Estimating COVID-19 Induced Rental Housing Hardship in 2021

Philip Garboden, Blogs, COVID-19, Housing,

BLOG POSTS ARE PRELIMINARY MATERIALS CIRCULATED TO STIMULATE DISCUSSION AND CRITICAL COMMENT. THE VIEWS EXPRESSED ARE THOSE OF THE INDIVIDUAL AUTHORS. WHILE BLOG POSTS BENEFIT FROM ACTIVE UHERO DISCUSSION, THEY HAVE NOT UNDERGONE FORMAL ACADEMIC PEER REVIEW.

By Philip ME Garboden and Isabelle Picciotto

Several months ago, UHERO estimated the impact of the COVID-19 induced economic downturn on renters in 2020. Much has changed since then, and generally not for the better. New outbreaks, both in Hawai’i and on the mainland, have worked against our hope for a steady and predictable economic recovery. As the Federal Government continues to delay passage of a second round of relief, it is prudent to update our simulation through the end of 2021.

As before, we utilize de-identified microdata from the American Community Survey (via iPUMS) to create a representative portrait of the State’s households that accurately reflects the number of multi-earner households, annual income, and employment by sector. We then randomly distributed UHERO’s forecasted 2021 job losses over the workers in each industry. After assigning job losses, we recalculated each household’s housing burden and report the percentage of households for whom COVID has jeopardized housing stability. To avoid aberrant random draws, we ran this process 100 times and averaged the results.

We assess four policy scenarios:

  1. Worst Case: In our worst-case scenario, we assume that households continued to receive regular or extended unemployment insurance (UI) benefits through the end of 2021 Q1 and nothing thereafter. While we believe that this scenario is unlikely, it represents the need in the absence of any additional intervention.
  2. UI Extension: In our second-to-worst-case scenario, we assume that households receive some form of extended unemployment benefits through the end of Q3.
  3. Small Boost: In the third scenario, we assume both an extension of UI benefits through Q3 and a modest $100 weekly increase in addition to those benefits.
  4. Modest Boost: In the fourth scenario, we assume an extension of UI benefits through Q3 and a $300 per week increase in addition to those benefits.

Two caveats are in order.

First, much remains unknown regarding Hawai’i’s economic future and all of our findings should be interpreted only as rough estimates. The pace of our State’s economic recovery depends on local, national, and international disease prevalence, and multiple unknowns regarding Federal assistance. Our goal is not to predict the future, but to provide a rough guide for contingency planning.

Second, our approach can only consider the impacts of job losses in the formal economy. It cannot consider informal income, support from friends and family, investment losses, or the effects on small business owners. We also do not assess pre-COVID needs, which were substantial. Our numbers are thus conservative.

All of the code used for this analysis is publically available on GitHub (including additional tables and figures). Our approach required us to make many assumptions. Those who disagree with those assumptions or are curious about how much they matter are encourage to branch our code and rerun our analysis (let us know what you find).

Results of the Analysis: Significant and Persistent Need Throughout 2021

Our primary metric is COVID induced changes in housing cost burden (the change in the ratio of housing costs to income since before the pandemic). Prior to the pandemic, many renter households already suffered from what housing researchers consider to be deleteriously high burdens, forcing them to underinvest in other forms of well-being simply to maintain their housing. While this is undeniably a problem in its own right, our focus is on households experiencing a drastic increase in their burdens, which they could not have anticipated when selecting their housing and are thus most likely to induce involuntary mobility.

In a nutshell, we project a significant number of households to have unsustainable changes in their housing cost burdens in all four quarters.

Despite the presence of standard UI benefits, the Scenario 1 Q1 results show 11,000 households with burden increases of at least 20%, 6,000 of which experience burdens over 40%. Given that Scenario 1 assumes UI expires after Q1, the need jumps sharply, to 17,000 households in Q2 despite UHERO’s projected job gains during that time. As the economy regains jobs over the rest of the year, the need falls, but not below 11,000 households.

Scenario 2 tells a similar story, but the need is moderated by the continuation of UI benefits through Q3. Scenarios 3 and 4 similarly reflect the benefits of additional income support, which reduces the need for targeted housing support. Nevertheless, even the most generous set of policy assumptions — a UI extension plus a $300 weekly boost — still projects 6,000 renter households in need of support after three quarters of economic recovery.

Rental Burden Change2021 Q12021 Q22021 Q32021 Q4
Scenario 1
No Change159,631159,633164,223167,852
1-10% Change 11,8666,3344,9153,847
10-20% Change4,4693,4222,6532,096
20-40% Change5,0123,7512,9792,376
40-200% Change6,33713,68111,0498,865
Scenario 2
No Change159,633161,643165,761167,838
1-10% Change11,86510,5668,0563,852
10-20% Change4,4644,0693,2482,099
20-40% Change5,0164,5563,6142,369
40-200% Change6,3385,9995,1478,866
Scenario 3
No Change161,224163,046166,816167,845
1-10% Change11,55210,3087,8983,853
10-20% Change4,7814,3303,3792,084
20-40% Change4,2863,9273,1562,364
40-200% Change5,4565,2094,5638,872
Scenario 4
No Change164,043165,530168,656167,843
1-10% Change11,1019,8797,5333,858
10-20% Change4,5814,1833,3132,092
20-40% Change4,6724,3703,6822,372
40-200% Change2,9272,8802,6458,854

As shown in the additional figures displayed on the GitHub page, these trends are largely consistent regardless of how we slice the simulation data. For example, our simulations suggest that in the absence of policy intervention, 10,000 additional households will find themselves newly burdened by the housing (meaning they now pay >30% of their income in rent).

Conclusion

These findings suggest several policy takeaways. First, it is clear that doing nothing is not a viable option. In the absence of additional support, at least 10,000 families will be in perilous housing situation in addition to the normal strains of Hawai’i’s housing market. With or without Federal assistance, we must find a way to support these households. While their well-being is of primary importance, failing to provide aid will likely also have long-term consequences for the State’s rental market if these job losses are translated into losses in the rental housing stock via conversion or sale.

The data also suggest that some households’ need will be quite extended, lasting beyond the end of 2021. This suggests a policy intervention that moves beyond a single influx of cash assistance (as welcome as that might be) to one that adjusts to household needs as economy recovers in unpredictable ways.