How Rebate Bans, Discriminatory MLS Listing Policies, And Minimum Service Requirements Can Reduce Price Competition For Real Estate Brokerage Services And Why It Matters

EAG Discussion Papers are the primary vehicle used to disseminate research from economists in the Economic Analysis Group (EAG) of the Antitrust Division. These papers are intended to inform interested individuals and institutions of EAG's research program and to stimulate comment and criticism on economic issues related to antitrust policy and regulation. The analysis and conclusions expressed herein are solely those of the authors and do not represent the views of the United States Department of Justice.

Information on the EAG research program and discussion paper series may be obtained from Russell Pittman, Director of Economic Research, Economic Analysis Group, Antitrust Division, U.S. Department of Justice, BICN 10-000, Washington, DC 20530, or by e-mail at russell.pittman@usdoj.gov. Comments on specific papers may be addressed directly to the authors at the same mailing address or at their e-mail address.

Recent EAG Discussion Paper titles are listed at the end of this paper. To obtain a complete list of titles or to request single copies of individual papers, please write to Janet Ficco at the above mailing address or at janet.ficco@usdoj.gov or call (202) 307-3779. Beginning with papers issued in 1999, copies of individual papers are also available from the Social Science Research Network at www.ssrn.com.

ABSTRACT

Institutional characteristics of a typical residential real estate transaction may create a causal link between commission rate reductions and service degradation. Such a link mitigates agents' incentive to compete on commission rates. Rebate bans, discriminatory MLS listing policies, and minimum service requirements strengthen the link thereby frustrating greater commission rate competition. Eliminating rebate bans, discriminatory MLS listing policies and minimum service requirements will increase commission rate competition and reduce wasteful rent seeking by agents.

1. INTRODUCTION

Between 1998 and 2006, inflation-adjusted commission fees¹ paid per home transaction increased from $10,012 to $13,261–a real increase of 32%.² During the same period, the number of licensed agents increased faster than the number of transactions leading to a 47% increase in the number of agents per home transaction.³ Many observers of the residential real estate brokerage industry puzzle over the apparent inconsistency of increasing agent commission fees coupled with dramatic increases in agent entry.

Some of these same observers have argued for the elimination of rebate bans, discriminatory MLS listing policies, and minimum service requirements as a way to promote more competitive commission fees. Yet, if the concern is a lack of price competition, opponents of rebate bans, discriminatory MLS policies and minimum service requirements are obligated to explain why removing these constraints will increase price competition. Without a plausible explanation, opponents' efforts may be rejected as unjustified and potentially more harmful than beneficial.⁴

We explain how rebate bans, discriminatory MLS policies, and minimum service requirements can reduce commission rate competition, and why more competition is important for economic efficiency. We argue that certain characteristics of the industry generate an inherent disincentive to compete through lower commission rates (i.e., rates are "sticky" or rigid downward).⁵ In an environment characterized by rising home sales prices, sticky commission rates automatically generate higher commission fees. Rather than driving down commission fees, additional entry inefficiently dissipates profits by spreading more thinly aggregate commission fees among a larger group of agents. Rebate bans, discriminatory MLS policies and minimum service requirements contribute to the disincentive to compete through lower commission rates. Removing these constraints will increase price competition and economic efficiency.

In the next section, we describe how commission fees are determined and how institutional characteristics of real estate brokerage services can create a disincentive to reduce commission rates. In section 3, we explain one particular characteristic, "steering," in more detail. In section 4, we explain how rebates, non-discriminatory MLS policies, and limited service agents can increase commission rate competition by mitigating the disincentive to reduce commission rates. Finally, in section 5, we show how more intense commission rate competition can improve economic efficiency in the market for real estate brokerage services.

2. THE LINK BETWEEN COMMISSION RATE REDUCTION AND SERVICE DEGRADATION

2.A. COMPLEXITIES OF COMMISSION RATE DETERMINATION

Commission rates are typically negotiated between the home seller and his listing agent and recorded in a "listing contract."⁶ The final commission fee to be paid by the home seller is calculated by multiplying the final sales price by the negotiated commission rate. Since a home seller typically has many listing agents to choose from, it is reasonable to assume that the commission rate paid by the home seller is determined by a procurement-like mechanism in which numerous agents may bid for the right to sell the house.⁷

Under textbook assumptions, one would expect such a mechanism to generate a competitive commission rate when many agents are vying for the listing contract. However, the institutional characteristics of residential brokerage services diverge considerably from textbook conditions.⁸ Among other things:

• Home sellers face a principal-agent problem. Agents' incentives may not align with their own, and home sellers cannot perfectly monitor their agent's effort in pursuing the home seller's interests.⁹
• Home sellers face considerably more uncertainty about the sales process than do agents. For example, home sellers may not know the market value of their home or the state of the housing market as well as agents. More generally, since most home sellers tend to sell very infrequently, they are not likely to know the ins and outs of the process.
• In the majority of cases, the commission fee is used not only to compensate the listing agent, but also a selling agent–the agent who procures the buyer, also known as the "cooperating agent" or "buy-side agent." This complicates the principal-agent problem further as it becomes a multi-tiered problem.
• Home sellers primarily desire two things: the shortest time on market and the highest net sales price (i.e. gross sales price less commission fee). These two desires compete with one another: other things being equal, a higher net sales price will result in a longer time on market.

Due to these institutional characteristics, the negotiated commission rate affects much more than how to split surplus between the home seller and the listing agent. Indeed, the negotiated commission rate also determines the amount of effort the listing agent and selling agents will exert in selling the listing. As a result, the outcome of the negotiation helps determine the total amount of surplus to be split in the first place.

