Red Dollar, Blue Dollar
Or, One Reason Why New Housing is Expensive
The Prologue
I once had a British finance professor who hammered the phrase “red dollar, blue dollar” into our heads. This stemmed from a little stunt on the first day of class. He ambled in, greeted us with that English pattern of speech full of ums and ers and well nows, and produced one of those star-spangled USPS Priority Mail envelopes. He then knocked two dollar bills out of the envelope onto the table in front of him. One was dyed blue, the other red. He then thundered, “red dollar, blue dollar!”
He turned to a guy in the first row and continued, “which’ll ya have?”
“Wha… what?” came the reply.
“Say I’m giving you a dollar. Do you want the red one or the blue one?”
“Oh, red I guess.”
“No!” He then turned to another student: “Do you want red or blue?”
“Blue?”
“No!” He tried a third: “Do you want red or blue?”
“Um, I don’t know… I don’t really care,” she replied.
“Correct! Why do you care if the dollar is red or blue. They’re both worth $1.”
“So do I get one?” the student pressed.
“Oh, no,” he replied, “these are my dollars.”
I’d spent the little episode realizing that our professor snatched Priority Mail envelopes from the post office instead of just buying his own manilla ones (or, rather, asking the department secretary for them). So, I hadn’t paid as much attention as I ought to have. But, it was a very good lesson!
The key idea is that if you’re investing your money, what do you care if you earn a dollar from a stock or from a bond or from the family business? Preferring red dollars to blue ones is irrational. And, it’s downright stupid to take one red dollar in place of two blue ones.1 This is very important to keep in mind when you consider how firms behave.2 If a company expects to earn less on a project than they would from just keeping their money in the bank, why would they bother with making something?
Now, what if we apply this spiel about red and blue dollars to housing? Specifically, I want to explain why most new housing is “luxury housing”.
The Game
We’ll look at New York City specifically and apply a very narrow frame: we’ll think only in terms of the minimum return in rent that would be required for a developer to be willing to build an apartment building.
Now, suppose you hate real estate developers and think that their greed is responsible for high rent. I’ll grant you that they may very well be unpleasant people; kindly salt of the earth folk don’t usually merit the headline “Who Didn’t Want Him Dead?” when their smoldering corpses are discovered in dumpsters. Nevertheless, that does not mean greedy developers cause high rents.
The wager is this: I will demonstrate that, so long as developers are not very very stupid, new housing won’t be affordable to the typical family. New housing does push overall housing prices down (more on that later, after our game) but the new houses themselves only pencil out if they target the relatively well-heeled. Notice that “not very very stupid” is a much weaker condition than “rapacious capitalist pig dog”.
The stakes are: If I win / convince you, you’ll talk about something more interesting than landlords at dinner parties for at least two months. If I lose, I’ll do nothing… after all, my writing is worth what you pay for it.
The Rules
I’ll assume a developer will only begin a project if he or she can expect to earn a 5% real yearly return on the total cost. This is a rule of thumb for returns on an asset, going back as far as Jane Austen novels.3 For our purposes, it’s a laughably low assumption (so no complaining later). The long run real return on an S&P 500 index fund (probably a much safer investment than attempting to build in NYC) is 6.5%. Moreover, I’m ignoring property taxes and operating costs; a more realistic figure with those included would be 10-17%.
I’m going to ignore the first-floor commercial rents on these buildings. Those wouldn’t come close to compensating for the low 5% assumption above, so you still may not complain.
I’ll also ignore the fact that, because NYC permits so little new housing, the market clearing rental price might be well above the absolute minimum rent that will make developers willing to build at all. This is another major concession on my part, so no crying allowed.
The overall effect of these rules is to 1) massively tilt the playing field in your favor, and 2) demonstrate that it’s just quite expensive to build new housing. It’s impossible to cover your costs if you accept a pittance in rent.
The Analysis
We’ll proceed with a few examples of apartment buildings in progress.4 None of these are mega-projects, high profile, or anything special, which is the whole point.
Let’s begin with this project in Gowanus, Brooklyn. The cost $1B and it will include 1,000 units. The average unit size is 850 sq ft,5 compared to a median new U.S. house size of over 2,000 sq ft.
Now, red dollar, blue dollar. Developers will not be willing to undertake this project if they can’t make at least a 5% return; they could just stick $1B in an S&P 500 index fund instead and earn 6.5%. This means we need to get $50,000,000 in rent each year, or $50,000 per apartment, or $4,167 per apartment per month. That $4,167 rent is affordable for a household with a $150,000 annual income.6
The (approximate) median household income in NYC is $77,000; the 80th percentile is $180,000. Under the most charitable of assumptions, these apartments cannot be broadly affordable. We don’t need to assume developers are greedy to get this result, only that they’re not morons. And, if we were using realistic assumptions, double the required rent.
