Jerusalem Demsas is an eloquent and forceful voice on housing policy. In a recent article, she asked this question: “How do we ensure that housing is both appreciating in value for homeowners but cheap enough for all would-be homeowners to buy in?” She answered her own question “We can’t.”
I think we can.
Demsas’s article is “The Homeownership Society Was a Mistake” in The Atlantic. In it, she diagnoses a major economic problem in America: our government often treats housing as an investment whose price should go up, while America needs housing to be a consumption good that is cheap to use. Her recommendation is that government should focus on making housing cheap and plentiful, “diminishing greatly” “its attractiveness as an investment.” And, if that wasn’t clear enough, she ends with “Let housing be a home; … Just don’t let it be an investment.”
Demsas’s problem is real, but her solution shocked me. It is both dangerous and unworkable.
Not treating housing as an investment is dangerous because housing is an investment. An investment moves value across time. If I spend money to build a house, that house is still valuable in 10 years. If I buy land, the land is valuable forever (barring disaster-movie disasters). In fact, housing is one of our country’s most valuable assets: land and buildings in America are worth more than the entire stock market. To not treat housing as an investment risks squandering that value.
Demsas’s solution is unworkable because we live in a democracy and the majority of voters own houses. Nationwide, 65% of housing units are “owner occupied” and homeowners are more likely to vote than renters. One study put it at 3% to 8% more likely. It is unreasonable for any elected official to go against homeowners.
Demsas’s solution can be ignored, but her problem should not be. Our government is currently prioritizing housing as an investment, rather than a consumption good, and that’s harming our country.
To understand the mechanism causing the harm, we need to look at housing as a risky investment. Land usually goes up in value, but it can go down. A dramatic example of that is Detroit from 2006 to 2009. Land prices plummeted. The price of the average single-family house on a lot fell by 55%. Some fell more.
The risk to homeowners is actually greater than that. A fall in land prices is usually caused by a worsening local economy. Over that same period in Detroit, unemployment rose to 16%. So, just when homeowners are most likely to lose their job, their asset falls in value.
Homeowners want to limit their risk. They elect officials who will protect their risky asset.
And, since homeowners are the majority, those elected officials go further by propping up that asset with tax benefits and competition-reducing zoning laws. The costs of those sweeteners are paid for by non-homeowners: we penalize renters. The penalties/favors drive low-wealth families to buy a home and go “all in” on a single risky asset. It is a matter of luck which ones profit from the bet.
By this mechanism, the Homeownership Society harms society. It penalizes the poor, encourages foolish risk, and generates inequality.
But the Homeownership Society is not all harms, or we would have gotten rid of it years ago.
Homeownership means permanence, which inspires residents to invest in their community. Homeownership means less risk, because any time you rent something there’s a risk that the renter damages it. Lastly, homeownership improves efficiency. Homeowners maintain their buildings better than renters. Homeowners are willing to invest in insulation, because they also pay the electrical bills.
Although Demsas doesn’t mention these benefits in the article, I’m sure she knows them. And I don’t think Demsas is against homeownership, per se. She is focused on the decisions of government officials to protect and prop up the price of housing. She sees their choice as lying on a single axis. They can choose high-priced housing or low-priced housing. You can see that in her self-question and self-answer: “How do we ensure that housing is both appreciating in value for homeowners but cheap enough for all would-be homeowners to buy in?” And seeing no options, she answers “We can’t.”.
I think we can. I think we can protect investments and lower the cost of housing. And, even more importantly, I think we can do it without denying the political reality that homeowners vote and that democratically elected officials serve (service?) the voters.
How do we get these benefits of homeownership without Demsas’s problems? Housing is two things: the land and the building on it. Most of Demsas’s negatives have to do with the land: it lasts forever (barring disasters) and swings up and down in value with the local economy. On the other hand, most of homeownership’s positives have to do with the building: lower risks, better maintenance, and better efficiency. If we can financially divorce the land from the building, we can get by Demsas’s impasse.
To do that, we need Financial Engineering. Financial Engineering is the art of writing a contract to allocate risks and financial incentives such that all parties benefit. An example contract that you’re familiar with is your insurance policy. You have risks you do not like: your house might burn down or your car might be hit by a garbage truck. The insurance company is willing to take on that risk if you pay a premium and a deductible. A financial engineer designed the contract with an “insured side” for you and an “insurer side” for the insurance company, to the benefit of both parties.
To solve Demsas’s problem and get past the impasse, we need a contract that pulls out all the “investment” qualities of the land and passes them on to an investor and leaves the “consumption good” qualities of the building with the homeowner.
