By Julieta Cuellar

 

^Abstract

The Eviction Lab’s recently released dataset of evictions in the United States provides rich opportunities for exploring the effect of state and local policies on eviction rates. Just cause eviction ordinances—local laws that outline what constitutes grounds for eviction—have gained traction as a policy solution for addressing the eviction crisis. This paper analyzes the relationship between just cause eviction ordinances and eviction rates and eviction filing rates in four California cities. A difference-in-differences matched case model suggests that there is a statistically significant, large, and negative difference between eviction rates and eviction filing rates before and after the passage of just cause eviction ordinances in the four treatment cities, as compared to the difference in these rates before and after the passage of just cause eviction ordinances in matched control cities. Cities that implemented just cause eviction laws experienced lower eviction, by 0.808 percentage points, and eviction filing rates, by 0.780 percentage points, than those that did not.

---

Introduction

The Eviction Lab at Princeton University recently compiled the first national database on evictions. The impetus for the database was the alarming findings of the 2016 book Evicted: Poverty and Profit in the American City by Matthew Desmond. Desmond drew on ethnographic research in Milwaukee, Wisconsin, as well as, data available from court records and his own “Milwaukee Area Renters Study” to highlight the large number of severely rent-burdened Americans facing chronic eviction. He characterized the issue of eviction, especially its effect on Black women, as equivalent to the phenomenon of mass incarceration on Black men (Desmond 2016). In other pieces, Desmond and his co-authors have identified a myriad of social problems that can be traced back to eviction: depression, poor health, and involuntary job loss (Desmond and Kimbro 2015, Desmond and Gershenson 2016).

Desmond’s work has inspired housing advocates and lawmakers to push for policies to address the eviction epidemic. Many solutions have emerged, ranging from those that focus on assisting tenants after the eviction process has begun (such as providing legal counsel to tenants in eviction court) to those that seek to address the supply of affordable housing.

One policy solution that has gained considerable attention that falls between these extremes is “just cause” eviction laws.

“Just cause” laws prohibit landlords from arbitrarily ending rental agreements and outline specific reasons for which landlords can evict tenants. 

Common just causes for eviction include: failure to pay rent, failure to abide by lease requirements, owner move-in, and owner seeking to permanently remove unit from housing market. Typically, without just cause eviction laws, landlords may evict tenants for any reason or no reason at all, with certain requirements only about the amount of notice that must be given. This status quo is called a “no-fault” eviction.

New Hampshire and New Jersey have statewide just cause eviction laws. But, most commonly, just cause laws are passed at the local level. Cities like Seattle, Washington and Berkley, California have had just cause eviction ordinances since the 1980s. In 2018, Philadelphia, Pennsylvania passed a just cause ordinance, while Boston, Massachusetts’ just cause measure was defeated after a three-year fight between housing advocates, property owners, and the City Council (Chakrabarti & Bologna 2018).

Given the push for these ordinances in response to the eviction crisis, this analysis seeks to test the effect of just cause eviction ordinances on eviction and eviction filing rates in the data made recently available by the Eviction Lab. Specifically, the analysis focuses on the impact of just cause eviction ordinances on eviction rates in cities where they were passed.

This paper begins with a brief literature review focused on eviction in the United States, followed by an outline of the research methodology of the paper and an analysis of the findings. It concludes with policy implications and limitations.

 

Literature Review

The Eviction Lab database was released in April 2018 but as of April 2019, no peer-reviewed research has been published using its content. However, press articles have used the Lab’s mapping tool and “eviction rankings.” Emily Badger and Quoctrung Bui of The New York Times pointed to the concentration of high eviction rates in the Deep South and the Rust Belt, focusing on Richmond, VA in particular (Badger and Bui 2018).

Among the few studies that examined eviction prior to Evicted, was J. Revel Sims’s exploration of eviction in Los Angeles between 1994 and 1999 (Sims 2015). Sims’s study identifies the causes of eviction. Sims selected four areas of Los Angeles that were eviction “hotspots” and identified the driving cause of each area’s evictions: government-sponsored gentrification-oriented development in Hollywood, government-sponsored speculative development in Downtown, real estate and finance intermediaries engaging in predatory practices in South Los Angeles, and immigrant growth machines in Koreatown.

Beyond the specific phenomenon of eviction, there is a long history of scholars studying housing displacement in the United States. Much of the early scholarship focused on the effects of urban renewal before turning towards the displacement wrought by gentrification. However, eviction specifically was not the focus of these studies.

