Two facts in New York seem inviolable. Housing prices always rise, and wherever you look, a shiny, new high-rise condo is being built. Therefore, the natural conclusion appears to be that high-rise construction is driving the affordability crisis.
However, this is a confusion of correlation with causation. Luxury construction and prices are indeed linked, but rising prices incentivize new construction, not the other way around. The real problem in New York is that vast swaths of land are “off-limits” for new housing. The areas that are the most able to build are those in the densest parts of the city, where the demand is the greatest and where rental tenants abound. As residents walk through their neighborhoods, they see new skyscrapers along with rising rents and blame the gentrifiers for their woes.
But the bottom line is this: housing supply and housing demand are in a constant battle against each other, and prices rise when the supply side loses to the demand side. New York’s problem is not that there is too much construction, but rather, there is not enough. Even though this does not “feel” right to the ordinary person, it’s the truth.
My research, as discussed below, shows that, all else equal, a 10% increase in new construction in a neighborhood is associated with about a 15% reduction in median sales prices and a 10% reduction in median monthly rents. So there is strong evidence that the laws of supply and demand function as they should.
Additionally, I also find that a 10% increase in median sales prices is typically met with only a 0.3% increase in new units. When looking at rents, I find a 10% increase in median rents is met with about a 2% increase in new units. In the horse race between supply and demand, new supply is utterly unable to keep up with demand.
The Economics of the Housing Market
There’s an old joke that I tell my students when I teach Introduction to Microeconomics. It barely gets a laugh, but I can’t help myself. It says that if you want to turn a parrot into an economist, you teach it to chirp the phrase, “Supply and Demand.”
The reason it’s funny—just to spell it out here—is because economists truly see the world through the lens of supply and demand. By making this distinction, one can better map out the incentives that drive human behavior and then look at how these incentives produce the social and economic outcomes we observe. Ultimately, the question is whether these incentives generate outcomes that are “good” or “bad” for society. If bad, then it suggests a role for policies to improve the collective well-being. Supply and demand analysis can also highlight where previous policies—however well-intentioned—have gone wrong.
The components of supply and demand have, in fact, been enshrined into “Laws” because they help us understand how and why markets function. Under specific conditions (which I’ll come back to below), these laws show that private markets are an efficient and reasonable way to allocate resources—they generate the lowest possible prices that bring forth supply. If some “shock” comes along (e. g., economic growth or an influx of population) to change these conditions, the market will adjust to bring things back to a long-run equilibrium. That is, markets, over time, can naturally restore themselves to a reasonable price level.
Contrary to popular sentiment, the housing market also behaves according to those laws. We can thus look to supply and demand to help us understand why housing prices have gotten so high and why new construction is not the cause.
The economics of supply begins with the Law of Supply: rising prices will incentivize more production. That’s why there’s gas in your car, milk in the fridge, and a computer at your desk. High prices mean higher profits, and they draw forth an army of producers. It’s no different in real estate: when there’s money to be made, developers will find a way to bring new units online.
As demand for housing increases, it causes prices to rise. In the short run, there’s not much that can be done to accommodate this new demand—pitting all against all in a game of low-vacancy musical chairs. But in the longer run, developers will build denser to maximize their return on their investment. If demand stops increasing, supply will catch up, and prices will moderate.
But cities are dynamic places. New York in the 21st century has proved to be attractive, resilient, and productive. Its population and employment base have been growing, incomes have risen, and its global stature has made it a lure for international investment. Because of New York’s renaissance, the demand for New York housing has been growing briskly over the last several decades, despite some ups and downs along the way.
As housing demand rises, it drives up prices. More consumers, many of whom have lots of cash, are chasing the same pool of units, and, in the short run, with supply relatively static, prices rise. But if the market was truly open and flexible, those rising prices would be met with new supply to moderate or reduce the price increases. But if demand keeps growing faster than supply, then all that new construction, on net, will not stop price hikes. In the end, it will appear as if new construction is driving up housing costs, when in fact, new construction can’t match the demand.
As I mentioned above, markets work well when specific conditions hold. Economists use the term “market failure” to indicate some wrench in the works that prevents a market from operating efficiently. Although the housing market works according to the laws of supply and demand, certain features make this market unique. Arguably, the most important is that a house or apartment exists at a fixed location. So, the geography of the property adds or detracts value. But the value of geography is inherently a social phenomenon, and other people’s actions affects us personally.
The problems that arise because of other people’s actions are called negative externalities. A high-rise gets built next to your building and puts you in the shadow. A new Walmart opens, and the roads become filled with reckless and angry drivers, who then hog all the parking. Naturally enough, residents turn to their local leaders to enact rules to stop or prevent these kinds of things.
