Real-Estate Agents

Should you use a real-estate agent to help buy or sell your home?  You'd be forgiven for thinking that the answer is a clear, "No."  No-agent websites have proliferated, and keeping an extra 3% on your house (the typical 6% commission is divided between buyer's and seller's agents) has some obvious appeal.  This post looks at the theoretical and practical issues of using a broker (many of which apply to using agents in other negotiations as well).

Some academics have argued that the housing market suffers a pretty serious level of "agency cost," i.e. your agent may be getting you a worse result because your interests are unaligned.  Consider this passage from Freakonomics:

A real-estate agent may see you not so much as an ally but as a mark...[A study found] that an agent keeps her own house on the market an average ten extra days, waiting for a better offer, and sells it for over 3 percent more than your house -- or $10,000 on the sale of a $300,000 house...The problem is that the agent only stands to personally gain an additional $150 by selling your house for $10,000 more, which isn't much reward for a lot of extra work.  So her job is to convince you that a $300,000 offer is in fact a very good offer, even a generous one, and that only a fool would refuse it.

Let's get to the heart of Levitt's point about incentives.  Your agent is getting a good commission, so she has a strong incentive to sell your house, but she doesn't have as good an incentive to sell your house for the highest price.  The commission structure rewards her for closing the deal as quickly and simply as possible, whereas you might be willing to wait or take some risk in order to get a higher price.

Here's how it generally works.  Your agent splits the 6% commission with the buyer's agent and then splits the remainder with his or her brokerage firm, so the actual commission is more like 1.5%.  1.5% on a $300,000 house is $4,500 which is a pretty good commission for one transaction, but the incremental return on boosting the sale price is much less.  For every dollar she adds to the sale price of the house she keeps just one and a half cents.

Let's take a specific example and see how this might play out.  Suppose you're interested in moving to a larger house within your neighborhood.  You're not in a hurry, but you're ready to move -- perhaps you and your spouse have decided to have kids and you want to add a couple of bedrooms and a yard.  Your current house has a market value somewhere in the range of $450K to $550K and after exploring your purchase options (and taking into account your existing mortgage) you conclude that your BATNA (staying in your current house) is preferable to any sale price that earns you less than $460K (after commission).  Here's what that implies for the value created by selling your house, for you and for your broker:

Sale Price	$450,000	$500,000	$550,000
Value to You	-$37,000	$10,000	$57,000
Value to Broker	$6,750	$7,500	$8,250

If we compare sale prices of $450K and $550K we see a huge difference for you -- from unacceptable to nearly $60,000 better than your BATNA.  The broker's commission changes much less -- at the unacceptable price it's 82% of what it is at the home run price.

This can work against you in two ways.  First, the broker has only modest incentive to do extra work (trying to find a better buyer).  Let's say you've got an offer for $500K but if the broker were to really work her network and invest another twenty hours of work she could find a buyer at $550K.  That's an hourly return to you of over $2,000 but for her it's less than $50 per hour.  She's likely better off spending that time cultivating new clients since the bulk of her commission comes from getting a sale at all rather than from maximizing price.

The second potential problem is risk aversion.  Intuitively it might seem that you're more risk averse, since it's your home, but in many cases the reverse is true.  Let's again consider our $500K buyer.  Suppose we think there's a 75% chance we can get that buyer to pay $550K if we hold out, but a 25% chance that we'll lose the sale.  In this situation, that's a very good bet for you.  You have a 75% chance of gaining $47,000 in value and a 25% chance of losing $20,000 in value.  For the agent it's a bad bet.  She has the same chance of gaining, but her gain ($1,250) is much lower than what she's risking ($7,500) so her expected return is negative.

This means that the agent has an incentive to encourage you to price your house at a lower-than-optimal (for you) price and to be less aggressive in negotiating.  If you have an offer for $500,000 (which is better than your BATNA, but not much), your broker may tell you that that's the best offer you're likely to get and you should take it.

Levitt argues that there's a straightforward and obvious cost to all this -- brokers push you towards a lower price in order to close the deal, while when it's their own house they hold out and get more money.  Does this mean that hiring a broker is a bad idea?

Not necessarily.  First of all, all of Levitt's "evidence" other than one study is anecdotal...and the plural of anecdote isn't data.  As for the study itself, while it does control for factors such as "location, age and quality of the house, aesthetics, and so on" there are two rather obvious factors it did not take into account: motivation for selling and where the seller was moving.

Opportunistic sellers are by nature more patient and more likely to respond to price opportunities than someone who has to sell.  Someone who has already bought a house or who is moving for a new job faces time pressure that may prevent them from holding out for a higher price.  Brokers, being immersed in housing, are presumably more likely to sell opportunistically (e.g. when the market is particularly hot for the type of property they own, or because a colleague with a suitable client is more aware of an agent's house than a random property on the market).  Since real-estate brokers tend to nurture community relationships over a long period of time it may also be that they are more likely to upgrade locally than move to another region, which again would let them choose their timing more patiently.  Thus, the differences Levitt notes could be explainable by factors he was unable to control for.

The reality is probably something not quite as bad as Levitt suggests, but still raises the question of whether hiring a broker is a good idea given that your incentives may be mismatched.  For most of us I believe the answer is still yes.  A broker's expertise is very useful in navigating the process of buying or selling a house and reducing the risk of pitfalls.  If you're selling, a broker can advise you on how to present your house most effectively, how to respond to buyer conditions and be alert to major pitfalls that could result in legal liability.  She may also be able to interpret statements from buyer's brokers more accurately than you would, since they know each other's signals, making her a useful partner even if you want to take the lead in negotiations yourself.

The more interesting question for me is how you might address the mismatch of incentives to get the best use of your broker.  A lot of sellers are negotiating with their broker but I suspect they're doing so in the wrong way.  The typical approach is to push for a lower commission.  This saves money but worsens the incentive mismatch and essentially relegates the broker to the role of low-skill intermediary whose only value-add is likely to be handling the legal paperwork.  If Levitt is right that broker effort adds to the final selling price then the last thing we want to do is remove that effort.

In many situations a more effective approach would be to suggest a higher commission rate but based off of a floor.  Suppose in our example above the seller's agent received a 30% commission (apart from the 3% that goes to the buyer's agent) on the purchase price less $450,000?  In other words, if the price is $450,000 (which you can presumably get without an agent's help) the agent would get nothing but would get twenty-five cents for each additional dollar?  Now a $450,000 price is worthless, a $500,000 price nets her $7,500 and a $550,000 price yields $15,000.  That extra twenty hours of work would now pay almost $400/hr, making the effort worth her while.  (These numbers are used only to illustrate the point; the general idea would be to set a floor that was easily achievable and a rate that would make the broker's commission equal at a 'normal' sale price.)

A side benefit to this approach is self-selection.  An agent who is confident she can get the highest price for your property is more likely to accept a commission structure like this, whereas one who thinks they're unlikely to do better than $500K may balk.

Finally, don't forget the power of talking.  Even if you decide you don't want to negotiate an unusual commission structure, talk to your agent about the incentive problem.  If you're comfortable holding out for a higher price make sure that your agent knows this and that she thinks of her mission to get you as high a price as possible rather than just completing the sale.