: Photo by E. Ravuri, June 25, 2019.

115

Figure 5.11A and B. In-fill along Fergus Street. In 2014 (the latest Google Streetview imagery for Fergus Street) this area was a well-cared for vacant greenspace (Figure 11A). By 2019, these up-scale single- family dwellings were erected (Figure 11B).

Source: Google Streetview, 2014.

Source: Photo by E. Ravuri, June 25, 2019.

116

Figures 5.12A and B. Renovation of building along Knowlton Street. Figure 12A is an image of a dilapidated building in 2018 but the streetscape and alley are in good and clean condition. Note the new windows that have been installed by 2019 and what appears to be a refurbishing of the siding (Figure 12B). Also, note the graffiti on the building to the right. This is a block face in the very early stages of gentrification.

Source: Google Streetview, 2018.

Source: Photo by E. Ravuri, June 25, 2019.

117

Interviews with Community Members: In December 2018 an email was sent to NEST (Northsiders Engaged in Sustainable Transformation) explaining the project and asking the agency to provide contacts for individuals who could speak on economic revitalization and demographic change in Northside between 2000 and 2016. During May/June of 2019, interviews with individuals familiar with the economic revitalization process and demographic changes that have occurred in Northside since 2000 were undertaken. The following individuals were interviewed:

Bob Carbon: Bob has lived in Northside since 2007 and has been an active member in NEST. He received a Master’s in Architecture from the University of Kentucky.

Carl Sterner: Carl has lived in Northside since 2015 and has been an active member in NEST. He received a Master’s in Architecture from the University of Cincinnati.

Stephanie Sunderland: Stephanie has lived in Northside for over two decades. She was one of the founding members of NEST and Citizens on Patrol.

The following questions were asked of each of the interviewees. Not all interviewees could answer each question.

1. Can you pinpoint a year (s) when you first started to see changes in Northside suggesting that

economic reinvestment and/or revitalization was occurring? Where did the process begin within

Northside? Be as specific as possible.

Stephanie: Stephanie first moved to Northside in 1982 and resided on Cherry Street until 1994 when she relocated to Clifton. She then returned to Northside in the early 2000s. Stephanie remarked that there was a tremendous change in the demographic and built environment of Northside in the late 1990s. She attributes this to the removal of low-income housing in places like Bond Hill, the West End, and the Fay

Apartments. This resulted in a large influx of low-income residents (both African-American and white).

118

Stephanie noticed a lot of gang activity and drug dealers moving into Northside. Not surprisingly, many middle-class residents in Northside left the community during the early 2000s as crime became endemic.

Several Northsiders then established Citizens on Patrol which sought to rectify the crime situation.

Stephanie was one of the founding members of NEST (2005) and served as the president until 2018.

NEST was further involved in the reporting of drug dealers, the recording of vacant properties, and ensuring that absentee landlords did not violate city building codes which paved the way for development in the neighborhood after 2008.

Carl: Carl has only lived in Northside for four years, but he attended graduate school at the University of

Cincinnati during the first decade of the twenty-first century and had some knowledge of the built environment in Northside during this time. He remarked that there has been a substantial amount of change over the past ten to twelve years. He noted the construction of new buildings (such as the Gantry

Development) and the renovation of the American Can Company into lofts/office space/restaurant. Carl said that the revitalization has not been limited to large-scale projects, but numerous single-family homes have also been renovated. Some of these single-family homes have been renovated by owner-occupiers.

Others have been purchased by ‘flippers’ who are looking to turn a profit.

Bob: Bob has lived in Northside since 2007. He recalls visiting Northside in the late 1990s and feeling unsafe on the side streets. Like Carl, Bob notes the construction of the Gantry Building and the renovation of the American Can Building as harbingers of revitalization in Northside. He also noted that many houses that were subdivided into one or more apartments have been converted back to single-family residences and renovated.

119

2. What effect did the Great Recession (2007-2009) have on economic reinvestment/revitalization in

Northside?

According to Stephanie, most of the 2000 to 2010 period was a part of the Great Recession. There was little investment in Northside prior to the 2008 recession, and this worsened during the recession.

Property values had declined significantly in Northside prior to 2008. Stephanie recalls in 2008 that several alleys in Northside were impassable because of the amount of trash deposited. Fortunately, NEST received funding to purchase some buildings in deplorable condition and demolished them. Their funding did pay for the relocation of tenants.

Surprisingly, none of my interviewees commented on the foreclosure crisis that accompanied the Great

Recession. This is surprising in that Northside was one of the City’s most adversely affected neighborhoods during this crisis (A Decade of Foreclosures in Hamilton County, 2012).

3. Can you comment on the impact of private development agencies versus non-profit/government

agencies in the reinvestment/revitalization process?

Each of the interviewees stated that the Gantry Development and American Can mixed-use developments have been positive for Northside in terms of bringing about revitalization. However, Bob was concerned about businesses in Northside as he sees them as behaving mostly in their own interests. Some of the business owners do not live in Northside and thus their commitment to the neighborhood (other than extracting profit for their service) is limited, according to Bob. Bob feels that NEST’s investment has been too focused on residential development in Northside and that the organization needs to expand into commercial investment.

Carl credits NEST with paving the way for private investment in Northside. Had it not been for the removal of blight and crime in Northside by NEST and Citizens on Patrol, Northside would not have been as attractive to further investment.

120

4. Has reinvestment/revitalization led to a change in the demographic composition (e.g. age,

race/ethnicity, socio-economic status) of Northside? If so, in what way? Have certain areas been

more affected than others? How did the Great Recession affect demographic composition in

Northside?

Stephanie has noticed an influx of younger population (20s-40s) since 2014. Conversely, Bob has noticed that there is a wide range of ages in Northside and that it has become difficult for younger families to afford to live in Northside. Bob only settled in Northside in 2007 and thus, his perception of change in the age structure, not surprisingly, deviates from Stephanie’s, whose experiences range from the early 1980s.

Each of the interviewees agreed that Northside was becoming whiter during the 2000s. Stephanie was the only respondent who lived in Northside during the 1990s and the early 2000s and witnessed the change of

Northside from a predominantly white community in 1982 to a mixed racial community by the 1990s.

Stephanie states that much of this was a result of the availability of Section 8 housing in Northside and the displacement of population in the West End.

5. Can you comment on the changes in types of businesses or improvements in already established

businesses in your community during the gentrification process?

Each of the interviewees commented that businesses catering to an up-scale clientele have opened in

Northside during the last decade. The most conspicuous establishments have been restaurants.

Littlefield’s, Melt, and Ruth’s Parkside Café are a few examples. Stephanie was the only respondent to comment on the lack of a grocery store in Northside (Save-a-Lot closed around 2013). Stephanie reports that a co-op grocery store may soon be constructed in the former Save-a-Lot lot, further evidence of up- scale businesses coming to Northside.

121

6. Is there anything else you would like to mention concerning economic revitalization and

demographic change in Walnut Hills?

The major concern for each of the interviewees was affordable housing. Bob remarked that it has become difficult for young families to afford housing in Northside due to escalating home and rental costs.

Stephanie mentioned that affordable housing for seniors has become a major issue in Northside. Carl mentioned the ‘kind of tiny house’, a dwelling unit around 1,000 square feet that would not only be environmentally friendly, but also affordable.

Conclusion: Tract 74 has seen tremendous change in the built environment and demographics between 2000 and 2016; in other words, gentrification. The tract went from one of exodus of both African-American and white population between 2000 and 2010 to an influx of whites between 2010 and 2016. The new white arrivals to Tract 74 had different socio-economic characteristics than their white counterparts who left the tract. This was inferred by the increases in educational attainment and median household incomes for whites over the 2000 to 2016 timeframe. Conversely, low-income African-Americans left the tract between 2000 and 2010 while middle-class African-Americans left between 2010 and 2016. The removal of low-income population of either race was largely a result of increases in median housing prices and median rents since 2000. Interviewees from NEST confirmed that affordable housing was one of the greatest concerns in Northside.

Google imagery revealed that Tract 74 underwent a substantial transformation in the built environment between 2007 and 2016. This transformation consisted of new large-scale developments such as the Gantry mixed-used complex and the renovation of American Can to mixed-use as well as numerous renovations of owner-occupied single-family housing. Hwang’s (2015) gentrification index placed Tract 74 at 0.41 in 2007 (a disinvested tract), changing to a tract that was in the late stage of early

122 gentrification by 2016 (a score of 0.56). Interviewees from NEST also stated that businesses underwent a gentrification process as more up-scale businesses moved into Northside after 2000.

123

Chapter 6 TRACT 55- MADISONVILLE

Introduction: This chapter examines the gentrification of Tract 55 in Madisonville between 2000 and 2016 and the subsequent displacement of the African-American population. The chapter begins with a short historical overview of Madisonville. Following Hammel and Wyly’s (1996) criteria for gentrification, aggregate characteristics from census data are compared between Tract 55 and the City of Cincinnati to demonstrate that gentrification occurred in Tract 55. Select economic, social, and demographic characteristics are then disaggregated to compare African-Americans with whites residing in Tract 55 for

2000, 2010, and 2016. Using Hwang’s (2015) criteria for determining gentrification in the built environment, a Google Streetview analysis is performed for select blocks in Tract 55. In the third part of the chapter, members of Community Development Corporations are interviewed to provide a subjective view of the gentrification process that was documented through the objective data sources of the Census and Google Streetview.

Madisonville contains three census tracts: 55, 56, and 108. Originally, Tract 108 was selected for detailed analysis because its African-American population was approximately the same percentage as that of Tract 74 (Northside), Tract 84 (College Hill), and Tract 19 (Walnut Hills). After detailed analysis, it was determined that gentrification of the built environment was not occurring in Tract 10816 and thus

Tract 55 was substituted for the analysis. Tract 56 did not experience gentrification as outlined by the

Hammel and Wyly (1996) criteria and thus was not considered as a possible replacement for Tract 108.

Tract 55 experienced both gentrification and displacement of the African-American population as

16 Tract 108 was a former industrial area which converted to an office complex in the late twentieth century. The tract’s population was only 722 in 2016 and it was decided that small changes in the population could have a disproportionate impact on the tract’s characteristics. For instance, Tract 108 experienced an influx of high-income individuals due to the construction of approximately 40 upscale housing units in a plan which has bee, as of 2000, forested area. This was an approximate 10 percent increase in housing units in Tract 108. The characteristics of these in-migrants were vastly different from the incumbent population making it appear that gentrification had occurred. Google imagery revealed that very little change occurred to the modest-sized single-family homes that were built in the earlier part of the twentieth century.

124 occurred in Tract 108, so on those criteria, the substitution seems to be appropriate. However, Tract 55’s

African-American population comprised 77.7 percent of the tract population in 2000 and was much higher than that of the African-African populations in 2000 in College Hill, Walnut Hills, or Northside, which was approximately fifty percent of the total population for each of the census tracts. This discrepancy in initial African-American population will thus need to be taken in to account.

125

Figure 6.1. Madisonville, Tract 55, Satellite Imagery.

126

A Short History of Madisonville: The area that would later become Madisonville was a Native American settlement that arose in the 1400s and was built by what is today referred to as the Fort Ancient moundbuilders17. The site attracted substantial interest from archaeologists during the late 1800s and early 1900s and was reputed to contain one of the largest archaeological collections of burials and middens in the Midwest (Brown,

1999). The area was first surveyed by William T. Darling in 1819 and platted for settlement by 1829

(Writer’s Program, 1943). By 1841, Madisonville had a population of approximately 400 residents.

Present-day Whetsel Avenue and Madison Road were old Native American Trails (Madisonville NBD

Renewal Plan, 2002) and provided the foundation for the first roads in present-day Madisonville.

Madisonville’s growth continued after the completion of the Marietta and Cincinnati Railroad in 1866

(Writer’s Program, 1943) and by 1898 Madisonville was connected to downtown Cincinnati by streetcar.

By the early 1900s, the Norwood Trough18 was the second most important concentration of industry in the

Mill Creek Valley and connected Madisonville to this important epicenter of industrial activity in the

Cincinnati area. In 1911, Madisonville was annexed to the City of Cincinnati (Writer’s Program, 1943).

Between the 1930s and 1960s, Madisonville was the business center for eastern Cincinnati (Madisonville

NBD Renewal Plan, 2002), but suburbanization and the construction of highways led to Madisonville’s demographic and economic decline during the 1970s (Maloney and Auffrey, 2013). By the 1970s, substantial blight had affected residential, business, and industrial areas of Madisonville and provided the impetus for revitalization of the community (Madisonville-Eastwood Community Plan, 1975). In the early 2000s, Madisonville was identified by “Go Cincinnati” (a plan on investment potential in Cincinnati neighborhoods by the City of Cincinnati) as a neighborhood with potential for economic development and this led to the “Quality of Life Plan” that framed the revitalization of Madisonville’s business, industrial,

17 Historical artifacts from this civilization (insert years of existence and geographical distribution here) are today housed in the Harvard Peabody Museum, the Smithsonian, the National Museum of the American Indian, and the Cincinnati Museum of Natural History. 18 The Norwood Trough was an abandoned pre-glacial river valley. Its lower elevation than the surrounding area allowed the railroads and the Norwood Lateral (highway) to be constructed here. Prior to the 1800s, it would have allowed people to move through the area by foot (Potter, 1996).

127 and residential properties (Quality of Life Plan, 2012). This plan was coordinated through the City of

Cincinnati but was vested with the interests of the Madisonville community.

Census Data: Tract 55’s total population declined by 7.4 percent between 2000 and 2016 (See Table 6.1). For the African-American population, the decline was 4.5 percent between 2000 and 2010 and a further decline of 33.9 percent between 2010 and 2016. Conversely, white population increased by 12.3 and 63.2 percent between 2000-10 and 2010-16, respectively. In 2000, African-Americans comprised 77.7 percent of Tract 55’s population, but by 2016, African-Americans only comprised 53.0 percent. Whites comprised 19.0 percent of Tract 55’s population in 2000, but 37.7 percent by 2016. According to Taeuber and Taeuber’s (1965) classification of racial change, displacement of African-Americans occurred between 2010 and 2016. While we cannot determine displacement from census data, interviews with long-term residents in Madisonville in 2016 confirmed that there had been an in-migration of whites and an out-migration of African-Americans over the past few years (Rinehart, 2016). This decline in the

African-American population is troubling in that Madisonville, as of 2007, had been noted as a stable racially integrated community for the past thirty years (Brown and Baize, 2009). In the following discussion, I will argue that the displacement was caused by both gentrification and the Great Recession.

128

Table 6.1. Population Change in Tract 55, 2000-2016 2000 2010 2016 2000 Percent 2010 Percent 2016 Percent of Tract of Tract of Tract Total Total Total Total 3,982 4,000 3,686

African- 3,096 2,956 1,954 77.7% 73.9% 53.0% American

Non- 758 851 1,389 19.0% 21.3% 37.7% Hispanic White Other 128 193 343 3.2% 4.8% 9.3%

Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

A Comparison of Tract 55 with the City of Cincinnati on Select Economic, Social, and Housing Characteristics, 2000-2016: Table 6.2 compares Tract 55 with the City of Cincinnati for median household income, the percentage of population over 25 with at least a bachelor’s degree (educational attainment), median housing values, and median rent in the years 2000, 2010, and 2016. A median household income below the city average in 2000 is what made Tract 55 eligible for gentrification (according to Hammel and

Wyly’s 1996 criteria). In 2000, Tract 55’s median household income was slightly below that of the City of Cincinnati (0.96). By 2016, Tract 55 had a median household income that was slightly above that of the City (1.02).

While there was little change in aggregate median household income for Tract 55 in comparison to the City between 2000 and 2016, that was not the case for educational attainment, which increased substantially in Tract 55 during this time. In 2000, only 15.4% of the population over age 25 in Tract 55 had at least a bachelor’s degree compared to 26.6% for the City. By 2016, Tract 55 had 35.4% educational attainment compared to 33.8% for the City. These increases in educational levels for both

Tract 55 and the City are not surprising in that the percentage of the U.S. population with a college degree has been increasing every decade since 1940 (U.S. Department of Education, 1993; U.S. Census, 2000

129 and ACS 2010). However, the doubling of the college-educated population in Tract 55 over only sixteen years indicates in-migration of higher-educated individuals into Tract 55 and/or the exodus of lower- educated individuals. The discrepancy between growth of median household income and growth of educational levels in Tract 55 is not surprising. Many highly educated in-migrants to gentrifying neighborhoods are in the early stages of their careers and have not had time to earn higher incomes

(Blasius, Friedrichs, and Ruhl, 2016) or are in occupations that require higher education, but do necessarily generate high incomes (e.g. artists, community activists, social workers).

Madisonville has been experiencing economic revitalization since 2000 and it is likely that it is in the early stages of gentrification. During this stage, individuals with less financial resources move into the neighborhood as cost of living tends to be lower (Clay, 1979). According to Smith (1979), gentrification will occur outside of the core of the city after central city neighborhoods have been revitalized.

Investments in Downtown and Over-the-Rhine had already been capitalized upon by 2010 leaving investors searching for other neighborhoods where the payoff will be greater. While not completed during the years examined in this study, two hundred million dollars has been set aside for investment in

Madisonville at the intersection of Madison and Red Bank Roads (Tweh, 2016). This is an expansion of

Medpace and will result in 250 multi-family housing units, 250,000 square feet of office space, 100,000 square feet of commercial space, and a 239-room hotel/conference center. It is likely that these investments will attract middle-class and professionals who will replace the early pioneers of gentrification (Clay, 1979).

In 2000, median housing values in Tract 55 were only 78.5 percent of those of the City. Both

Tract 55 and the City experienced increases in housing values between 2000 and 2010, likely due to the housing bubble. However, Tract 55 increased housing values by 18.1 percent between 2010 and 2016 while the City experienced a 7.2 percent decline in housing value over that same time period. By 2016, median housing values in Tract 55 had increased to 87.4 percent that of the City, still allowing homebuyers with limited financial resources to purchase housing in Tract 55 at a reasonable price.

130

Median rent increased substantially between 2000 and 2016 for both Tract 55 and the City.

However, the City’s 49.1 percent increase between 2000 and 2016 was below that of the 78.1 percent increase for Tract 55. Whereas median rent in Tract 55 was 89.4 of what it was in the City in 2000, it was

1.07 that of the City in 2016. While owner-occupied housing was still a bargain in Madisonville in 2016 in comparison to other parts of the City, increases in median rent may have put pressure on those with less financial resources and led to out-migration.

There is evidence to suggest that affordable housing in Madisonville may be experiencing pressure. According to a Community Impact Report using data from the MLS of Greater Cincinnati, the average sale price of houses in Madisonville in 2013 was only $65,736 but by 2017 the average sale price was $117,00019. This would seem to suggest that new housing and revitalization of aging housing stock had increased demand for housing in Madisonville. This is not surprising as millions of dollars of investment has been slated for Madisonville redevelopment (Tweh, 2016) during the past decade. Even with increases in median housing values in Madisonville, these prices are more affordable than median housing values in neighboring Hyde Park and Oakley and may have incentivized some homebuyers to purchase in Madisonville. The latest data from Zillow (2019) estimated median housing values of

$358,300 for Hyde Park, $270,100 for Oakley, but only $106,700 for Madisonville.

19 The data from MLS of Greater Cincinnati are for all of Madisonville which includes Tracts 55, 56 and 108.

131

Table 6.2. Median Household Income, Percentage Population over 25 with at least Bachelor’s Degree, Median Housing Values and Median Rent for City of Cincinnati and Tract 55, 2000-2016: Tract City Tract City Tract 55 City Cinti. 2016 55 2000 Cinti. 55 2010 Cinti. 2016 2000 2010 Med. Income $28,424 $29,493 $32,342 $33,681 $35,231 $34,629

Educ Bach + 15.4% 26.6% 22.1% 31.7% 35.4% 33.8%

Median $73,700 $93,000 $89,000 $129,700 $105,100 $120,300 Housing Value Median Rent $397 $444 $632 $593 $707 $662

Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Select Housing Characteristics for Tract 55, 2000-2016: Total housing units in Tract 55 increased from 2,052 units in 2000 to 2,381 units by 2016, an increase of 16.0 percent (See Table 6.3). Recall that the total population of the tract declined by 7.4 percent during this same period. It is likely that the increase in dwelling units was a result of the declining size of households. According to national trends concerning household size, 88 percent of household growth between 2005-30 will be from childless individuals (Center for Neighborhood Technology, 2011).

These trends are a result of the baby-boomers aging into the empty-nest/retirement stage as well as young adults who are delaying or foregoing child-bearing. Many individuals in these categories are attracted to densely settled, older neighborhoods closer to the city (Center for Neighborhood Technology, 2011).

Madisonville is noted as having at least fifteen different architectural styles (Madisonville NBD Urban

Renewal Plan, 2002) and this array of housing stock would attract a variety of persons. The increase in housing units came about largely through new construction of apartment buildings (both market-rate and subsidized) as well as single-family housing built by the Madisonville Community Urban Redevelopment

Corporation (MCURC)20. The Quality of Life Plan, initiated in 2008, focused on removing dilapidated housing from the inventory of housing units and replacing those units with new units (Demeropolis,

20 Interview with Bill Fischer, Interim Director of MCURC, May 21, 2019.

132

2018), which is in accordance with a long-range revitalization plan focused on the Madisonville Business

District (Go Cincinnati). This remediation of blight and the consequent reduction in crime by 25.2 percent between 2015-17 made Tract 55 more attractive to newcomers with greater economic resources

(Community Impact Report, 2013). As previously stated, revitalization efforts along the Whetsel/Madison intersection have attracted additional investment.

In 2000, 11.3 percent of dwelling units in Tract 55 were vacant. The high percentage of vacancies in 2000 was likely a result of blighted buildings that were not suitable for occupation but had not yet been targeted for demolition. Vacancy increased to 17.0 percent by 2010, likely due to the recession.

Surprisingly, vacancy rates were still 16.9 percent in 2016 after the concerted effort at removing dilapidated buildings by the City since 2008. This failure to reduce the percentage of vacant housing units suggests that recovery from the Great Recession was slow. Recall that the ACS data is an average of vacancy rates between 2012 and 2016 and thus much of the recent in-migration to Tract 55 was not recorded as of 2016.

Owner-occupied housing in Tract 55 underwent a major transformation between 2000 and 2016.

In 2000, 43.4 percent of the population was in owner-occupied housing. By 2010, only 34.3 percent of residents lived in owner-occupied housing suggesting that the Great Recession negatively affected homeownership in Madisonville. Madisonville had the third highest foreclosure rate in Cincinnati in 2008

(Brown and Baize, 2009), and by 2016, only 24.9 percent of residents were in owner-occupied housing.

Another factor that would initiate out-migration from the tract is the increase in median rent between

2000 and 2010 (59.2 percent). The percentage of the population residing in the same residence as the previous year was higher for 2010 (78.9%) than 2016 (74.6%). This suggests that the Great Recession prevented some individuals from moving into Tract 55 even though the tract was in the process of gentrifying. Even if housing was available, potential residents may not have had the economic resources to enter the neighborhood. We have no idea about the mobility rates of out-migrants from the tract during this time as the census does not track these individuals.

133

Table 6.3. Select Housing Characteristics for Tract 55, 2000-2016: 2000 2010 2016 Owner-occupied Housing 43.4% 34.3% 24.9% Percent Vacant 11.3% 17.0% 16.9% Number of Housing Units 2,052 2,423 2,381 Same Residence One Year Ago X 78.9% 74.6% Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Comparison of Median Household Income for African-Americans and Whites in Tract 55 with that of City: In 2000, African-American median household income in Tract 55 was 134 percent that of the

City of Cincinnati. Madisonville is noted as a middle-class community (Brown and Baize, 2009) and thus a higher median household income for African-Americans in Tract 55 than their counterparts in the City was not surprising (See Table 6.4). However, the financial vulnerability of the middle-class African-

American population residing in Tract 55 was visible by 2010. Whereas median household income increased for African-Americans in the City, it declined between 2000 – 2010 for those African-

Americans residing in Tract 55. This suggests that there was a slight exodus of middle-class African-

Americans from Tract 55 between 2000 and 2010. However, median household income of African-

Americans in Tract 55 was still 1.23 times that of the City in 2010. By 2016, median household income of

African-Americans in Tract 55 was only 0.75 times that of the City suggesting that a mass exodus of middle-class African-Americans occurred due to the housing crisis which affected Madisonville more acutely than the City as a whole (Brown and Baize, 2009). We assume that a mass exodus of middle-class

African-Americans occurred as the African-American population declined by one-third between 2010 and

2016. Conversely, in 2000, the median household income of whites in Tract 55 was substantially less

(only 75.9 percent) than their white counterparts in the City. By 2016, white median household income was 1.02 times that of their white counterparts in the City suggesting that an in-migration of higher- income whites to Tract 55 occurred between 2000 and 2016.

134

Table 6.4. Comparison of African-American and White Median Incomes for Tract 55 and City of Cincinnati, 2000-2016. 2000 2010 2016 City-Wide (African- $20,984 $22,816 $21,814 American) African-American (T 55) $28,196 $27,986 $16,375 City-Wide (White) $36,467 $47,067 $50,937 White (T 55) $27,692 $36,199 $51,800 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Comparison of Select Characteristics of African-Americans and Whites in Tract 55: Median household incomes in Tract 55 were comparable for whites and African-Americans in

2000 (See Table 6.5). Madisonville was considered a racially balanced middle-class neighborhood for decades and thus it is not surprising that income was comparable21. By 2010, white median household income was 1.29 times higher than that of African-Americans. Whereas median household income for whites increased by 30.7 percent between 2000 and 2010, median household income slightly declined for

African-Americans by 0.7 percent. By 2016, white median household income increased by another 43.1 percent. In contrast, African-American median household income plummeted by 41.5 percent (from

$27,986 in 2010 to $16,375 in 2016). As a result, by 2016 white median household income in Tract 55 was 3.16 times that of African-Americans. It is likely that this process of displacement of middle-income

African-Americans was initiated by the housing crisis. Homeownership rates for African-Americans in

Tract 55 declined from 48.8 percent in 2000 to 25.4 percent in 2016. Whites also experienced a decline in homeownership rates from 37.3 to 24.8 percent between 2000 and 2016, suggesting that the housing crisis in Madisonville also affected whites. However, the addition of up-scale rental units in Tract 55 would be more attractive to middle-class (mostly whites) who were not ready for the commitment to homeownership.

In 2000, only 13.1 percent of the African-American population over 25 had a college education versus 22.9 percent for whites. By 2010, this has declined to 9.1 percent of the African-American

21 Recall that Madisonville consists of three census tracts which have averaged about 50 percent African-American over the past several decades.

135 population and increased to 57.1 percent of whites in Tract 55. This suggests a substantial in-migration of well-educated whites to Tract 55.

The percentage of Tract 55’s population in the same residence as the previous year declined for whites from 72.4 percent (2010) to 65.6 percent (2016). By 2016, the recession had ended and the increase in white population likely led to recent in-migration to Tract 55. Conversely, African-American population in the same residence increased from 83.4 (2010) to 87.8 percent (2016). Recall that any

African-Americans who left the tract during this time would not be tracked in the census data unless they relocated to another residence in Tract 55.

Table 6.5. Select Characteristics for African-Americans and Whites in Tract 55, 2000-2016. Educational Same House 1 Median Attainment year ago Household Income African- 14.0% 87.8% $16,375 American 2016 African- 9.1% 83.4% $27,986 American 2010 African- 13.1% X $28,196 American 2000 White 2016 54.8% 65.6% $51,800 White 2010 57.3% 72.5% $36,199 White 2000 22.9% X $27,692 Source: ACS 2006-10 and 2012-16.

Google Streetview Analysis: Tract 55 consists of ninety blocks. Using the random number table, I selected 18 blocks for analysis according to Hwang’s (2015) methodology using Google Streetview22. This resulted in the analysis of 60 block faces. In 2009 (See Figure 6.2A), only 5 of 18 blocks showed evidence of gentrification (a gentrification score exceeding 0.50 by Hwang’s criteria). Median household income, educational attainment, and median housing values/rents all increased at rates above the City average

22 Hwang’s blocks for her Chicago study ranged from about 10 to 20 blocks per tract of which she randomly selected four blocks. I modified this and selected 18 blocks which would be about 20 percent of the blocks in the Tract which would be comparable to Hwang’s methodology.

136 indicating that Tract 55 underwent gentrification between 2000 and 2010. However, supporting this gentrification process with evidence from the built environment using Google Streetview is problematic.

Google Streetview imagery was available for 2007 for Northside, College Hill, and Walnut Hills but

Tract 55’s imagery did not begin until 200923 shortening the period of analysis for documenting gentrification.

By 2014/16, 13 of 18 blocks had experienced gentrification (See Figure 6.2B). The block with the greatest change in scores from 2009 to 2014/16 was 1003. This block went from disinvestment to middle- stage gentrification during this time. This is not surprising in that early gentrifiers often use ‘sweat equity’24 to revitalize housing stock (Clay, 1979). An examination of the sampled tracts for the Google

Streetview analysis reveals no spatial pattern to the gentrification process as of 2009. The gentrified block in the western part of the tract had a score of 0.89 and likely had never experienced disinvestment. The four blocks that gentrified in the eastern part of the tract are spread out and reveal no spatial pattern. The general idea from viewing the 2009 map is that disinvestment was rampant in Tract 5525.

Figure 6.2B displays gentrified blocks in Tract 55 as of 2014/16. Note that the gentrification process appears to be related to proximity of gentrified tracts in 2009 suggesting that blocks adjacent or near gentrifying areas are more likely to undergo gentrification than those further away (Hammel and

Wyly, 1996)26. The average gentrification score for Tract 55 in 2014/16 was 0.56 which placed it in the

23 A few block faces did have imagery from 2007 but 2009 was used as the base year because all block faces in the sample had imagery from that year. 24 Sweat equity is a term used to indicate that individual homeowners renovate their homes as time and money permit. 25 It appears that the western part of Tract 55 was oversampled (Do you mean the eastern side?). This is not unexpected in that the blocks in this part of Tract 55 are smaller than other parts of the Tract. About 2/3s of the blocks are located east of Stewart Avenue. Several of the blocks in the west were discarded in the sample as they contained private roads which were not accessible with Google Streetview or imagery for some of the roads was only available for one year making a comparison between two time stamps impossible. 26 The large number of blocks that were unsampled makes these observations largely unsubstantiated but does serve as an indictor of the gentrification process.

137 early to mid- gentrification stage. During this analysis, two block faces (one located in 4027 and one in

1030) presented worse conditions in the built environment in 2014/16 than in 200927.

27 The blocks did not experience declines in overall block score because other of the block faces were worse in 2009 than in 2014/16.

138

Figure 6.2A. Gentrified Blocks in Tract 55 in 2009.

139

Figure 6.2B. Gentrified Blocks in Tract 55 in 2014/16.

140

Table 6.6. Sampled Blocks in Tract 55 and Stage of Gentrification for 2009 and 2014/16 Key: Disinvestment below 0.50 Early between 0.50-0.57 Middle between 0.58-0.64 Late between 0.65-0.80 Turnover over 0.80

Block 2009 2014/16 Difference between 2007 and Score Score 2014/16 and Stage of Gentrification Block 1000 0.47 0.52 Went from disinvestment to early gentrification Block 1003 0.44 0.61 Went from disinvestment to middle-stage gentrification Block 1004 0.44 0.51 Went from disinvestment to early stage Block 1010 0.37 0.48 In disinvestment stage in each year, but improvement Block 1016 0.47 0.52 Went from disinvestment to early stage Block 1019 0.55 0.55 Was in early gentrification/ or non-disinvestment in each year Block 1026 0.33 0.48 In disinvestment stage in both years, but improvement Block 1030 0.44 0.44 No change on block Block 1032 0.61 0.67 Middle-stage gentrification or no disinvestment Block 3017 0.37 0.48 In disinvestment in both years, but improvement Block 3018 0.37 0.53 Disinvestment to early stage Block 3019 0.89 0.89 No disinvestment in either year Block 4008 0.51 0.55 In early stage of gentrification in both years Block 4010 0.75 0.81 In late stage gentrification or not disinvested to turnover stage Block 4015 0.41 0.52 Disinvestment to early stage Block 4017 0.41 0.48 In disinvestment in both years, but some improvement Block 4027 0.44 0.44 In disinvestment in both years Block 4028 0.44 0.52 From disinvestment to early stage

Tract 0.48 0.56 Average Source: Google Streetview and Hwang (2015).