Desmond’s Evicted and the articles that preceded its publication (Desmond 2012, Desmond and Kimbro 2015, Desmond and Gershenson 2016),  called attention to a national crisis, rather than individual city case studies of displacement.

The Eviction Lab’s database, in particular, depicts the magnitude of eviction at the national level. According to the Eviction Lab, twice as many people are evicted every day than die in car accidents (Brancaccio and Long 2018).

The Anti-Eviction Mapping Project (AEMP) has been compiling data on evictions in San Francisco since 2013. In a 2018 report in collaboration with Tenants Together, a coalition of local tenant organizations, AEMP published a statewide, California eviction report based on previously unreleased data from 2014 to 2017. The authors found that an average of 1.5 million people faced court evictions over the three-year period. They also found that evictions by-and-large happen quickly: 60 percent of cases were resolved within 30 days and in most counties, a large portion of eviction cases are settled because tenants do not file a response within the five-day deadline (Inglis and Preston 2018).

The actions of financial institutions can also increase eviction rates. For example, a June 2018 report by the AEMP and the California Reinvestment Coalition revealed how financial institutions drive evictions through “displacement financing.” This refers to the phenomenon by which banks loan to speculators, factoring in these borrowers’ tendency to flip residential buildings into higher rent buildings via eviction. Banks then loan out more money to these “serial evictors” by leveraging the property they originally flipped (Stein, McElroy, and Lashne 2018).

 

Another rare study of eviction in the United States was Elaina Johns-Wolfe’s analysis of eviction in Cincinnati and Hamilton County, Ohio between 2014 and 2017 (Wolfe 2018). Wolfe found that, similar to Desmond’s 2008 to 2009 study of Milwaukee, most landlords had legal representation in eviction proceedings, while the vast majority of tenants did not have representation. Wolfe also found that a small portion of landlords are behind the majority of evictions: of the 1,000 landlords that filed evictions in Hamilton County during the study period, ten were responsible for about 20 percent of all evictions. At the top of the list, with more than double the eviction filings of any other landlord, was the Cincinnati Metropolitan Housing Authority.

 

Research Design

This analysis uses data from the Eviction Lab at Princeton University. The Eviction Lab has collected, cleaned, and geocoded all recorded court-ordered evictions that occurred between 2000 and 2016 in the United States. The complete dataset consists of 82,935,981 court records from formal eviction records in 48 states and the District of Columbia. It is the most comprehensive dataset of evictions in the United States. These eviction records were combined with demographic information from the U.S. Census Bureau in order to add more detail about the individuals and communities affected. The database also includes state-reported, county-level information on landlord-tenant eviction cases filed in 27 states, the District of Columbia, and New York City.

The unit of analysis for this paper is the annual city eviction rate as well as the annual city eviction filing rate, as they are defined in the Eviction Lab database. The eviction filing rate is the ratio of the number of evictions filed in the city to the number of renter-occupied homes in the city. The filing rate counts all eviction cases filed in an area, including multiple cases filed against the same address in a single year. The eviction rate is the ratio of the number of eviction judgements in which renters were ordered to leave over the number of homes that received an eviction judgment. The eviction rate only counts a single address that received an eviction judgement.

This analysis will focus on cities that passed just cause eviction ordinances during the period covered by the Eviction Lab database (2000 to 2016). Just cause eviction ordinances differ in their scope across municipalities, so this analysis only considers ordinances that cover all rental units in a municipality. San Francisco, for example, is not included in this analysis, despite having a just cause eviction ordinance in effect. Its ordinance only applies to properties that have been issued a certificate of occupancy before 1979. Considering cities that have passed ordinances during the period covered by the database allows us to employ a difference-in-differences analysis.

While entire states have just cause eviction laws, this study focuses on California cities because the state has the highest incidence of cities with just cause eviction ordinances and only analyzing cities in the same state controls for factors that vary across states.

The cities that match these criteria are: East Palo Alto, Glendale, Oakland, and San Diego. Table 1 lists the years in which each city passed a just cause eviction ordinance.

Table 1 – Year Just Cause Ordinance Passed in Treatment Cities

City	Year Just Cause Eviction Ordinance Passed
East Palo Alto	2010
Glendale	2002
Oakland	2003
San Diego	2004

In the Eviction Lab dataset, the state of California consists of 21,390 cities. Their annual mean eviction and eviction filing rates are illustrated in Figure 1. Eviction filing rates are predictably higher than eviction rates. This is because not all evictions that are filed result in an executed eviction and, in the Eviction Lab dataset, the eviction rate only counts a single address once if that address received an eviction judgement. Both rates peak in 2008, coinciding with the onset of the Great Recession.

