David Leonhardt has an interesting piece on house prices but ends up making a serious logical error. He argues that house prices typically keep pace with income, meaning that they have risen more rapidly than inflation. He bases this assessment on the fact that the portion of income that has been devoted to to housing has remained constant over roughly the last 80 years.

There is a logical problem in this analysis. In principle, the issue is the movement of a the price of a house of the same quality, not the amount that people actually spend on housing. If the price of a house of the same quality rises in step with income, and the share of income devoted to housing remains constant, then this logically implies (i.e. there is no way around the conclusion), that the quality of housing has not increased over this period.

This would mean that the homes that people are buying today are no bigger or better than the homes that people bought 80 years ago. This contradicts an enormous amount of data and common sense. It is unlikely that anyone would seriously argue this case. Therefore, we can conclude that house prices have not kept pace with income growth.