141

Figures 6.3A and B are images from Block 3018 along Orlando Place. Figure A shows a parking lot that appears to be underutilized in 2009 (0.44 score for the block face). This location is close to the railroad tracks that separate Tract 55 and Tract 56. The two trucks in the picture suggest that products came by rail and were redistributed via trucks. Obviously, rail is not the major method of shipment in the

21st century and the highways probably made this location less appropriate as a distribution point. By

2016, a new building has been constructed (OAR Spay/Neuter Clinic), a new fence has been added and it appears that the parking lot has been resurfaced (0.72 score for block face).

142

Figure 6.3A. Parking Lot Along Orlando Place, 2009.

Source: Google Streetview, 2009.

Figure 6.3B. The Newly Built OAR Spay/Neuter Clinic along Orlando Place, 2016.

Source: Google Streetview, 2016.

143

Figures 6.4A and B are images of a residential area along Kenwood Road in Block 1016. This block face scored 0.33 in 2007 largely due to the houses that were in disrepair. Note the rusted porch roof on the house to the left. The lawn is mown but no additional adornments have been applied to the landscape. Sidewalks and driveways are in some disrepair. Note in 2016 that the house to the right has been demolished and a well-kept vacant space has taken its place. Also note the incremental change in the white house which received a fresh coat of paint on the porch. It also appears that a new sidewalk has been laid. This block face received a score of 0.55 in 2016.

144

Figure 6.4A. Disinvested Houses Along Kenwood Road in 2009.

Source: Google Streetview, 2007. Figure 6.4B. Vacant Lot and Façade Improvements Along Kenwood Road in 2016.

Source: Google Streetview, 2016.

145

Figures 6.5A-C show businesses along Whetsel Avenue in Block 4027. Figure A shows the corner business as being boarded-up in 2009. There is an existing convenience store and another store that is likely out of business, but not boarded-up. Figure B shows the same series of stores in 2016. Each of these stores has been boarded-up suggesting that business was worse in 2016 than it was in 2009.

However, it must be remembered that much of the revitalization plans for Madisonville’s Business

District were put in place after 2016 suggesting that this may have been a planned closure of businesses.

Figure C shows this strip in 2018 with new windows and it is likely that this area will have a different set of businesses from what it has replaced. Several residents of Madisonville are concerned that new businesses in the business district will be up-scale shops that will cater to the gentrifiers in Tract 55 as well as customers from outside the community but provide little products/services to those with less financial resources (Weingartner, 2016). This renovation of businesses along Whetsel Avenue is part of a larger project.

146

Figure 6.5A. Commercial Disinvestment Along Whetsel Avenue, 2009.

Source: Google Streetview, 2009. Figure 6.5B. Boarded-up Strip of Stores Along Whetsel Avenue, 2016.

Source: Google Streetview, 2016. Figure 6.5C. Newly Renovated Stores Along Whetsel Avenue, 201828.

Source: Google Streetview, 2018.

28 On June 13, 2019, I canvassed this block to document changes since 2016 in the built environment of Madisonville. These renovated stores were in the same condition as they were in 2018-unoccupied. The location of the stores is at most a five- minute walk from the Whetsel Avenue/Madison Road intersection with its mixed-use development project. It may be that the owners of these businesses are waiting for completion of the mixed-use project before opening their businesses.

147

Figures 6.6 A and B display imagery of a vacant lot on Owasco Street in 2009 and a newly-built residence as of 2016. It is likely that the vacant area in 2009 was a dilapidated structure that was removed when the City implemented its demolition of blighted properties in Madisonville. The 2009 gentrification score for this block face was 0.33. Note that the new house has been built in the style of the older houses on the street and is consistent with the style of the existing housing stock. The 2016 score for the block face was 0.61.

148

Figure 6.6A. Well-cared for Vacant Lot along Owasco Street, 2009.

Source: Google Streetview, 2009.

Figure 6.6B. Newly Built Single-Family Home Along Owasco Street, 2016.

Source: Google Streetview, 2016.

149

Figures 6.7A and B are images from Ward Street in Block 1003. The 2009 image displays a well- kept vacant lot, but the sidewalk is in bad shape. This block face scored 0.33 in 200929. In the 2014 image a community garden has been planted in the vacant lot. The sidewalks have been repaved and an attractive fence has been added. This community garden is one of 500 gardens in Madisonville that were created between 2014 and 2017 to allow residents and businesses to grow their own food (Madisonville

500 Gardens).

29 While the vacant plot of land was well cared for the rest of the block showed houses in disrepair which lowered the score for the block face.

150

Figure 6.7A. Vacant Lot Along Ward Street, 2009.

Source: Google Streetview, 2009. Figure 6.7B. Community Garden Along Ward Street, 2014.

Source: Google Streetview, 2014.

151

Update on Built Environment in Tract 55 as of June 2019:

On June 13, 2019, I canvassed Tract 55 to document changes in the built environment since 2016.

The most conspicuous change in the built environment was the mixed-use complex at Whetsel Avenue and Madison Road where large-scale development has occurred. Secondly, as I walked through the sampled blocks in my study, I noticed several newly-built houses. Demolition of dilapidated housing and in-fill has been a strategy employed by MCURC and private investors for enhancing the built environment in Madisonville.

Figures 6.8A shows a vacant lot at the intersection of Whetsel Avenue and Madison Road in June of 2018. By June 2019, the Ackerman Group had begun construction of its mixed-use development project (Figure 6.8B). This example continues to illustrate that Madisonville had made use of in-fill and new development to spur economic growth in not only the decayed business district but also throughout the community.

152

Figure 6.8A and B. Mixed-use development located at Whetsel Avenue and Madison Road (Ackerman Group).Before (2018) and After (2019).

Source: Google Streetview, 2018.

Source: Photo by E. Ravuri, June 13, 2019.

153

Figures 6.9A and B are the same house along Peabody Avenue in 2019 and 2014. While canvassing the block, I mistakenly identified this house as in-fill. When I checked Google Imagery, it was a house that had undergone a major façade improvement by 2019. What was interesting, was that the exterior of the house appeared in good condition in 2014, but additional investments were made to make the house even more attractive. Note that the porch roof was replaced by 2019. This could have been a problem with the structure or it may have been a style preference. These changes in a house that appeared in good condition could indicate that ‘flippers’ have invested in this dwelling or that the homeowner is capitalizing on his/her investment in the newly revitalizing area around the mixed-use Ackerman Project at Whetsel Avenue and Madison Road, which is only one block away.

154

Figure 6.9 B. 5805 Peabody Avenue, 2014.

Source: Google Streetview, 2014. Figure 6.9B..Renovated house, 5805 Peabody Avenue in 2019.

Source: Photo by E. Ravuri, June 13, 2019.

155

Figure 6.10. New in-fill houses along Ravenna Street.

Source: Photo by E. Ravuri, June 13, 2019.

156

Interview of Members of the Madisonville Community: In December 2018 an email was sent to MCURC (Madisonville) explaining the project and asking the agency to provide contacts for individuals who could speak on economic revitalization and demographic change in Madisonville between 2000 and 2016. During May of 2019, the following individuals were interviewed:

Byron Coaston: Life-long resident of Madisonville (since 1950s). Byron received a bachelor’s degree in

History from Northwestern University. He is an active participant in community revitalization in

Madisonville.

Bill Collins- Resident of Madisonville since 1989. Bill is a graduate of the University of North Carolina’s

History program. He has been studying the history of Madisonville for several decades.

Bill Fischer- Resident of Madisonville since 2007. Bill’s academic background is in Public Finance. He has worked for the City of Cincinnati for several decades in a variety of economic development projects.

He is currently the Interim Director of MCURC.

Chris Handlin- Resident of Madisonville since 1994. Chris is a retired Architect with a Master’s Degree from the University of Michigan.

Sara Sheets- Resident of Madisonville since 2003. Sara received a Master’s in Urban Planning from

University of California, Los Angeles. She was the director of MCURC until early 2019.

The following are questions were asked of each of the interviewees. Not all interviewees could answer each question.

157

1. Can you pinpoint a year (s) when you first started to see changes in Madisonville suggesting that

economic reinvestment and/or revitalization was occurring? Where did the process begin within

Madisonville? Please describe these processes. Be as specific as possible.

Interviewees largely acknowledged that revitalization in Madisonville began around 2008. This was when

MCURC was established. In 2012, the Quality of Life Plan was implemented and was a further catalyst to development.

Sara- Crime was an issue in Madisonville and before revitalization could proceed, this issue had to be addressed. Sara moved to Madisonville in 2003 and remarked that drug sales were a common occurrence on her street. Specifically, there were three abandoned houses where this activity was occurring and through working with the City of Cincinnati and Madisonville Police, these houses were torn down and the problem removed. Sara was instrumental in the founding of MCURC which was an organization that was created to address problems in Madisonville. Revitalization efforts really started to unfold after the

Quality of Life Plan (2012) was adopted. Sara sees this plan as a grassroots effort to revitalize

Madisonville with the assistance of funding from the City of Cincinnati.

Chris- Chris moved to Madisonville in 1994 and remarked that gang activity and crime discouraged community members from being out on the street. Like Sara, he witnessed drug deals in Madisonville’s blighted business district. There were not too many businesses in downtown Madisonville and Chris puts the nadir of the economic decline of Madisonville around 2000. Chris cites the Quality of Life Plan

(2012) as having a major impact on revitalization by focusing on education, crime reduction, community gardens, and health provision. Chris sees MCURC’s responsibility as stimulating the business district by acquiring property, controlling development and advocating for the community. After 2012, Chris noted that renovations to older buildings in Madisonville was becoming more commonplace. In the past few years he has noted that there has been a significant decline in the number of housing vacancies in

Madisonville.

158

Bill F.- Bill has lived in Madisonville since 2004. Before working for MCURC, he worked as a City

Manager on Economic Development. In the early 2000s, the City of Cincinnati was looking for several neighborhoods to invest in as catalysts to economic growth and revitalization. Spreading funds over 52 neighborhoods was an inefficient method for generating economic development. The task force

(Neighborhood Enhancement Program) examined available property in the neighborhood, the market for goods and services produced in each of the neighborhoods, and the transportation infrastructure (e.g. proximity to highways, public transportation). Four neighborhoods were selected as catalysts for development in Cincinnati through this process: Madisonville, Bond Hill/Roselawn, Walnut Hills, and

College Hill. Within Madisonville four blocks were selected for development and were located at the intersection of Whetsel Avenue and Madison Road. Bill notes that revitalization in the Madisonville

Business District has been underway since about the past five years.

Byron: Byron recalls the 1960s when downtown Madisonville was economically vibrant. He recalls that every corner had a gas station and there were numerous car dealerships along Madison Road. However, by the 1970s many businesses had closed or relocated to the suburbs ushering in the decline of the

Madisonville business district. By the early 2000s, numerous buildings in downtown Madisonville were demolished and nothing was erected in their place. By 2014 (about the last five years) Byron has noticed that a tremendous amount of development has focused on the business district.

2. What effect did the Great Recession (2007-2009) have on economic reinvestment/revitalization in

Madisonville?

While the Great Recession resulted in the closure of numerous businesses in Madisonville (as it did in the rest of the U.S.) other businesses arose to take the place of those businesses. Madisonville’s designation as an industrial zone declined significantly but manufacturing was on its way out long before the Great

Recession. This transition from a manufacturing-based to a service-based economy is a catalyst to

159 economic growth and in economic geography this is referred to as ‘creative destruction’ of economic sectors (Esslelizbichler, 2004).

Bill C. – Madisonville was a major core of manufacturing in Cincinnati until the late 1900s, but that has changed substantially. There is now (May 2019) a concentration of jobs in Madisonville that did not exist in 2007 (beginning around the Great Recession). Medpace and Fifth Third Bank Nationwide Operations are two major employers which have infused new economic life into Madisonville and are service occupations. These types of jobs attract a different workforce than the industrial workers of the 20th century.

Chris- As mentioned previously, Chris dates the economic collapse (from his time living in the neighborhood) of Madisonville around 2000 and thus does not attribute much affect from the Great

Recession as businesses and housing values were already so low. He did find that there were more programs and events to try to keep community cohesiveness alive during this time. This was also the time that MCURC was established which provided critical support to the community.

Bill F.- A few large businesses were lost during the Great Recession, but others evolved to take their place. Much of the build-up focused on the Red Bank Industrial Corridor but this revitalization plan was already in place prior to the Great Recession. The original plan was to bring industry back to the Red

Bank industrial corridor, but this did not come to fruition. Instead, an office corridor arose. This has likely been more beneficial to economic revitalization in Madisonville then bringing back industry. Along with the new office corridor, new housing has been built in the area, revitalizing the entire area.

Byron: The Great Recession forced many homebuyers into foreclosure and led to vacant housing stock throughout Madisonville. A program called ‘Weed and Feed’ was initiated during the Great Recession to help people renovate their homes (feed) while those buildings that were considered not salvageable were demolished (weed).

160

3. Can you comment on the impact of private development agencies versus non-profit/government

agencies in the reinvestment/revitalization process?

Sara- The non-profit sector (MCURC and Habitat for Humanity) has acted with the community vision in mind. That vision is a strong commitment to diversity which encourages a place for everyone in

Madisonville. Private developers (whether individuals or corporations) have had a positive impact on economic revitalization in Madisonville. There was too much blight and too many vacant homes for the non-profit corporations to effectively combat this issue. Another positive issue that has arisen from the non-profit sector is that 20 percent of housing units built by large private development corporations (that use public funds) need to be affordable units.

Chris- MCURC (non-profit) was instrumental in property acquisition which helped to control development and promote community advocacy. MedPace, a private investor, is a positive influence on development and a ‘good corporate citizen’ which has promoted new housing developments, restaurants, and retail. One negative consequence of the redevelopment occurring in Madisonville is that some private investors are holding property to see if they can turn a profit on the property in the future.

Bill F.- The large private investors such as Ackerman Group, Model Group, and Circle Development have been favorable for development in Madisonville. Ackerman Group was selected by MCURC as the developer for much of the renovations that are occurring in the Whetsel/Madison corridor and will be a mixed-use development containing affordable housing. Model Group focuses on affordable housing and is currently redeveloping Madison Villa. Circle Development is providing affordable housing options for the elderly and is involved in building nursing homes. The Episcopalian is a hybrid developer (non- profit/private) based on providing affordable housing. One issue is the number of absentee landlords who are lax on upkeep of properties in Madisonville and contribute relatively little to the economic and social development of the community.

161

4. Has reinvestment/revitalization led to a change in the demographic composition (e.g. age,

race/ethnicity, socio-economic status) of Madisonville? If so, in what way? Have certain areas

been more affected than others? How did the Great Recession affect demographic composition in

Madisonville?

Bill C.- There was a major exodus of the middle-class African-American population around 2008 and this was likely a result of the housing crisis. Conversely, the affordability crisis in housing has attracted the white middle-class who want to live on the East side but cannot afford housing in Hyde Park and Oakley.

Bill thinks the major question in the future is how to attract the African-American middle-class back to

Madisonville? Bill feels that one of the impediments to this is that public schools in Madisonville are not of high quality. Middle-class whites in Madisonville (not unlike other communities) send their children to private schools. According to Bill “it is easier to integrate communities more difficult to integrate schools”. Bill has also noticed that there has been an influx of Asians (specifically Indians) into

Madisonville in the past decade. He attributes this to the Fifth Third Bank Madisonville Operations which has a contract with Tata corporation in India to train individuals who then return to India with skills from the U.S.

Bill F.- Bill has noticed an influx of a younger, white population and a compositional change in demographics over the past few years. In 2013, Madisonville started the Jazz and Barbecue Fest which was a draw for the African-Americans in Madisonville and the surrounding area. Within the past year or so, Bill has noticed that the festival has attracted many white, millennials who have recently settled in the neighborhood.

Chris- The total decline in population has become noticeable over the past 25 years. Within the past few years, there has been an influx of whites in their thirties with children. Chris notes that he is often on the streets and parents are pushing their children in strollers. This was not a common occurrence in the past.

162

Byron: The socio-economic structure of Madisonville has changed substantially in the last five years.

Byron used to be surrounded by ‘poor’ people and now the middle-class has moved into Madisonville. He also notes that in-migrants are mostly white young adults and young families.

5. Can you comment on the changes in types of businesses or improvements in already established

businesses in your community during the gentrification process?

Bill F.- There has been a change in the types of businesses in Madisonville. Madisonville is beginning to attract up-scale businesses. One example is the Mad Llama which is a coffee shop on Madison Road which caters to a gentrified clientele. Lala’s is a small business which sells candies and is having a difficult time acquiring a storefront which charges an affordable rent. MCURC is attempting to help small-business owners who operate on a small budget and can’t afford to pay high rents. Many of the subsidies that formerly allowed small businesses to survive in Madisonville have become less available.

Bill stresses that MCURC is committed to equitable development in the business district. Up-scale businesses bring needed finances to the community but can displace small businesses but need to be included in the development plan as they bring much needed funds. There is also the concern that

Madisonville should not emulate the up-scale businesses that are found in Hyde Park and Oakley.

6. Is there anything else you would like to mention concerning economic revitalization and

demographic change in Madisonville?

Sara, Bill C., and Chris all stated that they were attracted to Madisonville when they moved to Cincinnati because of its diversity (socio-economic, race/ethnicity), its historical housing stock, affordability, and location.

One issue that arose through the interview was that of affordable housing, something I overlooked in framing the interview questions.

163

7. Topics specific to neighborhood: Affordable Housing

Bill F.- MCURC has built four affordable single-family housing units and will complete another five units by the end of 2019. MCURC has applied for funding to build another four units which are expected to be completed after 2020. However, these affordable houses are becoming too expensive to build and sell at a reduced price. Furthermore, even if MCURC sells the houses at an ‘affordable price’ the upkeep on the houses may be prohibitive for those of limited financial resources. Bill envisions another phase of housing that focuses on multi-family rental properties which increases density and fits in with the mixed- use strategy that Madisonville is pursuing in economic revitalization.

Sara- MCURC has constructed several affordable houses on vacant lots and is trying to maintain the low- income population by providing affordable housing opportunities. Any new developments that use public funds need to provide twenty percent of units that are affordable.

Chris- Rents continue to increase and it appears some individuals cannot afford to reside in Madisonville.

At the same time, private investors have bought properties and have yet to revitalize or sell them as they are waiting for prices to increase. ‘The Red’ is a current construction project that will produce a high-end housing complex but will not be suitable for those who can no longer afford to reside in Madisonville.

8. Additional Comments

Bill F.- Bill noted (and the Google Streetview analysis supports this) that revitalization in Madisonville has taken longer to come to fruition than that of the development that has occurred in Walnut Hills and

College Hill (two of the other four communities that received city dollars for reinvestment). He attributes this to the fact that Madisonville had less historical buildings than the other two communities and for those who are nostalgic for the architectural past, this would be less attractive for settling or for establishing a business. Most of the buildings in the business district in Madisonville along Whetsel and

164

Madison needed to be torn down because they were in such disrepair. This new development provides a sharp contrast to what one finds in a neighborhood like Walnut Hills which has restored many buildings.

Conclusion: As whites moved into Cincinnati suburbs after the 1950s, many African-Americans took advantage of good quality housing in affordable neighborhoods at the city-suburban interface. While in many neighborhoods in the U.S. this process led to segregation, in Madisonville racial integration has been maintained since the 1970s (Brown and Baize, 2009). Madisonville has welcomed African-

Americans since the 1870s and by the 1880s a community of African-American homeowners had arisen30.

Unfortunately, racial integration in Madisonville may be questionable in the future. In the 21st century, the white middle-class has reestablished its presence in Madisonville, likely a result of the dispersal of the

African-American population caught up in the financial downturn of the Great Recession31. Even though

Tract 55’s African-American population comprised 77.7 percent of the tract population in 2000, whites and African-Americans were equally matched in their economic/educational characteristics.

Tract 55 now has a well-educated, middle-class white population and a lower-educated, lower- income African-American population. The Quality of Life Plan (2012) is an attempt to rectify this situation. As of 2013, Madisonville was the first neighborhood in Cincinnati to focus on implementing

Form Based Code32 as a method to attract clientele and residents to Madisonville by providing mixed-use opportunities and a walkable neighborhood (https://opticosdesign.com/blog/madisonville-becomes-first- cincinnati-neighborhood-to-implement-new-fbc/). These new developments can also help ensure that the

30 Interview with Bill Collins, Madisonville Historian, May 6, 2019. 31 It is likely that higher income and better educated whites have also displaced or replaced their white counterparts with less income and education. 32 College Hill, Walnut Hills, and Westwood also adopted the Form Based Code after 2013 as a method of revitalizing their communities.

165 low-income population has an opportunity to remain in the neighborhood by providing affordable housing.

According to national trends, 77 percent of ‘echo boomers’ (those born between 1982- 1999) want to live in the urban core (Center for Neighborhood Technology, 2011). Even if the housing stock in

Tract 55 is not what gentrifiers are looking for in terms of older housing with architectural character that could be found in neighboring Hyde Park or Oakley, it does provide a low-cost alternative for those who do not want to reside in the suburbs or who cannot afford properties in Hyde Park or Oakley. According to the Center for Neighborhood Technology (2011) Madisonville has the eleventh highest concentration of jobs in Hamilton County making this location a good choice for those who want to minimize transport time and/or costs. Unfortunately, this concentration has relatively few employment opportunities in manufacturing, the traditional way for those with low educational levels to enter the middle-class. Thus, it is imperative that the emphasis on affordable housing remain a top priority for Madisonville as it goes forward in the revitalization process. An additional effort on the part of MCURC is to create a more diverse economy.

166

Chapter 7 TRACT 19-WALNUT HILLS

Introduction: The purpose of this chapter is to examine the effect that gentrification has had on the African-

American population residing in Tract 19 (Walnut Hills) between 2000 and 2016. The chapter begins with an overview of the history of Walnut Hills and then examines census data disaggregated by race to determine how social and economic characteristics have changed by race between 2000 and 2016. A

Google Streetview Analysis is undertaken of the built environment in Tract 19 to determine which stage of the gentrification process the tract was in as of both 2007 and 2014/16 according to Hwang’s (2015) criteria for gentrification. The third component of the chapter focuses on interviews with key members in

Walnut Hills who are familiar with the economic and demographic changes that have occurred over the past two decades. Lastly, the author canvases Tract 19 to document recent evidence of gentrification (as of June 2019) through photographs.

167

Figure 7.1. Walnut Hills, Tract 19, Satellite Imagery.

168

Short History of Walnut Hills: Walnut Hills, settled in 1793, consisted of a farm on the outskirts of Cincinnati until 1829 when the Lane Seminary was founded (Walnut Hills Development Foundation). Affluent residents were attracted to the area because of its proximity to downtown Cincinnati and its green hills. The community was a stop on the Underground Railroad and subsequently attracted African-Americans who settled in the community after the mid-1800s (Walnut Hills Community Plan, 1992). Walnut Hills was annexed to the

City of Cincinnati in 1869 (Walnut Hills Community Plan, 1992), and during the 1870s and 1880s streetcars connected Walnut Hills with downtown Cincinnati and led to further increases in the population. Peebles’ Corner (located at Gilbert and McMillan Avenues) was one of the busiest intersections outside of downtown Cincinnati in the late 1800s (Writer’s Program, Ohio, 1943).

Furthermore, merchants at Peebles’ Corner had customers coming from as far away as Pennsylvania and

Tennessee to purchase bulk products such as flour, sugar, coffee, and other groceries (Writer’s Program,

Ohio, 1943). However, by the early 1900s, the suburban exodus had begun and many of Walnut Hills’ large homes were subdivided into multi-family residences that housed lower-income individuals. By the

1930s, Tract 36 in Walnut Hills was 90 percent black while Tract 2133 hosted one of the greatest concentrations of Italians in the city; both groups largely resided in slum conditions (Konermann, 2015).

Census Data: Table 7.1 displays population data for Tract 19. The total population of the tract declined by 17.6 percent between 2000 and 2010. It then grew by 8.8 percent between 2010 and 2016. The African-

American population went from 53.8 percent of the tract composition in 2000 to 38.8 percent by 2016.

Conversely, the white population went from 40.3 percent of the tract total to 51.9 percent during this time.

Between 2000 and 2010, the white population increased by 15.6 percent while the African-American population declined by 34.0 percent suggesting mass displacement of African-Americans, perhaps of

33 Tract numbers in the 1930s do not match the tract numbers as of 2010. In the past several decades tracts have been aggregated due to decreases in population.

169 those least able to bear the increased cost of living in a gentrifying tract. Changes in the built environment that focused on upscale housing likely caused some of this displacement between 2000 and 2010 (Walnut

Hills Reinvestment Plan, 2016). In contrast to the significant decline of the African-American population during the first ten years of the 21st century, there was only a 1.9 percent decline in African-Americans between 2010 and 2016. This occurred even as the white population increased by another 10 percent. A recent news article suggested that Walnut Hills was undergoing revitalization without displacement34

(Copsey, 2017) and that there has been an effort to ensure that low-income individuals are not displaced during the gentrification process. This concern with displacement began in 1977 with the founding of The

Walnut Hills Redevelopment Foundation which focused on affordable housing. A new plan ‘The Walnut

Hills Reinvestment Plan’ of 2016 has been put in place to further support low-income housing programs.

Other attempts to combat dislocation of low-income individuals include a 2015 proposal to transform a vacant business into apartments with day-care facilities for parents returning to school (known as Scholar

House) and MORTAR, a non-profit start-up that supports minority business owners. The WHRF has created partnerships with the Walnut Hills Area Council and Walnut Hills Business Group to become a catalyst for further sustainable and positive change in Walnut Hills. Their efforts now focus on six areas:

1) enhanced public spaces; 2) civic dialogue; 3) enhanced social services; 4) public/private development partnerships; 5) cultural programing; and 6) protection of historic assets (https://walnuthillsrf.org/).

34 Walnut Hills consists of 4 census tracts. The loss of one-third of the African-American population between 2000 and 2010 indicates displacement occurred in Tract 19. Some of these displacees may have relocated to other census tracts within Walnut Hills.

170

Table 7.1. Population Change in Tract 19, 2000-2016

2000 2010 2016 Percent Percent Percent 2000 2010 2016 of of of Tract Tract Tract Total Total Total Total 1,754 1,445 1,572

African- 944 623 611 53.80% 43.10% 38.90% American

Non- Hispanic 706 738 816 40.30% 51.10% 51.90% White Other 104 84 145 6.90% 6.80% 9.20%

Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Comparison of Tract 19 and City of Cincinnati on Select Characteristics, 2000-2016: Table 7.2 displays median household income, the percentage of the population over 25 with at least a bachelor’s degree, and median housing values/median rent for Tract 19 and the City of Cincinnati.

Note that Tract 19’s median household income was only 91.4 percent of that of the city in 2000. This is what characterized Tract 19 as a tract eligible for gentrification: a 2000 median household income below the city average as specified by Hammel and Wyly (1996). As of 2016, Tract 19 had a median household income that was 140.0 percent of the city average, suggesting gentrification had occurred. The percentage of population over 25 with at least a bachelor’s degree increased substantially from 2000 to 2010 in Tract

19, going from 39.0 to 52.1 percent. Compare this with the much lower percentages for the City of 26.6 percent in 2000 and 31.7 percent in 2010. The relatively high percentage of educated population in Tract

19 by 2000 suggests that educated individuals with modest salaries had begun to move in. These in- migrants could be the early gentrifiers who tend to be young and have not reached their full earning potential or have jobs that do not command high salaries such as artists or social workers who are looking

171 for lower cost of living in neighborhoods close to the central city (Clay, 1979). Tract 19 made no further gain in educational attainment as of 2016 even as the measure increased across the City35.

Both Tract 19 and the City experienced substantial increases in median housing values between

2000 and 2010 (45.3 and 39.5 percent; respectively), likely a result of the housing speculation that occurred during that decade. However, the City experienced a 7 percent decline in housing value between

2010 and 2016 whereas Tract 19’s median housing values increased by 11.4 percent suggesting demand for living in a gentrified neighborhood kept the housing values from declining. Median rent was fairly comparable (9% higher) between Tract 19 and the City in 2000, but by 2010, Tract 19’s median rent was

42% higher than that of the city. It is likely that living in Tract 19 became too cost burdensome for lower- income individuals and initiated the displacement process. Note that whereas median rent increased in the

City between 2010 and 2016, it declined slightly in Tract 19. This decline in median rent in Tract 19 may be a result of the commitment to provide affordable housing in Walnut Hills (Walnut Hills Investment

Plan, 2016).

35 While Tract 19 clearly gentrified between 2000 and 2010 according to Hammel and Wyly’s criteria (1996), between 2010 and 2016, Tract 19’s increases in median household income, percentage of population over 25 with a bachelor’s degree, and median rent did not keep pace with that of the increase of the city on these three measures. Again, this is likely a result of the effort to resist displacement of low-income population in Tract 19.

172

Table 7.2. Median Household Income, Percentage Population with at Least Bachelor’s Degree, Median Housing Values and Median Rent for City of Cincinnati and Tract 19, 2000-2016: City City City Tract 19 Cinti. Tract 19 Cinti. Tract 19 Cinti. 2000 2000 2010 2010 2016 2016 Median $26,967 $29,493 $46,750 $33,681 $48,438 $34,629 Income Educational Attainment 0.39 0.266 0.521 0.317 0.506 0.338 Median Housing $107,700 $93,000 $156,500 $129,700 $174,400 $120,300 Value Median Rent 482 444 842 593 803 662

Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Select Housing Characteristics in Tract 19, 2000-2016: See Table 7.3 for this discussion. Total housing units declined by 7.5 percent between 2000 and

2010, going from 1,117 to 1,033 units. This suggests conversion from smaller to larger units; demolition of sub-standard units; and possibly newly constructed housing units that replaced older, less functional units. Vacancy increased from 147 to 297 units, a hundred percent increase between 2000 and 2010. In

2000, 13.2 percent of housing units stood vacant whereas by 2010 this was 28.8 percent. This corresponds with the onset of the Great Recession and those with less economic resources may have been forced out by foreclosure or inability to pay rent. Also, some apartments that were converted to up-scale rental units during the first decade of the 21st century may have sat empty during the recession. Median housing values increased by 45.3 percent during this time while median rent almost doubled (going from $482 to

$842 dollars per month). The increase in owner-occupied housing (from 23.5 to 42.0 percent) between

2000 and 2010 does not suggest that foreclosure was a major issue in Tract 19 as of 2010. In fact, it suggests that some rental units may have been converted to owner-occupied status. The significant decline in renters (-40.8%)36 suggests that renting became too cost burdensome for some of the population

36 The 40 percent decline in renters in Tract 19 between 2000 and 2010 suggests that the increase in homeowners is a result of the mass exodus of African-Americans; many of which were likely low-income renters.

173 and these individuals left Tract 19. A 2016 analysis of rental demand in Tract 19 found that 31.8 percent of rental demand was for apartments that rented for less than $500 dollars per month (Walnut Hills

Reinvestment Plan, 2016). One can infer that the demand for affordable apartments was even higher between 2000 and 2010 before the mass displacement of low-income population and the onset of the

Great Recession. It is also likely that in-migrants to tract 19 who decided to rent had smaller household sizes than those they replaced and accounts for some of the decline in the percentage of population renting between 2000 and 2010. This could occur as single individuals or childless couples moved into rental units that were formerly occupied by larger families (Walnut Hills Reinvestment Plan, 2016).

37Only two-thirds of the population was in the same residence in the year prior to the 2010 ACS suggesting a substantial turnover in population38. It is likely that the combined processes of gentrification and recession forced low-income individuals to leave their residences39 and these vacancies were occupied by new arrivals, most likely gentrifiers.

