This is a much belated reply* to a claim Matt Yglesias made a little while ago about buying a home to live in and as an investment.  Matt wrote:

If you own a first house in the DC area, and plan to eventually sell it and replace it with a larger house in the DC area, then overall trends in DC area real estate prices are irrelevant to you. Your investment will “pay off” if prices go up, but you’ll end up needing to spend more on your second house anyway. It all washes out. If you’re speculating in real estate you don’t intent to live in, things look different. But in terms of a place you inhabit, the relevant consideration is how its price evolves relative to the market average and your odds of doing better than average are 50 percent at all times.

The problem here is failure to consider the proper counterfactual opportunity cost. Say you’re the hypothetical current DC resident Matt is discussing, and in the future you plan on owning a house bigger house than you can currently afford. Matt says if prices go up then the future house you’re going to buy is more expensive so it’s a wash. But if prices go up and you don’t buy a house now than you still have to face those higher prices, but you won’t have the profits from your more valuable older home to offset the costs.

In fact, this is true even if you plan on renting in the future. This is because if prices go up you will incur higher housing costs regardless of whether or not you bought a house early on. If you’re going to live in DC in the future, then buying a house now is a hedge against rising house prices. When prices go up that investment can pay off, and the profit certainly doesn’t need to be a wash.

*Why am I replying to a short, old post on something of a minor point? Because Matt is writing a book about urban economics, and if correct a misunderstanding in the book before it gets published then Matt will have to credit me in his book. And then I’ll be in a book.