Over the years, New York City—like all cities around the world—has enacted a suite of regulations to help households against real and perceived externalities. The result has been relatively strict zoning regulations. A whole body of research shows that the more regulations in a housing market, the higher housing prices. In the name of reducing neighborhood harm—congestion, shadows, etc.—the city has helped creating a housing shortage.
As I have argued elsewhere, these zoning laws were implemented by planners who thought their zoning rules “felt” right to them. They were not based on weighing the costs of regulations against their social benefits. Instead, the rules were restrictive because planners could not tolerate population density and they prioritized single-family homes over apartment buildings. Once implemented, they have been virtually impossible to roll back. Today it’s illegal on 50% of the city’s residential land to tear down a one- or two-family home to build an apartment building—even for three or four families.
Furthermore, during World War II, when housing prices spiked due to a lack of new supply, the city enacted price controls by fixing rent levels to mitigate the burden. Over the years, the city couldn’t fully wean itself off them, and, today, New York has a rent stabilization program, which limits how much a landlord can raise the rent.
Naturally, this keeps people from moving. If they leave, even to downsize, they’d likely wind up paying more in rent. Today, 57% of rental units are subject to some form of price controls. The law says that rent stabilization stays in effect as long as the vacancy rate is below 5%. But since the program creates artificial housing scarcity, it perpetually justifies its own existence.
Because restrictive zoning and price stabilization are popular, leaders had to devise other ways to encourage new supply in the face of shortages. The answer has been a suite of all kinds of incentives, such as tax breaks, zoning bonuses for creating affordable units, or paying landlords to remain in the rent stabilization program. The net result is a policy tennis match between restrictions and subsidies, and no one wins because the two sides are equally matched. The city’s housing policies are like a person who is hot, so she turns on the air conditioning. Then she gets cold, so she turns on the heat at the same time.
Well-intentioned, but misguided, policies are not the only problem. Big cities worldwide face other types of “frictions” that prevent supply from coming online rapidly and smoothly. First, an old city is a built-up city, with many smaller lots holding low-rise structures from an earlier time. Because of high land and construction costs, the economics of new construction work much better on larger lots. But acquiring these lots is difficult and time consuming, which slows things down and raises the costs.
Furthermore, 66% of all structures in New York City are one- or two-family homes. These homes, along with the substantial number of condo and coop apartment buildings throughout the city, means that the vast majority of units is owner-occupied, even though two out of every three residents is a renter.
Homeownership, of course, has its benefits. But the downside is that it creates frictions, usually through NIMBYism, where residents try to stop new construction or upzonings from fear of harm to their property values or the quality of their neighborhoods (hence “Not in My Backyard”). For many, their homes are their largest source of wealth. Rather than seeing a house as shelter, they see it as a safety deposit box.
In truth, there is little research on which kinds of frictions are strongest in New York—government-imposed frictions, land acquisition difficulties, NIMBYism, or high costs. Likely they all play a role. But together, all these barriers make supplying new housing, especially—at the middle and low ends—hard to match the demand.
Housing Supply in the 21st Century So Far
But let’s be clear, all these frictions and barriers do not stop the city from building. In fact, Gotham has gone on something of a building spree in the last few decades, as prices have risen. However, given the city’s dynamism, while new units are being added, other units are being removed—either through demolition, conversion to other uses, or by combining units.
In Part I of this blog post series, I reviewed the data to demonstrate how Gotham’s housing stock has changed over the 21st century thus far. In 2002, the city had about 3.2 million housing units. By 2020, that number was just over 3.6 million units, a growth rate of about 13%. In the same period, New York’s total population grew by about 10%. While this seems like housing growth outpaced the population, several problems that actually make this a shortage of housing, not the least of which is that it had not caused the housing vacancy rate to budge upward. With low vacancies comes higher prices.
And when you drill down at the neighborhood level, you see large variation in construction. The vast majority of New York City over the 21st century either lost units or remained about the same. Of the approximately 29,000 city blocks, 75% saw either no change or a loss of units, with 4% of those losing six or more units (14% of blocks lost one or more units). Only 25% of blocks saw an increase in six or more units. In short, the block-level analysis shows that much of the city is quite static, unwilling or unable to pony up more housing.
The net growth in the housing stock has come not in a citywide effort but only in a few areas, and where this new housing is immediately absorbed due to the national and international demand.
Prices and Supply: The Evidence
A review of housing supply over the 21st century, however, is not complete without an analysis of how supply and prices are linked. For this, we need to turn to the economic theory discussed above, which separates supply and demand into separate components. Here, I bring the theory to the data. (The data sources and analysis can be found here.)
