Getting Past “Yes”: A Q&A on the Affordability Crisis (Part 3)
Many of the questions we keep encountering about the ambitious city-wide upzoning proposal called City of Yes for Housing Opportunity stem from a basic misunderstanding. Most people take “City of Yes” to be plain English, when, in fact, it’s government-speak for “City of No political will to meaningfully address the need for affordable housing.” In Part 1 of this series, we tackled the City’s efforts to pass off the ongoing affordability crisis as a general housing crisis, explaining the difference between the two and contrasting the ample evidence for the former with the mixed evidence for the latter. In Part 2, we refuted the City’s contention that there is widespread consensus among housing scholars and advocates that increasing housing supply at all income levels will improve housing affordability. Today, we explain in greater detail the limitations of this supply-based approach, which lies as the heart of the City of Yes proposal.
Do increases in housing supply make housing prices drop?
In Part 2, we saw that research into the impact of upzonings on housing prices has yielded mixed results. Some studies find an association between the relaxation of zoning constraints and increases in housing production; others find none. Some research identifies no association between upzonings and housing price changes; other research finds a positive associations between the two; and yet others find a negative one. If we now take upzonings out of the equation and just look at studies that focus squarely on the relation between increases in housing supply and in housing prices, the picture doesn’t get any clearer. Research findings on the matter are decidedly mixed, reflecting the contingent relation between these two variables.
Some evidence suggests that new residential development has disparate local effects on housing prices in different segments of the housing market. Damiano and Frenier’s (2020) Minneapolis study, for one, finds an association between new residential development and a 3.2% decrease in rents within 300 meters of new construction, but only in the highest-priced tertile of the market. In the lowest-priced tertile, the association is with a 6.6% increase. From a policy perspective, we should be primarily concerned with impacts on the bottom of the housing market, which serves most of the city’s rent burdened population. As pointed out in Part 1, the typical New York City household in 2023 paid 29.5% of its income towards rent—below the threshold of what might be classified as rent burdened. It was households earning less than the median income of $70,000 that had a median rent burden of 54%. This shows that it’s not households at the top of the income spectrum who need a break on rents. And yet, as Damiano and Frenier (2020) demonstrate, those households can be the primary (and exclusive) beneficiaries of increases in market rate housing.
Studies into the effect of new residential development on housing prices down-market tend to have a long term focus, because it is well understood that private markets do not upon construction directly provide housing for low income households in the absence of subsidies. Instead, it is hypothesized that new residential development indirectly increases housing affordability in two ways: 1) by luring higher income households out of lower-rent housing that they would otherwise occupy, and 2) by adding to the lower-rent housing stock as it deteriorates with age and becomes less desirable. Estimates of the rate at which these dynamics unfold differ. The rate of depreciation has been uniformly found to be low. A study by Margolis (1982), for instance, estimates the depreciation rate of rental housing at .39% per year. Bond and Coulson’s (1990) estimate is even lower. Assessments of rate at which rental units filter down to lower-income households, on the other hand, have varied more widely. An influential national study by Rosenthal (2014), for one, calculates that, in the absence of price inflation, rental units will filter down at a rate of 2.5% per year. That rate decreases in inverse relation to the rate of inflation. Given a more realistic (but still optimistic) 2% rate of inflation, for instance, the filtering rate for rental housing would drop to .5%.
To roughly illustrate the policy implications of findings such as Rosenthal’s, we will apply the filtering rate set forth in that study to the case of New York. Let us assume that a new rental unit enters the market renting for $3,500/month, which is the median asking rent for existing units (Comptroller’s Office, 2023). Assuming a rent burden of 30% (the median rent burden in 2023), that unit will be occupied by a household earning $140,000/year. According to the 2023 Housing and Vacancy Survey, there are about 856,800 rent burdened households in New York City. About 780,000 of them earn less than $100,000/year. At a filtering rate of .5%, it would take 58 years before the wealthiest of those households occupies our hypothetical unit.
Let us now assume that 25,000 new market rate units are produced every year going forward. This figure is based on the typical yearly number of permits authorized for new residential construction over the past two decades (Rent Guidelines Board, 2024, p.5). New housing permits provide a measure of how many housing units will be ready for occupancy within about three years of issuance. At that pace, it would take about 34 years for the number of new housing units in the market to equal the number of rent-burdened households earning less than $100,000/ year and waiting for those units to, someday in the distant future, filter down to them.(1)
Slow and steady wins the race, no?
