How Much Do We Really Spend on Public Housing?
Less than we used to. A brief analysis of HUD spending trends over the past 60 years.
Below I’m going to dive into some analysis of HUD spending trends over the past 60 years and what some of the funding in the reconciliation bill would do to change those trends, but first a quick update on where the housing package stands in the grand scheme of things.
Last week, Cea Weaver and I had an op-ed up at Politico on why it would be foolish to cut the housing package from the reconciliation bill. I was pleasantly surprised to see the article get tweeted out by the Senate Banking and Housing and the House Financial Services Committees, the two committees responsible for putting this package together.
The small amount of inside baseball information I have points to some members of the White House negotiating team eyeing the housing package for cuts to make room for other overdue welfare spending. This is seen as easy, partly because housing has gotten very little media attention, meaning few people know about it.
The committees responsible for the housing agenda are pushing back on this forcefully, including with a letter signed by the Democrats on the Financial Services Committee staking out their position—don’t cut the package. This may be small potatoes, but it is nonetheless noteworthy that no other piece of the agenda has gotten such a defense letter from the committee responsible for it.
Anyway, the brunt of the argument in the Politico piece is that the housing money is
counter-inflationary (lots of new housing production)
climate spending (in the same way that public procurement of electric vehicle fleets increases productive capacity in that sector, public capital spending on climate adaptive retrofits is sure to grease the wheels in that sector, too)
the biggest housing investment in generations (~$33 billion per year, for ten years as opposed to the one-time, one-year $14 billion investment in the Obama recovery bill)
The first two points are discussed at some length in the piece, but I want to go a bit more in depth into the last point with some analysis of HUD spending trends over the years.
How much do we spend on our main HUD programs?
Over at the Office of Management and Budget’s Public Budget Database, we have access to annual spending records for nearly every budget account within all of the various federal agencies, from 1962 to the present. This is a useful database that I’ve seen used to produce some interesting materials on housing spending, among other things.
But what I had not seen is a breakdown of our housing spending by program over the years as we have adjusted the funding levels and goals of those programs—particularly around the welfare reform era, which infected everything from cash assistance programs to public infrastructure investments like public housing.
The first plot here is a the total (inflation adjusted) outlays for the main set of public housing budget accounts: Section 8 (Project Based Voucher and Housing Choice Voucher) and Section 9 (Public Housing Capital Fund and Public Housing Operating Fund).
The large budget account that scales up from 1962 to 1998 was a single, large fund that paid for nearly everything in the public housing portfolio until reforms were passed in the late 90s (aside from the Public Housing Operating Fund which began in the 70s). Following these reforms, we have a transitional budget account (technically called the Housing Certificate Fund, but labeled here as a 1998-2005 transition fund, because effectively that’s what it was) that covered all of the Section 8 (PBV and HCV) spending before those programs got their own accounts.
Looking just at this chart, one should notice two things:
In real dollar terms, public housing spending has flatlined over the past two decades, following reforms in the late 1990s.
Section 8 programs have slowly grown, relatively speaking, while the Public Housing Capital Fund has shrunk (despite a one-time investment spike that shows up in 2010 as a result of the Obama recovery bill).
On the second point, there are two big reasons this is happening.
Many public housing authorities are moving units out of Section 9 and into Section 8 funding under the Rental Assistance Demonstration (RAD) program.
The two largest PHAs in the country, NYCHA (NYC) and CHA (Chicago), basically got shafted in 1998 when these reforms were passed: unlike every other PHA in the country, whose capital needs (or, ‘existing modernization needs’) are based on standard formula calculations, NYC and Chicago have their allocations per unit written in nominal dollar terms directly in the statute.
For NYCHA in particular, this has produced troubling outcomes. New York has some of the oldest public housing stock in the country, and capital costs have soared as the annual allocations from HUD have continually failed1 to meet the needs of those assets, and more importantly, of the people who live in them.
Using its 2001 funding level as a baseline, NYCHA estimated significant cumulative losses over the past two decades in its 2019 budget report. As the building stock continues to age without investments in capital repairs and upgrades, the total backlog is magnified—think of a homeowner who doesn’t spend $1,000 to fix their gutters and later winds up with a $10,000 water damage problem.
As NYCHA puts it in that 2019 budget report:
Funding available to NYCHA for capital improvements has not only failed to keep pace with its needs but has dramatically declined. From 2001 to 2017, annual federal capital grants have declined $74 million, or 18%, from $420 million to $346 million.
The current backlog is around $40 billion, which this bill could (potentially) fully fund, thanks to a clause that gives the HUD Secretary discretionary use of the lion’s share of the $80 billion for capital work.
But how much do we really spend?
But inflation-adjusted dollars are not the only way to measure spending. The OECD, for example, which compiles various social spending aggregate data from member countries, does not care much for spending in dollar terms and instead looks to spending as a percent of GDP as a more appropriate measure of the level of investment.
Below is the same data as in the first chart, but with the spending shown as a percent of GDP instead of as real dollars. This paints a very different picture. We still have Section 8 programs overtaking Section 9 programs for the reasons outlined above, but overall we see a strong and clear decline in spending over the past two decades: from 1999, at about 0.25% of GDP to 2019, at about 0.18%—not to mention the roughly 0.32% peak in 1997.2
So what’s going on? Well, in real dollar terms, our spending has flatlined. But as our economy has continued to grow, we have neglected to take advantage of the increased opportunity for investment, and we know our public housing programs have suffered for it.
So how much ought we spend?
Hopefully by this point you are wondering: what would the chart look like if we passed the housing package in the Build Back Better plan? Lucky for you, I’ve produced a plot for that as well. Below I’ve used the CBO’s economic projections for GDP over the next five years along with the funding levels and the directions for HUD outlined in the House proposal to project what the bill might do over five years.
For this, I anticipate HUD will scale up the new funding for the Section 8 programs each year, for five years, while the Public Housing Capital Fund will see a flat additional dollar contribution over the period, though it’s possible that money could be spent incrementally, too—it ultimately depends on which projects in the pipeline are shovel-ready.
Interestingly, the spending on this ensemble of public housing programs, while massive and historic, would basically just put us back on track with what we were spending, as a percent of GDP, back in 1999. It would definitely buck the downward trend we’ve been on for twenty years, but it’s not exactly radical, nor is it egregious, or indicative of an “entitlement society”.
It’s really just normal and basic public infrastructure and social safety net spending. And as the headline says in the Politico piece, Democrats would be fools not to do it.
https://www1.nyc.gov/assets/nycha/downloads/pdf/nycha-2019-budget-book.pdf. See the explanation and graphics on pages 19 and 20.
https://www.oecd.org/els/family/PH4-1-Public-spending-social-rental-housing.pdf. These results conform with results published by the OECD, which put US social rental housing spending at about 0.21% of GDP for 2020, though it is unclear which specific programs the US reports to OECD for their social housing questionnaire—it’s possible LIHTC units are counted, while Section 8 units are not, for example. Also see the Data and comparability issues section on pages 3 and 4, which explain some of the discrepancies for countries with significantly more social housing, but which is built and funded at the municipal or regional level rather than the national level, and thus does not get counted.