To this end, I created a data set that enables me to look at the relationship between changes in housing units and sale prices and rents from 2003 to 2019 across New York’s 59 Community Districts (CDs). If we look at scatter plots, as expected, we see a positive relationship between unit additions and prices (see the figure below). These graphs are simply a statistical demonstration of what most believe is the causal relationship between prices and construction, and which generates the confusion of correlation with causation.
However, the fundamental measurement problem is to isolate changes in supply that are unrelated to prices (so-called exogenous shifts in supply), so we can see how these shifts in supply impact prices. We can also do the reverse—if we can find exogenous shifts in prices—those that are affected by demand, independent of supply—we can see how builders respond to prices increases. We can then conduct a horse race between supply and demand.
The first step is to measure how supply is affected by non-price elements. To this end, I estimated how changes in the number of housing units across the Community Districts (CD) is impacted by zoning regulations, growth in households, and other determinants of housing supply. In this case, I find that growth in the number households, of course, leads to more supply and that, interestingly, the more area in a CD that is zoned for housing, the less subsequent growth in units it had. The reason is that it’s very difficult to acquire lots for teardowns in residential neighborhoods. It’s relatively easy to develop formerly industrial or commercial land, which provides less resistance to new construction.
I then looked at how median sales prices per square foot and median monthly rents (both adjusted for inflation) are affected by this increased supply. As mentioned above, the results show that, on average, across the city, an increase in 10% increase in units is associated with about a 15% reduction in real median sales prices! For rents, I find that an increase in housing units by 10% reduces median monthly rents by about 10%, all else equal. The lower responsiveness is probably due to rent stabilization.
Again, let me repeat, when we put the data into a statistical “centrifuge” that separates the demand side from the supply side, we can, in fact, see a strong role of adding housing supply into reducing prices, holding demand-side factors constant.
And, when we do the reverse by looking at how responsive supply is to price increase, we see that the city is, indeed, very unresponsive. My estimates show that an increase in housing prices by 10% only leads to a 0.3% increase in new supply, while a 10% increase in median rents leads to a 1.2% increase in units.
To summarize, increasing supply has a large effect on prices when we hold demand constant. But when prices rise due to increased demand, the city can only increase its supply by a relatively small amount. Thus, the statistical results show that the demand side is the big winner in the horse race.
I have discussed elsewhere the policies that are needed to increase housing supply across the income spectrum, so I won’t go into details here. I understand that homeownership and having a free-standing house in the suburbs is part of the American Dream, but the problem this creates in a bustling metropolis is that zoning vast acreage for single-family homes creates fixed-pie real estate, where those with money get a slice at the expense of those with modest means. It exacerbates income and racial segregation, increases income and wealth inequality, reduces walkability, raises carbon dioxide emissions, and prevents the rise of diverse vibrant, mixed-income neighborhoods. Restricting large areas for redevelopment packs new high-rise construction into the central areas in a way that exacerbates problems related to gentrification.
Homeowners, who want to preserve their wealth, and renters, who blame the rich for their high rents, have made common cause against fundamentally altering the housing markets they have created. Some states are tinkering on the margins by allowing more neighborhoods to convert single-family homes to two- or three-family homes. If enough people do that, it can be a great benefit to both the owners and those who rent in them. But it remains to be seen if that will make a dent in the larger problem in big cities across the U.S.
More fundamentally, we need to create what I call “buy-in nation,” where policies both incentivize new construction and create buy-in from those who are most resistant to it. We need to convert NIMBYists to YIMBYists with policies that protect residents against change or the fear of it. There are any number of programs that can help to do this if we are truly committed to ending the affordability crisis (see my prior posts here and here, for example).
However, the first step is to realize that housing markets work like all markets. More housing supply across the income spectrum will reduce the affordability crisis, and many of the roadblocks against new supply were created to stop this new supply in the first place. It’s time to fully honor the Laws of Supply and Demand.
Continue reading Part I of this series here.
 I want to be clear, however, that I am not saying that luxury high-rise construction alone will get us out of the affordability problem.
 In the old days, people would take in lodgers. It was not uncommon in tenement districts for a family to rent out their living room floor as a hostel for laborers. Today, not only is this illegal, but few would be willing to do it as it’s not compatible with what we consider an adequate lifestyle for the 21st century. But the truth is, it was a way for households to increase their income while providing cheap housing to those most in need of it. It allowed for more rapid social and economic mobility for those at the bottom rungs.
 In a high-cost city like New York, some form of construction cost subsidy is needed to make sure lower-income housing is built. But for these subsidies to work properly, they can’t be checked by restrictive zoning.
 It’s also worth pointing out that, as of 2020, there were only 1,015 elevator condo buildings throughout the entire city, comprising 0.12% of all structures–hardly a number that could cause all the perceived ills in the New York housing market.