On Marginal Revolution this morning, Tyler Cowen addresses the question “How bad is rent control when housing supply is artificially restricted by law?” He leads with this:

Many of you have been asking me about this NYT article on the pressures for rent control in Silicon Valley. If no (or few) new apartment blocks will be built anyway, what is wrong with rent control in that setting?

What, apparently, some of the questioners are getting at is that if the supply is constrained anyway, then there won’t be the bad effects on the supply side that economists talk about. Tyler points out that that’s not true. And one of his commenters, Kommenterlein writes: “The rental housing stock isn’t fixed – it will decline rapidly with rent control as rental apartments are converted into Condos and sold at market prices.”

So, yes, there are bad supply side effects of rent control.

There are also bad demand side effects of rent control and Tyler points to one of them:

At rent-controlled prices there will be excess demand for apartments. The “plums” will go to those who bribe, those who are well-connected, those who are skilled at breaking the law, and, to some extent, those who have low search costs. The latter category may include well-off people who hire others to search for them.

But he leaves out arguably the biggest demand-side effect: the misallocation of apartments to those who are already renting them. When the rent is not allowed to clear the market, the apartments no longer go to those who value them most. People who rented an apartment with three bedrooms, and had two kids, would, absent rent control, have some large probability of stepping down to a smaller apartment when their kids grow up and leave home: doing so is likely to save money. But now introduce rent control and that probability falls. They are more likely to keep that apartment because they won’t save money, or won’t save as much money, by stepping down. Instead, they will be in a queue with other apartment searchers.

Related to this, when I teach about rent control, I point out that there is a fundamental difference between rent control and the Nixon/Ford/Carter gasoline price controls. In the latter case, we were out there weekly competing with each other by lining up for gasoline. The simple fact that we got gasoline last week gave us no inherent advantage over others in getting it this week. We had to line up yet again. But, with rent control, the fact that we had a rent-controlled apartment last year gave us a huge advantage over those looking for apartments this year. The way rent controls are typically designed, those who continue to pay their rent have first say over the apartment when the lease comes up. There’s no inherent necessity for this. The rent control authorities could design the regulations so that every year, if a landlord wants to evict the tenant and put him on the same footing as everyone else, he can. Then a lot of the support for rent control, which is supported mainly by those who already have rent-controlled apartments, would dissipate. But, of course, the rent controllers don’t do that because one of their main goals is to help the people who already have apartments.

I also point out that a major reason that the gasoline price controls were eliminated was that gasoline buyers did not have the same privileged position for getting gasoline that current tenants have for keeping apartments. It’s easy, though, conceptually to imagine that the gasoline price controls are changed in a way that makes them more likely to last: if the amount supplied is 80% of the amount demanded, then allocate the supplies at the controlled price to those who will use that 80% (say, people with last names beginning with A to T) and allocate zero to the other 20%. This would be more like the way apartments are allocated now under rent control.

By the way, Tyler concludes:

Even if the current housing stock is fixed, rent control probably will create costs in excess of its benefits, and without significantly desirable distributional consequences.

I have no idea why the word “probably.” It’s not hard to show, using a downward-sloping demand curve and a vertical supply curve, deadweight loss from rent control. It’s a certainty.