The number of vacant units declined from 297 to 163 between 2010 and 2016. This decline cannot be a result of demolition of housing units or the conversion of double to single occupancy as only

11 units were lost. It is likely that the combined effects of the Great Recession and gentrification led to increased vacancy between 2000 and 2010 and then a recovery from this process once the recession ended. For example, Trevarren Flats40, located on East McMillan, consisted of three vacant buildings that were completely rehabilitated in 2016 and now consist of 31 luxury units (Walnut Hills Reinvestment

37 ESRI has a marketing program that categorizes geographical areas according to social, economic, demographic, and attitudinal data. According to the ESRI analysis, the Walnut Hills neighborhood is composed of City Commons (33%), Social Security Set (31%), Set to Impress (16%), Emerald City (11%), and Metro Renters (8%). Spatially, Set to Impress, Emerald City, and Metro Renters are concentrated in Tract 19. Their average household sizes are below that of the City Commons group (2.7 for City Commons versus 2.1-1.7 for the other three groups). 38 Theodus, Coulton and Pitingolo (2015) found in their study of poor neighborhoods in 10 U.S. cities between 2000 and 2010 that 40% of housing units experienced residential turnover within three years. Thus, a 1/3 turnover in residents who were residing in the same household as the previous year would be considered high. 39 The residence one year prior to the ACS cannot count those who were displaced from the tract. At best these data measure the percentage of recent in-migration to the tract or in some cases recent mobility from one residential unit in the tract to another unit. 40 Given that Trevarren Flats was not ready for occupation until 2016, few Walnut Hills residents would have been included in the 2012-2016 ACS data.

174

Plan, 2016). In 2014, Place Economics conducted an inventory of vacant buildings and their potential for rehabilitation. The agency recommended increasing the population density and the purchasing power in

Tract 19 (Southeast Walnut Hills) by converting vacant buildings that were structurally sound into apartments and businesses. Place Economics also suggested that vacant lots be considered for in-fill projects (Transforming Vacancy in Walnut Hills). Given the recentness of this strategy, the impact in terms of filling vacant buildings by 2016 has likely been minimal.

Owner-occupied housing declined from 42.0 to 30.3 percent suggesting that many of the vacancies as of 2010 were filled by renters by 2016. It is also possible that some homeowners were displaced after the recession ended. Working-class/middle-class homeowners may have been able to survive the recession by depleting their financial reserves, but this may have reached a limit by 2016.

Furthermore, while median housing value increased between 2010 - 2016 in Tract 19, it declined in the

City and it may be that better housing values were found in other areas of the City. In contrast, median rent had stabilized in Tract 19 while it increased in the City, making renting a more attractive option in

Tract 19. Increased options for subsidized housing would also contribute to the increase in renters

(Walnut Hills Investment Plan, 2016). Seventy-five percent of the population occupied the same residence as the previous year according to the 2010-16 ACS (up from two-thirds of the population in the

2010 ACS). The economic recovery from the recession would be less likely to uproot individuals from their homes and thus a decline in residential turnover is not surprising.

Table 7.3. Select Housing Characteristics for Tract 19, 2000-2016:

2000 2010 2016 Owner- occupied 23.50% 42.00% 30.30% Housing Percent Vacant 13.20% 28.80% 15.90% Number of 1,117 1,033 1,022 Housing Units Same Residence X 67.80% 75.00% One Year Ago Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

175

Comparison of City of Cincinnati and Tract 19 Incomes Disaggregated by Race, 2000-2016: Table 7.4 compares Tract 19’s median household income by race with that of the City of

Cincinnati. Note that median household incomes for African-Americans were only 81.2 percent of the

City’s median household income in 2000. By 2016, the two had converged (2% difference). This suggests that low-income African-Americans were displaced from Tract 19 and this displacement occurred between 2000 and 2010. Median household incomes for whites in Tract 19 grew at a slightly higher rate than their white counterparts in the city between 2000 and 2010 (33.8 versus 29.1 percent, respectively).

This suggests in-migration of higher income whites to Tract 19 and supports the classification of the tract as being in the process of gentrification during this time period. It must be noted that whites in Tract 19 already had median household incomes that were 25.5 percent greater than that of their white counterparts in the City in 2000. As previously mentioned, it is likely that the gentrification process was already underway in Tract 19 by 2000. Incomes were comparable between African-Americans in the City and

Tract 19 between 2010 and 2016 suggesting that displacement of low-income African-Americans, in terms of actual numbers, was not as pronounced as it was between 2000 and 2010.

Table 7.4. Comparison of African-American and White Median Household Incomes for Tract 19 and City of Cincinnati, 2000-2016.

2000 2010 2016 City-Wide (African- $20,984 $22,816 $21,814 American) African- $17,049 $20,625 $21,389 American (T 19) City-Wide $36,467 $47,067 $50,937 (White)

White (T 19) $45,750 $61,250 $65,250 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Select Characteristics for African-Americans and Whites in Tract 19, 2000-2016: Table 7.5 displays characteristics of African-Americans and whites in Tract 19 for 2000, 2010, and 2016. In 2000, median household incomes for whites was 2.7 times what it was for African-

176

Americans. This ratio increased to 3.1 by 2016 indicating that median household incomes between the two groups were diverging. It has already been noted that lower-income African-Americans likely left the tract between 2000 and 2010 and that accounted for the increase in median household incomes of

African-Americans by 2010. It is also likely that lower-income whites left the tract as the increase in median household income for whites between 2000 and 2010 (33.8% increase) was unlikely to be attributed to higher median household incomes of the 32 white individuals who were added to the tract.

As of 2010, the white population only had 65.1 percent of individuals who lived in the same residence as the previous year, suggesting that many were recent arrivals to Tract 19. As of 2016, whites had 75.0 percent of population that resided in the same residence as the year before. The increase in stability of the white population (greater percentage in same residence as previous year) suggests that the major influx of whites from gentrification occurred between 2000 and 2010 and these individuals remained in their residences after 2010. Note that African-Americans had a greater percentage of their population in the same residence as the prior year (74.7 percent) in 2010 than their white counterparts.

While it is possible that some in-migration of African-Americans occurred prior to the 2010 ACS and thus resulted in a turnover of African-Americans, it is likely that many of these individuals moved from one housing unit in Tract 19 to another housing unit and may be indicative of displacement within the tract41. African-Americans displaced from Tract 19 would not have been counted in the residence one year ago statistic because it is impossible to count the invisible as Atkinson (2000) states when talking about the difficulties of counting low-income displaced residents using census data. Surprisingly, there is no difference in percentage of African-American population in same residence as prior year between 2016

ACS and 2010 ACS, nor is there any real difference between the two racial groups in 2010. It is likely that a large percentage of poor African-Americans were displaced between 2000 and 2010 and those that

41 This intra-tract movement of low-income African-Americans from one rental unit to another was confirmed in my interview with Clarence Taylor, who is a member of WHRF.

177 remained had greater economic resources and/or subsidized housing which allowed African-Americans to remain in their residences.

The percentage of the population with at least a bachelor’s degree for whites increased from 63.3 to 75.7 percent between 2000 and 2016. Conversely, African-American educational levels hardly changed over the same time span, going from 14.5 to 15.4 percent. One bright spot is that whereas whites had an educational attainment statistic that was five times that of African-Americans in 2016, their median household income was only three times that of African-Americans. This suggests that African-Americans who remained in the tract were older, working-class, dual income earners who had attained higher incomes than their younger African-American counterparts who were displaced.

One way of comparing the educational/income gap is to examine the percentage of the population in the 25 to 44 age range, which is the prime age for gentrifiers (Moos, 2016). While the percentages fluctuate for whites between 2000 and 2016, at least 44.3 percent (2010) of those living in Tract 19 were in the 25 to 44 age range. Compare that with 29.9 percent for African-Americans in 2010. It is likely that a cohort of younger African-Americans (those who would have aged into the 25-44 age range) were displaced between 2000 and 2010. Surprisingly, by 2016, there was no real difference between the percentage of population between 25 and 44 for either group (51.8 percent for African-Americans and

52.5 percent for whites).

The gentrification process in Tract 19 had been underway for almost two decades by 2016. It is likely that some gentrifiers had aged into the 45+ age group and may have started families reducing the percentage of population in the 25-44 age range. As of 2016, gentrification in Tract 19 was further along than it was in 2000 and it may be that more families had moved into the tract. Early stages of gentrification are selective of younger, single, individuals while later stages attract the middle-class with families (Clay, 1979). A new trend in gentrification has shown that the pre-retirement/early retirement population has been attracted to gentrified areas (Hochstenbach and Boterman, 2018) and may account for some of the decline in the 25-44 age group for whites.

178

Table 7.5. Select Characteristics for African-Americans and Whites in Tract 19, 2000-2016. Same Median Educational House Household Attainment 1 year Income ago African- 15.40% 75.00% $21,389 American 2016 African- 12.80% 74.70% $20,625 American 2010 African- 14.50% X $17,049 American 2000 White 2016 75.70% 65.10% $65,250 White 2010 72.30% 73.40% $61,250 White 2000 63.30% X $45,750 Source: ACS 2006-10 and 2012-16.

Google Analysis: This section focuses on the built environment of Walnut Hills and its physical transformation through the gentrification process. Walnut Hills is noted for its high-quality historical housing stock which provides residents with a variety of housing types including single-family units of various sizes as well as multi-unit structures (Walnut Hills Reinvestment Plan, 2016). Walnut Hills is also involved in a large-scale project to revitalize the Gilbert Avenue and East MacMillan retail corridors (Walnut Hills

Reinvestment Plan, 2016), an attempt to recapture some of the economic prosperity of the early twentieth century.

Using a random number table, eight blocks were selected from Tract 19 for analysis of the built environment. Using Google Streetview, 27 block faces were analyzed to determine the extent that gentrification occurred between 2007 and 2014/1642. Unfortunately for the purposes of this research, much of the gentrification process in Tract 19 had been completed prior to 2007 and thus few blocks transitioned from disinvested to gentrified during this time43. Three blocks (2001, 2005, and 2007) had

42 See Methodology chapter for further information on this process. 43 A substantial amount of gentrification is now (as of May 2019) occurring along East McMillan between Kemper and Gilbert Avenue. This will be examined at the end of this chapter.

179

2007 scores that placed them in turnover status (+0.80). These three blocks experienced no significant changes by 2014/16 because they were either already gentrified or had never experienced disinvestment

(a block score below 0.50). Two tracts (1002 and 2002) were in the late stage of gentrification (0.78) in both years. One block (2000) was in the early stage of gentrification in 2007 and was still in the early stage by 2014/16. Two blocks (1000 and 1003) were in the disinvestment stage in 2007 but managed to transition into early gentrification by 2014/16. Block 1003 was the most disinvested of the blocks sampled and showed the greatest change between 2007 and 2014/16. Its score of 0.11 improved to 0.55 by 2014/16. Overall, Tract 19 went from late-middle stage gentrification in 2007 to late-stage gentrification by 2014/1644.

Figure 7.2A and B displays blocks that were sampled in Tract 19 in 2007 and 2014/16 using

Google Streetview. This sample includes seven blocks that share a border with Gilbert Avenue45. While only 8 blocks were sampled, a few generalizations can be noted. First, the southern part of the tract was further along in the gentrification process. This may be because of its proximity to downtown, but the small distance between the southern and northern parts of Tract 19 would not seem to warrant this conclusion. A 1975 study of Walnut Hills’ built environment showed that 29 percent of the buildings at

Peebles’ Corner (the most northern part of Tract 19) were substandard while only 13 percent of the buildings in the southeast of Walnut Hills (the remainder of Tract 19) were substandard (City of

Cincinnati, 1975). A 1992 follow-up to that analysis showed that Tract 19 (which included Peebles’

Corner) had 75.8 percent of its housing stock in good condition compared to less than 50 percent for other areas in Walnut Hills (Walnut Hills Community Plan, 1992). This difference in the building stock quality

44 Unfortunately, an earlier baseline is not available for Cincinnati. In Hwang’s (2015) analysis, she was able to compare fieldwork completed by Hammel and Wyly (1996) to track changes between 1995 and 2007 for Chicago and was able to provide a much more thorough analysis of stages in the gentrification process. 45 Many of the irregularly shaped blocks are part of Eden Park and do not have enough residential or commercial establishments to conduct the Google Streetview Analysis. If one of those blocks was selected in the sample, it was excluded and the next block from the random number table was selected. This method appears to have oversampled the Gilbert Avenue corridor at the expense of the rest of the tract. However, this corridor has been a major artery for movement to and from the City for the past 150 years and it seems appropriate that it should be over-sampled.

180 prior to 2000 may have attracted investment to areas of Walnut Hills with buildings that were in good condition and could initiate the process of gentrification. Maloney and Auffrey (2013) in their Social

Analysis of Cincinnati Neighborhoods found Tract 19 to be above the city average in 2005-2009 on indicators such as median household income and educational attainment in comparison to the other tracts in Walnut Hills (which were well below the City average). The second factor explaining the spatial distribution of gentrification in Tract 19 is that Eden Park (located in the Southern part of Tract 19) provides a recreational and cultural venue for those living in the area. While reaching Eden Park from the northern blocks does not require overcoming a substantial distance, there is a premium that will be paid by some to live in areas that have access to or views of this type of scenic/recreational/cultural amenity

(Mansfield et al., 2005).

181

Figure 7.2A. Gentrified Blocks in Tract 19, 2007.

182

Figure 7.2B. Gentrified Blocks in Tract 19, 2014/16.

183

Table 7.6. Gentrification Classification for Sampled Blocks in Tract 19, 2007 and 2014/16. Key46: Disinvestment below 0.50 Early between 0.50-0.57 Middle between 0.58-0.64 Late between 0.65-0.80 Turnover over 0.80

Difference between 2007 2007 2014/16 Block and 2014/16 and Stage of Score Score Gentrification Went from disinvestment to 1000 0.11 0.58 middle-stage gentrification Late stage gentrification; no 1002 0.78 0.78 change Went from disinvestment to 1003 0.47 0.55 early-stage gentrification Remained in early-stage 2000 0.51 0.53 gentrification in both years

2001 0.89 0.89 Turnover stage in both years

Late-stage gentrification; no 2002 0.78 0.78 change

2005 0.89 0.89 Turnover stage in both years

2007 0.89 0.89 Turnover stage in both years

Tract 0.67 0.74 Average

Source: Google Streetview and Hwang (2015).

Evidence of the progression through the stages of gentrification between 2007 and 2014/16 using

Google Streetview in Tract 19 is limited to Blocks 1000, 1002, and 1003. In most of the blocks, there was little evidence that gentrification occurred between 2007 and 2014/16. Figures 7.3A and B display subtle improvements along St. James Avenue (Block 1002). This block face had a 2007 gentrification score of

46 Based on Hwang’s (2015) Google Streetwise methodology for determining gentrification.

184

0.55. The block face did not substantially improve by 201647. The 2007 image of this house shows a somewhat unkempt yard. Windows on the structure likely need replacement and it appears that several of the windows have been boarded. The green trim along the windows needs repainting. Note the disrepair of the small room that juts out and the driveway that needs repaving. The 2016 image shows a freshly painted trim along the windows and windows appear to have been replaced. Landscaping along the house has created an attractive side yard. No change has occurred to the driveway.

47 While this house experienced an upgrade, it did not change the score for the block face. The house was chosen because it was the exception, not the norm for the block face.

185

Figure 7.3A. House on St. James Avenue in slight disrepair, 2007.

Source: Google Streetview, 2007.

Figure 7.3B. House on St. James Avenue with improvements, 2014.

Source: Google Streetview, 2016.

186

The next set of figures focus on block 1000. This block contains the intersection of Gilbert and

East McMillan Avenues (known as Peebles’ Corner), Kemper Lane and Curtis Street. It is the block that was the most disinvested as of 2007. Figures 7.4A and B show commercial gentrification. In 2007, a dilapidated business occupies the site. The business appears to be called ‘Great Looks’, but several letters were missing from the sign. By 2014, the business was demolished and a well-cared for vacant lot, now with an attractive fence, has replaced the building. The beauty supply store that was present in 2007 is still there but has received some façade enhancements. The business does not appear to be up-scale and likely caters to a working-class/lower-income population. This is evidence that gentrification has not completely reconfigured the commercial landscape to up-scale businesses and shows that Walnut Hills is committed to providing living, working, and retail opportunities to a wide range of socio-economic classes48.

48 Walnut Hills Redevelopment Foundation has implemented a façade improvement program for small businesses so that they are not displaced from the neighborhood.

187

Figure 7.4A. Disinvested Buildings along East McMillan Avenue, 2007.

Source: Google Streetview, 2007.

Figure 7.4B. Revitalized Business along East McMillan Avenue, 2014.

Source: Google Streetview, 2014.

188

Figures 7.5A and B show public space in Block 1000. The 2007 figure shows a former street that has been converted to a pedestrian walkway. Note the disrepair of the buildings on either side of the pedestrian way. While there is landscaping along the corridor, it is mixed in with the pavement. The 2014 image displays façade improvements for the buildings as well as a new fence. New greenery and barriers clearly indicate that this area is off-limits to motorized vehicles and is more inviting to pedestrians than the 2007 image.

189

Figure 7.5A. Public Space along East McMillan Avenue, 2007.

Source: Google Streetview, 2007.

Figure 7.5B. Revitalized Public Space along McMillan Avenue, 2014.

Source: Google Streetview, 2014.

190

Figures 7.6A and B show one of the buildings that is now Trevarren Flats, located on East

McMillan (Block 1000). The 2007 image shows a boarded-up lower entrance, but the upper floors appear to be of sound structure although some of the windows on the top floor appear to be broken. The 2016 image shows a façade improvement of the lower floor complete with large windows and an attractive entranceway. The block face score was 0.11 in 2007 but increased to 0.58 by 2016.

191

Figure 7.6A. Vacant Building Along East McMillan Avenue, 2007.

Source: Google Streetview, 2007. Figure 7.6B. Revitalized Apartment Complex-Trevarren Flats, 2016.

Source: Google Streetview, 2016.

192

Figures 7.7A and B show commercial gentrification along Gilbert Avenue in Block 1003. This block face had a gentrification score of 0.33 in 2007 and improved to 0.55 by 2014/16. This image is indicative of the process that occurred on this block face and not an exception as was noted in the Block

1000 image (St. James Avenue). In 2007, the building was in disrepair with nonfunctioning doors facing

Gilbert Avenue. There was evidence that U-Haul was operating the business in the 2007 image. The sign indicates the business and a truck to the side suggests that the business is still operational and has invested in a new look. In the 2014 image, new doors have been installed and the building has been freshly painted.

193

Figure 7.7A. Disinvested Business along Gilbert Avenue, 2007.

Source: Google Streetview, 2007. Figure 7.7B. Newly Revitalized U-Haul Along Gilbert Avenue, 2014.

Source: Google Streetview, 2014.

194

Figure 7.8. 2018 image of School being converted to apartments.

Source: Google Streetview, 2018.

195

Figure 7.8 displays the former Windsor Public School that has recently been converted to 44 market-rate apartments that were completed in 201649 (Walnut Hills Redevelopment Foundation). The school has been closed since 2004. While there have been no provisions for subsidized apartments in this project, the conversion of a vacant school to apartments would not displace any residents.

Current Revitalization Efforts in Tract 19: This analysis of revitalization and gentrification in Tract 19 was only in its formative stages by

2016. Figures 7.9-7.11 document the continued revitalization of the Walnut Hills Business District between Gilbert Avenue and Kemper Lane as of June 2019. It will be several years before census data will be available to tract the gentrification and displacement process that has occurred in Tract 19 since the revitalization of this area.

49 It is unlikely that these apartments were included in Tract 19’s housing inventory in the 2012-2016 ACS given that they were only ready for occupation in 2016.

196

Figure 7.9. Café Vivace. This is an up-scale coffee house in the renovated Hauck Building that opened after 2016. It is at the corner of East McMillan Avenue and Kemper Lane.

Source: Photo by E. Ravuri, June 7, 2019.

197

Figure 7.10. This building is located at East McMillan and Gilbert Avenues. It is being renovated and will contain apartments on the upper floors and retail on the bottom.

Source: Photo by E. Ravuri, June 7, 2019.

198

Figure 7.11A. Dilapidated Building replaced by new building along East McMillan Avenue that will host The Scholar House.

Source: Google Streetview, 2018.

Figure 7.11B. In-fill along East McMillan, 2019.

Source: Photo by E. Ravuri, June 7, 2019.

199

Interviews with Community Members in Walnut Hills: In December 2018 an email was sent to WHRF (Walnut Hills Redevelopment Foundation) explaining the project and asking the agency to provide contacts for individuals who could speak on economic revitalization and demographic change in Walnut Hills between 2000 and 2016.During May of 2019, interviews of individuals familiar with the economic revitalization and demographic change in Walnut

Hills were undertaken to provide a perspective on the changes taking place in the community.

Emily Sheckels Ahouse: Emily is the Executive Director of Walnut Hills Redevelopment Foundation

(WHRH). She has been in this position since 2018. Emily holds a Master’s degree in Urban Planning from Cornell University.

Mary-Cabrini Durkin: Mary-Cabrini has resided in Walnut Hills for sixty years and has been an active member of the WHRF.

Kathryne Gardette: Kathryne has resided in Walnut Hills for over twenty years and was a past president of the WHRF.

Clarence Taylor: Clarence has resided in Walnut Hills for over 50 years and has been active in the Walnut

Hills Community since he was 16.

The following are questions were asked of each of the interviewees. Not all interviewees could answer each question.

1. Can you pinpoint a year (s) when you first started to see changes in Walnut Hills suggesting that

economic reinvestment and/or revitalization was occurring? Where did the process begin within

Walnut Hills? Be as specific as possible.

For two of my informants, their 50 plus years in Walnut Hills allowed them to provide a timeline on their subjective experiences dealing with economic decline prior to 2000 and thus those experiences are recounted here before moving on to a discussion of economic revitalization.

200

Mary-Cabrini: When presented with question 1, Mary-Cabrini immediately asked what I meant by economic revitalization or the investment of ‘financial’ capital in Walnut Hills. She stresses that ‘human’ capital is just as an important component of a vibrant community as financial capital and Walnut Hills, even with its high poverty rates, still has a vibrant sense of community. This strong sense of community can be credited to the Victory Neighborhood Services Agency which was a major player in revitalization since the 1980s50. The agency created block clubs which helped build community spirit. These clubs allowed the block leaders to communicate to city leaders the concerns of residents on those blocks, becoming a powerful advocate for change in this community. Mary-Cabrini said that this relationship building was intergenerational and inter-racial.

In terms of recent revitalization (post 2000), Mary-Cabrini identified revitalization efforts in the business district (along East McMillan) and credits this to the Walnut Hills Redevelopment Foundation (WHRF).

During the late 1960s, East McMillan and Taft were turned into one-way streets while Interstate 71 cut

Walnut Hills into two, disrupting the social and mobility connections of the community. While Walnut

Hills’ economic decline had started prior to this time, the business district was further crippled by this change. The decision to turn East McMillan into a two-way street in 2012 marked a concerted effort to revitalize the business district in Walnut Hills and thus 2012 marks the turning point on the visible impact of the built environment along East McMillan.

Clarence: Clarence recalls prior to 2000 that several stores serving a lower-income population began to close their doors (e.g. Will’s Pawnshop, Louis’s Meats, a check-cashing place, and a jewelry shop).

Clarence recalls the opening of Kroger’s on East McMillan in 1976 and the closing of the store 40 years later. He doesn’t see the closing of Kroger’s as much of a death blow to the business district as the smaller businesses that failed during the latter part of the 1900s had already left the business district at its nadir. Revitalization in Walnut Hills began in 2012 as the WHDF started to purchase properties to attempt

50 Mary-Cabrini believes that the agency has disbanded within the past year or so.

201 to revitalize the business district. Most of these properties were from absent-tee landlords who did not keep the buildings in good shape. As a member of the WHDF, Clarence said that scarce financial resources required that the foundation choose to renovate the housing stock or focus on the business district as the first step in the revitalization process. Like Mary-Cabrini, Clarence sees the conversion of

East McMIllan from a one-way to a two-way street as a necessary component in the revitalization of the business district. Instead of being a through street on the way uptown/downtown or to Interstate 71, East

McMillan is now a more pedestrian-friendly area and is now a destination within and outside the community.

Kathryne: Like Clarence and Mary-Cabrini, Kathryne dates the changes in the built environment in

Walnut Hills to conversion of East MacMillan to a two-way street in 2012. However, she stated that it took several years before businesses began to prosper along East MacMillan.

Emily: Having only been involved with the WHRF for nine months, Emily can only speak to what she has heard concerning the revitalization efforts in Walnut Hills. She confirms that redevelopment focused on East McMillan business district with the recent phase being implemented in 2011. However, it was not until 2015-16 that tangible evidence of investment was visible in the built environment.

2. What effect did the Great Recession (2007-2009) have on economic reinvestment/revitalization in

Walnut Hills?

The consensus from the informants on the effect of the Great Recession was that it had minimal impact on the residents and businesses of Walnut Hills because poverty rates were already high prior to the onset of the recession and the business district was already in a state of decline. Foreclosures resulting from the housing crisis was not a major issue as most residents in Walnut Hills were renters. Recall that Walnut

Hills did not make the top ten neighborhoods in Cincinnati that experienced housing foreclosures (total

202 numbers) whereas the other three neighborhoods in this study all ranked in the top 5 for foreclosures in

2008.

3. Can you comment on the impact of private development agencies versus non-profit/government

agencies in the reinvestment/revitalization process?

Mary-Cabrini: The WHRF has been instrumental in the revitalization process and has attempted to ensure that redevelopment includes affordable housing and considers the needs of the community. Those developers (such as Model Group) who need public funds or need a variance often show up at the Walnut

Hills council meetings, but otherwise are not interested in building community cohesiveness. However, their presence in the community in appreciated as they need to provide a percentage of their housing units as affordable and they bring new businesses to the community. Private investors (either corporate or individual) have brought needed capital to Walnut Hills but have marginalized people; particularly the elderly, non-white, lower-income and those with disabilities. Mary-Cabrini sees this at an attack on community cohesiveness as those purchasing many of these properties are not invested in the community but purchase properties cheaply and then resell at a profit. The ‘flippers’ have no roots in the community and do not intend to stay and are detrimental to the mission of the WHRF. Likewise, homeowners of means (gentrifiers who reside in owner-occupied housing) often report their less well-off homeowning neighbors for violations in property maintenance causing friction between the two groups.

Kathryne: Kathryne presented an interesting twist on the reinvestment/revitalization process. Kroger’s decided to close its Walnut Hills store in 2017 due to its inability to make a profit for the past twenty-five years (Demeropolis, 2016). This was likely premature on Kroger’s part as by 2016 market-rate housing began to appear across the street from the former Kroger. According to Kathryne, there are now hot spots within Walnut Hills where consumers can order products from another Kroger’s and have them delivered to the parking lot of the currently vacant Kroger’s in Walnut Hills.

203

4. Has reinvestment/revitalization led to a change in the demographic composition (e.g. age,

race/ethnicity, socio-economic status) of Walnut Hills? If so, in what way? Have certain areas

been more affected than others? How did the Great Recession affect demographic composition in

Walnut Hills?

Mary-Cabrini: One of the great aspects of residence in the Walnut Hills community was that the racial interactions were mostly positive over most of Mary-Cabrini’s residence in the community (60 years).

That has been changing in the past decade. Mary-Cabrini states that many young, whites want to live in a mixed-neighborhood. She then comments that ‘they are un-mixing it!’. Mary-Cabrini also considers the millennials as a generation that does not relate to people next door. They are more interested in their cell phones than in engaging in meaningful community interaction. One unfortunate aspect of the revitalization process is that the adult children of African-Americans are not able to buy into the neighborhood because it has become too expensive or if they have inherited their parent’s house, they have trouble obtaining financial resources to maintain the property. This reduces the number of middle- class African-Americans who reside in Walnut Hills and dilutes the presence of African-Americans as well as increases the percentage of low-income and relatively high-income groups.

Clarence: As would be expected from his working with the WHRF, Clarence was cognizant of the fact that Walnut Hills was experiencing an influx of younger individuals most of whom are white. I asked

Clarence how African-Americans viewed the gentrification process and the in-migration of mostly white residents. Clarence responded by saying that it depends on who you consult. Poor African-Americans are threatened by gentifiers as they see them as displacing them from their residences. ‘Hard-working’

African-Americans (I interpreted this as middle-class) view it as positive as their property values increase.

Those on fixed-incomes worry about how this will affect property taxes.

204

Kathryne: The newcomers to Walnut Hills are mostly young whites (age 20s-30s), but she also notes that an increasing number of ‘empty-nesters’ have arrived in Walnut Hills since 2012.

5. Can you comment on the changes in types of businesses or improvements in already established

businesses in your community during the gentrification process?

Emily: The Paramount Square is being rehabbed (as of 2019) and will contain commercial enterprises on the first floor and apartments on the upper floors. The WHRF is concerned that there is a diversity of businesses that will cater to different clientele. There should not be an overconcentration of businesses that cater to gentrifiers. While Kroger’s closed before the emphasis on revitalizing the Walnut Hills business district, there are plans to open a small mini-grocery store in Fall 2019 that provide meats, fruits, and vegetables. Emily states this is necessary because Walnut Hills is technically a food desert and currently the only fresh food option for Walnut Hills is the Healthy Harvest Mobile Market which makes its rounds on Friday afternoons.

Mary-Cabrini: When I suggested that we meet at a coffee shop in Walnut Hills, perhaps Café Vivace, to conduct the interview, Mary-Cabrini indicated that she would rather meet at Parkside. Parkside sits across from Café Vivace51 on East McMillan Street. The contrast between the two coffee shops is immediately noticed. Parkside’s building is a typical nondescript building that was probably constructed in the 1950s with a large parking lot while Café Vivace sits in a renovated historical building and requires the patron to use on-street parking. Parkside caters to the incumbent population of Walnut Hills (as well as through traffic) and its prices and menu display nothing that suggests it is in a gentrifying neighborhood.

Conversely, Café Vivace is an up-scale coffee house. Before I could remark about the differences between the two coffee shops, Mary-Cabrini stated that she preferred Parkside because it has a sense of community around the place. We noticed that a few meetings were going on at Parkside and concluded that the business provided a community gathering venue. Conversely, my experience with Café Vivace

51 I met Clarence Taylor at Café Vivace for our interview two weeks prior to meeting Mary-Cabrini.

205 was that it was a place where ‘yuppies’52 could go and have coffee and work on their computers. There were very few patrons who had a companion.

6. Is there anything else you would like to mention concerning economic revitalization and

demographic change in Walnut Hills?

Affordable housing was a concern for all informants. Mary-Cabrini was the most vocal about this issue.

Mary-Cabrini: Mary-Cabrini considers herself a displaced person. She continues to move eastward to find a rental unit that she can afford. She was displaced in 2010 when her landlord suffered from the foreclosure crisis and then was displaced again in 2014 because her landlord sold the property to someone else. She laments that prices for rent have skyrocketed in the past decade. Mary-Cabrini relates that there is a deficit of 40,000 units of affordable housing in Cincinnati. One of the problems is that if investors do not need public dollars, they do not have to produce affordable housing. Mary-Cabrini does relate a few instances of the attempt to focus on affordable housing in Walnut Hills. The Alms, a subsidized housing complex run by absent-tee landlords that was in deplorable condition has been purchased by the Coalition for Affordable Housing and will be brought up to code and provide not only affordable housing but a safe environment for its residents. The Park-Eden is a senior public housing facility that is going to be renovated in the summer of 2019 and will help with keeping elderly individuals on fixed incomes from being displaced.

