On Tuesday, the Santa Monica City Council faced a difficult choice: whether to require 20-30 percent of units in new construction downtown be affordable, or whether to leave the requirement at only 15-20 percent.
This is a difficult decision, with no known right answer. But I believe that a constructive way to think about it can help take some of the tension out of the disagreements.
Most people in Santa Monica would like to see more affordable housing, so the higher requirement seems like a great idea. If it works.
Developers claim that the higher standard is too much, and that good projects will not pencil out if they are required to include that much affordable housing on site. But what if they’re bluffing? The affordable units are by definition subsidized: they’re below market rate, and the money that separates an affordable unit from a market-rate unit comes from someone’s wallet. When a city requires affordable units, that money comes from the developer’s wallet. So developers have a strong incentive to say that they couldn’t build any housing under the 20-30 percent requirement.
Because land is expensive to require and housing is expensive to build, there is clearly some level of required affordable units that doesn’t work. A 100 percent affordable development would not cover the costs of land and construction. But what about a 30 percent affordable development? We don’t know. The crucial part about the decision facing the City Council is the uncertainty involved. If the developers are bluffing, then it would be better to require 30 percent affordable housing and force the developers to reap lower profits—but still enough of a profit to get the projects built. But if the developers are not bluffing, then we won’t get any affordable housing built.
So the real question facing the City Council was not what to do, but rather this: Are the developers bluffing? In this situation, it’s wise to gather as much information as possible to resolve the uncertainty, and Tuesday night’s meeting featured a lot of information about the experience of other cities and about developers’ bottom lines that suggested that the developers are not bluffing. Even so, none of us knows for sure. The only certain thing is uncertainty.
Faced with irresolvable uncertainty, people have a naturally tendency to dig in. To take a guess and then commit to it. Some were convinced that the unprecedented 30 percent requirement is too much for developers to bear and that therefore the policy will backfire. Others—evidently a majority of the City Council—were equally convinced that the developers were bluffing. With the passage of the 30 percent requirement, we will now find out. But there is a much better way to deal with uncertainty, and that is to look at the costs of making a mistake.
Let’s say that we’re talking about a hypothetical project of 100 units. The City makes its decision first, and then later finds out whether the developers are bluffing or not. We don’t know up front if they’re bluffing, but we do know what will happen if they are:
If the City requires 20 percent affordable housing, it will get 20 units of affordable and 80 units of market-rate regardless of whether the developers are bluffing or not. But if the City requires 30 percent affordable housing, it will get 30 units of affordable housing and 70 units of market-rate housing, but zero housing of any sort if the developers are being truthful.
The costs of making a mistake are lopsided. A mistake on one side has much greater consequences than a mistake on the other side. The red boxes below show the mistakes—and again they will only be known as mistakes later. The magnitude of the mistake is defined by how much housing you give up when you make a policy and then guess wrong about:
The mistake that goes along with the City requiring 20 percent is zero for total housing; but the mistake that goes along with the City requiring 30 percent is a 100-unit reduction in total housing.
Of course, the biggest mistake in the face of uncertainty is to take a guess and then commit to it with complete confidence. Avoiding this mistake requires recognizing the uncertainty and quantifying it.
For example, let’s say that you’re 90 percent sure that the developers are bluffing. The question to ask is then: Would you rather have a 10 percent chance of losing 100 units, or a 90 percent chance of losing none? The expected value of the City requiring 20 percent affordable housing is 90 percent*0 whereas the expected value of the mistake associated with the City requiring 30 percent affordable housing is 10 percent*100 = 10. And since 10 is a bigger mistake to make than 0, even if you are really convinced that the developers are bluffing it still makes sense to not call their bluff. The stakes are too high.
But suppose that you care more about affordable housing than about market-rate housing. Suppose that you’ve read the report out of Berkeley about the effects of market-rate vs affordable housing on gentrification in low-income neighborhoods. That report says that, while both are necessary, affordable housing is twice as effective as market-rate housing at stopping gentrification.
So you value affordable housing at twice the rate of market-rate housing. That changes the calculus of the magnitude of the mistake:
Now if you’re 90 percent convinced that the developers are bluffing, the cost of a mistake when the City requires 20 percent affordable housing is 90 percent*5 = 4.5 lost adjusted-total units; whereas the cost of a mistake when the City requires 30 percent affordable housing is 10 percent*60 = 6 lost units.
Here’s the striking thing: even if you are 90 percent convinced that the developers are scamming us, it still makes sense to set the lower requirement. The stakes are too high.
Of course, if the City finds it’s made a mistake it can correct it later. But one mistake after another has been made for 40 years in California. How many more years of mistakes can we tolerate?
We can’t know if the developers are bluffing us or not. We can make wise decisions under uncertainty as long as we aren’t over-confident in how much we know about when someone else is bluffing.
Thanks for reading.