The latest Census Bureau Household Pulse Survey (HPS) was widely reported on with a headline takeaway that close to four million residents are likely to be evicted in the next two months. Also commonly mentioned in articles covering this latest release was August evictions data from Eviction Lab. Though not a common topic for the ALN blog, we thought we’d dig into the latest Household Pulse Survey a little bit.

Limitations of the Census Data Set

To begin with, some caveats regarding the data itself need to be laid out. For one, the Census Bureau makes it very clear in the data files that this survey data is experimental, and that caution should be used with the data due to potential sample size issues and large error terms. For this reason, this post will only utilize the national data because the state and metro-level subpopulation data is likely to be even more problematic.

Limitations in Reporting of Eviction Lab Data

As already mentioned, many of the articles covering the HPS also quoted August evictions data for specific markets from Eviction Lab showing increases between 50% and 100% in eviction filings for the month compared to their average. Some articles even quoted those jumps not as increases in filings but as increases in actual evictions – even though the data set tracks filings. Eviction Lab is a useful source despite being limited in geographic scope. However, there are at least two issues with quoting monthly comparisons before the end of August.

One is that regular consumers of Eviction Lab data will have noticed a common dynamic in the data. For a current month, early in the month eviction filings seem to spike compared to the historical average, but by the end of the month the spike has mostly or entirely disappeared relative to the average. This may be related to the methodology of the data aggregation and reporting, or it may be eviction filings have been especially front-loaded in months recently.

When the latest HPS was covered by various outlets, and still as of this writing a few days later, the latest August data available from Eviction Lab is August 6. In other words, these eye-popping growth numbers relative to the average are likely to dissipate somewhat in subsequent data releases for August.

Another issue has to do with reporting based on percent change relative to historical average filings versus reporting the number of reported filings. Dallas was a common example used in many articles, with August filings reported as being 200% higher than the historical average. It might be surprising to learn that those same August filings when observed as just total filings put the month on pace for the fewest filings since December of 2021.

“Likely to be Evicted Within Two Months”

This category of the Household Pulse Survey was generally used in headlines and garnered the most discussion and attention. An estimated 3.8 million people fell into the Very Likely or Somewhat Likely category for near-term eviction. Of that total, about 1.2 million were in the Very Likely category.

This is an estimate extrapolated from a sample of self-reported data and it is unclear whether respondents were referring to eviction filings already in progress or concern about eviction just due to a delinquent payment status. Making this more convoluted is the fact that by far the largest share of people in this category indicated being only one month behind on rent payments. 40% of those expecting to have to vacate within two months reported being only one month behind on rent. While no amount of rent distress is desirable, the alarmist nature of the headlines undermines the variety of issues facing renters and owners.

Perhaps as a result of not being majorly delinquent, half of the 3.8 million people estimated to be in this near-term displacement category are reflected in the data as having not applied for any state or local assistance programs. Among those that did apply, the largest share was reported in the data as still waiting on a response. For those that received a response, more households out of the 3.8 million were estimated to have been turned down for assistance than received it.

Characteristics of ‘At-Risk’ Renters

Of the approximately 3.8 million residents represented in the data, it is the prime working-aged cohort of 25-54 that accounts for the largest share. The 40 to 54 group edged out the 25 to 39 group by about 100,000 people and the two groups combined for three million of the 3.8 million.

It is the population with a high school diploma or GED as their highest educational attainment that are most represented in the data, outpacing even those without a high school diploma. Those with at least a bachelor’s degree make up the smallest subset in this at-risk pool.

Related to the educational attainment is the disparity between income levels. According to the HPS data, this near-term eviction risk is predominately associated with households earning below $25,000 per year. This category accounts for 42% of the 3.8 million residents. The share residents accounted for in each subsequent income category declines until the $150,000 and up subsets in which there are no estimated Very Likely eviction responses. This aspect of the data points to a more concentrated effect of possible eviction pressure from a price class perspective.

Interestingly, slightly more than half of the 3.8 million estimated residents were reported as having not lost employment income in the last four weeks – including other householders. This would seem to indicate a mix of people who expect to face eviction despite having maintained their income as well as a population that lost their employment income more than four weeks ago.

Takeaways

Without a high-quality nationwide data set on eviction filings, it can be hard to get a grip on expectations outside of anecdotes and monitoring occupancy data. This difficulty seems to have little effect on reporting of the topic. The Census Bureau Household Pulse survey is a useful and free tool, but one that should be used appropriately based on its shortcomings. The same can be said for Eviction Lab data and other sources making valuable attempts to close this knowledge gap for the industry.

Based on a more detailed look at the latest HPS, which had an unusually low response rate by its own recent standard, it seems reasonable to expect a problem less acute than the headlines and coverage related to the data expressed. Reasons for this expectation have been discussed, and include overall data quality questions, ambiguity around the self-reporting of eviction concern, two-thirds of the 3.8 million residents falling into the Somewhat Likely category for near-term eviction versus Very Likely, 40% of the estimated population reporting only being one month behind on rent, and others.

All of this is not to deny that the multifamily industry faces increased affordability challenges or that it is likely to continue dealing with the unintended consequences of artificially deflated evictions during the pandemic period. But perhaps a grain of salt is called for.

Disclaimer: All content and information within this article is for informational purposes only. ALN Apartment Data makes no representation as to the accuracy or completeness of any information in this or any other article posted on this site or found by following any link on this site. The owner will not be held liable for any losses, injuries, or damages from the display or use of this information. All content and information in this article may be shared provided a link to the article or website is included in the shared content.