That's a misleading metric for two reasons. First, "people" includes all the people who bought a house decades ago before the prices went up, and correspondingly paid the old prices instead of the current prices. And second, it assumes that if housing becomes more expensive what people all do is pay more for housing rather than being unable to afford it and then having to take a longer commute to access housing they can afford, live with their parents, become homeless, etc., which negatively impact the amount or type of housing they get instead of the amount they pay for it.
Meanwhile between 1984 and today median income increased by ~290% while median new home prices increased by ~425%:
https://fred.stlouisfed.org/series/MSPUS https://fred.stlouisfed.org/series/MEHOINUSA646N
And that's with the gap having reduced a bit in the last couple of years, since home prices have declined slightly since the peak in 2022 while median income continued its usual trend.
People are moving out of opportunity zones (and you know…losing opportunity) or moving in with roommates to keep the average percentage low.
Feels like such a basic factor to miss for what is supposedly an economists blog. That metric means nothing.
From https://www.jchs.harvard.edu/blog/rental-housing-unaffordabi...
> The net effect of this longer history is that renters today spend much more of their incomes on rent than they did in previous generations. The median renter household in 1960 spent less than a fifth of their income on rent. By 2022, housing costs consumed 31 percent of the median renter’s income. And in inflation-adjusted terms, the median rent is up 75 percent over these six decades while the median renter income has risen by just under 15 percent.