Also by Reinvestment Fund

Reinvestment Fund Provides Expert Witness Affidavit in Philadelphia Case on Evictions

In August, 2020, Ira Goldstein, President of Policy Solutions at Reinvestment Fund provided an expert witness affidavit for a court case in which the City of Philadelphia and intervenors (Tenant Union Representative Network and the Philadelphia Unemployment Project, represented by Community Legal Services) were defending an attempt to enjoin the emergency protections that were created for renters impacted by COVID.

The following is a portion of Mr. Goldstein’s affidavit. Click here for the complete affidavit.

The City and intervenors successfully defended these emergency protections for renters.


Affidavit of Expert Witness Ira J. Goldstein

I, Ira J. Goldstein, declare as follows: I hold a Ph.D. in Sociology, granted in 1986, from Temple University. My current position is that of President of the Policy Solutions group (“Policy”) at Reinvestment Fund. I have been at Reinvestment Fund since June 1999. Reinvestment Fund is a national mission-driven non-profit financial institution that creates opportunity for underserved people and places through partnerships. We marshal the capital, analytics, and expertise necessary to build strong, healthy, and more equitable communities.(1)

I am also a Lecturer at the University of Pennsylvania where, for the last 34 years, I have taught an undergraduate course in statistics and research methods for the Urban Studies Program. For several years, I also taught a graduate/undergraduate level, multidisciplinary course on housing policy.

In my position as President of Policy for RF, I am responsible for numerous grant and contract-related research projects. Among the portfolio of current and recently completed projects is a three-year applied research project focused on evictions in Philadelphia. This project is funded by the UK-based Oak Foundation, and the Philadelphia-based William Penn Foundation. Among our current/prior federal government clients are the United States Department of Housing and Urban Development, the Institute of Museum and Library Services and the United States Department of Justice. We also have provided housing and community development consultation to numerous state and local governments and Authorities including, in Pennsylvania, Philadelphia, Pittsburgh, Northampton County, Reading and many others.

I also crafted and executed an evaluation of Philadelphia’s Residential Mortgage Foreclosure Diversion Program. This work was funded by the Open Society Foundations (formerly known as the Open Societies Institute) and the William Penn Foundation. We issued several reports, the first of which was in February 2011 detailing methods to study these newly created court interventions. Later reports explored the impacts, outcomes and sustainability of agreements achieved through the court, including the percentage of people still in their homes 2.5+ years after resolution of the foreclosure action.(2)

Prior to holding my position at RF, I served as the U.S. Department of Housing and Urban Development’s Director of Fair Housing and Equal Opportunity for the federal Mid-Atlantic Region. Among other duties, I was responsible for directing enforcement of the federal Fair Housing Act (Act). I led a team of investigators across the Mid-Atlantic who investigated allegations of housing discrimination, and conciliators who endeavored to settle these discrimination cases, all as required by the Act and regulations implementing the Act. My publications include authorship and co-authorship of numerous reports, articles, book chapters and books regarding a broad range of housing, housing finance/mortgage lending, fair housing, and community development topics. I coauthored two books focused on these (and other) topics on the Philadelphia region. My resume is attached to this affidavit as Attachment 1.

I have been retained as an expert witness by the Intervening Defendants in this matter to: A. Provide my research findings about the trends and patterns of eviction filings in Philadelphia; B. Identify the demographic characteristics and zip codes most disproportionately impacted by COVID-19 in Philadelphia; C. Provide my professional opinion based on the data and research findings I collect; D. Provide my professional opinion about the impact to communities and Philadelphia residents if the Emergency Housing Protection Act (EHPA) is overturned; and E. Discuss my research findings regarding the successes and / or failures of Philadelphia’s Residential Mortgage Foreclosure Diversion Program.

A. Trends And Patterns Of Eviction Filings In Philadelphia

In 2017, Reinvestment Fund released its first Research Brief focused on trends and patterns of eviction filings in Philadelphia.(3) Based on court data collected by Philadelphia Legal Assistance, covering the period 2010-2015 inclusive, this Brief chronicled the near 20,000 eviction filings per year, representing a “rate” of approximately 7% (or more) each year studied.(4) This Brief also detailed eviction filing rates in Philadelphia census tracts by median household income level and racial composition. Generally, the filings rates were higher in lower income areas and in areas with higher percentages of Black residents. Differences in the eviction filing rates
in areas that were greater than 80% Black were typically three-times greater than the rates in areas that were under 10% Black.

A multiple regression analysis demonstrated that the percent of residents who are Black in a census tract had a substantial and statistically significant impact on the eviction rate. We concluded: “Indeed, net of an area’s median household income and tenure differences, each one percentage point increase in the percent African American/Black residents in a Census tract is associated with a 0.36 percentage point increase in the eviction rate.” (page 5).