No. Even if having to wait around for decades were not an issue, the plausibility and applicability of Rosenthal’s filtering rate still depends on several unlikely premises.
First, we cannot pretend that all new residential units would constitute net additions to the housing stock and that they would not occasionally take the place of existing housing units, some of which may be more affordable than their newly constructed counterparts. The elimination of such units can add to the need for affordable housing that the filtering of new units would ostensibly alleviate years or decades down the line.
Second, evidence suggests that the filtering rate, decreasing substantially in tighter ones (Liu et al, 2020). To give you a sense of the implications of this finding, consider that if the filtering rate in New York were a mere .25% lower, it would take over 100 years for our hypothetical new unit to be occupied by a household earning $100,000/year.
Third, we cannot assume that new housing units will continue to filter down at a steady rate until they become available to low-income households. Evidence shows that, beyond a certain point, landlords in weaker markets find it more economical to let aging units deteriorate until they exit the housing stock rather than continue renting them, and that, in stronger and gentrifying markets, they will upgrade those units, thereby reversing their filtering trajectory toward higher income households (Marcuse, 1985; Immergluck, 2018. Research provides empirical examples of negative filtering rates (Liu et al. 2022). Under those conditions, increases in market-rate units, no matter how great, would offer little relief to lower-income, rent burdened households.
Fourth, as calculated by Rosenthal, “filtering down” only refers to the process by which new housing units over time come to be occupied by lower income households. It says nothing about whether those households can afford that housing. Chapple and Zuk (2016) show this tension at play in San Francisco. Their study does offer evidence of downward filtering, showing that an association between new residential development and higher median rents flips to an association with lower median rents as the new units age. But It also finds an association between new construction and higher rent burdens among low-income households. This suggests that lower-income households tend to, out of necessity, occupy units that filter down to them at a faster rate than rents drop.
In sum, the City of Yes approach for addressing the affordable housing crisis through increases in overall housing supply would unfold something like this:
New housing units would enter the market at prices that can be expected to exceed the median asking rent of existing units, $3,500 (Comptroller’s Office, 2023). Then, it would take decades before they filter down to the vast majority of rent-burdened households, assuming that they filter down at all. And if they do filter down, they might not ease the rent burden of those households. What’s not to like?
Is there at least reason to believe that increasing market-rate housing supply slows down gentrification? The way the City explains it, the housing market sounds like a sort of waterbed; and if you build new housing in an expensive neighborhood, that sends ripples all over, suppressing gentrification in even far away places.
Well, waterbeds can be really comfortable. Housing markets, on the other hand, tend to be less so if you belong to a less-than-affluent household, live in a gentrifying neighborhood, and expect the private market to meet or respect your affordable housing needs. In part 4 of this series, we will examine the relation between new construction and gentrification.
Notes:
- Participation in Universal Affordability Preference (UAP), which grants a 20% density bonus, provided that those units resulting from that bonus are affordable to households earning 60% of area income, would not reduce the amount of time it would take new market-rate units to reach rent burdened households. While 20% of new units would go to those households each year, a correspondingly smaller number of market-rate units would enter the market than under a non-UAP scenario. Therefore, fewer new units would start filtering down each year toward the remaining households. It is true, however, that UAP units would go to lower-income households that filtering market rate units would likely never reach.
References:
Damiano, A., & Frenier, C. (2020). Build baby build? Housing submarkets and the effects of new construction on existing rents. Center for Urban and Regional Affairs Working Paper, University of Minnesota.
Immergluck, D., Carpenter, A., & Lueders, A. (2018). Hot city, cool city: Explaining neighbourhood-level losses in low-cost rental housing in southern US cities. International Journal of Housing Policy, 18(3), 454–478.
Liu, L., McManus, D., & Yannopoulos, E. (2022). Geographic and temporal variation in housing filtering rates. Regional Science and Urban Economics, 93, 103758.
Marcuse, P. (1985). Gentrification, Abandonment, and Displacement: Connections, Causes, and Policy Responses in New York City. Washington University Journal of Urban and Contemporary Law, 28, 195.
Margolis, S. E. (1982). Depreciation of Housing: An Empirical Consideration of the Filtering Hypothesis. The Review of Economics and Statistics, 64(1), 90–96.
Rosenthal, S. S. (2014). Are Private Markets and Filtering a Viable Source of Low-Income Housing? Estimates from a “Repeat Income” Model. American Economic Review, 104(2), 687–706.
Zuk, M., & Chapple, K. (2016). Housing Production, Filtering and Displacement: Untangling the Relationships.