The effect of the revitalization process on schools in Walnut Hills was also a concern. Clarence remarked that many school-aged children in Walnut Hills have moved around quite a bit due to displacement from their residences and this has had a destabilizing effect on their educational performance. Another concern

52 Yuppies are Young, Urban, Professionals. This term was coined in the 1980s, but is not heard so frequently today.

206 is that Douglas school was converted to a new school called the Spencer School and serves gifted children who are bused into Walnut Hills. Here are two groups of school children sharing the same community space (Walnut Hills) but have little contact with each other. In contrast, Mary-Cabrini laments that new arrivals to Walnut Hills often send their children to schools outside of Walnut Hills and this reduces the amount of time that these groups of students get to spend knowing each other and building community relationships. Kathryne believes that The Scholar House (when completed) will be beneficial to many low-income individuals in Walnut Hills.

Conclusion: Walnut Hills has undergone a major transformation in its built environment since 2012 when the decision was made to convert East McMillan from a one-way to a two-way street. This reconfiguring of the street provided a catalyst for commercial and residential development. Google Streetview Imagery revealed the dramatic change in the built environment from 2007 to 2016 along McMillan and Gilbert

Avenues as buildings were renovated or demolished and rebuilt. However, census data from 2000 and

2010 for Tract 19 showed that gentrification had already begun in Walnut Hills. The census data also showed that the African-American population declined by about one-third between 2000 and 2010 suggesting that displacement had occurred. Stemming this tide of displacement was an issue that was addressed by the Walnut Hills Redevelopment Foundation. This largely came about through public and private investors focusing on transforming vacant properties into residential units for a wide array of socio-economic levels. Between 2010 and 2016, displacement of African-American population in Tract

19 did not occur suggesting that these efforts at providing affordable housing may be working. However, each of my interviewees from Walnut Hills agreed that affordable housing is still a major issue to address as Walnut Hills goes forward in the development process.

207

Chapter 8 COLLEGE HILL

Introduction: This chapter seeks to explore the actions and/or policies that Tract 84 in College Hill adopted which kept the African-American population from being displaced during the gentrification process. The analysis relies on census data to gauge demographic and economic change for both African-Americans and whites in Tract 84 between 2000 and 2016. Tract 84 is also compared to the City of Cincinnati to determine if African-Americans were better or worse off than their African-American counterparts in the city before and after gentrification in the tract. The second part of the chapter uses Google Streetview to document changes in the urban environment between 2007 and 2014/16 to determine how pervasive the gentrification process was in Tract 84. The last component of the chapter is an analysis of economic and demographic changes in the tract as seen through the lens of individuals involved in the community over the past few decades.

The Neighborhood of College Hill consists of five tracts: 81, 82.01, 82.02, 84 and 111. Tracts 81,

82.01 and 111 had median household incomes exceeding the city average in 2000 and were not candidates for gentrification. Tract 82.2 did not gentrify over the study period and thus could not be selected for analysis. Tract 84 was selected for this detailed analysis as it had experienced gentrification between 2010 and 2016, as determined by employing the metrics used in this thesis (adopted from

Hammel and Wyly, 1996), but did not experience displacement of the African-American population. As a result, Tract 84 provides a different experience from the neighboring Tract 74 in Northside, which experienced gentrification during both 2000-10 and 2010-16 as well as displacement of the African-

American population between 2010-16.

208

Figure 8.1. College Hill, Tract 84, Satellite Imagery.

209

Short History of College Hill: College Hill was platted as a community in the late 1700s and mostly consisted of farmland and a few cottage-type industries (Smiddy, 2008). The community took its name from two small colleges that were established in the area in the mid- 1800s (College Hill Existing Conditions Study and Community

Plan, 1976). By the 1850s, aided by rail, the upper-class moved away from downtown Cincinnati and settled in College Hill which provided a healthier habitat, scenic views, and a more spacious living arrangement than being confined in the basin of the City of Cincinnati (Smiddy, 2008). Working-class populations then began to move to College Hill with the electric rail in the 1890s. Unstable hillsides have been an issue since settlement and have prevented College Hill from becoming densely populated. The community currently consists of mostly single-family residences, with most dwellings built during the

1950s and 1960s (College Hill Existing Conditions Study and Community Plan, 1976). College Hill’s most recent planning document is from 1976 leaving much valuable information pertaining to College

Hill’s economic and demographic conditions during the latter quarter of the twentieth century undocumented. Fortunately, Maloney and Auffrey’s (2013) Social Analysis of Cincinnati neighborhoods between 1970 and 2009 fills in some of this gap in the data. Maloney and Auffrey measured Socio- economic status (SES) as a combination of five factors: income, occupation, education level, family organization and housing. Each neighborhood in Cincinnati was ranked into one of four SES categories based on the above-listed characteristics53. College Hill experienced a major decline in its SES between

1970 and 2009 but was still categorized as being in category 3 out of 4 as its SES was still above the city average. This decline in SES over the four decades coincided with a decline in the built environment in select areas of College Hill. In 2002, the City of Cincinnati and College Hill, with the help of an Urban

Design Company (Michael Schuster Associates), developed a plan to revitalize the business district. The

53 Maloney and Auffrey (2013) used four categories in their analysis of the tracts in the City of Cincinnati. SES I referred to as ‘A High Problem Area’ which had high poverty rates and low educational levels. It consisted of many of the inner-city tracts, tracts along the Mill Creek and the Western part of the Ohio River. SES II was referred to as ‘Second Stage Neighborhoods’ which were mostly working-class neighborhoods. They still had some issues with poverty. SES III referred to as ‘Where Front Yards Begin’ were middle-class neighborhoods. SES IV referred to as ‘The Upper Quartile’ were the wealthiest neighborhoods in the City and were found in Eastern Cincinnati and around the University of Cincinnati.

210 plan focused on remediation of blighted properties, promoting mixed-use development, attracting new businesses, and providing a better streetscape (https://chcurc.com/). These objectives have not yet been completed as of the writing of this thesis.

Census Data: Table 8.1 displays population data for tract 84. The tract’s total population experienced a slight gain between 2000 and 2016 (3.1 percent). However, the racial composition of the tract changed significantly during this time. In 2000, the white population comprised 54.2 percent of the tract’s total while the African-American population comprised 42.7 percent. By 2016, the white population comprised only 41.0 percent of the tract versus 56.7 percent for African-Americans. The African-American population grew by 38.0 percent between 2000 and 2016 versus a decline of 25.1 percent for the white population.

However, the loss of the white population was experienced solely between 2000 and 2010.

Between 2010 and 2016, the white population increased by 3.9 percent. Conversely, the growth in the

African-American population happened between 2000 and 2010; it was stable between 2010 and 2016.

These population data suggest that African-Americans have not been displaced in Tract 84 during the process of gentrification that occurred between 2010 and 201654.

54 The racial turnover of population in Tract 84 between 2000 and 2010 suggests that the traditional invasion/succession process first discussed by Burgess et al. had occurred in this tract. Burgess et al. stated that whites move out to better housing opportunities which provides housing opportunities for African-Americans.

211

Table 8.1. Population Change in Tract 84, 2000-2016 2000 2010 2016 Total 2,015 2,100 2,078 African- 861 1,188 1,178 American

Non-Hispanic 1,093 819 851 White

Other 61 93 49 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS

A Comparison of Select Economic/Social Characteristics for Tract 84 and the City: Table 8.2 compares median household income, percentage of population over 25 with at least a bachelor’s degree and median housing values/median rents for Tract 84 and the City of Cincinnati for the three years in the analysis. These comparisons with the city allow us to determine when Tract 84 underwent gentrification. Tract 84 had lower median household incomes in 2000, 2010, and 2016 in comparison to the city55. However, while City median household incomes increased at a higher percentage than Tract 84 between 2000 and 2010 (14.2 versus 7.9 percent, respectively), Tract 84 outperformed the city in median household income growth between 2010 and 2016; a growth of 11.8 percent versus 2.8 percent for the city. There is no evidence that higher-income individuals (gentrifiers) had moved to Tract 84 in any large numbers between 2000 and 2010 as the growth in the median household income was below that of the City. Interpreting to what extent gentrification affected the growth in median household income for Tract 84 between 2010 and 2016 is more complicated. It is possible that Tract 84 did a better job at recovering from the recession and outpaced the City’s overall income growth during that period and gentrification did not occur. That hypothesis is not likely as educational levels (percentage of population over 25 with at least a bachelor’s degree) in Tract 84 increased substantially between 2010 and 2016.

55 Recall that to be a candidate for gentrification in this study, a tract needed to have median household incomes below the city average in 2000.

212

Tract 84 had 31.3 percent of its population over 25 with at least a bachelor’s degree (referred to as educational attainment throughout the remainder of this chapter due to its lengthy descriptor) in 2000 compared to 26.6 percent for the city. By 2010, only 27.7 percent of Tract 84’s over 25 population had college level attainment while the city had 31.7 percent on this educational metric. The decrease in educational attainment in Tract 84 between 2000 and 2010 suggests an exodus of educated population from the tract and/or an influx of less-educated individuals. Tract 84 was able to close the gap in educational attainment with the city by 2016 (32.4 versus 33.8 percent for the city). Between 2010 and

2016, Tract 84 increased its education attainment by 17.0 percent versus only a 6.6 percent increase for the City of Cincinnati. By 2016, Tract 84 was considered gentrified as the growth of median household income and educational attainment increased at a greater rate than the city’s56.

In 2000, Tract 84’s median housing values were 39.2 percent higher than the city’s median housing values (Table 2). Median housing values increased at a slower rate in Tract 84 in comparison to the city between 2000 and 2010. During this time, Cincinnati experienced a 39.5 percent increase in housing values while Tract 84 only experienced a 15.4 percent increase. This is not surprising as Tract 84 was not in the process of gentrification as of 2010 whereas this process was well underway in many

Cincinnati neighborhoods by this time (Simes, 2013). Median housing values declined at the same rate in

Tract 84 as they did in the City between 2010 and 2016 and were likely a result of the housing crisis and the recession57.

In 2000, median rent in Tract 84 was 23.7 percent higher than that of the city. However, median rent in the Tract increased at a slower rate than the city (11.7 versus 33.6 percent during this time). This low increase in median rent may have attracted those with less economic resources to Tract 84 by 2010

56 I modified Hammel and Wyly’s (1996) three factor gentrification criteria which called for a tract to have experienced an increase in 1) median household income; 2) percentage of population over 25 with at least bachelor’s degree and 3) median household incomes or median rents that were greater than that of the city for the period of analysis. Tracts could be included as gentrified if two out of the three criteria were met with growth rates that were at least 50 percent above that of the City. This is explained in greater detail in the Methodology Chapter. 57 ACS data is from 2012-2016 which may not be long enough after the recovery of the recession to show significant increases in median house prices.

213 and provided housing during the economic recession. Some of these individuals with few economic resources were likely female head-of-householders. Maloney and Auffrey (2013) found that College Hill had the seventh highest concentration of female head-of-householders in poverty in Cincinnati between

2005-200958. Between 2010 and 2016, rent in Tract 84 declined while it increased slightly in the city. By

2016, median rent in the city was 25.7 percent higher than that of Tract 84 making the tract an option for those with less financial resources, but certainly not a tract that could be considered as gentrified by using

Hammel and Wyly’s median housing value /median rent criterion. Instead, Tract 84 was considered gentrified by 2016 due to educational attainment levels and median household incomes that grew at rates above the city average between 2010 and 2016.

Table 8.2. Select Characteristics for Tract 84 and City of Cincinnati, 2000-2016.

Tract 84 City Cinti. Tract 84 City Cinti. 2000 2000 2010 2010 Median $26,599 $29,493 $28,700 $33,681 Income Educ 31.30% 26.60% 27.70% 31.70% Bach + Median Housin $129,500 $93,000 $149,500 $129,700 g Value Median $582 $444 $650 $593 Rent Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Housing Market Characteristics in Tract 84: Total housing units in Tract 84 increased from 1,003 to 1,103 between 2000 and 2010 (Table

8.3). This increase in housing units outpaced the increase in population (8.0 percent versus 3.1 percent) and may explain why median rent and median housing values did not grow as rapidly as they did in the

City. In 2007, Hickory Woods Townhomes were built on Hamilton Avenue and added 47 subsidized units

58 Maloney and Auffrey (2013) do not disaggregate their data by census tract so not each of the 555 female headed households in poverty in College Hill were necessarily located in Tract 84.

214 to the housing inventory of the tract (https://affordablehousingonline.com/housing- search/Ohio/Cincinnati/Hickory-Woods/10009700), while Ashwood Apartments (built in 1972), located on Ashtree Street, provide 152 units of subsidized housing

(https://www.apartmentfinder.com/Ohio/Cincinnati-Apartments/Ashwood-Apartments). These two subsidized housing options accounted for almost 20 percent of the housing stock in Tract 84 in 2010. By

2016, there were 1,083 housing units, a slight decline from 2010. A disinvested apartment complex located at Hamilton and Belmont was converted to luxury apartments by 2014

(http://thebridgesatbelmont.com/). This renovation did not increase housing units in the tract but provided apartments for those with higher incomes59. The slight reduction in housing units appears to have occurred after 2010 with selected demolition of properties that were in poor condition60.

Vacancy rates in Tract 84 in 2000 and 2010 were 10.5 and 11.0 percent, respectively. The stability of the vacancy rate between 2000 and 2010 suggests that housing foreclosure or forced removal from a rental unit was not a major issue in Tract 84. With at least 20 percent of housing stock subsidized, the displacement from the recession would be less likely to occur. It must be remembered that median rents had increased below the City average between 2000 and 2010 in this tract and this may have protected some individuals from being displaced.

Overall, there was an increase in owner-occupied housing between 2000 and 2010 (41.3 versus

45.5 percent, respectively). This may be because median housing values in Tract 84 did not experience high rates of growth between 2000 and 2010 (in comparison to the City) and provided an affordable option for the working-class/middle-class. Housing loans were also easier to obtain until 2008 when the housing market collapsed (Glaeser, Gottlieb, and Gyourko, 2010). In 2010, 86.3 percent of the population in Tract 84 resided in the same residence as the year before. This is a relatively low mobility rate in

6It may have even decreased housing units as luxury units are likely larger in size to what they replaced. The Bridges at Belmont website does not provide any information on sizes of apartments in the previous complex making this observation difficult to verify. 7This is my conclusion after using Google Streetview to examine block faces between 2007 and 2014/16 and noting removal of some housing stock.

215 comparison to the City of Cincinnati (74.0 percent in same residence as previous year). The increase in homeownership likely explains some of this low level of mobility as homeowners are less mobile than renters (Cilluffo, Geiger, and Fry, 2017). It is also likely that the Great Recession prevented individuals from seeking alternative locations because they were unable to sell their homes due to the possibility they would have to take a loss.

The vacancy rate increased somewhat by 2016 to 13.2 percent suggesting that some renters left the tract by 2016 for better opportunities elsewhere, perhaps leaving ‘sub-standard’ housing vacant.

Owner-occupied housing decreased from 45.5 to 41.1 percent between 2010 and 2016 and may been a delayed response to the recession61. Alternatively, the decline in owner-occupied housing could be following the macro-trend in the U.S. where it has been noted that the percentage of renters in 2016 was the highest it has been since 1965 (Cilluffo, Geiger, and Fry, 2017). The greater percentage of the population in the same residence as the previous year in 2010 suggests that the recession and housing crisis prevented persons from moving. This decline in mobility during the Great Recession was documented for the entire U.S. where moving rates were the lowest since the early 1980s (Molloy, Smith and Wozniak, 2011).

Table 8.3. Housing Characteristics for Tract 84, 2000-2016. 2000 2010 2016

Owner-occupied 41.30% 45.50% 41.10% Housing

Percent Vacant 10.50% 11.00% 13.20%

Number of 1,003 1,103 1,083 Housing Units Same Residence X[1] 86.30% 76.10% One Year Ago Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

8The 2016 ACS data is not compatible with the 2000 and 2010 data and thus is not discussed. This is because when disaggregated by race, only tenure status for the householder is available in the 2016 ACS, not the total population as is recorded in the 2000 and 2010 data.

216

City and Tract 84 median household incomes disaggregated by race: African-Americans in Tract 84 made less than their African-American counterparts in the City of

Cincinnati in all three years (See Table 8.4). This is particularly evident for 2010 when the median household income for African-Americans in the city was almost double that of African-Americans in

Tract 84. Whereas median household income for African-Americans were stagnant in the city between

2000 and 2016, they declined by about 12.0 percent between 2000 and 2016 in Tract 84. Since 2000,

Cincinnati has experienced substantial gentrification in the neighborhoods surrounding downtown (Over- the-Rhine, Pendleton, Mt. Auburn, and the West End). These neighborhoods also experienced displacement of their African-American populations (https://jackdelisionhist415.wordpress.com). It may be that College Hill was a recipient of some of the displaced. The central city still houses much of the poor African-American population, but African-Americans have been dispersed to other areas of the city since 2000 (U.S. Census Data).

In contrast, whites in Tract 84 had much higher incomes than whites in the City of Cincinnati in all three years, with whites in Tract 84 making 50 percent more than their white counterparts in the city by 2016. Compare this with 2000 and 2010 in which whites in Tract 84 made 14 and 20 percent more than their white counterparts in the city. Whether one examines the City or Tract 84 alone it is painfully obvious that white median household income increased substantially between 2000 and 2010, indicating that whites recovered from the recession to median household incomes that were substantially higher than in 2000. As we have seen, such was not the case for African-Americans.

217

Table 8.4. Median Household Income for African-Americans and Whites for Tract 84 and City of Cincinnati, 2000-2016. 2000 2010 2016

City-Wide (African- $20,984 $22,816 $21,814 American)

African-American $17,308 $13,469 $15,227 (T 84)

City-Wide (White) $36,467 $47,067 $50,937

White (T 84) $42,875 $62,344 $75,625 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Disaggregation of Select Characteristics in Tract 84 by Race: When median household income is disaggregated by race, it is obvious that the African-American population in Tract 84 did not perform well in the growth of median household income in comparison with whites (Table 8.5). Between 2000 and 2010, the median household income for African-Americans declined by 28.5 percent while it grew by 48.0 percent for whites. The significant increase in median household income for whites between 2000 and 2010 suggests that there was an exodus of lower-income whites and a possible influx (although not enough to compensate for the loss) of higher income whites into the tract. Some whites may have left in the early 2000s to take advantage of available housing loans during the run-up to the housing crisis of 2008. The decline in African-American median household income between 2000 and 2010 suggests an influx of lower-income African-Americans into the tract to fill vacancies left by whites. These assumptions are supported by the following: the numerical loss of the white population between 2000 and 2010 (274 persons) was comparable to the gain in African-Americans

(327) during this time. Median household income for African-Americans increased between 2010 and

2016 but was still below what is was in 2000. For whites, median household income increased by 21.3 percent between 2010 and 2016. While median household income increased for both African-Americans and whites between 2000 and 2016, white median household income was almost five times that of

218

African-American income in 2016. In 2000, the ratio of white to African-American median household income was only 2.48.

The percentage of white population with educational attainment declined from 39.9 percent to

36.5 percent between 2000 and 2010. Recall that the loss of the white population from the tract during this time was about 274 individuals. The assumption was that these were whites with low-income levels and one would assume low-educational levels. It is perhaps necessary to explain that process in greater detail. It must be acknowledged that young adults are more likely to have attained a college education than their older counterparts (Ryan and Bauman, 2016) but have yet to obtain positions that lead to higher incomes. These young adults are also more mobile than their older counterparts (Speare, 1975) and may have left the tract between 2000 and 2016 which would simultaneously decrease educational levels while increasing median incomes. Conversely, older whites who remained could have been working-class, perhaps two-worker households, who had higher incomes from long-term stable jobs but had educational attainment below their younger counterparts. My original age for a ‘young adult’ was 25-44 but the demographics of the white population in Tract 84 call for a re-configuration of that definition. According to U.S. Census data, the median age of whites in Tract 84 in 2000 was 73.1 and 63.0 in 2010. This very high median age is partly a result of Twin Towers, a senior living complex with 116 units mostly catering to the white population (http://lec.org/twin-towers/location/). This decline in median age does not support the hypothesis that working-age whites left the tract in large numbers. However, as of 2010, there were

160 less whites over age 70 residing in Tract 84 than there was in 2000 (U.S. Census 2000 and 2010)62.

This would not only decrease the number of whites in Tract 84 but decrease median age and increase median incomes of whites as the most elderly likely had the least economic resources63.

As for the decline in percentage of white population with educational attainment between 2000 and 2010, that mystery remains to be solved. Between 2010 and 2016, educational attainment for whites

62 Of course, not all individuals over age 70 were residing in Twin Towers. 63 This should have also raised the educational attainment as it would be presumed that the elderly would have less education than younger adults.

219 increased from 36.5 to 44.4 percent. The increase in educational attainment and the increase in median household incomes suggests an influx of higher-educated whites as Tract 84 outpaced the city’s growth in educational attainment and median household income between 2010 and 2016. The median age of whites only decreased to 59.9 by 2016 (from 63.0 in 2010), but the percentage of the white population in the 25-

44 age range which is the prime age for gentrification (Hochstenbach and Boterman, 2018) was 15.4% in comparison to about 7.0 percent for 2000 and 2010 (U.S. Census, 2000 and 2010). This suggests an influx of young, educated individuals64 or that whites in the 15-34 age range residing in Tract 84 in 2010 aged into the 25-44 age range and completed college.

Educational attainment for African-Americans declined between 2000 and 2010 from 13.3 to 11.0 percent, respectively. Unlike their white counterparts, African-Americans experienced a concomitant decline in median household income. As stated previously, it is my hypothesis that there was an influx of less-educated African-Americans into the tract. I draw this hypothesis from the following: 1. an increase in the number of African-Americans in the tract between 2000 and 2010 and 2. a percentage decrease in median household income lending support to this hypothesis. The educational attainment level increased slightly for African-Americans between 2010 and 2016 (from 11.0 to 14.8 percent, respectively).

Ninety-four percent of the white population in Tract 84 resided in the same residence as the previous year, according to the ACS 2006-10 count. This is not surprising in that the years leading up to the 2006-10

ACS corresponded with the recession and would make mobility more difficult. The stability of the vacancy rates (for the total population) and increase in homeownership for whites in Tract 84 from 60.2 to

63.4 percent between 2000 and 2010 (U.S. Census 2000 and ACS 2006-10) suggests that whites were not forced out of the tract due to inability to pay for housing. It may be that whites with less financial resources left the tract during the early 2000s for housing opportunities elsewhere as loans were easier to acquire during this time for those with less financial resources or for first-time homeowners (Glaeser,

Gottlieb, and Gyourko, 2010). According to the ACS 2012-2016, residence in the same house as the

64 Note that income did not increase as rapidly as educational levels and is likely a result of these individuals being in early stages of their career.

220 previous year for whites declined to 82.2%. This suggests that some whites who were constrained from moving during the financial crisis were able to relocate and were replaced by whites with higher incomes.

This would also explain the growth in median household income between 2010 and 2016.

The percentage of population in the same residence as the prior year for African-Americans during 2006-2010 (79.9%) was higher than it was for 2012-2016 (71.1%), but lower than whites for both years. While African-Americans saw a substantial increase in their rates of homeownership, from 20.4 to

33.1,65 African-American homeownership rates in Tract 84 were well below that of whites. Thus, it is not surprising that there was more turnover (a lower percentage residing in the same residence as the previous year) for African-Americans as renters are more mobile than homeowners (Moore, 2017). In addition, subsidized housing would have protected some African-Americans from being displaced between 2006-

10 and likely accounts for the higher percentage of population in the same residence as the previous year for 2010 in contrast to 2016. Between 2012-2016, the economic recovery may have allowed more movement for African-Americans as it did for their white counterparts.

Table 8.5. Comparison of Select Characteristics for African-Americans and Whites in Tract 84, 2000- 2016. Median Educational Same House 1 Household Attainment year ago Income African-American 2016 14.80% 71.10% $15,227

African-American 2010 11.00% 79.90% $13,469

African-American 2000 13.30% X $17,308

White 2016 44.40% 82.20% $75,625 White 2010 36.50% 94.10% $62,344 White 2000 38.90% X $42,875 Source: ACS 2006-10 and 2012-16.

65 The 2000 and 2010 data are 100 percent counts. The 2010 data was collected in 2010, not in the year prior to the ACS 2006-2010 survey.

221

Google Analysis: Using Google Streetview on four blocks selected randomly from Tract 84 I found that 3 of these blocks had not experienced the disinvestment process by 2007; a precursor to the gentrification process.

In other words, the housing stock was in good condition as of 2007 in 3 of the 4 blocks sampled. The size of blocks in the Tract 84 sample varied from 311 acres (Block 1000) to 5.3 acres in Block 1003 (CAGIS shapefiles). For example, Block 1003 had very little variability in the types of housing and the condition of houses. All buildings were single-family homes positioned on medium-sized lots defined as about 0.2-

0.25 acres (Hamilton County Auditor). Not surprisingly, the scores for the block faces did not deviate much in either 2007 or 2014/16 (See Figure 8.2).

In contrast, Block 1000, the largest block in Tract 84, showed much variability. Within the same block is found subsidized housing, senior housing, luxury apartments, and large well-cared for houses on large lots (+0.25 acres according to Hamilton County Auditor). The amount of green space in this block is also extensive. Two out of the four block faces in Block 1000 were categorized as in disinvestment in

2007. The block was not considered disinvested in that the average for the four blocks was above 0.50.

By 2014/16, the two disinvested block faces had gentrified moving the block from 0.55 in 2007 to 0.63 in

2014/16.

While there are only eight blocks in Tract 84 and only four of those were sampled, there is a spatial distribution to the gentrification process. It is focused on the Hamilton Avenue/Belmont intersection. Hamilton Avenue provides a conduit to the uptown neighborhoods, including the University of Cincinnati and other regional job centers as well as downtown. While not part of Tract 84, mixed-used developments have recently been constructed or are in varying stages of planning/construction in adjacent

Tracts 81 and 82.01 (http://chcurc.com/wp-content/uploads/2014/09/chcurc-annual-meeting-presentation-

2016.pdf). The previous chapter examined the gentrification of Tract 74 in Northside and noted a substantial increase in median housing prices and median rents. It may be that the gentrification process has moved closer to the city-suburb interface as the cost of living in Northside for gentrifiers increased.

This would be in keeping with Smith’s (1979) hypothesis that capital moves into neighborhoods further

222 from the central city as the gentrification process becomes more economically profitable to move into underexploited areas.

223

Figure 8.2. Gentrified Blocks in Tract 84, 2007 and 2014/16.

224

Table 8.6. Sampled Blocks in Tract 74 and Stage of Gentrification for 2007 and 2014/16.

Key66:

Disinvestment below 0.50 Early between 0.50-0.57 Middle between 0.58-0.64 Late between 0.65-0.80 Turnover over 0.80

2014/16 Difference between 2007 and Block 2007 Score Score 2014/16 and Stage of Gentrification Went from early stage/not disinvested 1000 0.55 0.63 to middle stage[1] No evidence of disinvestment in 1002 0.75 0.81 2000; slight improvements led to turnover by 2016 No evidence of disinvestment in either year; consists of moderate to 1003 0.89 0.89 spacious owner-occupied housing on suburban-sized lots Disinvestment to minimum 1007 0.4 0.44 improvements Tract 0.64 0.69 Average

Source: Google Streetview and Hwang (2015).

Examples of Non-disinvested/Disinvested/Gentrified Areas in Tract 84: Tract 84 is largely a residential neighborhood and thus there is no discussion of gentrification in commercial areas. Figures 8.3A and B display images of representative houses from block 1003 (the block with the highest gentrification score) and block 1007 (the only disinvested block in the sample) for

2014/1667. Two houses from Bellmeadows Drive are shown in Figure 8.3A. These are medium-sized ranch houses made of brick and are in excellent condition from a visual inspection of the outside. Note the spacious front yards and the well- crafted landscape. A sidewalk provides extra security for

66 Based on Hwang’s (2015) Google Streetwise methodology for determining gentrification.

67 There was very little change regarding the gentrification/disinvestment process in these blocks between 2007 and 2014/16 and thus only the 2014/16 images are displayed.

225 pedestrians but is probably unnecessary as this is a residential side street. This block received a 0.89 score in both 2007 and 2014/16 which indicates that it never went through a process of disinvestment.

Figure 8.3B is an image of representative housing in Block 1007 (Kirby Avenue) which consists of smaller ranches with vinyl siding. The roofs do not appear to be in as good of condition as the roofs in the

Bellmeadows image. The landscape is well-cared for indicating that residents have pride in their yards.

The driveway is unattractive due to the mix of paving types but is not in a state of disrepair. Sidewalks provide pedestrian safety as this is a major transportation artery. Note that garages were not part of the original construction of the houses. The house in the left-hand part of the image appears to have a car-port while the garage in the right-hand image is separately located at the back of the house. This Block received a score of 0.40 in 2007 and minimal improvement in 2014/16 (0.44). The below 0.50 score

(indicating disinvestment) largely is a result of the Hwang (2015) classification which gives higher scores to brick/stone buildings as these materials withstand the process of aging better than other materials such as vinyl.

226

Figure 8.3A. Spacious ranch houses on Bellmeadows Drive-Block 1003.

Source: Google Streetview, 2014.

Figure 8.3B. Small Ranch Houses on Kirby Avenue- Block 1007.

Source: Google Streetview, 2014.

227

Figure 8.4 is an image of Hickory Wood Townhomes (Block 1000) which is a subsidized housing option with 47 units. It was built in 2007 and is located on Hamilton Avenue. In contrast, Figures 8.5A and B display Bridges at Belmont which is a luxury apartment complex located at the intersection of

Belmont and Hamilton Avenues. These housing options, located in the same block (different block faces), illustrate the difficulty of analyzing the process of gentrification over a large territory. One might argue that the newly constructed Hickory Wood Townhomes are indicative of gentrification; a gentrification that has considered the effect of displacement on incumbent populations by providing housing opportunities. These townhouses are spacious, well-constructed, contain a yard and attempt to provide a middle-class life-style to those with limited financial resources.

Approximately one mile up the road, the luxury apartments at Bridges at Belmont show a boarded building with a fence to prevent unauthorized access to the building in 2011. In 2014, the building had been totally rehabilitated, the parking lot improved, and a stylish fence added to the property. Each of these properties is located on the bus line which improves accessibility to Uptown/Downtown and both provide high-quality housing at reasonable prices for their respective populations. In fact, Bridges at

Belmont website specifically states that it provides luxury apartments at affordable prices. Bridges at

Belmont’s location about one-third of a mile south of the gateway entrance to College Hill Business

District has likely influenced the rehabilitation of this property. The College Hill Community Urban

Development Corporation (CHCUDC) has raised millions of dollars through grants and donations over the past 10 years to revitalize downtown College Hill (CHCUDC, 2016). The issue is whether some type of mixed-income housing would have been a better option than the spatial separation of low-income and middle- income renters68.

68 The plans for the mixed-development complex focused on Marlowe Avenue (Tract 81) which has allotted low- income housing options is evidence that College Hill is pursuing a revitalization process that is inclusive of all economic groups.

228

Figure 8.4. Hickory Woods Townhomes located on Hamilton Avenue (Block 1000).

Source: Google Streetview, 2014.

229

Figure 8.5A.Vacant apartment complex (2011) on Belmont Avenue (Block 1000).

Source: Google Streetview, 2011.

Figure 8.5B. Newly Remodeled Luxury Apartments on Belmont, 2014.

Source: Google Streetview, 2016.

230

Figures 8.6A and B show small houses that were probably occupied by the working-class in the middle of the twentieth century. These images are located on the Southern end of Glenview Avenue.

Images from 2007 show these properties to be in slight disrepair (property on left) to more substantial disrepair (right). No attempt at beautification of the front yards is apparent in 2007. The 2014 image shows that the property at the left-hand side of the image was remodeled. It has been freshly painted and it appears to have a new roof. The front yard is simple but appears well cared for. The property on the right was demolished and a vacant lot with wildflowers has taken its place. It is uncertain whether this is an example of gentrification or an example of recovery after the housing crisis and economic recession.

The housing is not typically what gentrifiers target. Gentrifiers typically are interested in newly remodeled or built apartments/condominiums and housing that is architectually pleasing (Florida, 2005).

