Tech hubs across the country have had a lot of success with job creation and productivity growth over the past few decades, but they've seen a pretty striking divergence of fates when it comes to housing affordability. Those that allowed more residential construction (relative to current housing stock), like Austin and Raleigh, NC, tended to fare pretty well. Those that allowed less, like San Francisco and San Jose, have done much more poorly. This chart, from Jed Kolko at Trulia, says it all:

‍

Raleigh, NC, has seen its population nearly double from 222,000 in 1990 to 423,000 today, and as the chart shows, it's more than doubled its housing stock in order to keep up. Austin, likewise, has grown about 70 percent since 1990 and allowed its housing stock to increase by more than 90 percent over that same time frame. As booming tech hubs (and successful, popular places in general), these cities have stayed shockingly affordable.

Contrast that with Middlesex county, where the population has grown by 22 percent but the housing stock only grew by 14 percent. Or San Francisco, where the population increased by 14 percent but only 11 percent more units were built. By square foot, San Francisco is almost five times more expensive than Austin. Housing affordability is an extremely complex issue, but if you want to make the point in the simplest possible terms you can't do much better than this chart: build more, pay less.

But that's not the whole story. What this doesn't address is the impact of price on overall (or latent) demand, a subtle but incredibly important concept. Cities where the housing stock grew faster than the population managed to keep their prices relatively low, but they were successful not because they kept ahead of population growth, but because their housing stock kept ahead of latent demand. The important question isn't "how many people moved there?" but "how many people would like to move there?"

The more successful you are at keeping prices low, the larger that number's going to be, and, in turn, the more you must build. That may sound like a catch-22, but most cities have some finite level of demand even at fairly low prices, and cities can aid each other by serving as reasonable substitutes — I may prefer to live in San Francisco, but I might consider San Diego if the price is right. If San Diego is expensive too, I might as well bite the bullet and go with the city I'll enjoy more.

The key, and what so many cities have sadly failed to do, is to stay ahead of the market, wherever the market is for your specific city, and thereby maintain some reasonable level of rental vacancy. Once you fall behind on those measures, things start to spiral out of control as low vacancies drive up rents, and high rents drive up land prices. Those land prices then require developers to ask higher rents to recoup their investment, and you end up in a situation where it may be nearly impossible to bring rents and home prices back down. To stay with the market requires constant reassessment, and it's why development goals like "we will build 5,000 new units this year" are basically meaningless. The absolute number of units you build is far less important than whether it's actually enough to satisfy demand.

Tech hubs aren't the only places in the country struggling with affordability problems, and this general relationship of "build more, pay less" is likely to hold true for virtually any growing city. Even Houston, with its horrible sprawl-dependent growth patterns, has at least managed to stay affordable by building enough units to house anyone that wants to live there. They've captured a lot of talented but price-sensitive residents by providing a good value, even if it means putting up with sweltering heat for half the year and blowing tons of money on car dependency. If a lot of those people had been able to afford San Francisco, San Jose, or San Diego instead, they'd be making more money, doing much less harm to the environment, and enjoying much more pleasant weather. But density's a dirty word, so that didn't happen.

[This article was originally published by Better Institutions.]