Figure 1 – California Eviction Rates and Eviction Filing Rates 2000-2016

Within the subset of the Eviction Lab dataset that covers the selected California cities, there are 5,127 missing values for eviction rate and eviction filing rate. There are no instances in which an observation has an eviction rate missing but not an eviction filing rate missing or vice versa. These missing values are concentrated in the period from 2000 to 2010.  The only city with missing values for eviction rate or eviction filing rate is East Palo Alto for the years 2000 and 2001. The analysis treats these observations as missing values rather than imputing them given the low number of missing values and the fact that they occurred a decade before the treatment (East Palo Alto’s just cause eviction ordinance was passed in 2010). There are no missing values for the demographic and economic variables.

In order to test whether just cause eviction laws are effective in addressing eviction rates, I will use a difference-in-differences, case-controlled model, in which each of the four cities with just cause eviction ordinances is matched with a control city or cities that are similar across a number of factors, except that they did not pass just cause eviction ordinances. Other models considered were a difference-in-differences model that analyzed eviction rates in municipalities bordering each other, where one municipality passed a just eviction ordinance and the other did not. This model was dismissed because differences across municipalities, such as income level and property value which can result from differing municipal laws, could not be controlled for.

The variables used to match control and treatment cities are:

• Mean population, 2000-2016
• Mean poverty rate, 2000-2016
• Mean percent renter occupied, 2000-2016[i]
• Mean median gross rent, 2000-2016[ii]
• Mean median household income, 2000-2016[iii]
• Mean median property value, 2000-2016[iv]
• Mean rent burden, 2000-2016[v]

These factors were chosen based on their likelihood to affect evictions and eviction filings. Population serves as a proxy for the size of the housing market, while mean percent renter occupied serves as a proxy for the size of the rental market. Gross rent and property value attempt to capture the costliness of the city, both for renters and homeowners. Finally, the rent burden and poverty rate serve as a proxy for the size of two populations that are the most vulnerable to eviction—those that are low-income and those that already spend more than 30 percent of their income on housing. While Black individuals disproportionately face eviction, according to Desmond’s study, race and ethnicity are not included as factors because it was unclear how the influence of race and ethnicity for specific groups translated to the California setting.

Some cities had only one close match, as in the case with Oakland. Others, such as San Diego and Glendale, had more than one matched city but additional matched cities were eliminated for a variety of factors. For example, San Francisco, which matched with San Diego, was eliminated because it has a just cause eviction ordinance that covers only part of the rental stock in the city. Garden Grove and Santa Rosa were also matches with Glendale but were eliminated due to missing eviction and eviction filing rate data in the early 2000s, the period in which Glendale passed its just cause eviction ordinance. Given the small sample size and availability of other match cities, these cities were eliminated rather than imputing eviction and eviction filing rates. As a result of this process, only East Palo Alto had multiple match cities. See Appendix 1 for Tables 1-4, which contain the observed values of these variables for matching treatment and control cities.

The treatment cities and their matched control cities are listed in Table 2. Figure 2 shows their eviction rates between 2000 and 2016.

Table 2 –Treatment Cities and Matched Control Cities

Treatment City	Matched Control City/Cities
East Palo Alto	Ashland Imperial Beach Lawndale Marina Seaside
Glendale	Chula Vista
Oakland	Long Beach
San Diego	San Jose

While the economic and housing factors used to match cities control for most factors, ideally the eviction and eviction filing rates between treatment and matched control cities would also be on similar trajectories before the eviction ordinances were passed. This is the case for Oakland, Glendale, and San Diego. In the case of East Palo Alto, the trend between its eviction rates and matched cities’ eviction rates is less similar. This suggests that there were other macroeconomic trends underway in these cities that are not controlled for. However, the fact that East Palo Alto has multiple match cities attempts to address this issue.

Figure 2 – Eviction Rates Across Treatment and Match Cities 2000-2016

The model used to compare the difference in eviction and eviction filing rates in treatment cities and matched control cities is: 

Where

ŷ is the annual eviction rate or annual eviction filing rate;

x₁ is a dummy variable for post-treatment (post-treatment = 1);

x₂ is a dummy variable for treated (treated = 1);

x₁ * x₂ is the difference-in-differences interaction between post-treatment and treated;

 γ is the year fixed effect.