While not far from freeway exits and the amenities of the University of Cincinnati and downtown, its lack of walkability to nearby amenties would make it an unlikely choice for gentrifiers. Nor is there public transportation in proximity to this block face, making it additionally less attractive to gentrifiers.

231

Figure 8.6A. Small houses on Glenview Avenue in State of Disrepair, 2007.

Source: Google Streetview, 2007.

Figure 8.6B. Rehabbed House on Glenview Avenue, 2014.

Source: Google Streetview, 2014.

232

Gentrification appears to be in the very early stages in select blocks/block faces in Tract 84.

While the census data suggest the process is occurring, the evidence from the built environment is subtle.

While not located in Tract 84, the redevelopment of the Business District in College Hill since 2006 has likely had a profound impact on the gentrification of parts of Tract 84. The recency of this revitalization has probably not yet translated to substantial increases in median rent and median housing prices in Tract

84 as of 2016. Block 1007 was the only block that was disinvested in 2007, remaining so in 2014/2016.

This disinvestment will likely not remain so for long. Investment in Tracts 78 and 79 in western

Northside have been progressing swiftly over the last few years. At the time of this writing, Kirby Avenue was experiencing a major revitalization project focused on better control of stormwater/sewage flow

(http://www.projectgroundwork.org/downloads/westfork/WestForkWatershedFactSheet.pdf), which will likely affect property values in Block 1007 after its completion.

An Update on the Built Environment in Tract 84 in 2019: On June 29, 2019, I canvassed the sampled blocks in Tract 84 that were part of the Google

Streetview analysis. There was little change in the built environment from 2016 to 2019. Recall from the

2007 to 2014/16 Google Streetview analysis that only the Bridges Apartment Complex, located at

Hamilton and Belmont Avenues, received noticeable renovations over the study period. I returned to that corner to document any further changes in the built environment. Figure 8.7 shows two abandoned dwellings in proximity to the renovated apartment complex. Both buildings appeared in poor condition and it did not appear that renovation would be successful. I found it surprising that these two dwellings had not been removed from the landscape by CHCURC to beautify the landscape and to prevent any loitering or criminal activity. However, the lawn was mown showing that there was at least some concern for the appearance of these lots. Given that this location at Hamilton and Belmont Avenues appears to be prime real estate, it may be that the owners of the property are waiting for a higher selling price.

233

Figure 8.7. Two abandoned dwellings at the corner of Hamilton and Belmont Avenues.

Source: Photo by E. Ravuri, June 29, 2019.

234

I also returned to Block 1007, the only sampled block was not gentrified in 2007 or 2014/16. I was interested in documenting the Metropolitan Sewer District project occurring across the street from the houses on this block face. I noted no major changes in the dwellings on the block face between 2016 and

2019. Figure 8.8 shows the project. Given the association of wastewater and up-scale residences, I do not foresee Block 1007 as becoming gentrified in the near future.

235

Figure 8.8. Construction Site for Metropolitan Sewer District Project located on Kirby Road.

Source: Photo by E. Ravuri, June 29, 2017.

236

Interviews with Community Members: In December 2018 an email was sent to CHCURC (College Hill Community Urban Redevelopment

Corporation) explaining the project and asking the agency to provide contacts for individuals who could speak on economic revitalization and demographic change in College Hill between 2000 and 2016.

During May of 2019, interviews of individuals familiar with the economic revitalization and demographic change in College Hill were undertaken to provide a perspective on the changes taking place in the community.

Ed Loyd: Ed grew up in College Hill in the 1970s and 1980s. He returned to College in 2006. He earned a degree in History and Political Science from the University of Cincinnati. Ed is keeper of the College

Hill History Archives.

Beth McLean: Beth has resided in College Hill since 1988. She founded the College Hill Gardeners in

2002 to beautify areas around vacant buildings and vacant lots. Beth has also served on the College Hill

Community Redevelopment Corporation (CHCURC) for over a decade.

Terry Owens: Terry has lived in College Hill since 2012. She is a retired Cincinnati Public Schools

Teacher. She now owns the Silk End Textiles in the College Hill Business District. She served as the

College Hill Business Association’s President for four years.

Doug Trimmel: Doug moved to College Hill from Cleveland in 1974. He was an engineer before retirement and today spends his time renovating his 200- year old house on Belmont Avenue. He is actively involved in the revitalization of College Hill.

1. Can you pinpoint a year (s) when you first started to see changes in College Hill suggesting that

economic reinvestment and/or revitalization was occurring? Where did the process begin within

College Hill? Be as specific as possible.

237

Each of the interviewees had moved to College Hill prior to 2000 and could document that the business district was vibrant up until the early 1990s. Several interviewees responded that one could shop in the business district and get most of one’s daily needs. Interviewees identified the intersection of North Bend and Hamilton Avenues as the location where revitalization began69.

Terry: Terry moved to College Hill in the early 1990s. She laments the loss of Graeter’s Ice Cream and several florists by the 1990s and noticed that the business district of College Hill was in a state of decline.

Terry marks the turning point for revitalization of the business district around 2005-06 and credits this to the City’s monies appropriated for façade improvements to businesses in College Hill. Terry bought a business in 2012 (Silk End Textiles) and took advantage of the second façade improvement by getting new windows/awnings on her building and getting the building painted.

Beth: Beth moved to College Hill in 1988 and remarked that the business district was still vibrant at that time. By the 1990s, signs of decline began as businesses closed and vacant buildings became prevalent.

Beth credits the revitalization of the business district in College Hill to CHURC with its investments in vacant properties. While the original revitalization of College Hill focused on attracting businesses, recent emphasis has been on providing housing. The Dixon Building (corner of Hamilton and North Bend) is currently being renovated and will consist of 17 apartments owned by CHURC. The bottom floor contains the College Hill Coffee Company which is a prosperous business in the business district. Marlow

Court (founded by Episcopalians) is a recently built complex that provides affordable housing for seniors while Hollywood Properties is a private developer of market-rate apartments.

69 The string of businesses at the Southern end of College Hill along Hamilton Avenue have experienced little evidence of gentrification.

238

Ed: Ed grew up in College Hill during the late 1970s and 1980s. Ed remembers the business district as a thriving economic area. Ed left Cincinnati in the 1990s, but periodic visits to his parent’s house revealed that College Hill’s business district was basically defunct by the end of the 1990s. Ed moved back to

College Hill in 2006 before the business district was revitalized. He noticed the change in the business district around 2012 when CHURC began to purchase properties and direct redevelopment efforts. Ed laments the closing of businesses that he visited during his childhood (Hollywood Theatre, Graeter’s Ice

Cream), but has found a new array of businesses that he frequents (Tortilleria Garcia, Brink’s, Marty’s

Hops and Vines). Most of these businesses have opened in the past 5 years.

Doug: Doug was the only one of the four interviewees from College Hill that focused more on revitalization in housing than in the business district. Doug has noted since 2013 that there has been a major influx of homeowners into College Hill. Much of this has to do with the affordability of housing stock in College Hill.

2. What effect did the Great Recession (2007-2009) have on economic reinvestment/revitalization in

College Hill?

The Great Recession had more of an impact on housing than it did on businesses as the business district in

College Hill was at its nadir by the early 2000s. Many homeowners lost their homes during this time and transient populations increased in College Hill.

Three houses contiguous to Ed’s property (all on Astor Place) went into foreclosure during the Great

Recession. These were middle-class homeowners who became overextended and could not pay their mortgages. To this date (May 2019), as far as Ed knows, the area around Cary and North Bend experienced a wave of foreclosures and the houses are still unoccupied. Doug’s view of the Great

Recession was that it affected those with less of an economic ‘cushion’ to weather downturns in the economy (those who purchased smaller houses which they could marginally afford). Doug recalls the

239 effect of the Great Recession on his street (Belmont Avenue) where houses/rental units were abandoned and problems with crime and upkeep of the property became major problems. Terry related that four or five houses on her street went into foreclosure during the Great Recession. These experiences would be expected given that College Hill experienced a greater number of housing foreclosures than any other neighborhood in Cincinnati between 2006 and 2012

(http://www.marcconline.com/files/DCHO_06_12_2013_1_Foreclosure-Report-2012_WebCopy.pdf .).

3. Can you comment on the impact of private development agencies versus non-profit/government

agencies in the reinvestment/revitalization process?

The interviewees revealed that CHURC had the greatest effect on revitalization in College Hill. Terry stated that the College Hill Investment Corporation (CHIC), a non-profit organization that consisted of six women, began in the early 2000s and has bought properties along Hamilton Avenue that were vacant as an attempt to revitalize the community. According to Beth, private developers have been less interested in real estate in College Hill because many of the parcels are not large enough for large-scale developments that would be cost effective.

4. Has reinvestment/revitalization led to a change in the demographic composition (e.g. age,

race/ethnicity, socio-economic status) of College Hill? If so, in what way? Have certain areas

been more affected than others? How did the Great Recession affect demographic composition in

College Hill?

Each of the interviewees remarked that there has been an influx of population under age 40 into College

Hill in the last decade. Many of these individuals grew up in College Hill and moved elsewhere and decided to move back. Ed was one of those individuals who returned to College Hill in his 30s and

240 purchased his parent’s house (the house he grew up in). This return to College Hill was sparked by affordability of housing in the neighborhood.

Doug recalls the 1970s when College Hill had a high percentage of elderly population, but this has changed in the past ten years. Both Ed and Doug commented on the recent arrival of young families to

College Hill.

When asked about racial change in College Hill, each respondent stated that College Hill has always been racially diverse and that this was a point of pride for its residents. Ed remarked that on his street

(Hillcrest) that there were only white homeowners when he was growing up (late 1970s/1980s). This is no longer the case and he views this as a positive change. None of the interviewees revealed that displacement of the African-American population70 was a major issue, but a few noted that the Great

Recession had a greater impact on African-American homeowners than white homeowners.

5. Can you comment on the changes in types of businesses or improvements in already established

businesses in your community during the gentrification process?

Ed remarked that the E-Z market was a dilapidated business that was constantly a nuisance to the community (loitering, littering, noise). CHURC purchased this property, demolished the building and turned the site into a pocket park. Ed has also noticed that businesses along Hamilton Avenue are much more up-scale than they were when he was a child. These businesses include Brink’s, Red Rose Pizza,

Marty’s Hops and Vines, and the College Hill Coffee Company. Terry’s business (Silk End Textiles) is indicative of gentrification and appeals to customers inside as well as outside of College Hill.

70 Transient populations were articulated as a major problem and is likely that many renters were displaced during the Great Recession.

241

A few of the respondents noted that the off-street parking behind several of the businesses at Hamilton and North Bend has been instrumental in attracting a larger clientele to these businesses. Each of the respondents noted the loss of Kroger’s and Shuller’s Wigwam (restaurant) in the past decade. However, according to my respondents College Hill will not allow big-box stores and fast-food restaurants to be erected in the business district which does not fare well for the opening of another grocery store. Mixed- use development is the keystone to revitalization in College Hill and thus perhaps a small grocery store could be added in this complex. It is hard to imagine that those areas will not soon be developed into businesses.

6. Is there anything else you would like to mention concerning economic revitalization and

demographic change in College Hill?

Each of the respondents stressed that College Hill is attractive due to its diversity (racial, age, socio- economic status) and they would not have it any other way. Doug responded that the diversity of housing types in College Hill allows everyone to be an individual.

Each of the respondents also noted that College Hill is central to everything in Cincinnati. It is close to the interstate, downtown, and the University of Cincinnati. Several of the respondents noted that College Hill had the best of the urban and rural worlds. One is surrounded by greenspace (e.g. Laboiteax Woods, Isaac

Weiss Farm).

A major complaint of the respondents was that much of the population in College Hill is transient (mostly low-income renters). These individuals are not invested in the community. My respondents did not suggest that these individuals be removed from College Hill, but that a program that turned some of these rental units into affordable owner-occupied units could be beneficial for all groups.

242

Conclusion: Changes to the built environment as shown through Google Streetview in Tract 84 were not congruent with the educational/economic changes that occurred in the Tract from 2010 to 2016. Hammel and Wyly’s (1996) gentrification indicators of growth in median household income, growth in percentage of population over 25 with at least a bachelor’s degree and growth in median housing value/median rent all exceeded the growth for the City of Cincinnati; identifying Tract 84 as undergoing gentrification between 2010 and 2016. However, other than the renovation of Bridges at Belmont, little change occurred in the built environment that may be associated with gentrification.

This lack of economic investment in the built environment in Tract 84 (although there was significant investment in neighboring Tract 82.01) may explain why the tract did not experience displacement of the African-American population between 2010 and 2016. The census data indicate that the white population in Tract 84 by 2016 was better educated and wealthier than whites in 2000 or 2010.

This likely came about from an exodus of lesser educated, poorer whites between 2000 and 2010 and the influx of better educated and wealthier whites from 2010 to 2016. This turnover in the educational/economic characteristics of white in versus outmigrants may be a result of two factors. First, the housing crisis adversely affected College Hill and likely led to an exodus of individuals who could not pay their mortgages. Secondly, preferences for residing in urban neighborhoods have increased over the past several decades and combined with an array of high-quality housing stock (recall that most blocks were not disinvested in 2007 or 2014/16), likely attracted newcomers. Another consideration is that while median rents and median housing values grew at higher rates between 2010 and 2016 than the City of

Cincinnati, the growth in Tract 84 was much lower than in other neighborhoods undergoing revitalization.

As the interviewees from College Hill noted, much of the influx of residents to College Hill over the past decade has been in response to availability of affordable housing

243

Chapter 9 A COMPARISON OF THE FOUR CENSUS TRACTS

Introduction: The overarching objective of this thesis was to determine if there was a difference in the policies enacted by different neighborhoods that prevented or facilitated the displacement of the African-

American population during the process of gentrification in the City of Cincinnati between 2000 and

2016. Four tracts in four neighborhoods were selected for this analysis: Tract 19 in Walnut Hills, Tract 55 in Madisonville, Tract 74 in Northside, and Tract 84 in College Hill. The emphasis on the displacement of the African-American population does not minimize the effect of gentrification on white individuals on fixed-incomes who have been displaced or who are struggling to continue to acquire the financial resources to remain in their current residence. An example of this displacement includes Mary-Cabrini

Durkin71 from Walnut Hills. An elderly, white female with limited financial resources, Ms. Durkin was displaced from her apartment several times in the past decade due to rising rents caused by economic revitalization. Unfortunately, it is difficult to track displaced individuals as there are no official statistics in Cincinnati which track this process72. Atkinson (2000) considered the tracking of displaced individuals as counting the invisible- an almost impossible task. Thus, the focus on the displacement of African-

American population is an artifact of available data; residential displacement can be tracked by documenting the change in the number and percentage composition of a tract’s African-American and white populations between two years and inferring possible reasons for this change73. This information is available at the tract level in the U.S. decennial censuses and in the five-year ACS data.

This chapter will focus on a comparison/contrast of changes indicative of gentrification in the four selected neighborhoods to attempt to uncover why displacement of the African-American population

71 Interview with Mary-Cabrini Durkin, May 2019. 72 New York City has a database that allows tracking of displaced individuals, but this tracking system is not found in most cities. 73 The American Housing Survey (AHS) is an excellent source that tracks movement of households and the reasons for those moves longitudinally. Unfortunately, that data source is nationally representative of mobility in the U.S. and is not detailed enough to examine change at the neighborhood level.

244 occurred in two of these tracts (Tracts 55 and 74) and not in the other two (Tracts 19 and 84) between

2010 and 2016. The discussion will also examine possible reasons why displacement of African-

Americans occurred in Tract 19 between 2000 and 2010 but did not occur between 2010 and 2016. The methods of analysis to interpret these processes include: a review of census data, visual examination using

Google Streetview imagery, and interviews with community members. Only the census data is objective.

Each tract is compared on the same demographic, social, and economic characteristics to see how these characteristics changed between 2000 and 2016. These changes are the result of differences in the characteristics of both in-migrants and out-migrants (as well as non-migrants) to/from each of the tracts over the period of analysis. An analysis of changes in the built environment is accomplished by using

Google Streetview imagery between 2007 and 2016. Unfortunately, providing an objective overview of changes in the built environment would be too time consuming as one would need to view every block in the four census tracts. About twenty percent of the total blocks (fifty percent in the case of Tract 84) provided an appropriate sample size according to Hwang’s (2015) criteria. As the choice of blocks was generated using a random number table, the process was as objective as possible. The number of blocks in the census tracts ranged from 8 in Tract 84 (College Hill) to 90 in Tract 55 (Madisonville). The Google

Streetview visual analysis is only one indicator of the gentrification process and it should be viewed as such. The most subjective part of the analysis was the interviews with community members. Each of these individuals was highly involved in the social, political, and economic functioning of their communities, but had different perspectives on affordable housing, displacement, and the revitalization efforts in their communities.

The three data sources (Census, Google Streetview, interviews with members of CDCs) ideally would have covered the 2000 to 2019 timeframe, but this was not possible. Census data at the tract level is lagged by a few years74. Thus, the study of changes in demographic, social, economic and housing

74 The 2017 census data at the tract level was released earlier in 2019, but my analysis had already been carried out with the 2016 data.

245 characteristics by tract ends in 2016 leaving three years of data unaccounted for. Secondly, Google

Streetview imagery only became available in 2007 leaving changes in the built environment for most of the first decade of the 21st century undocumented. Like the census data, there is a lag in Google

Streetview and the latest images of the built environment are from 2018 requiring the analyst to document the most recent changes on her own by canvassing the neighborhoods and taking updated photographs of

Google Streetview images shown in 2014/16. Interviews with community members were initially intended to align with the 2000 to 2016 study period. This became impossible to control; interviewees were more cognizant of the latest changes in their neighborhoods than during the 2000 to 2016 period. In almost all cases, interviewees needed to go back before 2000 and discuss their personal experiences in the neighborhood and/or other historical components of the neighborhood to provide context for what happened in the built environment post-2000.

Census Data: Population Change Table 9.1 displays total population change and change in racial composition for the four census tracts (Tracts 19, 55, 74, and 84) in this study between 2000 and 2016. Tract 55 (Madisonville) and Tract

19 (Walnut Hills) experienced a 7.4 and 10.4 percent decline in total population, respectively, between

2000 and 2016. Tract 74 (Northside) underwent the greatest decline in total population between 2000 and

2016 (-24.4 percent)75. Tract 84’s (College Hill) population was stable between 2000 and 2016.

While the total population in the four tracts has declined or stabilized under the gentrification process occurring between 2000 and 2016, that is not the emphasis of this study. The remainder of this section examines changes in the number and percentage composition of African-Americans and whites in each of the four tracts to determine if displacement of the African-American population occurred. Recall that Taeuber and Taeuber (1965) define displacement between two timeframes as 1) decline in number of

75 The population decline in each tract occurred solely between 2000 and 2010 and then increased between 2010 and 2016 but was still below the total in 2000.

246

African-Americans in the tract; 2) an increase in number of whites in the tract; and 3) a decrease in the percentage composition of the tract’s population that is African-American of at least 5.0 percent.

Table 9.1. Population of Tracts by Race, 2000-2016.

Tract Tract 55 Tract 74 Tract 84 19Walnut Madisonville Northside College Hills Hill Total Population, 1,754 3,982 2,210 2,015 2000 Total Population, 1,445 4,000 1,598 2,100 2010 Total Population, 1,572 3,686 1,670 2,075 2016 African-American 944 3,096 975 861 Population, 2000 African-American 623 2,956 646 1,188 Population, 2010 African-American 611 1,954 356 1,178 Population, 2016 White Population, 706 758 1,182 1,093 2000 White Population, 738 851 861 819 2010 White Population, 816 1,389 1,115 851 2016 African-American 53.8% 77.7% 44.1% 42.7% Percent, 2000 African-American 43.1% 73.9% 40.4% 56.6% Percent, 2010 African-American 38.9% 53.0% 21.3% 56.7% Percent, 2016 White Percent, 2000 40.3% 19.0% 53.8% 54.2% White Percent, 2010 51.1% 21.3% 53.9% 39.0% White Percent, 2016 51.9% 37.7% 66.8% 41.0% Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

247

Little change in racial composition occurred in Tract 55 (Madisonville) between 2000 and 2010.

There was an increase of whites and a decline of African-Americans, but this did not qualify as displacement as the percentage change in African-American composition in the tract declined by less than

5 percent. However, Tract 55 experienced a 33.9 percent decline (-1,002) in African-American population between 2010 and 2016, suggesting mass displacement of African-Americans. Over the same period, the tract experienced an increase of almost 100 percent in its white population (+538). Thus, according to

Taeuber and Taeuber (1965) Tract 55 experienced displacement between 2010 and 2016. Evidence from the built environment (Google Streetview) and a review of planning documents indicated that this increase in white population and decrease in African-American population occurred during a time of reinvestment in the community; in other words, gentrification.

Tract 74 (Northside) experienced a decline in both number of African-Americans and whites between 2000 and 2010 disqualifying it as a candidate for displacement according to Taeuber and

Taeuber’s (1965) methodology. Disinvestment in Northside between 2000 and 2010 was rampant. Recall

Stefanie Sunderland who remarked about the large number of vacant houses and the proliferation of crime at this time. Stefanie recalls that middle-class home-owners (regardless of race) left Northside during the early 2000s. As Smith (1979) claims in his rent gap thesis, areas that have been disinvested for a long time are ripe for redevelopment. Northsiders Engaged in Sustainable Transformation (NEST) was organized in 2004 and this provided a catalyst for some limited development which attracted a small number of middle-class, educated individuals to Northside which elevated its status to a gentrified tract by

201076. Between 2010 and 2016 the white population increased substantially and rising costs of living

(discussed below) likely caused displacement of the incumbent African-American population. Tract 74’s

76 Recall that the Hammel and Wyly (1996) gentrification designation required that the 2000 median household income below the city average to be eligible for gentrification. Three additional criteria were then applied 1) a 2000-10 growth of median household income above the city average; 2) a 2000-10 growth in percentage of over 25 population with at least a bachelor’s degree above the city average; and 3) a 2000-10 growth in median housing value or median rent above that of the city average. The same procedure for determining gentrification is followed for the 2010-2016 timeframe.

248

African-American population declined by a further 44.9 percent between 2010 and 2016. Stefanie

Sunderland noted that absentee landlords simply stopped renting properties when NEST insisted that they be brought to code. It is likely that this had the unintended effect of displacing some low-income individuals (white and African-American) while rising costs of living from the revitalization of Northside likely contributed to displacement. Tract 74 ended the 2000-16 timeframe with the lowest percentage of

African-American population of the four tracts (only 21.3 percent) and establishing Tract 74 as experiencing displacement of African-Americans between 2010 and 2016.

Tract 19 (Walnut Hills) experienced a decline of 35.3 percent in its African-American population between 2000 and 2016, with most of that decline (34 percent) occurring between 2000 and 2010. In

2000, African-Americans comprised 53.6 of the total tract population, but this was reduced to 38.9 percent by 2016. Census numbers suggest that the increase in the white percentage of the total population of Tract 19 was more a result of the decline in African-American population than it was an increase in white population (at least between 2000 and 2010). It is hard to imagine how the increase of 32 whites between 2000 and 2010 could have caused the decrease of over 300 African-Americans from the tract.

Even between 2010 and 2016, the white population only increased by ten percent (78 whites added to the tract)77. Tract 19’s exodus of African-American population between 2000 and 2010 suggests that some other factor was driving this process other than the in-migration of whites (gentrifiers). From my interviews with CDC members of the Walnut Hills Redevelopment Foundation (WHRF) it was evident that private investors were buying up many of the large houses (used as rentals) that had been subdivided earlier in the century and returning them to single-family, owner-occupied homes during the late 1990s.

African-American displacement likely occurred in the 2000 to 2010 period as these multi-family rental units were converted to single-family homes that attracted higher-income individuals and presumably housed smaller household sizes than the rental units that they replaced. This seems to make sense in that

77 Interviewees documented the substantial increase of whites in Tract 19, Walnut Hills (as of May 2019) suggesting that future ACS data will reveal a major influx of whites into Tract 19 and may result in an exodus of African-Americans.

249 over 300 less African-Americans resided in Tract 19 in 2010 than in 2000 while an addition of only 32 whites accrued to the tract during this time. After 2010, developers focused on converting vacant buildings to apartments; this may explain why Tract 19 did not experience displacement of African-

American population as it did between 2000 and 2010.

Tract 84 (College Hill) went against the trends of the other three tracts. Its African-American population increased from 42.7 percent of the tract total in 2000 to 56.7 percent in 2016. College Hill was the only one of the four neighborhoods that did not begin the gentrification process between 2000 and

2010; thus, the displacement of the African-American population between 2000 and 2010 would not be expected. Displacement typically occurs later in the gentrification process as increasing numbers of newcomers compete with the incumbent population for housing or the influx of residents cause an increase in the cost of living (Bates, 2013).

From the analysis of change in the composition of the tract population in the four study neighborhoods between 2010 and 2016, it appears that a large influx of whites78 to the tract over a short period of time resulted in displacement of the African-American population. Note that Tract 84 (College

Hill) and Tract 19 (Walnut Hills) only experienced an increase of 32 and 78 whites, respectively, between

2010 and 2016. Neither of these tracts experienced displacement of African-Americans during this time.

Conversely, several hundred whites moved into Tract 55 (Madisonville) and Tract 74 (Northside) between 2010 and 2016 and a large exodus of African-Americans ensued in both tracts79. This is to be expected; as demand for housing increases, costs will increase – as costs of living increase in a gentrifying tract those with less financial resources need to seek housing elsewhere. In addition, substandard housing is often demolished or abandoned by absentee landlords80 during the process of

78 Since tracts had different total populations in 2010, I created a rate of white population change between 2010- 16 to the total tract population in 2010. The formula is (white total 2016-white total 2010)/total population 2010. Tracts 84 and 19 had white change rates of 1.5% and 5.4%, respectively, versus white change rates of 15.2% for Tract 74 and 14.6% for Tract 55. 79 Each neighborhood (including College Hill) has experienced declining populations in the past several decades, largely a result of the suburbanization process. 80 Interview with Stefanie Sunderland, May 2019.

250 gentrification which may limit the availability of dwelling units for low-income populations and provide the impetus for relocating elsewhere in the City.

Comparison of Median Household Incomes by Tract for African-Americans and Whites: Changes in the numbers and percentage composition of African-American and white populations is the only indicator that can be used to determine if displacement occurred using census data. However, it is also necessary to see how the economic, social, and housing characteristics of African-Americans and whites changed over the study period in the four neighborhoods. This will allow us to infer characteristics of in-migrants and out-migrants.

Table 9.2 examines the median household income, by race, for the four census tracts for the years

2000, 2010, and 2016. From the raw data in the table one can calculate that Tracts 74 and 19 experienced the greatest increase in median household income between 2000 and 2016 (71.3 and 79.6 percent; respectively) suggesting an influx of individuals with higher incomes and likely an exodus of individuals with lower incomes.

Tract 19: Tract 19’s African-American population experienced increases in median household income from 2000-10 suggesting that those African-Americans with the least financial resources left the tract.

Alternatively, the increase could be due to inflation between 2000 and 2010. Looking to how population changes in Tract 19 affected median household income it should be noted that Tract 19 lost 35.3 percent of its population between 2000 and 2010. The 79.6 percent increase in median household income between

2000 and 2010 suggests that higher-income individuals moved into the tract. Recall that only 32 whites were added to Tract 19 between 2000 and 2010. There was likely a churning of white population during this time as lower-income whites, just like their lower-income African-American counterparts were priced out of living in Tract 19. In this sense churning refers to the replacement of low-income whites with higher-income whites. Unfortunately, the census data does not allow us to measure this process.

Tract 74: In Tract 74, African-Americans experienced a 58.6% increase in median household income between 2000 and 2010 and by 2010, African-American median household income was higher than that

251 of whites. This is not surprising in that Stefanie Sunderland’s testimony already stated that middle-class whites were leaving Northside in the early 2000s due to crime whereas lower-income individuals (mostly

African-American) left after 2004 when NEST applied pressure on authorities to ensure that housing codes were enforced on properties owned by absentee landlords. The 38.9 percent decline in median household income for African-Americans in Tract 74 between 2010 and 2016 was likely a result of the exodus of middle-class African-Americans who were caught up in the housing crisis.

Tract 84: Median incomes of African-Americans declined by 22.2 percent for Tract 84 between 2000 and

2010, but the African-American population increased suggesting an influx of lower-income African-

Americans to Tract 84. Recall that at least 20 percent of Tract 84’s housing units were subsidized/affordable units and would provide housing opportunities for low-income individuals being forced to leave their residences in other neighborhoods during the Great Recession.

Tract 55: Median household incomes for African-Americans in Tract 55 were relatively stable between

2000 and 2010 but declined by 40 percent by 2016, likely a result of the out-migration of middle-income homeowning African-Americans due to the housing crisis. From above, recall that displacement did not take place between 2000 and 2010, but a widespread displacement did occur between 2010 – 2016.

All Tracts: African-Americans in Tracts 55 and 74 had comparable incomes to their white counterparts in 2000. Gentrification appears to have been a blow to middle-class African-Americans in each of these tracts as the trajectory of these two groups’ median household incomes diverged after 2010. Whites in

Tract 19 and 84 had much higher incomes than their African-American counterparts in 2000 and the gap has only gotten larger in the years since. It is hard to imagine that whites in College Hill or Walnut Hills are in competition with African-Americans for housing given their differences in median household income; in addition, the provision of subsidized housing in these tracts likely stopped some displacement from occurring. Unless investors (individuals or corporate) renovate properties that currently have low- income individuals residing in them, it is hard to see how this process of displacement would occur in areas such as Tracts 19 and 84 which clearly have two separate housing markets co-existing.

252

The divergence in the median household incomes between whites and African-Americans in these four tracts is part of a larger trend in the U.S. in which the middle-class has steadily declined over the past few decades (FRED81, 2016; NPR82, 2016). Much of this erosion of the middle-class was due to the decline in manufacturing employment which allowed both whites and non-whites with relatively low educational levels to obtain jobs which provided middle-class wages. African-Americans were more adversely affected by the decline in manufacturing beginning in the 1960s than their white counterparts as a much greater percentage of African-American workers were employed in manufacturing (Wilson,

1987). It is widely known that whites have had better access to educational opportunities throughout the twentieth century than their African-American counterparts. As of 2016, 33.8 percent of whites over age

25 had at least a bachelor’s degree compared to only 20.0 percent for African-Americans (U.S. Census,

2016). Education (academic or vocational) is now one of the few ways to obtain middle-class status in the

U.S. in the twenty-first century (Autor, 2011) and unfortunately, is much more difficult for African-

Americans because of the legacy of discrimination in the U.S. educational system (Cook, 2015).

The Great Recession further damaged the middle-class as many jobs in manufacturing and other low-wage industries were eliminated (Autor, 2011). Again, African-Americans were more affected by the

Great Recession as a greater percentage of this population was employed in those industries most affected by the recession. While the transition to a post-industrial economy has largely been completed and the

U.S. has recovered from the Great Recession, one of the greatest issues of national concern is that of housing affordability for the middle-class. Between 2012 and 2016, housing prices in the 30 largest metropolitan areas in the U.S. increased by 26 percent versus a median household income increase of only

1.6 percent (Richardson, 2017). In 2012, about 58.0 percent of housing in Cincinnati was affordable to the middle-class, but by 2016, this figure declined to 48.0 percent. By 2016, only 10.0 percent of homes in

81 The share of adults in the U.S. middle-class has declined from 61.0 percent in 1971 to about 50.0 percent by 2015. 82 The Gini Coefficient is a measure that scores income inequality from 0 (total equality) to 100 (total inequality). The Gini Coefficient in the U.S. has increased from 34.5 in 1979 to 41.5 by 2016 illustrating that income has become more inequitable in the U.S. over the past four decades.

253

Cincinnati were affordable to African-Americans whereas this figure was 60.0 percent for whites

(Richardson, 2017). Thus, whites in Cincinnati (as well as other cities) have the economic clout to outbid their African-Americans counterparts for preferred neighborhoods in the City83. Interviews with members of the CDCs in all four neighborhoods noted that housing affordability for the middle-class was a major issue in their neighborhoods that would need to be addressed in the future. Recall that Tract 84 (College

Hill) only began the gentrification process between 2010 and 2016 whereas the other tracts began the process between 2000 and 2010. Several of the College Hill interviewees noted that Northsiders had moved to College Hill over the past few years because of the housing affordability crisis in Northside caused by the revitalization of the neighborhood.

