Seems like it must depend a lot on how much pent up demand there is for housing. Add 10% to the Bay Area and it's not going to make a dent. Add 10% in Detroit and you just add 10% more vacant units, but probably little change in rents because they're already bottomed out. They might actually increase the average rent in this case if you build more desirable units that have a mortgage to pay off.

> Zero chance this is linear. A 20% increase would be expected to create more than a 2% decrease.

I think this effect is probably complicated by generated demand.

Think about a city like SF or Seattle. Adding a lot of vacancies may not drop prices much for a while. Instead, it'll suck in a lot of people from the surrounding area. It's only after there is enough new construction to overcome that effect that people will see price declines in the city.

Of course, if SF or Seattle is sucking in people from the surrounding areas, then the price of real estate in those areas will decrease substantially. It's all about what you measure.

Exactly, came her to say this. It obviously can't be linear, its just the slope at a point, but that didn't stop the headline writer. For "every" 10%, really? I'm pretty certain that for even just the next adjacent 10%, rents wouldn't just fall 1%.

I bet broad-based 10% increases in a city are rarely even seen in the data. The authors are just interpreting their regression model coefficients for the media.

What happens when you double the housing supply? Who knows. Not a single data point in that regime.

Only a 500% increase in supply to halve rents.

250K people with $250K incomes want to live near Mission Dolores Park.

0% of YIMBY plans plan for reality.