The post-treatment variable is coded for the specific year each treatment city passed its just cause eviction ordinance, as each treatment city has a different year that delineates between pre- and post-treatment. The year fixed effects controls for additional factors that would have affected all cities in any given year, outside of those already controlled for through the matching process, such as state and federal policies or economic climate.

 

Findings and Discussion

The results of the difference-in-differences regression, given below in Table 3, are quite powerful. They suggest that passage of a just cause eviction ordinance has a negative statistically significant effect on eviction and eviction filing rates.

Table 3 – Difference-in-Differences Regression Analysis

Standard errors in parentheses  *** p<0.01, ** p<0.05, * p<0.1

The coefficient on post-treatment*treated suggests that there is a statistically significant (p < 0.01), negative difference between eviction rates before and after the passage of just cause eviction ordinances in the four treatment cities compared with the difference in eviction rates in the same period in matched control cities. Specifically, there is a -0.808 percentage point difference between eviction rates before and after the passage of just cause eviction ordinances in the treatment cities, as compared to the difference in eviction rates in matched control cities before and after the passage of just cause eviction ordinances in treatment cities. Given that the eviction rates in treatment and matched control cities range from 0.07 to 4.32 percent, the magnitude of this difference is remarkable.

The difference between eviction filing rates before and after the passage of just cause eviction ordinances in the treatment cities, compared to the difference in filing rates in matched control cities during the same time period is also statistically significant (p < 0.01), but of a smaller magnitude, -0.780 percentage points. The R-squared on both regressions (0.317 and 0.248) is rather high given the few variables included. This suggests that the treatment (passage of just cause eviction ordinances) explains much of the variation in eviction and eviction filing rates between matched control and treatment cities.

Despite these results, the overall trends in eviction rates in treatment and control cities in Figure 2 present a less straightforward picture. There is no significant drop in eviction rates in the treatment cities after the passage of the just cause eviction ordinance, except in the case of Glendale, nor do the trends among treatment and control cities change dramatically.

 

Policy Implications

These results suggest that just cause eviction ordinances have a significant and noticeable effect on eviction and eviction filing rates. Given the budget limitations of many states and municipalities to fund other solutions to the eviction crisis, passage of just cause eviction ordinances appears to be a relatively low-cost, effective policy solution.

Just cause eviction laws have received attention in major cities where residents are concerned about gentrification (e.g. Boston and Philadelphia). Rather than focusing exclusively on large cities, this study shows the desired effect on eviction rates in a diverse group of cities. There is evidence that smaller cities and more suburban communities are starting to also consider just-cause eviction ordinances, such as Montgomery County, Maryland. This is particularly salient given the national trend of impoverished populations moving from urban to suburban communities (Kneebone and Berube 2014).

While this study focused on just cause eviction ordinances on their own, these laws are frequently paired with rent-control laws. This is because just cause eviction ordinances protect residents from being evicted arbitrarily, but do not protect tenants if they are unable to afford their rent payments.

In a context of rapidly rising housing prices, just cause eviction ordinances on their own do not fully protect a city’s most vulnerable residents.

Limitations

The results of this analysis provide a compelling case for the passage of just cause eviction ordinances to lower eviction rates. However, there are several limitations to this analysis.

Several housing justice organizations, among them the AEMP, raised concerns with the Eviction Lab data in an August 2018 Shelterforce piece (Aiello et al 2018). They called attention to the efforts of local organizations in a number of cities to gather data on evictions, and that due to a lack of transparent and equitable relationship-building on the part of the Eviction Lab, this data collected by local organizations was not included in the final dataset. This has resulted in vastly different figures on eviction rates for these cities. For example, the authors cite that the data from AEMP and Tenants Together show that over 3,000 evictions occurred in San Francisco in 2014 and 2015, more than double the Eviction Lab’s count.

More generally, evictions also occur outside of the judicial system. These informal evictions—landlords induce renters to leave through monetary incentives or illegal lockouts—are not accounted for in the dataset or this analysis. According to the Eviction Lab, these informal evictions are likely more common than the court-ordered evictions included in the dataset.

It is possible that extra-judicial evictions increase after just cause eviction ordinances are passed, rendering the true effect of these laws smaller or close to zero. 

This issue can be addressed by using information obtained through qualitative interviews with tenants, such as the data collected by AEMP and Tenants Together. Unfortunately, the more comprehensive data collected by these organizations on California only covers a period from 2014 to 2016. This time period is to narrow to allow for the same type of analysis conducted in this paper.