83 Annual population change for central cities in the fifty largest cities in the U.S. between 1970 and 2013 revealed that the annual growth rate for central cities was the fastest during the 1990-2000 timeframe (1.20 percent) and the 2010-2013 (1.10 percent). Annual growth in the suburbs of these metropolitan areas revealed a 1.80 percent growth between 1990 and 2000 and only a 0.90 percent growth between 2010-2013 (Acolin, Voith, and Wachter, 2016).

254

Table 9.2. Median Household Income for African-Americans and Whites by Tract, 2000-2016.

Tract 19 Tract 55 Tract 74 Tract 84 Walnut Madisonville Northside College Hill Hills African- $17,049 $28,196 $22,857 $17,308 American,2000 African- $20,625 $27,986 $36,250 $13,469 American, 2010 African- $21,389 $16,375 $22,143 $15,227 American, 2016 White, 2000 $45,750 $27,692 $27,333 $42,875 White, 2010 $61,250 $36,199 $32,361 $62,344 White, 2016 $65,250 $51,800 $49,286 $75,625 Total, 2000 $26,967 $28,424 $25,932 $26,599 Total, 2010 $46,750 $32,342 $31,976 $28,700 Total, 2016 $48,438 $35,231 $44,429 $32,100 Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Median Housing Values and Median Rents:

See Table 9.3 for this discussion. Tract 19: Median housing value increased from $107,700 to $ 174,400 (a 61.9 percent increase) between

2000 and 2016 giving Tract 19 the highest median housing values as of 2016 for the four tracts in this study. Median rent saw a substantial increase ((74.0 percent) between 2000 and 2010 going from $484 to

$842. Recall that Tract 19 experienced a decline of one-third of its African-American population between

2000 and 2010, likely a result of the increases in median rent. Given that 84.5 percent of African-

Americans in Tract 19 lived in rental housing in 2000, it is likely that many of these individuals sought lower cost housing in other areas of Cincinnati. Surprisingly, median rent between 2010 and 2016 declined by 4.6 percent. This decline in median rent suggests that some mechanism suppressed increases in rent. These could be subsidized housing opportunities or some other provision for affordable housing.

255

Conversely, this stabilization of rent between 2010 and 2016 may be a result of a lull in the renovation/construction of high-end apartments. Kathyne Gardette (WHRF) stated that it took several years after the 2012 conversion of East MacMillan to a two-way street for development dollars to flow into Walnut Hills. Trevarran Flats, the Hauck Building, and Windsor Flats are all high-end apartments in

Tract 19 that would not have been included in the calculation of median rents for 2016 as they were not ready for residents at this time. Most importantly, the stabilization of median rents between 2010 and

2016 probably kept displacement of African-American population from occurring in Tract 19 during this time. By 2010, much of the low-income population in Tract 19 would have moved elsewhere and only those with higher median incomes would be able to remain.

Tract 74: Median housing prices in Tract 74 increased from $56,000 to $115,000 (a 100.5 percent increase) between 2000 and 2016. Of the four neighborhoods in this analysis, Tract 74 had the lowest median housing values in 2000 (See Table 9.3). By examining Smith’s (1979) rent gap theory, it is obvious that capital invested in Northside had the greatest potential to turn a profit as it was the most disinvested neighborhood of the four. Median rent in Tract 74 increased from $484 to $810, a 65.3 percent increase; giving the tract the highest median rent of the four tracts by 2016. Note that the African-

American population declined by 63.5 percent between 2000 and 2016. Some of this decline undoubtedly was a result of African-American renters being unable to afford residing in Tract 74, but it is also likely that some middle-class African-Americans were also displaced from the tract. In 2010, median household income for African-Americans in Tract 74 was $36,250 versus $22,143 in 2016 suggesting that middle- class African-American homeowners were also priced out of the housing market.

Tract 55: Tract 55 had the second lowest median housing values in 2000. The tract only experienced a

42.6 percent increase in median housing value between 2000 and 2016, but, median rent increased from

$397 to $707 (a 78.1 percent increase). Along the same lines of reasoning as advanced in the discussion of Tract 74, the lower median housing values and median rents in Tract 55 would create the next largest potential return on capital (of the four neighborhoods in this analysis) and attract investors and residents.

256

The one-third decline in African-American population in Tract 55 between 2010 and 2016 suggests that many African-Americans were unable to afford housing in Tract 55 by 2016. This is not surprising given that this decline in African-American population occurred around the same time as the housing market crisis/Great Recession.

Tract 84: Housing values in Tract 8484 were the highest of the four tracts in 2000 ($129,000), but only increased by 5.8 percent between 2000 and 2016. Returning to Smith’s (1979) rent gap theory, it might be said that the housing stock in Tract 84 had not depreciated to such an extent that a profit could be made on investing in the housing stock in this tract. While median rent increased between 2000 and 2010 in

Tract 84, it declined to below its 2000 value by 2016 suggesting that more individuals were taking advantage of some type of rent subsidy. In my interviews with members from the College Hill CDCs, it was found that several middle-class individuals from Northside moved to College Hill due to rising housing costs in Northside. It is not unreasonable to hypothesize that lower-income African-Americans taking advantage of rent subsidies would also move from Northside to College Hill. Recall the increase in

African-American population in Tract 84 from 861 to 1,188 (327) between 2000 and 2010 and the loss of

329 African-Americans from Tract 74 during this same time.

84 Tract 19 (Walnut Hills) had unaffordable median rents as of 2016 but this was a legacy of the increase that occurred between 2000 and 2010.

257

Table 9.3. Median Housing Value and Median Rents for Tracts, 2000-2016.

Median Median Median Median Median Median Housing Housing Housing Rent 2010 Rent 2010 Rent 2016 Value, Value, Value, 2000 2010 2016 Tract 19 $107,700 $156,500 $174,400 $484 $842 $803 Walnut Hills Tract 55 $73,700 $89,000 $105,100 $397 $632 $707 Madisonville Tract 74 $56,000 $105,500 $115,000 $494 $606 $810 Northside Tract 84 $129,500 $149,500 $137,000 $582 $650 $528 College Hill Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Vacancies and Homeownership85: See Table 9.4A-C for this discussion. Each of the tracts had comparable percentages of vacancies in 2000 (between 10 and 13 percent). However, by 2010, Tract 19 (Walnut Hills) had a substantially higher vacancy rate than the other three (28.8 percent). Much of the renovation in Walnut Hills since

2010 has focused on vacant buildings; these do not displace individuals. This type of redevelopment perhaps explains how the neighborhood gentrified without displacement of the African-American population in Tract 19 between 2010 and 2016.

85 Walnut Hills was the least affected (in terms of numbers of foreclosures) neighborhood of the four during the housing crisis that accompanied (or caused) the Great Recession (A Decade of Foreclosures, 2012). Part of this is to be expected in that Walnut Hills had home-ownership rates that were below that of the other three neighborhoods (only 23.5 percent compared to 37 plus percent for the other three in 2000 for Tract 19). Between 2006 and 2012, College Hill had 586 foreclosures; Madisonville, 504; and Northside 450 (A Decade of Foreclosures, 2012). These foreclosures would have caused a tremendous amount of turnover in population in these communities. The number of foreclosures in Walnut Hills between 2006 and 2012 was 167.

258

Tracts 19, 74 and 84 experienced increases in the percentage of owner-occupied housing between

2000 and 2010 while a decline in owner-occupied housing occurred in Tracts 55. Many of these homeowners likely were subject to sub-prime lending and may have later lost their homes after 2010.

Tract 19: A decline in the number of owner-occupied housing units for African-Americans declined between 2000-10 and 2010-16. However, white homeownership increased in each of the time periods giving Tract 19 an overall increase of 39 owner-occupied units between 2000 and 2016 (a 17.8 percent increase). The overall increase in homeownership corresponded to a major exodus of the African-

American population during the decade and an increase in vacancy suggesting that these individuals were renters, artificially inflating homeownership rates.

Tract 74: For Tract 74, white homeownership increased to 57.5 percent (from 51.1 percent) between

2000 and 2010, suggesting that whites took advantage of home loans in the early 2000s and were able to hold on to those homes during the recession. Conversely, African-American homeownership in Tract 74 decreased from 31.1 percent to 29.7 percent between 2000 and 2010. While the percentage decline between 2000 and 2010 is minimal, this resulted in a decline in the number of African-American owner- occupied homes from 91 to 66 between 2000 and 2010 (or a 27.5 percent decline in the number of

African-American households in Tract 74). Even though African-American owner-occupied dwellings declined from 66 to 58 between 2010 and 2016, homeownership rates for African-Americans increased from 29.7 to 37.2 percent. This was largely a result of the exodus of African-American renters from Tract

74 between 2010 and 2016. In the early 2000s, demolition of housing in the West End led to many

African-Americans using Section 8 housing vouchers to settle in Northside (Northside Comprehensive

Land Use Plan Update, 2014), who then presumably moved on after cost of living increased in Tract 74.

Tract 84: Homeownership rates in Tract 84 increased for African-Americans from 20 to 33 percent between 2000 and 2010, but then decreased slightly between 2010 and 2016. In terms of total owner- occupied housing units, Tract 84 increased from 336 to 386 between 2000 and 2016. African-American owner-occupied units went from 70 in 2000 to 164 in 2016. Most of this growth in owner-occupied

259 housing for African-Americans occurred between 2000 and 2010. It is likely that African-Americans purchased housing that was more modest and were able to survive the housing crisis. Conversely, white owner-occupied housing declined from 271 units in 2000 to 213 units in 2016 and these may have been more expensive houses in which owners became overextended and could not meet mortgage payments.

Recall that one the interviewees from College Hill stated that three houses contiguous to his property were repossessed during the economic crisis. These were white homeowners in up-scale properties.

Tract 55: Madisonville experienced a major decline in homeownership (from 43.4 in 2000 to 34.3 percent in 2010 and a further decline to 24.0 percent by 2016). Both groups experienced a major decline in the percentage of population residing in owner-occupied housing (from 45.9 percent for African-

Americans and 37.4 percent for whites in 2000 to about 25.0 percent in 2016 for both groups. There were

790 owner-occupied units in Tract 55 in 2000, but only 492 by 2016. Given the increase in vacancy between 2000 and 2010/2016 (from 11.3 to 17.0 percent), it is likely that many of these dwellings went into foreclosure and remained vacant by 2016. Disaggregating these data by race reveals that the housing foreclosure crisis in Tract 55 was more detrimental to African-Americans than whites. African-American owner-occupied homes declined from 599 in 2000 to 418 by 2010 and then to 250 (a loss of 349 owner- occupied units) by 2016. This suggests that the African-American population in Tract 55 was not able to recover from the housing crisis, even by 2016. There was a decline in the total number of dwelling units occupied by African-American homeowners, from 1,305 to 983 (-322) between 2000 and 2016 suggesting a mass exodus of homeowners. White owner-occupied homes in Tract 55 decreased from 178 to 100 between 2000 and 2010 but recovered to 206 owner-occupied units by 2016. Thus, we can conclude that the decline in the overall percentage of white owner-occupied housing between 2010 and

2016 is likely a result of greater numbers of whites choosing to rent versus owning a home. Occupied dwelling units by whites increased from 476 to 830 between 2000 and 2016. The number of housing units

(not occupied housing units) in Tract 55 increased from 2,052 to 2,423 (20 percent) during this time and

260 many of these new units in the housing stock may be apartments which would have decreased the percentage of population in owner-occupied homes.

Table 9.4A. Percentage of Vacant Dwellings in Four Census Tracts, 2000-2016.

Percent Vacant, Percent Vacant, Percent Vacant, 2000 2010 2016 Tract 19 13.2% 28.8% 15.9% Walnut Hills Tract55 11.3% 17.0% 16.9% Madisonville

Tract 74 11.1% 10.1% 19.7% Northside Tract 84 10.5% 11.0% 13.2% College Hill Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Table 9.4B. Homeownership Rates for African-Americans and Whites in Four Census Tracts, 2000-2016.

Home-owner Home-owner Home-owner Home-owner Home-owner Home-owner African- African- African- White 2000 White 2010 White 2016 American American American 2000 2010 2016 Tract 19 13.8% 16.8% 9.7% 34.8% 50.% 44.8% Walnut Hills Tract 55 45.9% 35.2% 25.4% 37.4% 18.1% 24.8% Madisonville Tract 74 31.1% 29.7% 37.2% 50.6% 56.3% 43.8% Northside Tract 84 20.6% 37.2% 30.7% 50.3% 45.2% 55.6% College Hill Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

261

Table 9.4C. Number of Occupied Housing Units by Race.

Tract 19 Tract 55 Tract 74 Tract 84 Walnut Madison- College Hills ville Northside Hill

Total Units[1], 2000/Owner-occupied 943/222 1,821/790 837/366 898/336

Total Units, 2010/Owner-occupied 736/309 2,010/689 664/341 982/356

Total Units, 2016/Owner-occupied 859/261 1,979/492 837/375 940/386

Housing Units, AA HH, 2000/Owner-occupied 455/63 1,305/599 293/91 339/70

Housing Units, AA, HH 2010/Owner-occupied 327/55 1,188/418 222/66 379/141

Housing Units, AA, HH, 2016/Owner-occupied 320/31 983/250 156/58 534/164

White, HH, 2000/Owner-occupied 443/154 476/178 526/266 539/271

White, HH, 2010/Owner-occupied 390/196 551/100 465/262 440/199

White, HH, 2016/Owner-occupied 453/203 830/206 614/269 383/213

Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Same Residence as Previous Year: While not ideal, residence one year prior to the ACS is one way to track movements of population using census data (See Table 9.5). It must be acknowledged that those displaced from a tract (out- migrants) are not counted in these ACS data. However, these data capture recent in-migration to the tract.

It is well-documented that renters and those with lower-incomes are more mobile than their homeowning and higher-income counterparts (Coulton, 2012). African-Americans generally have higher mobility rates

(less percentage of population in same residence as the prior year) than their white counterparts as

African-Americans tend to have lower homeownership and lower-incomes than whites (Coulton, 2012).

Results from the four tracts from 2010 and 2016 show that these gentrifying tracts are not congruent with

262 national averages (See Table 9.5). African-Americans in the four tracts tended to have lower mobility

(higher percentage of population in the same house as the previous year) than their white counterparts.

This makes sense in that displacees are not counted in these statistics. While African-American displacement (as defined in this thesis) did not occur in Tracts 19 and 84 between 2010 and 2016, gentrification and rising costs of living in these tracts likely curtailed in-migration of lower-income

African-American population. Each of the tracts had percentage of whites in the same residence as the previous year that declined between 2010 and 2016, likely a result of recent in-migration as of 2016 from gentrification.

Note that in 2010 Northside had the lowest percentage of African-Americans in the same residence as the previous year (only 45%) of all the neighborhoods. This indicates one of two possibilities: 1) recent in-migration of African-Americans to Tract 74 or 2) movement of incumbent residents within the tract to other housing units. The decline in African-American population between

2000 and 2010 does not support hypothesis one. It is likely that many African-American renters moved within Tract 74 when money was scarce during the Great Recession. Recall that displacement of African-

Americans in the West End led to many of these individuals settling in Northside (Northside

Comprehensive Land Use Plan Update, 2014). It may also be that many African-Americans doubled-up on housing with friends and relatives during these difficult economic times. By 2016, 97.2 percent of

African-Americans in Tract 74 were in the same residence as in the previous year. Recall that by 2016 the

African-American population in Tract 74 had declined to only 20 percent of the tract’s population and it is likely that these individuals had taken up residence in subsidized dwellings or that they were in better economic shape post-recession.

263

Table 9.5. Percentage of Population that was in Same Residence as Previous Year for African-Americans and Whites, 2010 and 2016. African- African- White, 2010 White, 2016 American, 2010 American, 2016 Tract 19 74.7% 75.0% 73.4% 65.1% Walnut Hills Tract 55 83.4% 87.8% 72.5% 65.6% Madisonville Tract 74 45.3% 97.2% 81.2% 72.7% Northside Tract 84 79.9% 71.1% 94.1% 82.2% College Hill Source: American Factfinder. 2000 Census; 2006-2010 ACS; and 2012-2016 ACS.

Google Streetview: To determine the effect that gentrification has had on the built environment of the four neighborhoods, it is necessary to examine the housing stock, the commercial/industrial districts, the use of vacant areas, and the streetscape. Even though the business districts in three of the four tracts received major reinvestment dollars during the 2010 – 2016 time period, it is housing that was the most conspicuous feature in the built environment and thus housing receives the most emphasis in this discussion.

Housing: Most of the Google imagery for the four tracts focuses on housing and this is not surprising in that Americans are fond of the single-family home. The overall condition of housing in a tract is related to when housing was built. It is assumed that older housing will have more wear and tear than newer housing. This is a very general statement in that the durability of housing relates to types of building materials used in construction. For instance, vinyl is of less durability than brick or stone. As previously discussed in this chapter, reinvestment in housing stock occurs when the rent gap depreciates to a certain point where it becomes profitable to reinvest in these areas. This could result in demolition of dilapidated structures or renovation of structures in relatively sound condition. The percentage of new housing in a tract (post 2000) is a combination of many factors which include but are not limited to: 1) date of

264 settlement of tract; 2) materials used in construction of housing; 3) policies on removal/renovation of substandard housing stock; 4) spatial capacity to incorporate new housing; and 5) recent demand for housing stock in that tract.

Table 9.6 displays the percentage of housing units built during four time periods for each of the tracts. Note that Northside’s (Tract 74) housing stock had the greatest percentage (85.3%) of housing units built before 1939 while Walnut Hills (Tract 19) trailed with only 60.7 percent built prior to 193986.

Both communities experienced a housing boom during the late 1800s/early 1900s and much of their housing stock dates to this time (Cincinnati Plan, 1925). Conversely, Tracts 55 and 84 had only about 28 percent of their housing units built prior to 1939 and given that College Hill and Madisonville are located further from downtown Cincinnati, it is not surprising that their housing stock is not as old as that of

Walnut Hills or Northside. Between 1940 and 1969 only 11.3 percent of Tract 19 and 7.9 percent of Tract

74’s housing stock was built. This was the heyday of suburbanization and densely populated communities like Walnut Hills and Northside would have been less attractive to those seeking single-family housing.

In contrast, 37.2 percent of Tract 55’s and 25.0 percent of Tract 84’s housing stock was built during this time. Between 1970 and 2000, Tract 19 had 21.0 percent of its dwellings built. Remediation of blighted areas and the elimination of housing stock to accommodate Interstate 71 would have eliminated aging stock and provide an impetus for the building of new dwelling units87. Not surprisingly, Tract 74 (2.4%) and Tract 19 (7.2%) had a very small percentage of their housing stock built between 2000 and 2016.

These two communities have made concerted efforts to renovate existing dwelling units and had numerous buildings and whole historical districts that were able to draw upon historical preservation dollars (National Register Historic Districts). Madisonville also had historical preservation status for its two historical districts, but Madisonville’s revitalization focused on demolition as many of the buildings

86 The census aggregates any housing built prior to 1939 into one category. 87 In 1974, Cumminsville (Northside) was divided into two communities: South Cumminsville and Northside when I- 74 severed the neighborhood. Much of Northside’s infrastructure was left intact, but South Cumminsville lost some of its infrastructure, including housing.

265 that were in disrepair and thus, a much larger percent of its housing stock was built after 200088. Tract

84’s housing stock built between 2000 and 2016 was 22.4 percent (the highest of the four tracts).

Table 9.6. Percentage of Housing Stock Built at Different Times in Four Census Tracts.

Percent of Percent of Percent of Percent of Housing Built Housing Built Housing Built Housing Built Prior to 1939 1940-1969 1970-1999 2000-2016 Tract 19 60.4% 11.1% 21.0% 7.2% Walnut Hills Tract 55 28.8% 37.2% 20.0% 14.0% Madisonville Tract 74 85.3% 7.9% 4.4% 2.4% Northside Tract 84 27.3% 25.0% 25.2% 22.4% College Hill Source: American Factfinder. 2012-2016 ACS.

Stage of Gentrification: Google Streetview Visual Analysis (GSVA) revealed that the four tracts were at different stages of gentrification (See Table 9.7).

Tract 19: Tract 19 (Walnut Hills) scored the highest on Hwang’s (2015) gentrification index in 2007/09.

Even though Walnut Hills’ built environment was in decline, Tract 19 was better off than the tracts in the rest of Walnut Hills and had some of the most structurally sound buildings which were of architectural and historical interest (Walnut Hills Reinvestment Plan, 2016). Most of the surveyed blocks in Tract 19 were in good shape (block face average score exceeding 0.50) in 2007/09. Tract 19’s score on the gentrification index was 0.67 indicating that the tract was in the late stage of gentrification by 200789.

Tract 19 had a score of 0.74 on the gentrification index in 2014/16 and most of the increase in the gentrification index came about through improvements made in the business district. Vacant buildings

88 Interview with Bill Fischer, May 2019. 89 There is no baseline for the condition of the built environment using Google Streetview prior to 2007, so it is possible that Tract 19 had never degenerated to disinvested (gentrification index below 0.50).

266 along East McMillan were converted to luxury apartments which helped to increase the score as well.

Walnut Hills’ public space on East McMillan was reconfigured between 2007 and 2014/16 to provide a more inviting place for individuals to relax and socialize, again contributing to an increase in the gentrification index for this block.

Tract 84: Tract 84 (College Hill) presented the next highest score on the gentrification index in 2007/09

(0.64). Tract 84 is not part of the business district in College Hill and most of the Google Streetview analysis focused on changes in residences and the streetscape. Dwelling units in Tract 84 vary greatly in their size, condition, and value. Housing in the northern part of the tract is large, attractive, and situated on sizable lots. Conversely, small houses in disrepair were found in the southern part of the tract. Tract 84 only slightly improved its gentrification index score to 0.67 by 2014/16 and this was largely the result of the renovation of an apartment complex into up-scale apartments on Belmont Avenue. These apartments are the closest dwelling units to the revitalized business district, located about one- mile North on

Hamilton Avenue and may be the first sign of gentrification in the tract90.

Tract 55: Tract 55 (Madisonville) scored 0.48 in 2007/09 on the gentrification index, ranking it as a disinvested tract. Like Walnut Hill’s business district, Madisonville’s business district had been in decline for decades (Madisonville Business District Plan, 2002). However, by 2016, the business district in

Walnut Hills’ Tract 19 showed evidence of revitalization. In contrast, the built environment of the business district in Madisonville declined between 2007/09 and 2014/16. Block 4027 along Whetsel

Avenue revealed a single boarded-up shop in 2007; the entire row of shops was boarded-up by 2016.

However, this setback in revitalization along Whetsel Avenue91 was countered by improvements in the built environment by removing dilapidated housing stock and were evident in Google Streetview by

90 Recall that the gentrification index assesses the conditions of the built environment at a certain time. Hwang’s (2015) methodology does not allow a designation of gentrification in 2000 as the Google imagery was only available beginning in 2007. Tract 84’s median housing value in 2000 was $129,000 compared to $93,000 for the City of Cincinnati suggesting that the housing stock was in good condition. However, median housing value does not determine eligibility for gentrification according to Hammel and Wyly’s (1996) criteria. Since Tract 84 had a median household income below the City’s average in 2000, it was a candidate for gentrification. 91 By 2018, the boarded-up stores along Whetsel Avenue were renovated and were evidence of gentrification.

267

2014/16. Madisonville’s community gardens provided a meeting place for community members and a way to ensure that vacant parcels were well taken care of. By 2014/16, Madisonville scored 0.56 on the gentrification index indicating that it went from disinvested in 2009 to early gentrification in 2016.

Tract 74: Tract 74 (Northside) had the lowest score on the gentrification index in 2007/09 (0.41).

However, the gentrification process proceeded very rapidly in Northside. Much of Northside’s investment has come from the Gantry Development and the transformation of American Can Factory into mixed-used development by private investors. By 2016, Tract 74 had moved up to the early-stage of gentrification

(0.56). Like Madisonville and Walnut Hills, Northside embraced a mixed-use revitalization plan which focused on new construction as well as revitalization of older buildings (Northside Comprehensive Land

Use Plan Update, 2014). Northside deviates from the other three communities in that the revitalization of single-family housing stock has been prolific. In some instances, entire block faces experienced a metamorphosis in less than a decade.

Table 9.7. Scores on the Gentrification Index for the Four Census Tracts.

Gentrification Gentrification Index 2007/09 Index 2014/16 Tract 19 0.67 Late-stage 0.74 Late-stage Walnut Hills gentrification or gentrification non-disinvested Tract 55 0.48 Disinvested 0.56 Early middle- Madisonville stage gentrification Tract 74 0.41 Disinvested 0.56 Early middle- Northside stage gentrification Tract 84 0.64 Middle-stage 0.67 Early late-stage College Hill gentrification or gentrification non-disinvested Source: Google Streetview, 2007/09 and 2014/16 and Hwang (2015).

268

We could also use median house values as a proxy for level of disinvestment in the built environment for 2000. Tract 8492 had median housing values that were 1.39 times that of the City’s while

Tract 19 had median housing prices 1.16 higher than the City. Thus, it was not surprising that most of the housing stock in Tracts 19 and 84 that was viewed through Google Streetview in 2007 was in good condition. Tracts 74 and 55 had median housing values that were only 60.2 and 79.2 percent, respectively, of that of the City in 2000. In Smith’s (1979) language, the rent gap in Tracts 55 and 74 made investments by individuals or corporations more profitable in these two tracts. These investments were undoubtedly aided by the Go Cincinnati Plan (2009) which funneled millions of dollars into the revitalization of three areas in Cincinnati (the Madisonville Corridor, the Southern Mill Creek Area which includes Northside, and Seymour/Reading Corridor) to provide a catalyst for revitalization. This process proceeded more rapidly in Northside than in Madisonville and largely explains why Northside’s built environment experienced the greatest increase in the gentrification index between 2007 and 2016.

Summary of Interviews: The purpose of the interviews with community members was part of a technique referred to as triangulation in which three independent sources of data are used to determine if each supports the conclusions gained from the other two sources. The interviews provide qualitative evidence that when used in conjunction with the visual evidence of the built environment from Google Streetview and the quantitative data from the census are able to document the link between gentrification and change in racial composition in the four tracts. I was also interested in whether the interviewees felt that the policies and practices enacted by public agencies and private investors were instrumental in combating the displacement of the African-American population.

Triage: While this study focuses on four census tracts, it was impossible to view the gentrification/displacement process solely at the tract level as the process was initiated at the

92 There is no evidence to suggest that Tract 84 would have had a score on the gentrification index that was below 0.50 in 2000.

269 neighborhood level93. One common thread that was uncovered in the interviews was that economic revitalization needed to focus on targeted areas. Galster et al. (2006) in their study of revitalization in

Richmond, Virginia found that targeted investment in a few select neighborhoods that showed the most promise for revitalization (what they called ‘triage’) would be a catalyst to further development in the city. The redevelopment of downtown Cincinnati in the early 2000s followed this model and focused on revitalization in Over-the-Rhine (Brenneman, 2018). This targeted investment concentrated scarce resources into a highly visible areas that would provide a catalyst for further revitalization. College Hill

Community Urban Redevelopment Corporation CHCURC), Madisonville Community Urban

Redevelopment Corporation (MCURC), and the WHRF all followed this model of development and have revitalized their business districts.

Ownership: Before revitalization efforts could begin in the neighborhoods, my interviewees needed to constantly complain to the authorities to redress problems such as loitering, crime, and vacant buildings.

By applying this pressure, order was restored to their communities. However, it was not enough for CDCs to be social or political activists for the neighborhood; they needed to be able to guide development in a way that they viewed as beneficial to the community. Thus, this entailed acquiring an economic stake in the community by becoming property owners (buildings, vacant lots, etc.). For instance, CHURC purchased dilapidated buildings that were a source of crime and public nuisance and either redirected to them the type of business that would be more suitable to the neighborhood or demolished the building if it was beyond repair. This policy was also followed in Northside as Stefanie Sunderland noted that NEST purchased businesses that were a public nuisance or a public safety issue, such as rental properties that were not up to code.

93 Originally, I intended to focus only on Tracts 19, 55, 74, and 84, but I relaxed that requirement and focused at the neighborhood scale which was much easier for the interviewees to speak to. The advantage of this approach is to gain an overall understanding of the changes taking place at the neighborhood scale that would have effect on specific tracts. A disadvantage of this approach is that the respondent may focus his/her attention on revitalization occurring in other tracts within the neighborhood and not specifically speak to changes occurring in the gentrifying tract. This was only an issue in College Hill where Tract 84 was to contiguous, not within, the College Hill Business District (Census Tract 82.02) which has received the most revitalization.

270

Displacement: In terms of the displacement of African-American population due to the revitalization process, several of the interviewees appeared hesitant about answering this question94. A few of the interviewees commented about the prominent role that Madisonville, Walnut Hills, and College Hill played in the original settlement of the African-American population. These were neighborhoods that were friendly to African-Americans dating to before the Civil War and each had established middle-class,

African-American communities. All interviewees stressed the importance of their communities as being places for a diversity of persons. However, the physical presence of a group in a community is not enough to ensure that the group’s social, political, and economic concerns are met. I did not get the sense that the

African-American community in Madisonville, College Hill or Northside had as active a role in community activism as their counterparts in Walnut Hills95. Until the 1960s, the Walnut Hills business district was dominated by African-American businesses96. Thus, it is not a surprise that African-American

(as well as white) members of the WHRF are involved in empowering97 African-Americans in their community. This may partly explain why displacement has not occurred recently (2010-16) in Tract 19.98

Those middle-class African-Americans who have stayed in Tract 19 have been very vocal in community politics99. Activism on the behalf of the African-American population is less noticeable in College Hill, likely because of the high percentage of subsidized housing units in Tract 84. This top-down approach to providing housing for low-income populations (of any race) is likely detrimental to community building

94 Most of my interviewees were white and I may have unintentionally put them on the defensive. 95 Recall that the interviewees were volunteers and were not a representative sample of the average community member of their neighborhoods. 96 Interview with Kathryne Gardette, May 2019. 97 For example, Kathryne Gardette from the WHRF, provides one-on-one sessions with women in Walnut Hills who want to improve their education/job opportunities in the community. 98 Only two of my interviewees were African-American. Both were from Walnut Hills and both expressed that empowerment was the key to resisting displacement. 99 Two processes have led to the demise of central city African-American neighborhoods in U.S. Cities. According to Wilson (1987), desegregation and suburbanization was the first process. It allowed middle-class African-Americans to move to the suburbs, concentrating lower-income African-Americans in inner city neighborhoods. Secondly, the deindustrialization process destroyed the major form of employment for middle-class African-Americans without advanced education. The decline of manufacturing jobs did not involve a dispersion of middle-class African- Americans to other neighborhoods in metropolitan areas; instead, it led to the elimination of a large part of the African-American middle-class.

271

(HUD, 1996). It relieves the recipient of the responsibility for providing for himself/herself and often causes resentment from the non-poor segments of society (Tighe, 2010)100. Several interviewees from

College Hill remarked that transient (low-income renters of any race) were an issue, mostly seen as a source of nuisances, and that they did not take care of their rental properties.

Affordable Housing: Probably the most prominent theme to emerge from the interviews was the need for affordable housing. This question is not race specific but applies to all low/middle-income individuals. In the words of Mary-Cabrini: ‘affordable for whom’? Recall that Mary-Cabrini has been displaced several times since 2000 from her apartment in Walnut Hills. Kathryne Gardette spoke about Habitat for

Humanity and the good it has done in Walnut Hills. However, several homeowners have subsequently lost their homes because once the original tax abatements are gone they can’t pay the taxes on these properties. Bill Fischer (Madisonville) has also noted that many who buy into affordable housing in

Madisonville can’t provide the upkeep needed to ensure that these properties don’t end up in disrepair. An

EPA (2015) report on affordable housing in Northside noted that many homeowners can’t afford the upkeep on their properties and end up with numerous code violations and eventually sell to real estate investors who renovate the houses and sell at higher prices, furthering contributing to the problem of affordable housing. To deal with the decrease in affordable housing in Northside, Carl Sterner replied that

NEST has recently invested in the concept of ‘tiny’ houses which would be more affordable for low- income residents.