The development includes rent stabilized units, but that doesn’t change the fact that 1,000 apartments are costing $1B to build. If the rent stabilized units’ rents are capped, the developer will need to be able to charge a higher rate for the other ones. Remember, if he doesn’t get an average rent of $4,167 per unit, he’s an idiot.
Here’s another example, on the Upper East Side of Manhattan this time: 89 apartments at an average size of 1,500 sq ft. Large by NYC standards but hardly extravagant as the U.S. goes (in fact, about 700 sq ft below the median). Costs are roughly $165MM, meaning that, to make 5%, the average annual rent per unit needs to come in at about $92,500, or $7,708 per unit per month. This is affordable for the lucky household earning well north of $250K each year.
Now, there are occasionally cheaper projects. This one comes in at $200MM7 for 465 units at an 800 sq ft average. For a 5% return, this would require $21,500 per year per unit, or about $1,800 per month, which would be just affordable for the median NYC household.
So, overall, if we make a bunch of generous (wildly unrealistic) assumptions, we can occasionally find a project that could pencil out to be affordable for a median household. That still leaves 50% of households out of luck. Also, for all of these examples, remember that we ignored property taxes (very high in NYC), ignored operating costs, ignored depreciation, and assumed that developers would be happy making slightly less than they could expect to if they just put their money in the stock market – i.e., assumed they’d be happy taking on more risk to earn less money. So again, if you’d like to be somewhat more realistic, double the rent figures.
The Moral
New housing is expensive (partly8) because it is expensive to build, not because real estate developers are price gouging. Some of this is for good reasons (fire resistance say); some is for silly ones (look up ULURP, I double dog dare you). We could do something about the silly reasons, but it will probably never be cheap to put up nice new buildings in the middle of New York. It’s just always going to be the case that a shiny new apartment will cost more than one that’s 20 years old, not all that different than the case of any other good. We don’t expect new cars to cost less than used ones.
The point here isn’t to argue that billionaire NYC developers are unfairly maligned. It’s to note that “developers are sometimes bad and greedy” isn’t a helpful way to think about housing development in general; new houses would be expensive even if their builders were angels (red dollar, blue dollar). And if developers would like better PR, I’d suggest they find hobbies other than murder,9 and refrain from hiring prostitutes to seduce and film their brothers-in-law.10
The (Important) Epilogue: New Housing Still Lowers Overall Rents
If new housing is targeted toward the affluent, how does it lower rents for the middle class and poor? Well, the high-level answer is supply and demand: if you increase the number of housing units available, all else equal, rent prices will fall (that’s the econ 101 answer). Exactly how does that work? If you’d like some dry reading, look up the economic literature on vacancy chains. Otherwise, take my explanation: If an affluent family moves into a new apartment, they vacate a nice but older apartment. Another family moves into the vacated old apartment. And so on. It’s a bit like how when hermit crabs find a new empty shell, they line up from largest to smallest, then adorably race to new shells all at once. Each hermit crab ends up in a somewhat nicer house. We’re not so different from the crabs.
Assuming risk is the same. Obviously, you’re going to want a higher payout if you run a chance of losing money. This is why stocks (on average) return more than T-bills (i.e., one year US Treasury bonds). Stocks are riskier so they have to pay more, or everyone would just buy the safe T-bills. It’s the same reason logging pays well: it’s dangerous work.
Any type of firm, from the mom-and-pop bookstore you pay 200% mark-ups at to the evil faceless grocery store you blame for ruining your marriage.
Piketty notes this in Capital in the 21st Century: “In the novels of Jane Austen and Honoré de Balzac, the fact that land (like government bonds) yields roughly 5 percent of the amount of capital invested … is so taken for granted that it often goes unmentioned. Contemporary readers were well aware that it took capital on the order of 1 million francs to produce an annual rent of 50,000 francs.”
This is a Substack post, not a lecture, so I’m just choosing a few examples pretty much at random where it’s easy to see what the construction costs are. There’s a limit to how much homework I like doing in my free time.
Assuming 85% of the residential space goes to apartments, with the remaining 15% allocated to hallways and such. This is a generous assumption. Realistically, you’re only looking at 70-75% of space being rentable.
I’m defining affordable as no more than 1/3 of gross annual income, a widely used rule of thumb.
The linked article doesn’t mention the price of the air rights purchase. It was about $25MM.
Also, housing is generally expensive because we purposely restrict the supply. That’s not a smart thing to do, and we should stop doing it. But, that’s not the topic of this post.
Charles Kushner, who arranged to have a tape of the encounter sent to his sister. He was pardoned for this by one Donald J. (for Jehoshaphat?) Trump, himself a NYC developer who had brief public notoriety for his walk on role in Little Rascals.