Social Econ Blog

This is pretty funny.

Some economists did a pretty rigorous comparison of the price of homes sold using a for-sale-by-owner website to the price sold using a real estate agent. They find that owners who sell their own homes receive the same price, on average, as owners who sold comparable homes using real estate agents. This suggests that if owners value their time and effort at less than the value of a commission (which can be a lot of money), they should strongly consider selling without the help of a real estate agent (at least in Madison where the study was conducted).

The National Association of Realtors apparently does not like this conclusion (although it is not obvious why). They prefer their own "study" which shows that sellers who use real estate professionals receive a 27% premium over those who sell on their own. Of course:

Those studies had their own asterisks. The 2005 survey was based on buyers’ written responses, rather than actual records of transactions, and included data only from people who chose to reply — 7,813 responses out of 145,000 questionnaires mailed out.

The association also did not disclose data on house size, lot size and some other factors that could affect price differences between the sales techniques. The association does not consider that a weakness of the study, though.

“When you’re looking at this large of a survey, the aggregate numbers smooth those things out,” Mr. Molony said. “We feel it’s representative.”

Yes, kids, sample bias and omitted variables bias are not problems if your sample is large enough. I wish I'd had Walter Molony in graduate school. Life would have been so much easier.

[Note -- It is disappointing that, in an otherwise reasonably good article, the author let's Mr. Molony's quote stand. He sets up the key issues in the previous two paragraphs, but drops them after letting Mr. Molony speak -- even though the response offered by Molony is totally wrong. Thus, readers not well versed in statistics might finish the article thinking that the NAR study used a valid method and provides useful information -- when it does not.

Sadly, this practice of letting obviously wrong statements go unchallenged is common practice among journalists indoctrinated to be "objective", and it leads me to join Brad Delong in asking, "why oh why can't we have a better press corps?"]