New data keeps coming in that shows that increases in housing supply tend to be followed by declining rental rates, even in the cities facing the highest demand. After a boom year for apartment construction in 2016, rents are falling in New York City, San Francisco, and Washington, DC.
- Median rents for one-bedrooms across New York City fell by 9.1% in the past year.
- Even in San Francisco — the most productive city in the country — a burst of new supply in 2016 has led to falling rents. Estimates put rent year-over-year decline in prices at 1% to 9%.
- Rising vacancy rates and quarter-over-quarter declines in Seattle’s rental rates are a sign that it’s leaving a period of double-digit annual increases.
- A decade of increasing construction rates in DC has leveled off rent prices.
- Los Angeles has not seen the apartment boom that has benefitted renters in other expensive cities, and its rental rates are the fastest rising in California.
- Scott points out the striking correlation between housing construction and house prices in in-demand cities.
This trend of declining rents is some very preliminary evidence against Tyler Cowen’s claim that densification in expensive cities will result in economic growth, but not falling rents as more buildings draw in more business activity and talent. A New York developer echoes Tyler’s view: “There are so many units, but then they all get eaten up. That is the way New York works.”
Aren’t these all fairly short-term trends that follow a severe price spike and corrective construction boom … and are therefore consistent with the narrative that over the medium to long term, supply can slow or prevent rent growth but can’t really drive rents down until there’s a drop in demand?
Please review Pittsburgh. More new apartments than people over the past 6 years, rising rents.
That’s right, in a sense. When the price spikes, then the developers expect that they can get paid to build new housing. Who builds housing expecting to lose money, after all? The new construction enabled by high prices is necessary but insufficient. Wherever housing remains difficult to build, prices for new housing will remain high.
The problem we have is the lack of Affordable To Build housing. In the major cities, there are endless payments and processes before you can get an entitlement to build. I just heard that San Francisco has 60,000 units entitled to build, but only 6,000 under construction. There could be personal reasons for some of them, but the majority are because after all the negotiations over fees and “community benefits,” the housing will cost more to build than it will receive in sales and rents. And then much needed housing doesn’t get built.
After World War II, Japan increased its housing stock by 65% in about 15 years without leaning on developers and investment bankers. They did it by allowing individual homeowners to replace their homes with newer, bigger homes. We need to eliminate racist zoning rules and allow individuals across the cities to tear down and rebuild their homes. Use capitalism to create an elastic supply of housing and end the housing crisis.
Thanks, Theodore. It does make sense to me that reducing per-unit or per-square-foot development costs (regulatory costs, labor costs, whatever) would make it possible for supply to lower prices instead of just preventing prices from rising.
I’m just arguing that supply by itself generally can’t lower prices in the medium to long term, so we probably shouldn’t claim that it can.
If increasing supply and falling rents is a supply-side success, is increasing supply and increasing rents a supply-side failure?
As to Seattle here is a more recent article (titled “After brief slowdown, Seattle-area rents surge back up again; when will it end?”: http://www.seattletimes.com/business/real-estate/after-brief-slowdown-seattle-area-rents-surge-back-up-again-when-will-it-end/
Prices decrease over the long run via filtering through depreciation of the housing stock. If enough new supply is added to satisfy overall demand, rents drop in the existing stock over the long-run as it depreciates…