Two years later, Reinvestment Fund released another Research Brief updating and elaborating the analysis contained in the 2017 Brief.(5) This Brief, based on data obtained from Philadelphia Municipal Court, detailing trends and patterns through 2019, noted an overall decline in the volume of filings and rates. It also chronicled a persistent differential in rates by census tract income and racial composition.

In the 2017 Brief, we concluded that the factors most likely contributing to the decline in filings included declining unemployment and rising levels of labor force participation, rising access to medical insurance, and eviction process reforms and funding instituted by Philadelphia City Council and the Kenney Administration.

Having noticed the large number/percent of eviction filings associated with tenants of the Philadelphia Housing Authority – rates which, over the last five years studied ranged between 9.5% and 12% – the analysis sought to statistically control for public housing in understanding the relationship between an area’s racial composition and evictions. We concluded: “With the higher filing rate for PHA units noted above, we do also observe that the presence of 50 or more PHA units is associated with elevated filing rates. Our analysis also shows that racial composition has a persistent, statistically significant effect on filing volume, independent of other key factors. In
fact, net of differences in census tract incomes, tenure mixes, and concentrations of public housing units, for each 10% increase in percent African American in the census tract, there are 7.6 more eviction filings …” (page 9).

Attachment 2 presents a map of the eviction filing rate for 2018-2019; areas that are over 50% Black are outlined in blue and areas that are over 25% Hispanic are crosshatched. This map shows that most areas with elevated eviction rates across the city of Philadelphia are also places where the majority of the residents are Black. That pattern is not as clear for areas that are above average in their percent of Hispanic residents. Reinvestment Fund conducted a set of interviews to understand the lower than expected filing rates (i.e., lower than expected because of the low level of income in Hispanic areas and rates of owner occupancy) in Hispanic areas. Those interviews
were with staff of non-profit organizations located and/or serving the Hispanic community. Several themes emerged from those interviews. First, these local practitioners made clear that Hispanic residents in these areas do in fact get evicted, but by our counting formal evictions (i.e., court filings), we do not see a fully accurate representation of the eviction of Hispanic households in the data. “Interviewees described a reliance on family and social networks in Philadelphia’s Hispanic communities in order to find housing. This affects landlord-tenant dynamics: the personal connection between landlords and tenants may make landlords slower to evict and may encourage tenants to “self-evict” when they cannot pay in order to protect the relationship that helped them get the unit in the first place. Informal hosting or “couch surfing” is also reportedly common, where there is no lease or expectation of long-term residency … Finally, interviewees pointed to a long-running lack of connection between city government and Hispanic communities, in part due to cultural expectations regarding the role of government. Landlords are less likely to go to court in a dispute and tenants are less likely to turn to the city for help with concerns about code violations or other housing rights. This dynamic may be exacerbated by the inadequacy of Spanish-language resources.” (page 14) Thus, given that our research generally relies on formal eviction filings in court, we are unable to fully describe the eviction of Hispanic households because many of the Hispanic renters live in informal arrangements and are subject to an extra-judicial eviction process.

B. Demographic Characteristics and Zip Codes Most Disproportionately Impacted By COVID-19 In Philadelphia

The City of Philadelphia Department of Public Health publicly reports data on COVID-19; they report data by race/ethnicity. As of July 27, 2020,(6) the Black COVID- 19 positivity rate was 206 per 10,000 Black residents; it was 153 per 10,000 Hispanic residents and 90.1 per 10,000 White residents. Thus, the Black positivity rate is more than double that of White residents and the Hispanic rate is more than 50% higher than the White rate.

Hospitalizations are similarly disparate by race/ethnicity. The Black hospitalization rate due to COVID-19 was 55 per 10,000 Black residents; it was 38.4 per 10,000 Hispanic residents and 19.8 per 10,000 White residents. Lastly, the COVID-19 death rate is also disparate by race/ethnicity. The Black death rate was 12.7 per 10,000 Black residents; it was 6.5 per 10,000 Hispanic residents and 8.5 per 10,000 White residents.

Maps of the COVID-19 test positivity rate per 10,000 Philadelphians (shaded by zip code) and hospitalization rate per 10,000 Philadelphians (also shaded by zip code) are found under Attachment 3. While many of the areas manifesting elevated COVID-19 positivity rates and hospitalization rates are areas that are predominantly Black, several are not.(7) One of the areas shaded among zip codes with the highest positivity rate is 19126 (Oak Lane); it is 83% Black. Another of the highest positivity rate zip codes is 19142 (Elmwood / Southwest Philadelphia); it is 85% Black. Zip code 19126 also has among the highest rates of hospitalization.

