Friday, September 15, 2006

Do Real Estate Brokers Shirk?

Steve Levitt is an incredibly smart guy (he has a Clark Medal, the prize for best economist under 40, to prove it) and, according to the people I know who know him, a very nice guy. Freakonomics is fun to read, and I liked it enough to use it for a seminar I taught last summer.

A reason the book is so much fun is that it is provocative, and challenges us to think about things from perspectives we had not before considered. It asks why drug dealers live with their mothers; whether Roe v Wade was responsible for the falling crime rate in the 1990s; whether Sumo Wrestlers cheat; and whether real estate agents shirk.

Levitt sets up the real estate agent-house seller relationship as a classical principal-agent problem: the agent is like anyone else: he has an incentive to do as little work as possible in exchange for as much commission as possible. Therefore, Levitt conjectures that real estate agents have an incentive to get sellers to take the first offer they receive from buyers. Because commissions are based on the entire price of the house, the fact of a sale gives agents a large benefit. If the seller refuses an offer, the agent has to do more work, but gets very little added compensation. For example, if there is a $490,000 offer on a house that could ultimately sell for $500,000, and the agent gets a 1.5 percent cut, from the standpoint of the agent, the benefit of waiting is only $150. The agent will almost certainly have to do a lot more work for the $150, and so has an incentive to encourage the buyer to take the $490k.

This would be a powerful argument for an incentive problem if agents were playing a one-shot game. But they are not--they are playing a repeated game. Agents rely on listings to make money, and listings come from referrals. Agents--good ones anyway--have an incentive to make sellers very happy, because happy sellers drive future business.

So now let's get to the evidence that Levitt provides to show that agents shirk: he shows that agents leave their own houses on the market longer and sell them for higher prices than the houses that they sell for others. This is an interesting fact, but is not necessarily explained by agents' shirking. Rather, it could be that the houses agents own themselves are different in unobserved characteristics from the houses that they sell to others. Or it could be that agents are different as human beings--they are more risk taking, and more entrepreneurial. For all we know, sellers ignore the advice of their agents and sell more quickly than they should, because they are relieved at the prospect of a sale and don't care so much about the marginal benefit they would get for waiting. In order to disentangle this, we would need to know something about the unobserved characteristics of people who become real estate agents--and of people who don't.

Full disclosure: while I was writing my dissertation in the late 1980s, I worked doing research for the Wisconsin Realtors Association. I had fun while I was there. Believe me, the Realtors are different with respect to their attitude toward risk taking from the rest of us.

6 comments:

Anonymous said...

Wouldn't it also make sense that the incentive element of the P-A relationship is removed from the equation when an agent sells their own house, and therefore the incentive to hold out for more money is the singular incentive. In a sale of a 3rd party's property the agent takes only a small percentage of the incrimental increase in selling price. When selling their own house, they keep dollar one and every dollar thereafter. Conclusion: the incentive to hold out is greater, and if agent's fees were 100% of purchase price, they might hold out longer in 3rd party sales as well.

Richard Green said...

Yes, this is exactly Levitt's point, and it may well be correct. In fact, it seems to me that everyone would be better off is agents were compensated using some kind of consignment system, so the marginal compensation to the agent is greater, holding average compensation constant.

But the incentive is not singular--when one sells her own house, she is not worried about referrals. And any empirical tests of differences must consider the possibility of selection bias.

Anonymous said...

Everone takes the emotional element of the sale out of the equation.Time is money for the seller and the buyer.If the transaction system was more integrated causing a more seamless transfer of data,commission dollars would be more valuable.

Anonymous said...

The model then must perceive a correlation between future income stream from referrals and ability to deliver the "ask". Clearly a lower "ask" can be assumed to result in a quicker sale. Then wouldn't there be a component that weights (i) the income derived from a higher "ask", against (ii) the income from a future referral? If so, then wouldn't the "ask" for a house for a 3rd party be lower than that for a similar asset owned by an agent (there is no referral benefit derived)? Is it possible that the shirking behavior is best measured at the point of pricing, rather than once priced and marketed? Or at the very least the shirking behavior needs to factor the pre-pricing decision into the equation.

real estate in Philippines said...

I agree with your points completely. In real estate investment, just like in equity investment, one has to find a value situation or it's just not worth it. I also agree that it's arguably harder and more time consuming for investors to find that value in real estate.

Deirdre G

shailena said...

Well, it’s amazing. The miracle has been done. Hat’s off. Well done, as we know that “hard work always

pays off”, after a long struggle with sincere effort it’s done.
-----------------------------
johndouglas
Brokers