≡ Menu

A Note on Perfectly Inelastic Supply of Rental Housing and Welfare Losses

I like David Henderson’s reaction, on EconLog, to Tyler Cowen’s recent post on rent control.

One of the commenters on David’s post is “Brian,” who writes:

With a fixed supply (not realistic), producer surplus becomes consumer surplus along with a shortage. But no DWL [deadweight loss – Don Bx]

I’m not arguing for the rent control, but the “costs” in that scenario are different than under perfect inelasticity.

Below the fold is a slightly amended version what I wrote as a comment at EconLog in response to Brian’s comment.


You correctly note that the assumption of a perfectly inelastic supply of rental housing is not realistic. But I believe that it is incorrect to say that, were this assumption to hold in reality, rent control would create no deadweight loss.

In what Deirdre McCloskey calls “the first act” there is indeed, unlike with an upward-sloping supply curve of rental housing, no deadweight loss. It’s all transfer from landlords to tenants. But unless we assume also a perfectly inelastic demand for rental housing, rent control will in the second act cause current and prospective renters to waste resources as each competes to increase his or her chances of securing one of the rent-controlled apartments. Losses – ones in every relevant way comparable to conventional deadweight losses – ensue.

Picture a standard supply-and-demand graph, but one in which the supply curve is perfectly vertical. With the demand curve shaped as it usually is – namely, sloping downward to the right – rent control will still cause a shortage of rental housing. Tenants will therefore compete more vigorously – now using only non-price means of competition – to secure these rental units. The use of resources in such non-price competition – for example, wining and dining landlords, racing or keeping constant vigilance to sign up for newly vacated rental units, whatever – is a waste of resources from society’s perspective.

It’s true that these wastes are conventionally called “rent-seeking wastes” rather than “deadweight losses.” But I’ve always taken Gordon Tullock’s point to be that rent-seeking-‘rectangle’ losses are no less real or important losses than are deadweight-loss (or Harberger) ‘triangle’ losses: both losses act as deadweights on society. Valuable goods and services that would otherwise have been produced remain unproduced because of the interventions that cause these losses. Both losses emerge because government interventions render uses of resources that were once profitable now unprofitable. Both losses ‘measure’ the reduction in valuable output. Both losses ‘measure’ the reduction in the size of the economic pie.

I don’t wish to defend too strongly my claim that rent-seeking losses should be called and classified as “deadweight losses” (although I do believe that that claim is defensible). But I do wish to insist that the absence of conventionally defined deadweight losses does not mean that an intervention, such as rent control, effects only a transfer and, therefore, causes no real losses. The resulting non-price competition among potential renters results in losses.

Note here that I here mention neither rent-seeking efforts by tenants to secure rent-control regulation from government nor rent-‘protecting’* efforts by landlords to fight rent control. These rent-seeking and rent-‘protecting’ efforts only increase the social losses from rent control beyond the losses that I identify above.


* A more-appropriate term here is “‘property-protecting’ efforts” or “‘rights-protecting’ efforts.”

UPDATE: Do also read David Henderson’s comment on this post.

Much of the problem here is caused by the convention of discussing groups of people – “tenants,” “landlords,” “producers,” “consumers” – as if all of the individuals in each of these groups of people share, within that group, a unitary interest.  (Note: David’s comment is a wise reminder that these collective terms are fraught with danger.) But in fact not all tenants share a unitary interest.  Ditto for all other such groups that we often discuss collectively.

So, for example, when wealth is transferred from “landlords” to “tenants” (in the form of a forced lower rental rate), not all tenants gain, and those who do gain do not gain equally. Indeed, even if tenants as a group gain as a result of rent control, some tenants might – likely do – lose as others reap all of the gains.  And not all landlords lose equally; some perhaps do not lose at all even though landlords as a group lose; some might even gain.