Form Based Code: Lastly, form-based code was mentioned by several of the interviewees as having a positive impact on economic revitalization and being a way to incorporate affordable housing into the revitalization plans of these communities. However, form-based code was only adopted by the first neighborhoods in Cincinnati in 2013101 and thus it is not expected that this has had much effect on the displacement of low-income populations in the tracts between 2000 and 2016.

100 Most of this evidence is inferred from Tighe’s review of the literature on public opinion and affordable housing. 101 Form-Based Code in Cincinnati was first initiated in College Hill, Madisonville, Walnut Hills, and Westwood.

272

Summary of Major Differences in Tracts Undergoing Displacement versus Those with No Displacement: One of the most important lessons from this comparison of displacement of the African-American population in these four tracts is that the displacement process can be staunched or even reversed if an inventory of vacant buildings is undertaken to discern if these buildings could provide dwellings for all socio-economic levels. This is the process that seems to have occurred in Walnut Hills between 2010 and

2016. However, it is displacement102 that is driving vacancy rates. Tract 19 lost one-third of its African-

American population between 2000 and 2010, and not surprisingly, the vacancy rate increased from 13.2 percent in 2000 to 28.8 percent by 2010. Walnut Hills purposively adopted policies for converting vacant buildings into apartments or owner-occupied housing that also provide affordable housing (Transforming

Vacancy in Walnut Hills, 2015). The vacancy rate was then reduced to 15.9 percent by 2016 and no displacement of African-American population occurred in Tract 19 between 2010 and 2016.

Tract 74 may be experiencing the displacement-vacancy process experienced by Walnut Hills

(between 2000-10) between 2010 and 2016. Tract 74 lost 45.0 percent of its African-American population between 2010 and 2016 and not surprisingly, its vacancy rate increased from 10.1 to 19.7 percent during this time. Whether Tract 74 continues to lose African-Americans will likely depend on its housing policies. If absent-tee landlords continue to raise rents on their buildings or renovate them for purchase by the middle-class, then those with lower incomes (African-American and others) will likely be displaced.

If Northside should adopt similar policies to those of Walnut Hills which require some provision for lower-income residents then displacement may be avoided.

Another major factor that seems to be driving gentrification and displacement103 is the investment in neighborhoods that had the greatest potential for profit; those that were the most disinvested in 2000.

102 This is true for both African-Americans and whites. The issue with these data is that vacancy cannot be disaggregated by race. One must infer these processes by examining changes in total households in a tract by race. 103 The research methodology for this thesis measured displacement by looking at the differences in numbers and percentages of African-American population between three timestamps. The assumption was that low-income African-Americans were pushed out of their residences due to rising costs of living. Marcuse (1985) advanced his

273

As stated previously, Tracts 55 and 74 had the lowest median housing values in 2000 of the four tracts and investment in those tracts would yield higher rates of return. This was the focus of Go Cincinnati Plan

(2009) which used these neighborhoods as catalysts for revitalization. While MCURC and NEST have attempted to provide affordable housing for those under threat of displacement, the rapid revitalization of these tracts in a relatively short period of time has not allowed these agencies to keep pace with the need for affordable housing.

Recall that Northside and Madisonville experienced displacement of African-American population between 2010 and 2016. During this time their built environments transitioned from disinvested (below 0.50) to early/middle-stage gentrification (0.56) suggesting that the two processes of displacement and revitalization of the built environment were related. The inflow of investment likely caused an increase in median rent in Tract 55 and Tract 74 and caused displacement of low-income populations. While Over-the-Rhine was not a focus of the Go Cincinnati Plan, earlier tax breaks given to corporations to revitalize Over-the-Rhine likely led to widespread dislocation of low-income population as the affordable housing stock in OTR declined by 73 percent between 2002 and 2015 (Brenneman,

2018). Given the substantial increases in median rent in Tracts 55 and 74 between 2000 and 2016, it is likely that affordable housing has significantly declined. This was echoed in the responses given from interviewees from the Madisonville and Northside CDCs.

In contrast to targeted reinvestment in Madisonville and Northside imposed on these communities from outside, Walnut Hill’s revitalization of the business district was imposed from within the community

(with city and private developments). Kathyne Gardette stated that Walnut Hills had to pressure the City into conducting a study on the feasibility of converting East McMillan back to a two-way street in 2012.

Each of my interviewees from Walnut Hills noted that this was crucial to the revitalization efforts in

Walnut Hills and paved the way for public and private investment. Much of this development focused on

thesis of ‘exclusionary displacement’ which prohibits low-income individuals from entering a neighborhood because cost of living is too high.

274 utilization of vacant buildings (which does not displace residents) or provision of new affordable housing such as the Scholar House. Between 2000 and 2010, Tract 19 was at the gentrification stage that Tracts 55 and 74 were in 2010-16 and suggests that middle104 stages of gentrification correspond with displacement

(Bates, 2013), further suggesting that MCURC’s and NEST’s focus on providing affordable housing in the coming years will staunch the flow of displacees. Or, it may be that the most vulnerable have left the neighborhood and physical displacement of low-income populations will be less noticeable. However, the social, political, and economic marginalization of low-income population could result in isolation of the low-income population within their communities and lead to contention between gentrifiers and the incumbent population. Recall that homeowners in Tract 84 had issues with the ‘transient’ (mostly

African-American) renter population while incumbent residents of Walnut Hills were often at odds with their new neighbors, mostly over housing and yard maintenance.

Each of the CDCs in the four communities are attempting to be equitable in the development process. However, housing affordability is a national problem and will likely require federal, state, and local resources to solve the problem. Hamilton County was at a deficit of 40,000 affordable housing units in 2017 (Housing Affordability in Hamilton County, 2017). A recent report has noted that 61 percent of

African-American renter households and 39 percent of African-American homeowners in Hamilton

County are cost-burdened105 (Crowder, 2018). CDCs in neighborhoods undergoing revitalization likely need to seek more public/private partnerships to finance affordable housing. Madisonville, College Hill, and Walnut Hills have adopted (since 2013) form-based code106 and this is one method of ensuring the availability of affordable housing (City of Cincinnati, 2013). Unfortunately, the timeframe (2000-2016) for this specific study does not allow us to determine the effect that form-based code has had on racial

104 Early stage gentrification is less likely to displace because revitalization is incremental and not attracted many newcomers (residents or businesses). Late stage gentrification is less likely to displace as those with the least financial resources have already been displaced. 105 Cost-burdened households are ones that spend more than 30 percent of their income on rent/mortgage and utilities. 106 Northside went through the process of determining the feasibility of adopting form-based code, but as of 2019, has not adopted the practice.

275 change in the four neighborhoods. The 2020 decennial census should be able to capture demographic, economic, and housing changes since the implementation of form-based code policies.

276

Chapter 10 CONCLUSION:

This thesis examined the gentrification process in the City of Cincinnati between 2000 and 2016.

Gentrification is a process that has unfolded over the past half century in not only the largest cities in the

U.S., but medium and smaller cities (Richardson et Al., 2019). Like suburbanization, gentrification has led to a massive change in the built environment as well as a change in the economic and demographic composition of population residing in the cities of the U.S. Suburbanization was (and often still is) an exclusionary process as it did not allow those with less economic resources (often minorities) into new suburban neighborhoods. The urban sprawl created by the desire of middle-class Americans to pursue the

American Dream of owning a single-family dwelling and automobile in suburbia gave rise to massive destruction of greenspace, introduced tremendous quantities of air pollution from automobiles into our atmosphere and resulted in the inefficient use of public services by deconcentrating the population. Much of the suburbanization process was based on consumer preference for low density living aided by the capital motive: that is cheap land on the outskirts of the city. However, as early as the 1960s, some

Americans were disenfranchised with the sterile environment of suburbia. Most planners are familiar with

Malvina Reynolds’ (1962) tune ‘Little Boxes’ which rallies against this sterile environment of suburbia with lyrics such as ‘they’re all made out of ticky tacky and they all look just the same’. This rejection of suburbia may have given rise to the first ‘gentrifiers’ in American cities. As we move toward the third decade of the twenty-first century it appears that this process of gentrification will continue to occur in cities in the U.S. The reutilization of urban space makes economic and environmental sense. The major issue with gentrification is that of displacement of the low-income population. Planners and community development corporations must seek to rectify the displacement problem.

Gentrification theories revolved around two major debates. Ley (1986) argued that cultural preferences attracted certain individuals to urban neighborhoods with amenities such as up-scale restaurants, cultural venues, greenspace, and access to downtown employment. Smith (1979) argued that the rent gap was responsible for attracting investment in disinvested neighborhoods. For Smith, most U.S.

277 cities had building stock and infrastructure in relatively good condition that could be further capitalized upon. Even if the building stock was dilapidated, the urban space in many of these disinvested neighborhoods was at a premium as it was close to the downtown. Dilapidated buildings could be demolished and new structures built. Hamnett (1991) argued that both the above conditions were necessary for gentrification to occur. Capitalists (represented by Smith) supplied a product (renovated urban space) and gentrifiers consumed this product (Ley). The four case studies for Cincinnati support

Hamnett’s argument that both cultural preferences and financial motives were important in economic revitalization/gentrification in College Hill, Madisonville, Northside and Walnut Hills. Interviews from members of CDCs in the four neighborhoods revealed that accessibility to transportation (highways/bus routes), downtown amenities/employment attracted newcomers to their neighborhoods. Others stressed walkability of the neighborhoods and the historical architecture of the neighborhood as attractive to residents. Greenspace, artist workshops, up-scale restaurants were also attractive to newcomers as well as financially secure long-term residents. Support for Smith’s financial motive came from interviews with

CDC members as well as neighborhood master plans. Tracts 74 (Northside) and 55 (Madisonville) were in deplorable condition in 2000 and it was through concerted efforts of the City of Cincinnati that investments were made to economically revitalize these areas. This public investment then paved the way for private investment. While Tract 19 (Walnut Hills) was in much better condition than the other three census tracts in Walnut Hills, the former business district on East McMillan (Peebles Corner) was thoroughly disinvested in the early 2000s (Walnut Hills Reinvestment Plan, 2016). The amount of profit to be gained from investment in this corridor was enormous and attracted public and private investment.

Similar projects were implemented in the College Hill business district and revitalized that area (CHURC,

2014). Each of these financial investments would have resulted in wasted resources, if a percentage of

Cincinnati’s population did not find living, working, and recreating in these neighborhoods attractive.

278

Displacement: One of the most challenging aspects of measuring displacement caused by gentrification is that there are few ways for tracking displaced individuals through census data. Recall that Atkinson (2000) referred to this as ‘counting the invisible’. However, it is possible to track change at the aggregate level: that is changes in racial composition or socio-economic status at the census tract level. Change in racial composition was the approach taken in this thesis. At the tract level for the City of Cincinnati it was demonstrated using statistical analysis that there was not a significant relationship between gentrification and racial displacement during the 2000-2010 or the 2010-2016 timeframe. However, when examining the displacement process over the sixteen- year timeframe, it was noted that displacement was 4.6 times more likely to occur in a gentrified than in a non-gentrified tract and this was significant at the 0.05 level.

This means that several tracts not undergoing gentrification experienced displacement of the African-

American population. There are too many outside factors that need to go into this analysis to draw any firm conclusions from this. For example, non-gentrifying tracts may be losing African-Americans to tracts in the suburbs. We saw that this occurred as the African-American population in Cincinnati went from a concentration in the tracts surrounding downtown in 2000 to a dispersed distribution in the inner- ring suburbs by 2016. Changes in low-income housing provision likely moved low-income African-

Americans to the inner-ring suburbs and transferred poverty to these neighborhoods (Housing

Affordability in Hamilton County, 2017). Another question I cannot answer is where the displaced

African-Americans from gentrifying tracts have gone. It is likely that middle-income African-American displacees have selected different neighborhoods than their low-income African-American counterparts.

The next step was to examine four neighborhoods in Cincinnati that were identified as undergoing gentrification between 2000 and 2016. Tract 74 (Northside) experienced gentrification between 2000-10 and 2010-16, but only displacement between 2010-16. Tract 55 (Madisonville) experienced gentrification and displacement only between 2010-16. Tract 19 (Walnut Hills) experienced gentrification between 2000-10 and 2010-16, but only displacement between 2000-10. Tract 84 (College

Hill) experienced gentrification between 2010-16 but did not experience displacement during this

279 timeframe. For Tract 19, we saw that community engagement through the Walnut Hills Redevelopment

Foundation (WHRF) and a concerted effort to renovate vacant buildings staunched the exodus of African-

Americans between 2010 and 2016. In Tract 84 (College Hill) the availability of subsidized housing seemed to prevent displacement. Conversely, speculation by absent-tee landlords and investment by private developers caused sky-rocketing housing prices in Tract 74 and likely led to displacement of the

African-American population between 2010 and 2016. Likewise, Tract 55’s private investment has caused affordability issues for those with limited incomes. The added burden of the housing crisis eliminated a middle-class segment of African-American homeowners from Tracts 55 and 74.

Did displacement follow the gentrification stage models that were introduced in the literature review? Clay (1979) found that in Stage 1 gentrification that displacement did not occur. Stage one gentrification occurs as individuals move into a neighborhood and use ‘sweat equity’ to increase property value. This seems to have occurred in 2000-10 for Northside. Interviewees from NEST revealed that crime was rampant in Northside in the early 2000s, but Citizens on Patrol and the newly formed NEST made it at least semi-attractive for some risk-taking ‘early gentrifiers’ to be attracted to the neighborhood.

However, if this study had examined the displacement of the low-income population (regardless of race) a different conclusion may have been reached. The loss of whites and African-Americans of lower-income from Tract 74 suggests that displacement did occur with the gentrification process between 2000 and

2010. This is not a surprising finding as a high percentage of white Northsiders were Appalachian in the early 2000s and these individuals tended to have higher poverty rates than their non-Appalachian counterparts (Maloney and Auffrey, 2013). In College Hill (up the road from Northside) we saw that there was an influx of African-Americans between 2000 and 2010 suggesting that African-American displacees from Northside moved to Tract 84. My College Hill respondents informed me that they knew several

Northsiders who moved to College Hill in the past few years because of housing affordability issues in

Northside. I do not know if these ‘displacees’ were African-American or white, but this suggests that as neighborhoods proceed further in the gentrification process that they move into active displacement.

280

These are Clay’s (1979) second and third stages of gentrification and Lees, Slater, and Wyly’s (2008) fourth stage of gentrification in which substantial economic investment in the neighborhood causes rising housing costs and displacement.

Walnut Hills provided an interesting contrast as it experienced African-American displacement and gentrification between 2000 and 2010 but did not experience displacement between 2010 and 2016.

Again, it may be that most of the low-income African-American population had been removed from Tract

19 by 2010. This is not hard to imagine. Walnut Hills consists of four census tracts and Tract 19 had substantially higher costs of living than the other three tracts (U.S. Census Bureau). Low-income African-

Americans (or whites) could move a short distance (to another tract) and still be connected into Walnut

Hills political, social, and economic system. While this is true, moving to a new apartment or house is an expensive, time-consuming, and stress provoking event that usually does not occur without provocation.

While median housing prices continued escalating in Tract 19 between 2000 and 2016, median rents had stabilized suggesting that the rental affordable housing issue was being addressed.

Tract 55 (Madisonville) only began gentrification between 2010 and 2016 and experienced displacement of the African-American population. However, it is impossible to disentangle the effect that the Great Recession had on African-American homeowners. It was noted that Madisonville was an epicenter of jobs in the eastern part of Cincinnati and a massive amount of investment went into the

Madisonville Corridor after 2012 (City of Cincinnati, 2009). This investment in Madisonville runs counter to Clay’s (1979) stages of gentrification where in stage 1 individual ‘risk takers’ enter a disinvested tract and gradually improve the built environment. Madisonville’s gentrification appears to be driven by major corporations (with the backing of the City of Cincinnati) and skipped the first 1-2 stages in the gentrification stage model. Stages 3 and 4 stress the investment of banks and corporations as being prevalent in the gentrification process. Thus, it is no surprise that massive displacement of the African-

American population in Tract 55 occurred between 2010 and 2016 as these investments poured into Tract

55.

281

The question becomes, how can gentrification occur without long-term residents of these neighborhoods being displaced? How do those with limited political, social, and economic clout protect themselves from those with more resources from claiming this ‘new’ urban space? It should be obvious that planners don’t want to reverse fifty years of desegregation for cities to become segregated once more with the gentrification process.

Possible Solutions to Combat Displacement: My interviews with community development corporation members cited affordable housing as the number one concern in their neighborhoods. Thus, this discussion will focus on policies that may be utilized to combat displacement. Stopping gentrification, or revitalization, is not the solution to this problem. Equitable development should be pursued.

Before exploring potential solutions, I’d like to examine a poorly thought out solution to the gentrification and displacement problem. New York City’s Mayor Bloomsburg activated an Inclusionary

Zoning policy in Greenpoint-Williamsport, Brooklyn, a newly gentrifying neighborhood, in 2005

(Stabrowski, 2015). The purpose of the project was to convert industrial space to affordable housing to accommodate low-income population who could not afford to remain or to settle in Greenpoint-

Williamsport because of increased housing costs caused by gentrification. However, it soon became clear that more market-rate housing was being built than affordable housing and that the displacement of low- income population had not ceased. Instead of CDCs organizing resistance to displacement, they turned their efforts to securing apartments in other areas of the city for the displaced. From a political efficacy perspective, this was a colossal failure. Long-term residents were not interested in services that facilitated their removal from the neighborhood. Each of the documents that I have reviewed (HUD, 2018; NYU

Furman Center, 2017; Ellen, 2012) stressed that CDCs must be advocates of the low-income population and ensure that they are not displaced from the neighborhood.

282

Another program of limited use appeared to that of providing housing vouchers to the low- income population to prevent displacement in gentrifying neighborhoods. Dyson and Varady (2018) found in their study of the gentrifying neighborhood of Over-the-Rhine in Cincinnati that low-income residents relocated to adjacent neighborhoods to Over-the-Rhine during the early 2000s. While the authors did not address for the reasons for this out-migration of low-income population, it was likely that the housing vouchers were insufficient to cover the cost of housing in Over-the-Rhine.

Housing and Urban Development (HUD 2018) focused on three policies to preserve affordable housing in cities in the U.S. that are applicable to this study. It is important to note that HUD was not interested in developing a housing strategy before the 1990s because the agency thought of disinvestment as being the major factor driving displacement of low-income individuals, not gentrification. The first strategy was to preserve already existing affordable housing stock. A new program begun in 2013 was the

Rental Assistance Demonstration that moved public housing to what HUD referred to as a more stable funding program by allowing public housing agencies to leverage funding from external resources. Going forward, this may be a policy that Cincinnati could enact to counter displacement of the low-income population. Secondly, HUD encouraged greater development which would include all levels of housing and expand consumer choice in the housing market. These include federally based policies that support transit-oriented development, reduction in minimum lot size, mixed-use development and inclusionary zoning policies. Some of these policies would not be applicable to Cincinnati’s situation. For instance, transit-oriented- development which includes rapid movement by rail would be difficult to enact in

Cincinnati without substantial economic investment. This does not mean that Cincinnati should ignore its major streets that are connected through the bus routes and could further concentrate on mixed-use development. Thirdly, HUD encourages engagement of community residents. The entire community needs to support the idea of affordable housing. This community engagement needs to be cross-race and cross-class for full effect.

283

The community development corporations in the four case study neighborhoods have been instrumental in guiding economic revitalization and promoting equitable development in their communities. CHCURC, NEST, and MCURC were all founded by concerned members of their respective communities in the early 2000s while WHRF was formed in the 1970s but restructured its focus in the 2000s to address equitable development. While the original concerns of these organizations may have been related to blight and crime, the CDCs expanded their roles to address the provision of housing for low-income individuals. One of the critical lessons learned is that CDCs can partner with for- profit corporations and are in the best position to have these corporations respond to their demands for creating affordable housing. In fact, I was reminded many times during my interviews with CDC members that any private development company using public funds for housing development is required to provide twenty percent of the units as affordable.

One of the leaders in gentrification and displacement research involves a collaboration between

UC Berkeley and the Federal Reserve Bank of San Francisco. They are interested in predicting which neighborhoods will gentrify and the application of strategies to ensure that the displacement of the low- income population does not occur. They have examined several strategies to allow public, private, and non-profit organizations to collaborate to create growth that is inclusive and equitable for all socio- economic groups. This database has been expanded to other cities (unfortunately it has not made it to

Cincinnati). For example, in San Francisco’s Chinatown, Inclusionary Zoning was enacted to preserve affordable housing. For Chinatown, lower height limits on buildings dissuaded developers from this area.

Inclusionary Zoning would not work in Cincinnati as it is applicable to cities with rapidly growing populations and escalating land values. However, Walnut Hills has used Form Based Code to ensure that new buildings adhere to height restrictions and reflect the existing architectural style. Most importantly

Form Based Code has allowed mixed-use development and the Walnut Hills Redevelopment Foundation has insisted that a percentage of the housing units in the new development be devoted to affordable housing (Walnut Hills Reinvestment Plan, 2016). Currently, only four Cincinnati neighborhoods have

284

FBC (City of Cincinnati, 2013) and thus mixed-use development may not be possible in some of the other neighborhoods in Cincinnati undergoing gentrification and displacement.

Land Banks: The use of city-owned land to combat displacement in gentrifying neighborhoods was a strategy advanced by the NYU Furman Center (2016) and Gentrification and The National Low-Income Housing

Commission (2019). One of these policies ensures that parcels of land devoted to low-income homeowners remain within the city trust and cannot be purchased by land speculators. Cincinnati had

26,000 units of vacant housing as of 2016 (U.S. Census Bureau) which accounted for 16.5 percent of its housing stock. These units could be placed in a land trust which would ensure that property did not go to the private sector. Indeed, Hamilton County appears to be working along these lines. The Hamilton

County Land Revitalization Corporation was established in 2011 and is now managed by the Port

(Hamilton County Land Bank, 2019). One program enacted was REACH which focused on rehabilitation and the construction of new buildings. Walnut Hills was a beneficiary of this program as of 2016.

Building on Transport Systems: Transit-oriented-development (TOD) could be a development tool in preventing displacement of low-income population if done properly. UC Berkeley and the San Francisco Federal Reserve Bank

(2015) found in their study of Diridon Station in San Jose that combining TOD with impact fees, public- private partnerships, and tax increment financing attracted investment and did not displace the low- income population. Another case study from the UC Berkeley/San Francisco Federal Reserve Bank

(2015) collaboration focused on The Monument Corridor in Concord, California. In this case, an overlay was used to increase density along the transit corridor (TOD) and to provide an incentive for the building of more affordable housing and preventing displacement of the low-income population. While a fixed rail strategy may not be appropriate for Cincinnati at this time, it is possible to think ‘outside the box’ and

285 promote economic revitalization along Cincinnati’s major corridors into downtown by enhancing public transportation. Columbus, Ohio’s public transportation system only consists of buses, but this has not discouraged the city from pursuing affordable housing strategies that take advantage of the public transportation system to ensure more equitable development in the city. Zoning changes are underway to allow centrally located affordable housing that will move low-income individuals to employment opportunities (Cohen, 2019).

Fruitland, Oakland had success with using transit-oriented-development to prevent displacement of the low-income Latino population (Schneider, 2018). The station opened in 2003 and 25 percent of the housing in the development complex was affordable. Moreover, there was also an effort to focus on services that were needed for the resident Latino population instead of solely focusing on up-scale businesses. Saldana and Wykowski (2012) call for a proactive approach to racial equity in TOD planning and claim that it is not enough to provide affordable housing along transit stations, although this is a good start. The low-income population must have access to employment opportunities that pay livable wages.

The problem with current TOD is that both low-wage service and high-wage professional industries concentrate along the rail line and there is a dearth of well-paying employment opportunities for individuals with lower educational or skill levels.

East McMillan and Hamilton Avenues are two of the most used bus routes in the City (Cincinnati

Bus Coalition, 2019). Madison Road could easily join that designation as it is one of the busiest employment enclaves outside of downtown Cincinnati (City of Cincinnati, 2009). Many of my interviewees in Walnut Hills, Northside, and College Hill remarked about accessibility to employment, entertainment and other amenities as being a major benefit to residing in their respective neighborhoods.

This accessibility included being located on the bus routes as well as being in proximity to the highways.

Mixed-use development in these areas would allow individuals of all socio-economic levels more options for employment, shopping, recreation and housing. Furthermore, each of the neighborhoods have welcomed private development into their communities and mixed-use development at higher densities

286 could preserve affordable housing. By further enhancing transport networks all socio-economic levels would benefit and the possible densification of housing could be undertaken. It is critical for the City of

Cincinnati to ensure that private developers bear a proportion of the cost for affordable housing.

Each of the neighborhoods in the Cincinnati study has the potential to attract dense development along the transit corridor and to provide affordable housing. Walnut Hills has done this with Scholar

House and the Paramount Building, once renovated, will also have affordable housing. The Episcopalians have done this with Madison Villa, a renovated affordable senior housing complex in proximity to the new Ackerman Group project at the intersection of Madison Road and Whetsel Avenue. Northside has recently (2018) constructed Knowlton Place for seniors. Furthermore, in April 2019, zoning codes were changed in Northside to allow the Northside Transit Center to begin construction (Swartsell, 2019) which could lead to the construction of more affordable housing near this critical transportation node. College

Hill not only has Twin Towers but also has Marlowe Court for seniors. These increases in affordable housing opportunities for seniors is not surprising given the aging of the American population and the political clout of baby-boomers, the youngest who reached age 55 as of 2019. The emphasis on funding senior housing probably reflects the perceptions of the public that these individuals deserve to be ensured affordable housing as many are now on fixed incomes and should not be displaced from their long-term residences in their communities. Those of working age (of any racial category) are often demonized for not being responsible for ensuring their livelihoods and there may be less sympathy for providing housing options for these individuals107. This should not deter the City of Cincinnati from offering tax deductions to private investors in exchange for the provision of a certain amount of affordable housing.

107 This was my perception after talking to several of the members from the CDCS in the four neighborhoods. There seemed to be a ‘worthy’ low-income and a not ‘worthy’ low-income. The elderly were categorized as ‘worthy’. To avoid discomfort in the interview, I did not ask for much elaboration on these direct or indirect comments as this was not the purpose of the interview.

287

Vacancies and Affordable Housing: Cincinnati is a ‘shrinking city’ which is defined as a city that has lost much of its population since

1950 (Mallach, 2011). Most of these cities are associated with the manufacturing belt. However, while population has declined in the City of Cincinnati since 1950, much of the building stock has remained in good condition (City of Cincinnati, 2009). Each of the four neighborhoods in this study had at least 10.0 percent vacancy rates in dwellings in 2000 (U.S. Census Bureau). While a percentage of this housing stock is likely dilapidated, renovation of vacant buildings and in-fill are two excellent strategies for making use of vacant land and vacant buildings. Walnut Hills had the best policy for dealing with vacant buildings. While the Hauck Building, Trevarren Flats, and Windsor Flats cater to mostly high-income renters, this renovation of vacant buildings did not displace anyone. The American Can Factory in Tract

74 (converted industrial plant) is also an example of utilizing building stock in good condition for housing. While this is a market-rate apartment complex, it did not displace any residents from Northside.

Members of CHCURC stated that the Dixon Building (vacant on upper floors) located at Hamilton

Avenue and Northbend Road was in the process, as of 2019, of being renovated to accommodate 17 affordable apartments. Madisonville has done an excellent job of in-fill housing on properties where dilapidated housing previously stood. However, it is uncertain if these demolitions displaced individuals or if they were vacant buildings. It is also likely that many low-income individuals still cannot afford to purchase these new in-fill houses even at reduced prices. Bill Fischer relayed to me that MCURC might change its affordable housing strategy and owner-occupied housing may consist of smaller units that are part of a four-plex. Doug Trimmel of College Hill also noted that one way to address affordable housing issues and turnover of population in College Hill was to renovate buildings with four apartment units into four owner-occupied housing units for low-income individuals (similar to MCURC’s four-plex strategy).

Carl Sterner stated that NEST was exploring options on forgivable interest loans for low-income/mid- income homeowners to combat rising housing costs and keep those individuals from losing their homes and being displaced from the community.

288

Converting vacant buildings to housing is an excellent strategy to combat the displacement of population but owners of vacant buildings are not necessary interested in the welfare of the low-income population. One excellent strategy to employ would be to require building owners to pay vacancy taxes on their vacant buildings (National Low-Income Housing Commission, 2019). However, the policy is controversial. According to the PEW Center, vacancy taxes work better in cities with rapidly increasing land values and not so well in cities with declining or stagnant land values such as Cincinnati (Povich,

2017). Another policy to deal with vacant buildings/lots has been the tax on vacant, impervious land that was first implemented in Philadelphia in 2014. If 75 percent of a vacant property is located on impervious surface (e.g. a parking lot) taxes are increased by 15 percent. Investors can not only avoid the high taxes but receive further tax deductions if they provide 25, 50, or 70 percent affordable housing on these sites.

Again, this policy works best in cities with increasing population (Dealing with Gentrification, 2019).

A Comment on Google Streetview and Census Compatibility: A vital question to answer concerned whether census data and visual imagery from Google

Streetview of the built environment accurately indicated gentrification in these four neighborhoods.

Tracts 74 and 55 showed the most substantial change in the built environment between 2007 and 2016.

This is not surprising as these two tracts were the most disinvested of the four in 2000. Google Streetview imagery fared less well with the analysis for Tract 19. It is likely that large parts of Tract 19 had already undergone the gentrification process before 2007 and that I was unable to document this process using

Google Streetview as the imagery was not available prior to 2007. While Tract 84 had median household incomes that placed it in the disinvested category as of 2000, its built environment was in good condition illustrating that the census data and the Google Imagery did not align.

Future analysis of gentrification at the tract level and coordinating this with Google Streetview will continue to be a productive exercise. One caveat is in order. The Google Streetview imagery only became available in 2007 with many streets not having coverage until 2009. The 2020 decennial census

289 will provide an excellent opportunity to compare census data on social, economic, and demographic indicators with the built environment. As I am finishing this thesis in July 2019 major renovations/construction projects are occurring in Walnut Hills and Madisonville while Northside is preparing for its Bus Transit terminal which should have a profound impact on the demographics in these communities after 2020

290

REFERENCES: Aalbers, Manuel B. Introduction to the Forum: From Third to Fifth Wave Gentrification. Tidjschrift voor Economische en Sociale Geografie, 2019. Forthcoming. Aka, E. Gentrification and Socioeconomic Impacts of Neighborhood Integration and Diversification in Atlanta, Georgia. National Science Journal. 35 (1): 1-13, 2010. Alonso, William. A Reformulation of Classical Location Theory and its Relation to Rent Theory. Papers of the Regional Science Association. 19(1): 22-44, 1967. Apartment Finder Website: Ashwood Apartments.https://www.apartmentfinder.com/Ohio/Cincinnati- Apartments/Ashwood-Apartments Atkinson, Rowland. The Hidden Costs of Gentrification: Displacement in Central London. Journal of Housing and Built Environment. 15(4); 307-326, 2000. Autor, David. The Polarization of Job Opportunities in the U.S. Labor Market: Implications for Employment and Earnings. Community Investments. 23(2): 11-41, 2011. Bates, Lisa. Gentrification and Displacement Study: Implementing an Equitable Inclusive Development Strategy in the Context of Gentrifcation. 2013. www.portlandoregon.gov/bps/article/454027, Last accessed September 30, 2018.