As shown in Table 1, zip codes with higher COVID-19 positivity rates are home to a renter population that is predominantly Black. That pattern reverses for Hispanic renters who comprise larger percentages of the renter population in lower positivity rate zip codes. The typical zip code in areas with higher COVID-19 positivity rates has substantially higher rates of residents who are medically uninsured; in fact, the typical uninsured rate in high positivity rate zip codes in nearly double the rate found in low positivity rate zip codes.

Also shown in Table 1, zip codes with higher COVID-19 hospitalization rates are home to a group of renters who are more predominantly Black. This is a more extreme pattern than observed for the positivity rate. The percentage of renters who are Black in low hospitalization rate areas is 5.3% and 76% in high/very high hospitalization rate areas. The pattern for Hispanic renters is different, but it remains that the percent of renters who are Hispanic is generally higher in zip codes with higher COVID-19 hospitalization rates. The typical zip code in areas with higher hospitalization rates has substantially higher rates of residents who are medically uninsured; in fact, the typical uninsured rate in high COVID-19 hospitalization rate zip codes is more than double the rate found in low positivity rate zip codes.

The typical zip code’s eviction rate varies by the COVID-19 positivity rate. Zip codes with the highest positivity rates typically manifest a 9.1% eviction rate; that compares to 3.3% for the zip codes with the lowest positivity rate. The pattern is even more extreme when considering the COVID-19 hospitalization rate. In areas with the lowest COVID-19 hospitalization rates, the typical zip code has an eviction rate of 1.4%. By comparison, the typical eviction rate in areas with the highest hospitalization rates is 6.5 times higher, at 9.1%.

Maps of the 2018-2019 eviction rate per census tract are presented under Attachment 4 with an overlay indicating areas with the highest COVID-19 positivity and hospitalization rates.(8) Most areas with elevated eviction rates fall within zip codes with the highest COVID-19 positivity rates. A similar pattern is observed when evictions are presented with the hospitalization rate. Many parts of north, northwest, west and southwest Philadelphia with high eviction rates fall within those zip codes with the highest hospitalization rates. Researchers at the Federal Reserve Bank of Philadelphia recently released an analysis identifying the extent to which neighborhoods in Philadelphia, throughout the Bank’s district, and for the nation as a whole, had residents who would be most economically impacted by COVID-19 (hereafter, “economic risk”).(9) In order to understand economic risk as the pandemic unfolded they “…measure[d] the share of all employment that is at risk because it requires close physical proximity and was therefore likely to be negatively impacted by social distancing requirements and associated shutdown orders.” (page 1) Census tract economic risk scores were provided by the study’s authors to Reinvestment Fund for purposes of this affidavit.

Data show that areas at elevated levels of economic risk are home to substantially greater shares of Black and Hispanic renters compared to areas with lesser economic risk. For example, the percent of renters who are Black is 20% in the lowest economic risk areas and 48% in highest economic risk areas; it is over 60% in the second highest economic risk areas. The percent of renters who are Hispanic is 5.9% in the lowest economic risk areas and 27.8% in the highest economic risk areas.

Economic risk is also associated with whether residents have health insurance. The typical census tract in low economic risk areas has 3.8% of its residents medically uninsured, while 11.3% of the residents in the highest economic risk areas are similarly  uninsured.

Areas identified by the Federal Reserve Bank researchers at higher economic risk are also areas that generally appear to be higher in COVID-19 test positivity rate, and especially, hospitalization rate. Attachment 6 contains maps of risk with overlays denoting places with the highest levels of COVID-19 test positivity and hospitalization rates.

Table 2 presents data on the population, medical insurance status and evictions by economic risk category. Areas manifesting greater risk of COVID-19 economic impact also have higher eviction filing volumes and rates of evictions. The lowest risk areas have 4,025 evictions in 2018-2019 with an eviction rate of 2.7%. The highest risk area have 9,234 evictions with an eviction rate of 8%. The areas in the second highest risk category have 9,397 eviction filings for a rate of 8.2%.

Maps of the Federal Reserve Bank’s estimation of economic risk due to COVID-19 and evictions are provided under Attachment 7.

C. My Professional Opinion Based On The Data And Research Findings

The conclusion I draw from the connection between the risk of COVID-19, race, medical insurance coverage and evictions is that: Areas that are most at risk for the adverse economic consequences of COVID-19 are also home to greater shares of Black and Hispanic renters. They are also areas where Census data shows that the residents are more likely to be medically uninsured.(10) Lastly, these are areas where the contemporary eviction rates are substantially elevated – elevated in a way that is disproportionately adverse in substantially Black areas. If the Federal Reserve Bank’s researchers are correct about the economic risk to residents of these areas, it is likely
then that as the pandemic unfolds, these predominantly Black and Hispanic areas will be economically adversely impacted and more likely to be more substantially and adversely impacted by evictions. The situation is exacerbated by the fact that these are also areas where the percent of residents without medical insurance is elevated, raising the specter that residents may be unable to seek medical advice and treatment.(11)