In addition, this paper’s methodology used matched control cities across a number of relevant variables, but this list of variables is hardly exhaustive. In particular, this analysis did not control for demographic characteristics. This analysis also did not compare the effects of different just cause eviction ordinances, or the differences between cities with only just cause eviction ordinances and those that also couple these with rent control ordinances. This study is also limited by the fact that it only focuses on California. Due to particularities of the California housing market and state laws, these findings may not be generalizable to the rest of the country. Finally, this analysis does not allow for a comparison of the effect of just cause eviction ordinances across states—which could be useful given how state laws affect and interact with local housing laws.

 

Conclusion

This analysis of eviction filing rate and eviction rate data in California is unprecedented. It suggests that just cause eviction ordinances have a statistically significant negative effect on eviction and eviction filing rates. Specifically, eviction rates dropped by 0.808 percentage points and eviction filing rates dropped by 0.780 percentage points after passage of just cause eviction ordinances.

Future studies of just cause eviction ordinances should consider how the effectiveness of this policy solution is amplified or altered by its combination with other measures such as rent control and other local and state laws.

Future studies that seek to recommend action for a particular municipality considering a just cause eviction ordinance. should start by seeking out data collected or reports published by local housing organizations. Local groups have local knowledge and context that can increase the accuracy of collected data.

At the moment, this analysis is limited by the number of cities that have passed these ordinances. As the Eviction Lab database grows, data collection by local organizations such as AEMP increases, and more cities pass just cause eviction ordinances, this kind of analysis can be expanded and deepened to better understand the effects of local laws on eviction. With more data from cities and states, we can begin to compare their effects in different housing markets and different parts of the United States.

---

About the Author

Julieta Cuéllar is a 2019 Master in Public Affairs graduate of the Woodrow Wilson School at Princeton University. She would like to thank Prof. Martin Gilens, Sanata Sy-Sahande, Graham Simpson, and the JPIA editors for their assistance in this project. She can be reached at [email protected] (Linkedin Profile)

Notes 

References

Aiello, Daniela; Bates, Lisa; Graziani, Terra; Herring, Christopher; Maharawal, Manissa; McElroy, Erin; Phan, Pamela and Gretchen Purser. “Eviction Lab Misses the Mark.” Shelterforce, August 22, 2018.

Badger, Emily and Quoctrung Bui. “In 83 Million Eviction Records, a Sweeping and Intimate New Look at Housing in America.” The New York Times, April 7, 2018.

Brancaccio, David and Katie Long. “Millions of Americans Are Evicted Every Year—and Not Just in Big Cities.” Marketplace, April 9, 2018.

Chakrabarti, Meghna and Jaime Bologna. “How Boston’s Big Attempt at Rental Law Reform Failed.” WBUR, May 16, 2018.

Desmond, Matthew. Evicted: Poverty and Profit in the American City. New York: Crown, 2016.

Desmond, Matthew. “Eviction and the Reproduction of Urban Poverty.” American Journal of Sociology 118, no. 1 (2012): 88-133.

Desmond, Matthew and Carl Gershenson. “Housing and Employment Insecurity Among the Working Poor.” Social Problems 63, no. 1 (2016): 46-67.

Desmond, Matthew and Rachel T. Kimbro. “Eviction’s Fallout: Housing, Hardship, and Health.” Social Forces 94, no. 1 (2015): 295-324.

Johns-Wolfe, Elaina. “‘You are Being Asked to Leave the Premises’: a Study of Eviction in Cincinnati and Hamilton County, Ohio, 2014-2017.” Legal Aid Society of Cincinnati, 2018, https://www.lascinti.org/wp-content/uploads/Eviction-Report_Final.pdf.

Kneebone, Elizabeth and Alan Berube. Confronting Suburban Poverty in America. Washington, D.C.: Brookings Institution, 2014.

Inglis, Aimee and Dean Preston. “California Evictions are Fast and Frequent.” Tenants Together, May 2018, https://static1.squarespace.com/static/52b7d7a6e4b0b3e376ac8ea2/t/5b1273ca0e2e72ec53ab0655/1527935949227/CA_Evictions_are_Fast_and_Frequent.pdf

Sims, J. Revel. “More than Gentrification: Geographies of Capitalist Displacement in Los Angeles 1994 – 1999.” Urban Geography 37, no. 1 (2015): 26-56.