I call this contract a “land-value contract”. With it, the homeowner will (effectively) sell the land under their house to the investor. If a property is valued at $300,000 for the house and $200,000 for the land, the homeowner signs the “seller side” of a land-value contract and receives $200,000. The investor signs the “buyer side” and delivers the $200,000. If the homeowner ever sells the property, the contract will end and the homeowner will pay the investor the current price of the land, which may be more or less than $200,000.
The investor is giving up $200,000 for this contract and taking on risk, so they must receive something in return. Thus, the homeowner pays a premium to the investor. The premium depends on the value of the land and is, effectively, the “ground rent”, the rental cost of the land under the house. The homeowner has, effectively, sold the land to the investor and now rents it back.
For housing policy geeks, this is almost Georgism. Homeowners pay rent for their land and those payments go to the nation’s investors (rather than the government). The contract can also be seen as a market-driven land trust. Homeowners own their home and pay rent for the land under their house, but those payments are done at market rates (not subsidized)
Homeowners would need a smaller down payment, since they are borrowing less and the risk is less. Homeowners would need to make a payment, the ground rent, that would change from year to year, but they already experience that with property taxes.
Investors should be interested in these contracts. A study of housing in multiple counties from 1950 to 2015 showed that the capital gain from housing has a risk-adjusted-return close to that of stock indices. And housing is weakly correlated with the stock index. The very wealthy, who own most stocks, could diversify by buying land-value contracts and the less wealthy, who currently put too much into a single piece of real estate, could diversify into stocks.
How does the land-value contract solve the problems of the Homeownership Society? It may pass the land value on to investors, but the land-value contract is still a large risky asset and won’t investors vote to protect against its risk? True, but the investors are different from the low-wealth homeowners going “all in” on a single risky investment. Investors diversify, by holding equities and by holding land-value contracts from many cities. And when someone invests in multiple assets, they care less about the ups and downs of any particular asset.
Land-value contracts will allow current homeowners to sell the land under their house and invest that money in stocks and in the land under every house in America. They would have to pay the ground rent for their house, but they would also receive the dividends and ground rents from other properties. Changes in the local ground rent are a risk, but it is a dramatically smaller risk than what homeowners hold right now. This lower risk weakens the motivations of elected officials to prop up those assets, at the cost of others.
Land-value contracts will allow current homeowners to sell the land under their house and invest that money in stocks and in the land under every house in America. They would have to pay the ground rent for their house, but they would also receive the dividends and ground rents from other properties. Changes in the local ground rent are a risk, but it is a dramatically smaller risk than what homeowners hold right now. This lower risk weakens the motivations of elected officials to prop up those assets, at the cost of others.
Land-value contracts could be created today by a company. A bank could sell them, combined with lower-downpayment mortgages. But it would be hard to launch this financial product. First, the bank would have to recruit new investors and educate them about the investment. Second, the bank would have to roll out the program nationwide, rather than in one city, to enable diversification. Lastly, and most importantly, there is no independent organization calculating the ground rent and land values in the contract. Customers would be scared to sign a contract where the bank determined that rate. Customers accept variable-rate mortgages, but that rate is set by an outside(ish) entity.
It would be easier for the government to launch the market. The government is independent(ish) and can compute the ground rent and land values. The government can also mandate the data (housing prices and rents) required to calculate those values. The federal government has an incentive to create this market: it implicitly backs mortgage-backed bonds and could reduce its risk by mandating the signing of land-value contracts by mortgage holders.
(Finance experts will point out that financial contracts on real estate already exist. The Chicago Mercantile Exchange (CME) has futures and options on the Case-Shiller Housing Index of 10 cities. These are based on the price of the land and the house on it. That, and other factors, make it difficult for homeowners to know how to offset their land’s value using these futures. These existing contracts are not the same as land-value contracts; ease-of-use and exact offsetting of risk matters.)
If you have access to The Atlantic, I encourage you to read Demsas’s article with the concept of land-value contracts in mind. The article is dense with arguments and supporting data. I believe the ability to separate the land’s value from the building’s value illuminates the arguments. Land-value contracts add a new axis, allowing us to see that cities prosper when (un-propped-up) land values increase and building costs decrease.
I value Demsas’s article for bringing attention to a problem. She wants housing to be a consumption good, like a TV or car. With land-value contracts, I think a house becomes exactly that. And it is achieved without denying the economic realities of housing and the political realities of our democracy. With land-value contracts, we can fix the Homeownership Society.