Eviction Lab updates December 15, 2020

Eviction Moratoria have Prevented Over a Million Eviction Filings in the U.S. during the COVID-19 Pandemic

  • Joe Fish, Emily Lemmerman, Renee Louis, and Peter Hepburn
  • The Eviction Lab

America faced a rental housing crisis long before the onset of the COVID-19 pandemic. In a typical, non-pandemic year, landlords would file 3.7 million eviction cases across the United States. As a result of eviction moratoria, court closures, and temporary expansions to the social safety net, far fewer eviction cases have been filed in 2020. How many fewer? And what does that tell us about the potential for a surge in eviction filings in early 2021?

Leveraging data drawn from the Eviction Tracking System (ETS), we quantify “missing” eviction filings as the difference between the number of cases that likely would have been filed in a typical year and the number that were actually filed during the pandemic. All told, we estimate that protections rolled-out during the pandemic have prevented at least 1.6 million eviction filings across the United States, cases that—in the absence of further eviction protections—may be pushed into 2021. This would create a surge in housing courts and putting an untold number of families at risk of displacement.

We estimate that protections rolled-out during the pandemic have prevented at least 1.6 million eviction filings across the United States, cases that—in the absence of further eviction protections—may be pushed into 2021.

Focusing on the 18 cities and six states we observed in the ETS, in Figure 1 we plot total weekly eviction filings as a percentage of historical average.1 For nearly every week since the pandemic began, filings have been far below average.

Figure 1. Weekly Eviction Filings in Eviction Tracking System Sites Relative to Historical Average
figure_1_20201204 Created with Sketch. 0% 25% 50% 75% 100% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov CASELOAD RELATIVE TO HISTORICAL AVERAGE

Weekly eviction filings are aggregated across sites tracked by the Eviction Tracking System (ETS) as of 11/30/2020. Historical averages are calculated for the same seven-day periods in previous years. The vertical red dashed line in mid-March marks the start of the COVID-19 pandemic. The bars in orange are the two weeks between the CARES Act eviction moratorium and the CDC eviction moratorium when no federal moratorium was in effect.

Between March 15th and November 30th of a typical year, we would expect to see 520,000 eviction filings across the current ETS sites. This year we have observed 172,000 filings, or roughly one-third of normal. That leaves nearly 350,000 missing filings during the pandemic. Missing filings can be understood as a measure of the overall success of emergency eviction bans at the local, state, and federal levels, as well as CARES Act stimulus payments and unemployment expansion. The spike in evictions witnessed in August between the time the CARES Act eviction moratorium expired and the CDC moratorium went into effect (symbolized by the orange bars in Figure 1) indicate that what kept evictions lower-than-average during the pandemic was legal prohibitions on displacement.

Figure 2. Overall Missing Filings across ETS Sites
agg_fig (1) Created with Sketch. 172,000 0 % 25 % 50 % 75 % 100% FILINGS SINCE MARCH 15 Observed Filings 348,000 Missing Filings

These missing filing accrued most quickly in the early months of the pandemic. In Figure 3 we plot the cumulative number of eviction filings across all ETS sites over the course of 2020, as well as the historical average.

Figure 3. Cumulative Eviction Filings across ETS Sites in 2020 and in Historical Average
figure_3_20201204 Created with Sketch. Jun May Apr Mar Feb Ja n Dec Nov Oct Sep Aug Jul 0 200,000 400,000 600,000 CUMULATIVE NUMBER OF FILINGS 2020 HISTORICAL AVERAGE

Prior to the start of the pandemic, eviction filings in 2020 across all ETS sites were almost exactly on pace with the historical average: about 50,000 cases filed each month. Between March and the end of July—when the CARES Act was in place and all sites were, for at least some time, under state or local eviction moratoria—new filings fell to an average of 2,100 per month, or 4.2% of the normal rate.

Since the start of September—with the CDC eviction moratorium in place, but with more state and local eviction moratoria expired—this rose to over 7,000 cases filed per month. While this rate remains significantly below historical average, it is nonetheless over triple the rate of new filings observed when the CARES Act and the local and state moratoria were at their strongest. Strict moratoria reduced eviction filings dramatically. As protections have been weakened, more cases have been filed.

The share of missing eviction filings varies considerably across cities and states covered by the ETS. We explore this variation in Figure 4, plotting the percentage of expected cases that have been filed (the blue bar) and the percentage of missing filings (the red bar). The number in white is the absolute number of missing filings.