Stein, Kevin; McElroy, Eric and Carla Leshne. “Disrupting Displacement Financing in Oakland and Beyond.” California Reinvestment Coalition, June 2018, https://static1.squarespace.com/static/52b7d7a6e4b0b3e376ac8ea2/t/5b1eb1950e2e72b49f5a1679/1528738430348/Disrupting+Displacement+Financing.pdf

Appendix 1

Table 1 – Descriptive Variables for East Palo Alto and Matched Cities*

Variable	East Palo Alto	Ashland	Imperial Beach	Lawndale	Marina	Seaside
Mean Population	30,280	22,144	26,781	32,194	21,042	33,127
Mean Poverty Rate	15.73	16.02	14.55	14.66	11.29	11.05
Mean Percent Renter Occupied	59.17	65.61	68.71	67.22	58.32	57.73
Mean Median Gross Rent	1,137.94	1,030.12	1,008	1,128.71	1,005.47	1,229.71
Mean Median Household Income	48,521.88	45,409.29	43,501.41	44,783.65	49,522.41	51,261.53
Mean Median Property Value	427,658.80	311,247.10	334,476.50	339,964.70	392,417.60	389,652.90
Mean Rent Burdened	34.88	31.78	31.98	33.45	30.39	33.16
Mean Percent White	8.18	19.23	38.25	18.08	37.40	34.16
Mean Percent African American	18.02	18.20	4.32	9.26	8.82	9.76
Mean Percent Hispanic	60.20	39.90	45.88	59.87	26.25	39.09
Mean Percent American Indian	0.21	0.53	0.45	0.25	0.36	0.48
Mean Percent Asian	3.29	17.63	6.90	9.10	17.65	9.76
Mean Number of Renter Occupied Households	3,992	4,794	6,444	6,431	3,839	5,854

 

Table 2 – Descriptive Variables for Glendale and Matched City*

Variable	Glendale	Chula Vista
Mean Population	195,757	219,571
Mean Poverty Rate	12.31	9.77
Mean Percent Renter Occupied	62.53	41.23
Mean Median Gross Rent	1,105.41	1,075.47
Mean Median Household Income	49,979.59	58,452
Mean Median Property Value	538,158.8	354,876.50
Mean Rent Burdened	35.35	34.03
Mean Percent White	59.85	24.23
Mean Percent African American	1.36	4.34
Mean Percent Hispanic	18.29	55.07
Mean Percent American Indian	0.18	0.25
Mean Percent Asian	16.19	12.76
Mean Number of Renter Occupied Households	44,879	30,071

 

Table 3 – Descriptive Variables for Oakland and Matched City*

Variable	Oakland	Long Beach
Mean Population	401,797	465,024
Mean Poverty Rate	16.88	17.93
Mean Percent Renter Occupied	58.51	58.81
Mean Median Gross Rent	963.65	940.88
Mean Median Household Income	48,714.18	47,420.47
Mean Median Property Value	415,723.5	394,200
Mean Rent Burdened	31.05	31.63
Mean Percent White	25.76	30.21
Mean Percent African American	29.24	13.35
Mean Percent Hispanic	24.46	39.52
Mean Percent American Indian	0.35	0.32
Mean Percent Asian	15.62	12.45
Mean Number of Renter Occupied Households	91,537	96,733

 

Table 4 – Descriptive Variables for Oakland and Matched City*

Variable	San Diego	San Jose
Mean Population	1,298,308	946,935
Mean Poverty Rate	11.30	8.10
Mean Percent Renter Occupied	51.17	40.27
Mean Median Gross Rent	1,148.18	1,357.65
Mean Median Household Income	58,774.18	78,455.65
Mean Median Property Value	415,447.10	554,741.20
Mean Rent Burdened	31.04	30.04
Mean Percent White	46.52	31.19
Mean Percent African American	6.72	2.96
Mean Percent Hispanic	27.81	31.72
Mean Percent American Indian	0.29	0.25
Mean Percent Asian	14.98	30.54
Mean Number of Renter Occupied Households	246,114	121,737

* Italicized variables were not used to match treatment and control cities but are listed here for additional comparison.

Table 5 – Difference-in-Differences Regression Analysis: East Palo Alto & Match Cities

 

Table 6 – Difference-in-Differences Regression Analysis: Glendale & Match City

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Table 7 – Difference-in-Differences Regression Analysis: Oakland & Match City

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

 

Table 8 – Difference-in-Differences Regression Analysis: San Diego & Match City

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1