D. Impact To Communities And Philadelphia Residents If The EHPA Is Overturned

In a public health crisis like the COVID-19 pandemic, stable housing (like social distancing, washing your hands, and wearing a mask) is a vital health and safety measure. People who may be infected and must defend an eviction filing action in court run the risk of infecting their landlord, attorneys, court staff and other tenants facing eviction (let alone the people they come in contact with on the journey from home to court and vice versa). If they are not themselves infected, they are placed at risk of infection by others who are. And, if the eviction proceeds to a Writ of Possession for a family with a COVID-19 infected family member, that family must then find another place to live, again, risking friends or families with whom they may double-up, or in the most dire situation, other residents of the City’s homeless shelters, encampments, or in the streets. And for those families without an infected member, the forced move puts them at risk of contracting an infection. The data therefore suggest that destabilizing housing by lifting protections against evictions disproportionately subjects Black and Hispanic people and people of lower income to enhanced risk of not only contracting the COVID-19 virus but also suffering a more dire consequence, such as death.

E. Philadelphia’s Residential Mortgage Foreclosure Diversion Program

As noted in paragraph 5, Reinvestment Fund studied Philadelphia’s Residential Mortgage Foreclosure Diversion Court (Diversion Court). The Diversion Court was established by Order of the court in April 2008 with an expiration of December 31, 2009;(12) a subsequent Order issued on December 17, 2009 made it a permanent fixture of the court. The Diversion Court process worked with homeowners and lenders/servicers after the foreclosure action was filed in the Court of Common Pleas.

The process was essentially an opt-out by the homeowner. That is, if eligible, the Diversion Court was available to them unless they decided not to participate in the process. If the homeowner chose to participate, the lender/servicer was required to participate.

Homeowners were given access to a free housing counselor who would perform one or more of the following functions: counsel the homeowner, identify resources that might be available to resolve the foreclosure, support/represent the homeowner in negotiations with the lender/servicer.

Between 2011 and 2014, we released three reports. The first report detailed methods for evaluating the impacts of Diversion Court.(13) In 2011, we also prepared a report detailing learnings from the first approximately two years of case filings. (14) Our third report, prepared in 2014 was an update to the first and detailed six years of activity . (15)

The primary learnings from reviewing approximately 60,000 court orders representing almost 30,000 foreclosure cases processed by the Diversion Court are, over the period studied: (1) Typically 65%-70% of all eligible homeowners participated in the process; (2) Of those that participated, in excess of 30% achieved some form of Agreement between the homeowner and her mortgage lender/servicer; (3) Diversion Court added relatively little time to the overall foreclosure process – 54 days on average and two formal contacts with the court; ( 4) Of those that achieved an Agreement, more than 85% remained in their homes approximately two-plus years post-Agreement; (5) We explored whether there was any bias in one’s ability to access the Diversion Program or to achieve an Agreement. Rigorous multivariate statistical analyses in the 2011 and 2014 studies uncovered no racial bias. We stated in the 2014 rep01i: ” … residents of Philadelphia are able to access the Diversion Program and achieve agreements with their lender/servicer without regard to the racial/ethnic composition of the neighborhood within which they live or the income level of that community.” (p. 7)


(1) A complete description of Reinvestment Fund can be found at its website:

(2) See, for example: or

(3) See:

(4) The “rate” was defined as the number of eviction filings as a function of the number of rental units as defined by the US Census American Community Survey for the same year.

(5) See:


(7) Important to take note that the zip code area in Northeast Philadelphia (zip code 19136) with the highest tier positivity rate is the Holmesburg area, and it likely is elevated because of the presence of a prison population.

(8) We point out that the eviction data is summarized at the census tract level while the COVID-19 positivity and hospitalization rates are summarized at the zip code level. In general, zip codes are much larger than census tracts.

(9) See Eileen Divringi and Davin Reed, 2020. “Which Neighborhoods and Households Will Be Most Impacted by COVID-19?”. Accessed from:

(10) The Kaiser Family Foundation notes that those most likely to be medically uninsured across the United States are low income families with at least one worker in the family. They note also that adults and people of color (compared to non-Hispanic White people) are more likely to be without medical insurance. Kaiser also posits that the primary reason that people are not insured is the cost of coverage. (see: In Philadelphia, the number of medically uninsured residents had fallen since the passage of the Affordable Care Act, however in 2018, the Census reports an uptick in the number of medically uninsured.

(11) The Kaiser Family Foundation (Ibid.) states “Studies repeatedly demonstrate that uninsured people are less likely than those with insurance to receive preventive care and services for major health conditions and chronic diseases.” Public health professionals tell us that those major health conditions and chronic diseases contribute to more adverse results of contracting COVID-19.