Beauregard, R.A. Trajectories of Neighborhood Change: The Case of Gentrification. Environment and Planning A. 22: 855-874, 1990. Bereitschaft, Bradley. Neighborhood Change Among Creative-Cultural Districts in Mid-Sized U.S. Metropolitan Areas, 2000-10. Regional Studies, Regional Science. 1(1): 158-183, 2014. Betancur, John. Gentrification and Community Fabric in Chicago Urban Studies. 48(2): 383-406, 2011. Blasius, Jorg, Jurgen Friedrichs, and Heiko Ruhl. Pioneers and Gentrifiers in the Process of Gentrification. International Journal of Housing Policy. 16(1): 50-69, 2016 Bourne, Larry S. The Demise of Gentrification? A Commentary and Prospective View. Urban Geography. 14(1): 95-107, 1993. Boyd, Michelle. The Downside of Racial Uplift: Meaning of Gentrification in an African America Neighborhood. City & Society. 17 (2): 265-288, 2005. Brenneman, Brianne. Gentrification Disguised as Urban Revitalization, 2018, https://agorajournal.squarespace.com

Bridges at Belmont Apartments Website. http://thebridgesatbelmont.com/ Brown, James A. The View from Madisonville: Protohistoric Western Fort Ancient Interaction Patterns. American Anthropologist. 101 (2): 409-410, 1999. Brown, Elizabeth and Tony Baize. Impediments to Fair Housing Choice in Hamilton County, Ohio. 2009. CAGIS. http://cagismaps.hamilton-co.org/cagisportal

291

Center for Neighborhood Technology. Broadening Urban Investment to Leverage Transit (BUILT) in Cincinnati, 2011 http://www.reconnectingamerica.org/assets/Uploads/BUILTCincinnati-FINAL.pdf Chan, Sewin. Spatial Lock-in: Do Falling House Prices Constrain Residential Mobility? Journal of Urban Economics. 49(3): 567-586, 2001. Chapple, Karen. Income Inequality and Urban Displacement: The New Gentrification. New Labor Forum. 26(1): 84-93, 2017. Chronopoulos, Themis. African Americans, Gentrification, and Neoliberal Urbanization: The Case of Fort Greene, Brooklyn. Journal of African American Studies. 20: 294-322, 2016. Cilluffo, Anthony, Abigail Geiger, and Richard Fry. More U.S. Households are Renting than at Any Point in 50 Years. Pew Research Center, 2017. https://www.pewresearch.org/fact-tank/2017/07/19/more-u-s-households-are-renting-than-at-any-point- in-50-years/ Cincinnati Bus Coalition, Last Accessed July 5, 2019. https://betterbuscoalition.org/metro-data City of Cincinnati. Plan Cincinnati: A Comprehensive Plan for the Future, 2012. Last accessed January 15, 2019. https://www.cincinnati-oh.gov/buildings/assets/File/final_plan_cincinnati_document_11-21- 12.pdf City of Cincinnati. The City of Cincinnati: 2015-2019 Consolidated Plan and 2015 Annual Action, 2015. Last accessed January 15, 2019. https://choosecincy.com/Cincinnati/media/Cincinnati/Com%20Dev%20PDFs/Substantial-Amendment- 2015-2019-Consolidated-Plan-West-Price-Hill-Area-Program_1.pdf?ext=.pdf City of Cincinnati. Cincinnati Form-Based Code, 2013, https://www.cincinnati-oh.gov/planning/news/final-draft-of-cincinnati-form-based-code/

City of Cincinnati. Go Cincinnati: Growth and Opportunities Study for the City of Cincinnati, 2009, https://www.cincinnati-oh.gov/planning/assets/File/go_cincinnati_summary.pdf

City of Cincinnati. National Register Historic Districts. https://www.cincinnati-oh.gov/buildings/historic-conservation/national-register-historic-districts/

City of Cincinnati. Official Plan of the City of Cincinnati, 1925, https://www.cincinnati- oh.gov/planning/assets/File/1925%20Official%20Plan%20of%20the%20City%20of%20Cincinnati.pdf Clay, P. Neighborhood Renewal: Middle-Class and Incumbent Upgrading in American Neighborhoods. Cohen, Oriya. How Housing Matters, 2019,

292 https://howhousingmatters.org/articles/can-transportation-oriented-affordable-housing-boost-economic- mobility-minimize-displacement/ Coleman, Candace. Gentrification in the Wake of the Subprime Mortgage Crisis. 2012. College Hill Community Urban Development Corporation Website.https://chcurc.com/ College Hill: A Diverse and Neighborly Community. http://chcurc.com/wp-content/uploads/2014/09/chcurc-annual-meeting-presentation-2016.pdf

College Hill Planning Document, 1976. https://www.cincinnati- oh.gov/planning/assets/File/College%20Hill%201976.pdf Cook, Lindsey. U.S. Education: Still Separate and Unequal. U.S. News, 2015, https://www.usnews.com/news/blogs/data-mine/2015/01/28/us-education-still-separate-and-unequal

Cooke, Thomas J. Residential Mobility of the Poor and the Growth of Poverty in Inner-ring Suburbs. Urban Geography 31(2): 179-193, 2010. Copsey, Hillary. Can You Have Development Without Gentrification? Some in the Walnut Hills Neighborhood of Cincinnati Say Yes, 2017, https://beltmag.com/redevelopment-without-gentrification/ Coulton, Claudia J. Using Data to Understand Residential Mobility and Neighborhood Change, 2012 http://www.whatcountsforamerica.org/portfolio/using-data-to-understand-residential-mobility-and- neighborhood-change/#_ftn6 Crowder, Jr. James A. All-In Cincinnati: Equity is the Path to Inclusive Prosperity, 2018, https://nationalequityatlas.org/sites/default/files/All-In_Cincinnati_10-17-18e.pdf Crowder, Kyle and Scott J. South. Race, Class, and Changing Patterns of Migration between Poor and Nonpoor Neighborhoods. American Journal of Sociology. 110(6): 1715-1763, 2005. Davidson, Mark. Displacement, Space and Dwelling: Placing Gentrification Debate. Ethnics, Place & Environment. 12(2): 219-234, 2009. Davidson, Mark and Loretta Lees. New-Build ‘gentrification’ and London’s Riverside Renaissance. Environment and Planning A. 37: 1165-1190, 2005. Dealing with Gentrification: Tax on Vacant, Impervious Land, 2019, http://dealingwithgentrification.org/tax-on-vacant-land/ A Decade of Foreclosures and the Crisis Continues: A Study of Foreclosures in Hamilton County, Ohio in 2012. http://www.marcconline.com/files/DCHO_06_12_2013_1_Foreclosure-Report-2012_WebCopy.pdf

De Marco, A., & Hunt, H. Racial inequality, poverty and gentrification in Durham, North Carolina. Chapel Hill, NC: UNC School of Law, North Carolina Poverty Research Fund, 2018. Demeropolis, Tom. Madisonville Gets Another Chance, 2018

293 https://www.mcurc.org/wp-content/uploads/2018/08/Madisonville-gets-another-chance-Business- Courier-7_20_18.pdf Ding, Lei and Jackelyn Hwang. The Consequences of Gentrification: A Focus on Resident’s Financial Health in Philadelphia. Cityscape: A Journal of Policy Development and Research. 18 (3): 27-55, 2016. Ding, Lei, Jackelyn Hwang and Ellen Divringi. Gentrification and Residential Mobility in Philadelphia. Regional Science and Urban Economics. 61: 38-51, 2016. Dyson, Ryan and David P. Varady. Using Housing Vouchers to Create Stable Mixed-Income Gentrifying Neighborhoods: A Case Study of Over-the-Rhine, Cincinnati. Journal of Community Practice. 26 (3): 257-282, 2018. Ellen, Ingrid Gould. Can Gentrification be Inclusive? 2017, https://www.jchs.harvard.edu/sites/default/files/a_shared_future_can_gentrification_be_inclusive_0.pdf Ellen, Ingrid Gould and Katherine M. O’Regan. How Low- Income Neighborhoods Change: Entry, Exit, and Enhancement. Regional Science and Urban Economics. 41(2): 89-97, 2011. EPA. Supporting Equitable Development: Building Blocks for Sustainable Communities, 2015, http://www.northside.net/northside-community-council/wp- content/uploads/2013/01/CinncinatiNextStepsMemo_Final.pdf Esslelizbichler, Jurgen. The Geography of Job Creation and Destruction in the U.S. Manufacturing Sector, 1967-1997. Annals of the Association of American Geographers. 94(3): 602-619, 2004. Federal Reserve Bank of Cleveland. Cincinnati Recession. www.clevelandfed.org Florida, Richard. The Great Reset; How New Ways of Living and Working Drive Post-Crash Prosperity. Harper: New York City, 2010. Florida, Richard. Cities and the Creative Class. Routledge: New York, 2005. Freeman, Lance M. Commentary: 21st Century Gentrification. Cityscape: A Journal of Policy Development and Research. 18(3): 163-168, 2016. Freeman, Lance. Displacement or Succession? Residential Mobility in Gentrifying Neighborhoods. Urban Affairs Review. 40 (4): 563-491, 2005. Freeman, Lance and Frank Braconi. Gentrification and Displacement in New York City in the 1990s. Journal of the American Planning Association. 70 (1): 39-52, 2004. FRED. Federal Reserve Bank of St. Louis Online Database. https://fred.stlouisfed.org/series/SIPOVGINIUSA Frey, W.H. 1979. Central city white flight: Racial and nonracial causes. American Sociological Review. 44(3): 425-448.

Furman Center (NYU). Gentrification Response: A Survey of Strategies to Maintain Neighborhood Economic Diversity, 2016, http://furmancenter.org/research/publication/gentrification-responses-a-survey-of-strategies-to-maintain- neighborhood-ec

294

Gale, Dennis E. Middle Class Resettlement in Older Urban Neighborhoods: The Evidence and the Implications. Journal of the American Planning Association. 45(3): 293-304, 1979. Galster, George, Peter Tatian and John Accordino. Targeting Investments for Neighborhood Revitalization. Journal of American Planning Association. 72(4): 457-474, 2006, Glass, Ruth. Introduction to London: Aspects of Change. Centre for Urban Studies: London, 1964. Glaeser, Edward L., Joshua Gottlieb, and Joseph Gyourko. Did Credit Market Policies Cause the Housing Bubble? 2010.https://users.nber.org/~jdgottl/housing_2010_final.pdf Glaeser, E.L. and Gyourko, J. 2005. Urban decline and durable housing. Journal of Political Economy. 113(2): 345-375.

Google Streetview. https://www.google.com/maps/@43.4336959,-83.9696384,14z Greenwood, Michael J. and Gary L. Hunt. Jobs Versus Amenities in the Analysis of Metropolitan Migration. Journal of Urban Economics. 25 (1-16), 1989. Grier, G. and E. Grier. Urban Dispacement: A Reconnaisance. In Back to the City: Issues in Neighborhood Renovation. Edited by S. Laska and D. Spain. Oxford: Pergamon Press, 252-268, 1980. Grub, W.N. 1982. The Flight to the Suburbs of Population and Employment, 1960-1970. Journal of Urban Economics. 11(3): 348-367.

Hackworth, Jason. Post-Recession Gentrification in New York City. Urban Affairs Review. 37(6): 815- 843, 2002. Hackworth, Jason and Neil Smith. The Changing State of Gentrification. Tijdschrift voor Economische en Sociale Geographie. 92(4): 464-477, 2001. Hall, Matthew, Kyle Crowder, and Amy Spring. Neighborhood Foreclosures, Racial/Ethnic Transitions, and Residential Segregation. American Sociological Review. 80(3): 526-549, 2015. Hamilton County Auditor’s https://wedge1.hcauditor.org/search_results Hamilton County Landbank. Last Accessed July 5, 2019. https://www.cincinnatiport.org/what-we-offer/neighborhood-residential-commercial-revitalization/ Hammel, Daniel J. and Elvin K. Wyly. A Model for Identifying Gentrified Areas with Census Data. Urban Geography. 17 (3): 248-268, 1996. Hamnett, Chris. The Blind Men and the Elephant: The Explanation of Gentrification. Transactions of the Institute of British Geographers. 16(2): 173-189, 1991. Hampton, Shane. 60 Years of Urban Change: Midwest. 2014. Last accessed January 15, 2019. http://iqc.ou.edu/2014/12/12/60yrsmidwest/ Hartley, Daniel and Daniel Kolliner. Neighborhood Gentrification during the Boom and After. 2014. https://www.cleveland.org Hickory Woods Townhomes Website.

295 https://affordablehousingonline.com/housing-search/Ohio/Cincinnati/Hickory-Woods/10009700 History of Cumminsville, 1792-1914. https://archive.org/details/historyofcummins00unse/page/n6 Hochstenbach, Cody and Willem Boterman. Age, Life Course and Generations in Gentrification Processes. In Handbook of Gentrification Studies. Edited by Loretta Lees and Martin Phillips. Edgar Elgar Publishing, 2018. Holthaus, David. Our Forgotten Neighborhoods: South Cumminsville, 2017, https://www.wcpo.com/longform/our-forgotten-neighborhoods-south-cumminsville

Hoover, E.M. and Vernon, R. 1959. Anatomy of a Metropolis. Garden City, N.Y.: Doubleday Anchor.

Housing Affordability in Hamilton County, 2017, https://www.xavier.edu/communitybuilding/documents/Housing_Affordability_Hamilton_Co_rev03.02.2 017.pdf

Hoynes, Hilary, Douglas L. Miller and Jessamyn Schaller. Who Suffers During Recessions? Journal of Economic Perspectives. 26(3): 27-48, 2012. HUD. Displacement of Lower-Income Families in Urban Areas Report. 2018, https://www.huduser.gov/portal/sites/default/files/pdf/DisplacementReport.pdf

HUD. American Housing Reveals Rise in Doubled-Up Households During Recession, 2014, https://www.huduser.gov/portal/pdredge/pdr_edge_research_012714.html HUD. Beyond Shelter: Building Communities of Opportunity. The United States Report of Habitat II, 1996, https://www.huduser.gov/Publications/PDF/natlrprt.pdf Hwang, Jackelyn. Gentrification, Race, and Immigration in the Changing American City. Ph.D. Dissertation. 2015. Hwang, Jackelyn and R.J. Sampson. Divergent Pathways of Gentrification: Racial Inequality and the Social Order of Renewal in Chicago Neighborhoods. American Sociological Review. 79(4): 726-751, 2014. Hyra, Derek and Jacob S. Rugh. The U.S. Great Recession: Exploring its Association with Black Neighborhood Rise, Decline, and Recovery. Urban Geography.37(5): 700-726, 2016. Integra Life Sciences, 2013, https://www.cincinnatiport.org/wp-content/uploads/Industrial-Revitalization-Integra-Life-Sciences.pdf Jackson, Jonathan. The Consequences of Gentrification for Racial Change in Washington, D.C. Housing Policy Debate. 25(2): 353-373, 2015.

296

Kennedy, Maureen and Paul Leonard. Dealing with Neighborhood Change: A Primer on Gentrification and Policy Choices. The Brookings Institute and PolicyLink, 2001. Kinder Institute of Urban Research at Rice University. 2018. https://kinder.rice.edu/sites/g/files/bxs1676/f/documents/KI%20Research%20Report_Gentrification%202 .pdf Kirkland, Elizabeth. What’s Race Got to Do with It? Looking for the Racial Dimensions of Gentrification. The Western Journal of Black Studies. 32 (2), 2008. Konermann, Alyssa. Will Walnut Hills Rise Again?, 2015, https://www.cincinnatimagazine.com/citywiseblog/will-walnut-hills-rise-again/ Kuehn, Daniel. Metropolitan Job Growth Patterns in the Great Recession, 2011, https://www.urban.org/sites/default/files/publication/27636/412421-Metropolitan-Job-Growth-Patterns- in-the-Great-Recession.PDF

Landis, John D. Tracking and Explaining Neighborhood Socio-economic Change in U.S. Metropolitan Areas Between 1990 and 2010. Housing Policy Debate. 26(1): 20=-52, 2016. Lee, Barrett A., Daphne Spain and Debra J. Umberson. Neighborhood Revitalization and Racial Change: The Case of Washington, D.C. Demography. 22(4) 581-602, 1985. Lees, Loretta. Urban Renaissance in an Urban Recession: The End of Gentrification? Environment and Planning A: Economy and Space. 41: 1529-1533, 2009. Lees, Loretta, Tom Slater, and Elvin Wyly. Gentrification. Routledge: New York, 2008. Lees, Loretta and Liz Bondi. De-gentrification and Economic Recession: The Case of New York City. Urban Geography. 16(3): 234-253, 1995. Lees, Loretta. Rethinking Gentrification: Beyond the Positions of Economics or Culture. Progress in Human Geography. 18(2): 137-150, 1994. Le Gates, R. and C. Hartman. The Anatomy of Displacement in the United States. In Gentrification in the City. Editors Neil Smith and P. Williams, 1986. Ley, David. Alternative Explanations for Inner-City Gentrification: A Canadian Assessment. Annals of the Association of American Geographers. 76(4): 521-535, 1986. Li, Xiaojiang, Chuansong Zhang, Weidong Li, Robert Ricard, Qingyan Meng, and Weixing Zhang. Assessing Street-Level Urban Greenery Using Google Street View and a Modified Green View Index. Urban Forestry and Urban Greening. 14: 675-685, 2015. London, Bruce, Barrett A. Lee, and S. Gregory Lipton. The Determinants of Gentrification in the United States: A City-Level Analysis. Urban Affairs Quarterly. 21(3): 369-387, 1986. London, Bruce, Donald S. Bradly and James R. Hudson. Introduction: Approaches to Inner-City Revitalization. Urban Affairs Quarterly. 15(4): 373-380, 1980.

297

Long, L. 1988. Migration and Residential mobility in the United States. Russel Sage Foundation: New York. Loukaitou-Sideris, Anastasia, Silvia Gonzalez and Paul Ong. Triangulating Neighborhood Knowledge to Understand Neighborhood Change: Methods to Study Gentrification. Journal of Planning Education and Research. 1-16. Madisonville 500 Gardens, https://greenumbrella.org/resources/GU%20Initiatives/Food%20Policy%20Council/500%20Gardens_Lay er_02%20FINAL!.pdf Madisonville Plan, 2016, https://www.cincinnati.com/story/money/2016/03/09/200m-investment-planned-madisonville/81486302/ Madisonville Becomes First Cincinnati Neighborhood to Implement New FBC, 2013 https://opticosdesign.com/blog/madisonville-becomes-first-cincinnati-neighborhood-to-implement-new- fbc/ Madisonville Industrial Corridor Urban Renewal Plan, 1992, https://www.cincinnati-oh.gov/planning/assets/File/Madisonville%20URP%201991.pdf Madisonville Community Plan, 1975. https://www.cincinnati- oh.gov/planning/assets/File/Madisonville_Eastwood%20Community%20Plan%201975.pdf Mallach, Alan. Re-engineering the Urban Landscape: Land Use Reconfiguration and the Morphological Transformation of Shrinking Industrial Cities. In Engineering Earth. Stanley Brunn (Ed.). 1855-1883. Springer: Dordrecht, 2011. Mallach, Alan and Lavea Brachman. Ohio’s Cities at a Turning Point: Finding the Way Forward. 2010. Last accessed January 15, 2019. https://www.brookings.edu/wpcontent/uploads/2016/06/0518_ohio_cities_mallach_brachman.pdf Maloney, Michael and Christopher Auffrey. The Social Areas of Cincinnati: An Analysis of Social Needs: Patterns for Five Census Decades, 2013. Last accessed January 12, 2019. http//www.socialareasofcincinnati.org Mansfield, Carol, Subhrendu K. Pattanayak; William McDow; Robert McDonald; and Patrick Halpin. Shades of Green: Measuring the Value of Urban Forests in the Housing Market. Journal of Forest Economics. 11(3): 177-199, 2005. Marcuse, Peter. Gentrification, Abandonment, and Displacement: Connections, Causes and Policy Responses in New York City. Urban Law Annual: Journal of Urban and Contemporary Law. 28: 195- 240, 1985. Mathema, Silva. Gentrification: An Updated Literature Review. 2013. Last accessed February 18, 2019. https://prrac.org/pdf/Gentrification_literature_review_-_October_2013.pdf

298

Mather, Mark and Linda A. Jacobsen. Household Change in the United States. Population Reference Bureau. 2012. Last accessed January 15, 2019. https://www.prb.org/us-household-change/ McKinnesh, T., R. Walsh and K. White. Who Gentrifies Low-Income Neighborhoods? Journal of Urban Economics. 67, 1180-193, 2010. Molloy, Raven, Christopher L. Smith and Abigail Wozniak. Internal Migration in the United States. Journal of Economic Perspectives. 25 (3): 173-196, 2011.

Moore, Derick. Renters Moving at Historically Low Rates. U.S. Census Bureau, 2017. https://www.census.gov/library/stories/2017/12/lower-moving-rate.html

Moos, Markus. From Gentrification to Youthification? The Increasing Importance of Young Age in Delineating High Density Living. Urban Studies. 53(4): 2903-2920, 2016. Murphy, Alexandra K. and Scott W. Allard. The Changing Geography of Poverty. Focus. 32(1): 19-23, 2015. Naik, Nikhil, Scott Duke Kominers, Ramesh Raskar, Edward L. Glaeser, and Cesar A. Hidalgo. Computer Vision Uncovers Predictors of Physical Urban Change. PNAS. 2017. Nakajima, Makoto. The Diverse Impact of the Great Recession. 2013. Last accessed January 15. https://www.philadelphiafed.org/-/media/research-and-data/publications/business- review/2013/q2/brq213_the-diverse-impacts-of-the-great-recession.pdf?la=en National Low- Income Housing Commission. Gentrification and Neighborhood Revitalization: What’s the Difference? 2019, Newman, Kathe and Elvin K. Wyly. The Right to Stay Put, Revisited: Gentrification and Resistance to Displacement in New York City. Urban Studies. 43 (1): 23-57, 2006. Newman, Sandra J. and Michael S. Owen. Residential Displacement: Extent, Nature, and Effects. Journal of Social Issues. 38 (3): 135-148, 1982. Northside: The Urban Village. Last accessed March 15, 2019. http://www.northside.net/

Northside Comprehensive Land Use Plan Update, 2014, https://www.cincinnati-oh.gov/planning/assets/File/nlupu_final_document_6-18-14_reduced(1).pdf NPR. A Portrait of America’s Middle Class by the Numbers, 2016, https://www.npr.org/2016/07/07/484941939/a-portrait-of-americas-middle-class-by-the-numbers Owens, Ann and Jennifer Candipan. Racial/Ethnic Transition and Hierarchy Among Ascending Neighborhoods. Urban Affairs Review. 2018. Park, R.E., Burgess, E.W. and McKenzie, R.D. 1925. The City. Chicago: The University of Chicago Press.

299

Potter, Paul Edwin. Exploring the Geology of the Cincinnati/Northern Kentucky Region. Kentucky Geological Survey: Lexington, 1996. Povich, Elaine S. Can Extra Taxes on Vacant Land Cure City Blight?, 2017, https://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2017/03/07/can-extra-taxes-on- vacant-land-cure-city-blight Quinn, James W., Stephen J. Mooney, Daniel M. Sheehan, Julien O. Teitler, Kathryn M. Neckerman, Tanya K. Kaufman, Gina S. Lovasi, Michael D.M. Bader and Andrew G. Rundle. Neighborhood Physical Disorder in New York City. Journal of Maps. 12 (1): 53-60, 2016. Revington, Nick. Gentrification, Transit, and Land Use: Moving Beyond Neoclassical Theory. Geography Compass. 9(3): 152-163, 2015. Richardson, Jason, Bruce Mitchell, and Juan Franco. Shifting Neighborhoods: Gentrification and Cultural Displacement in American Cities, 2019, https://community-wealth.org/content/shifting-neighborhoods-gentrification-and-cultural-displacement- american-cities Rinehart, Bill. Finding Racial Balance in Madisonville, 2016, https://www.wvxu.org/post/finding-racial-balance-madisonville#stream/0 Rose, Damaris. Rethinking Gentrification: Beyond the Uneven Development of Marxist Urban Theory. Environment and Planning D: Society and Space. 1: 47-74, 1984. Ross, Jaimie. The Advent of Form-Based Codes: A Critical Time to Ensure Mixed Income Communities, 2012, https://www.flhousing.org/wp-content/uploads/2012/05/Form-Based-Codes.pdf Rundle, Andrew G., Michael D.M. Bader, Catherine A. Richards, Kathyn M. Neckerman, and Julien O. Teitler. Using Google Streetview to Audit Neighborhood Environments. American Journal of Preventive Medicine. 40(1): 94-100, 2011. Ryan, Camille L. and Kurt Bauman. Educational Attainment in the United States, 2015. https://www.census.gov/content/dam/Census/library/publications/2016/demo/p20-578.pdf

Saldana, Rebecca and Margaret Wykowski. Racial Equity: New Cornerstone of Transit-Oriented Development. Race, Poverty, & the Environment. 19 (2): 13-17, 2012. Schneider, Benjamin. How Transit-Oriented Development Can Prevent Displacemnt, 2018, https://www.citylab.com/equity/2018/04/how-transit-oriented-development-can-prevent- displacement/556373/ Schill, Michael H. and Richard P. Revitalizing America’s Cities: Neighborhood Reinvestment and Displacement. Albany, New York: State University of New York Press, 1983. Shaw, Kate S. and Iris W. Hagemans. Gentrification Without Displacement and the Consequent Loss of Place: The Effects of Class Transition on Low-income Residents of Secure Housing in Gentrifying Areas. International Journal of Urban and Regional Research. 39(2): 323-341, 2015.

300

Simes, Randy A. Gentrification Occurring in More Than Cincinnati’s Center City Neighborhoods. UrbanCincy. 2013, http://www.urbancincy.com/2013/12/gentrification-occurring-in-more-cincinnati-neighborhoods-than- just-those-in-center-city/ Simple Random Number Table. ph=280&expw=424&q=random+number+table&selectedindex=0&ajaxhist=0&vt=0&eim=1,2,6 Slater, Tom. The Eviction of Critical Perspectives from Gentrification Research. International Journal of Urban and Regional Research. 30 (4): 737-757, 2006. Smith, Lisa. Walnut Hills Wants Development but not at the Cost of its Residential Neighborhoods, 2019, https://www.wcpo.com/news/transportation-development/move-up-cincinnati/walnut-hills-wants- development-but-not-at-the-cost-of-its-residential-neighborhoods Smith, Neil. The New Urban Frontier: Gentrification and the Revanchist City. New York: Routledge, 1996. Smith, Neil. Toward a Theory of Gentrification: A Back to the City Movement of Capital, Not People. Journal of the American Planning Association. 538-548, 1979. Solari, Claudia D. Affluent Neighborhood Persistence and Change in U.S. Cities. City & Community. 1(4): 370-388, 2012. Spear, Alden, Sidney Goldstein and William H. Frey. Residential Mobility, Migration, and Metropolitan Change. HarperCollins Distribution Series: New York, 1975. Stabrowski, Filip. Inclusionary Zoning and Exclusionary Development: The Politics of Affordable Housing in North Brookyn. International Journal of Urban and Regional Research. 1-17, 2015. Sumka, H.J. Neighborhood Revitalization and Displacement: A Review of Evidence. Journal of American Planning Association. 45 (2): 480-487, 1979. Sutton, Stacey. Gentrification and the Increasing Significance of Racial Transition in New York City, 1990-2010. Urban Affairs Review. 1-31, 2018. Swartsell, Nick. Northside Transit Center Poised to Move Forward, 2019, https://www.citybeat.com/news/blog/21064565/cincinnati-city-council-approves-key-zoning-changes-for- northside-transit-hub Taeuber, Karl E. and Alma F. Taeuber. Negroes in the Cities: Residential Segregation and Neighborhood Change. Chicago: Aldine, 1965. Theodus, Brett, Rob Pitingolo, and Claudia J. Coulton. Housing Unit Turnover and the Socioeconomic Mix of Low-Income Neighborhoods. The Annals of the Academy of Political and Social Science. 660(1): 117-135, 2015. Thomas, June Manning. The Cities Left Behind. Built Environment. 17 (3/4): 218-231, 1991. Tighe, J. Rosie. Public Opinion and Affordable Housing: A Review of the Literature. Journal of Planning Literature. 25(1): 3-17, 2010.

301

Transforming Vacancy in Walnut Hills: Relocal Analysis, 2015. https://www.placeeconomics.com Tweh, Bowdeya. $200M Investment Planned in Madisonville, 2016, Twin Towers: A Life Enriching Community Website. http://lec.org/twin-towers/location/ Urban Displacement Project, 2015, https://www.urbandisplacement.org/ Urban Migration & Gentrification: Stories from Cincinnati’s Over-the-Rhine. https://jackdelisiohist415.wordpress.com U.S. Bureau of Census. 1950 Census of Housing: Volume Last accessed January 15, 2019. https://www2.census.gov/library/publications/decennial/1950/housing-volume-1/36965082v1p5ch03.pdf U.S. Census Bureau. U.S. Decennial Census and American Community Surveys 2006-2010 and 2012- 2016. Last accessed January 15, 2019. https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml U.S. Department of Education. 120 Years of American Education: A Statistical Portrait, 1993. https://nces.ed.gov/pubs93/93442.pdf U.S. Department of Housing and Urban Development. Comprehensive Housing Market Analysis: Cincinnati, Ohio-Kentucky-Indiana, 2017. Last accessed January 15, 2019. https://www.huduser.gov/portal/publications/pdf/CincinnatiOH-comp-17.pdf U.S. Department of Labor Statistics, 2012, https://www.bls.gov/ Verick, Sher and Iyanatul Islam. The Great Recession of 2008-2009: Causes, Consequences, and Policy Responses, 2010, http://ftp.iza.org/dp4934.pdf Vigdor, Jacob L. Does Gentrification Harm the Poor. Brookings Wharton Papers on Urban Affairs. 133- 182, 2002. Walnut Hills Redevelopment Foundation, https://walnuthillsrf.org/ 2017 Walnut Hills Reinvestment Plan, 2016. https://www.cincinnati- oh.gov/planning/assets/File/Walnut%20Hills%20Reinvestment%20Plan%20Final.pdf Walnut Hills Redevelopment Foundation, 2013, https://walnuthillsrf.org/throwback-thursday-the-hauck-building/ Walnut Hills Community Plan, 1991,

302 https://www.cincinnati- oh.gov/planning/assets/File/Walnut%20Hills%20Community%20Plan%201992.pdf Walnut Hills Urban Design Plan, 1975, https://www.cincinnati-oh.gov/planning/assets/File/Walnut%20Hills%20UDP%201975.pdf https://walnuthillsrf.org/wp-content/uploads/2018/05/Walnut-Hills-Market-Analysis-Strategy.pdf Warde, A. Gentrification of Consumption: Issues of Class and Gender. Environment and Planning D: Society and Space. 9: 223-32, 1991. Weingartner, Tara. Are Changes Underway in Madisonville Revitalization or Gentrification?, 2016 https://www.wvxu.org/post/are-changes-underway-madisonville-revitalization-or-gentrification#stream/0 West Fork Watershed Fact Sheet. http://www.projectgroundwork.org/downloads/westfork/WestForkWatershedFactSheet.pdf White, Harrison C. Multipliers, Vacancy Chains, and Filtering in Housing. American Institute of Planners Journal. 37(2): 88-94, 1971. Wilson, William Julius. The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press, 1987. Woodward, Colin. How Cincinnati Salvaged the Nation’s Most Dangerous Neighborhood, 2016, https://www.politico.com/magazine/story/2016/06/what-works-cincinnati-ohio-over-the-rhine-crime- neighborhood-turnaround-city-urban-revitalization-213969 Writer’s Program, Ohio. Cincinnati: A Guide to the Queen City and its Neighbors. Cincinnati: Niesen- Hart Press, 1943. Wyly, Elvin K. and Daniel J. Hammel. Islands of Decay in Seas of Renewal. Housing Policy Debate.10(4): 711-771, 1999. Zillow, Median House Value. 2019, https://www.zillow.com/madisonville-cincinnati-oh/home-values/

303

304