Fake Numbers

In this morning's New York Times, David Leonhardt declares that “the official numbers on house prices—the last refuge of soothing information about the real estate market on the coasts—are deeply misleading.”  The reality, says Leonhardt, is much worse that the official stats show.That's bad news for the consumer, particularly for the homeowner who has purchased any time since the big real estate run up that started in about 2000. 

But it also raises an interesting question for scholars who make data an integral part of their work:  how could the numbers be off?  Leonhardt focuses on two factors.  1) The sales data reflect only the prices for houses that sold, not for the many, many more houses that are sitting on the market unsold.  2) The data are not for all houses, but only for a subset of the market.

These are not small differences.  Officially, Boston prices are up about 1%.  Leonhardt says that local experts report that the number is down 20%.  The official Boston index, however, excludes all mortgages larger than $417,000.  The last point knocked me over.  If the drop (or rise) is not uniform across different price levels, then the data have been off all along.

The first problem (sales data reflect only prices for houses that sold) should straighten itself out over time because the big, unsold inventory will begin to push those prices down.  Of course, so long as the inventory is big, there is more downward price pressure coming up. 

But the second problem says there is a systematic kink in the data that is little known and whose impact is almost impossible to measure.  That revelation sends a shiver down my spine.  How many other take-it-for-granted numbers have this kind of twist?

Comments

[thetruth]

Don't be such a pessimist. Redefining how one reports statistics is a valuable and proven method of improving the nation's economic welfare.

That's how we solved both the unemployment problem and the inflation problem. We simply changed the way we report the unemployment numbers, and now nobody complains about unemployment, even though if we reported it the old way, it would seem as bad as it was during some recessions.

And we changed the way we report inflation numbers, so that when the cost of a house triples in the space of a few years, we chant, "Horray! Prices have gone up, so that means that inflation has decreased!".

So if we simply change the method of reporting house prices to discard as outliers any houses whose prices have gone down, we can have an improved economy, rising property values, and we can all be happy.

Don't laugh. It works. No one notices.

Rising house prices decrease inflation. Rising house prices have always decreased inflation. Mmmm gin!