A commenter on my earlier post writes,
In San Francisco, over 70% of rental units are rent-controlled. This allows a lot of low-income people to live there, pushing down the median income. Meanwhile, there are no price controls for home sales, and these are pushed up by the shortage of market-rate units.
If you excluded everyone in a rent-controlled unit, I think the price-income ratio would look more reasonable.
Arithmetically, this explanation works. The median income would be artificially low relative to the median house price. But what about the economics?
I expect rent control, along with building restrictions, to lower the quality and quantity of housing services supplied. It also creates a big incentive for owners to convert rental properties into homes for sale. Over a long enough run, these supply effects tend to push up rents. I am a bit reluctant to believe that rent control achieves an equilibrium in which housing is affordable for renters and not for buyers. I am more inclined to side with the view that rent control backfires in the long run, leading to high rents.
Of course, one can argue that there will still be a gap between prices and rents, particularly if conversion of rental units is impeded by regulation. In that case, the only option available to property owners would be to limit maintenance. If the quality of rental units has tended to deteriorate, then that would support the commenter’s explanation.
Incidentally, another commenter suggested that the ratio of wealth to income might be high in cities with a high ratio of house prices to income. This would be very plausible if we were talking about average ratios. It is somewhat less plausible for median ratios, because median wealth tends to be pretty low.