Figure 4. Distribution of Missing Filings across ETS Sites

Some cities and states have far more successfully reduced eviction filings than others. Jurisdictions in the cities in our sample have enacted a wide variety of protections for renters during the pandemic. Some of these measures were extremely effective in reducing eviction filings. For example, in Boston—subject to Massachusetts’s statewide eviction moratorium between April 20th and October 17th—eviction filings since March 15th were at less than 10% of historical average. Houston, by contrast, was subject to a much shorter eviction moratorium that did allow for new filings. In that city, we observed nearly half the number of normal filings.

The differences in states’ success in reducing eviction filings is even more striking. Minnesota has had a statewide eviction moratorium in place since March 16th; only 10% of normal cases have been filed. Neighboring Wisconsin ended their moratorium on May 26th and has seen almost 60% of normal filings since the start of the pandemic.

ETS data collection covers 18 cities and six full states, home to one in every four renter households in the United States. Housing markets and socio-demographic composition of ETS areas resemble those of the United States as a whole (see Appendix Table). The majority of renting households are, however, absent from our calculations of missing eviction filings. To provide a national estimate of the total number of missing eviction filings during the COVID-19, pandemic we turn to regression analysis, extrapolating what we observed in the ETS to all other counties for which we have historical eviction filing data.2 This allows us to cover an additional 27.5 million renter households, bringing overall coverage to 38.5 million out of 43 million renter households (89.5%).

Strict moratoria reduced eviction filings dramatically. As protections have been weakened, more cases have been filed.

Extrapolating to the rest of the nation from the 18 cities and 6 states in our sample requires significant assumptions. We estimate that there have been 800,000 eviction filings across all non-ETS counties between March 15th and November 30th.3 In a typical year over this period, these counties would have experienced over 2.1 million eviction filings, thus leaving 1.3 million missing filings. Across the country, as many as 1.6 million eviction cases may be missing from 2020—and potentially pushed to early 2021.

On January 1, 2021, when the CDC eviction moratorium expires, renters will have the fewest protections available to them since the start of the COVID-19 pandemic. At the time of writing, only eight states have moratoria that are scheduled to extend into the new year.4 Eviction moratoriums were effective at lowering the rate of eviction filings during the pandemic. However, eviction filings are rising as we approach the end of the year, and tenants facing the threat of eviction appear to be falling further behind on rent.

Representativeness of ETS:
ETS U.S.
Estimate SD Estimate SD
Number of Renter HHs 11,076,250 17,124 43,669,988 21,378
Median Rental Housing Age 64.5 167 49.6 86.1
Median Rent ($) 1105 293 1,076 317
Percent Children Renting HH 30.20% 0.09% 32.80% 0.03%
Percent Female Renting HH 18.80% 0.05% 19.00% 0.03%
Poverty Rate 24.50% 0.01% 25.60% 0.00%
Vacancy Rate 10.80% 0.03% 11.10% 0.01%
Renting Householder Head Race
Black 22.30% 0.05% 19.70% 0.02%
Latinx 14.80% 0.04% 19.80% 0.02%
Other 6.70% 0.00% 7.80% 0.05%
White 56.10% 0.06% 52.60% 0.02%

Note: With the exception of the first row, all estimates are means. Means for median rent and rental housing age were calculated using a weighted mean, weighted by number of renting households for the county level. Standard deviations were calculated using the margin of error provided by the Census Bureau for the American Community Survey.



  1. Historical averages vary between sites. A full listing of baseline years can be found here. Please also see the ETS methodology report.
  2. Data collected from ETS sites can be used to predict eviction filing patterns in the rest of the U.S. To do so, we fit a Poisson regression model in areas covered by the ETS predicting county-week eviction filings in 2020 as a function of historical eviction filing patterns, socio-demographic factors, and pandemic-related policy interventions. Specifically, the predictors were poverty rate; number of renter households; historical eviction filing numbers for the county-week; an indicator of whether or not the state was under a moratorium during that week; and an interaction between the final two variables. We use statewide indicators for eviction moratoria and make no distinction between different types of moratoria. We use robust standard errors and cluster at the county level. We utilized the trained model to predict out of sample the number of filings that occurred in each county for which we had valid baseline data. Once we generated these predictions, we aggregated by county and compared the cumulative number of predicted filings with the historical baseline. We take this difference to be the number of “missing” filings within each county. Finally, we aggregate by county to create the total number of missing filings across all counties.
  3. The 95% confidence interval for this prediction is (762,000, 853,000) and is based on robust standard errors, clustered at the county level
  4. These states are California, Illinois, Kansas, Minnesota, Nevada, New Jersey, Vermont, and the District of Columbia. This number may rise, however, if other states such as Connecticut, Hawaii, Montana, New Mexico, New York, Oregon, and Washington continue their moratoria, which are either set to expire by January 1 or have no clearly stated expiration date.
Filed under research
rendered by = themes/THEMENAME/layouts/updates/single.html