Given the complexity of the situation, it is difficult to characterize fully the outcome of the negotiation between home seller and listing agent. Rather than doing so, in the remainder of this section, we describe two interrelated features – the principal-agent problem and steering – that we believe will influence the negotiation and likely prevent commission rates from being bid all the way down to an agent's reservation value.

The key observation is that these two features generate a relationship (or at least a perceived relationship in the minds of those negotiating) that seems to be fundamental to the determination of commission rates: lowering the commission rate makes it more difficult for a home seller to sell their home.¹⁰

2.B. DISINCENTIVE TO LOWER COMMISSION RATE DUE TO PRINCIPAL-AGENT PROBLEM

As mentioned above, home sellers primarily desire two things: the shortest time on market and the highest net sales price (i.e. gross sales price less commission fee). As a result, a listing agent might attempt to win a listing contract by offering the home seller a lower commission rate than her rivals. Other things being equal, a lower commission rate increases the net sales price obtained by the home seller. However, a critical feature of agent services is that offering a lower commission rate will not leave other things equal.

The typical listing contract does not perfectly align the incentives of the listing agent (the "agent") with the interests of the home seller (the "principal"). For example, most listing contracts are structured such that the agent bears the majority of the costs of selling the home but only receives a fraction sales price. Therefore, the listing agent will not be willing to undertake as much costly effort in selling the listing as the home seller would like.¹¹ If the listing contract also included an agreement on the amount of effort the agent would undertake, this problem would not arise. Since monitoring agent effort is exceedingly difficult, it is not practical to write effort clauses into listing contracts. As a result, the commission rate ends up being the home seller's sole lever for encouraging agent effort. The only way to encourage more effort is through a higher commission rate.¹²

The link between the commission rate and agent effort implies that, a lower commission rate will either lower the net sales price (i.e., a lower gross sales price) or increase the time on the market. As a result, rather than negotiating for the lowest commission rate possible, a home seller has the incentive to shade the commission rate upward to induce a more optimal level of effort from the listing agent. While the home seller could hold out for a commission rate that would just allow a listing agent to break even, he has a disincentive to do so since agent effort would be too low.¹³ This inherent disincentive mixed with a high level of home seller uncertainty can lead to home sellers who are very reluctant to negotiate ever lower commission rates.

It is important to recognize that home sellers do not negotiate with listing agents in a vacuum. A home seller is competing with all other current home sellers to sell his home. As a result, not only will a home seller be reluctant to negotiate the lowest rate possible, he will also be concerned about negotiating a rate that diverges from the rates negotiated by other home sellers for fear that his agent will exert relatively less effort in selling his home. The fear and uncertainty of diverging from whatever other listing agents are getting, or the same agent is getting for a different listing, could explain a convention we understand to prevail in local markets in which home sellers and listing agents often agree on the "going rate."

Finally, compounding the home seller's uncertainty is the concern he likely has about his agent's relationship with other agents whom she will need to rely upon for bringing buyers to the transaction. In other words, due to the nature of the listing contract, the home seller faces a multi-tiered principal agent problem. We turn to the second tier next.

2.C. DISINCENTIVE TO LOWER COMMISSION FEE DUE TO STEERING

As described above, in the majority of transactions, the listing agent uses a portion of the commission fee to compensate the selling agent. As a result, the listing agent faces a principal-agent problem of her own. The key reason is a selling agent behavior commonly known as "steering," which occurs anytime a selling agent discourages a buyer from purchasing certain listings.¹⁴

If a listing agent attempts to list a home with a lower commission rate than his rivals, selling agents may respond by steering buyers away from that home.¹⁵ Steering by selling agents will reduce the pool of potential buyers. Other things being equal, a smaller pool of potential buyers will result in a greater time on market or a lower gross sales price. In this sense, steering by selling agents degrades the listing agent's service, and the listing agent will have a disincentive to reduce her commission rate. While we have described this disincentive from the listing agent's perspective, to the extent the home seller is aware of the possibility of steering, he, too, will worry that negotiating a lower-than-prevailing rate could jeopardize the prospects of selling his home quickly and at a good price.¹⁶

Principal-agent problems and steering can create a direct link between the commission rate a home seller pays and the quality of agent service he obtains. Specifically, in the presence of principal- agent problems and steering, a lower commission rate can lead to greater time on the market and lower gross sales price.¹⁷ Since degraded quality will make it difficult for agents to attract additional clients, agents face a disincentive to compete by offering a lower commission rate. ¹⁸ Greater commission rate competition will result when there exist ways to cut commission rates without service degradation.

In the next section we provide a more detailed look at steering. We focus on steering since it is often alleged by industry participants, and there exists evidence that it actually occurs. Moreover, explaining the details of steering, allows us better to demonstrate the specific effects of rebate bans, discriminatory MLS listing policies, and minimum service requirements.

3. STEERING

3.A. STEERING IS A POSSIBLE STRATEGY

Selling agents' ability to steer buyers results from a number of institutional characteristics of the residential real estate brokerage industry.

• Listing agents depend on selling agents to sell listings. Since the early 1920's agent represented residential real estate transactions typically have been carried out through cooperative marketing systems known as multiple listing services ("MLS") in which listing agents pool together their listings. Under this cooperative marketing system, listing agents rely heavily on the cooperative efforts of selling agents to maximize the exposure of their listing and, ultimately, close the transaction. While it is possible for listing agents to sell listings without the cooperation of selling agents, in the majority of transactions, listing agents reach closing with the cooperation of a selling agent.¹⁹
• Easy to identify commission rate reductions by listing agents. In residential real estate transactions, the listing agent is responsible for negotiating commission rates with home sellers. Selling agents receive a portion of the listing agent's commission fee conditional on procuring a buyer for the transaction. Listing agents are required to submit to the MLS the portion, or "co-broke," of their commission that will be awarded to the selling agent. As a result, any price cut by a listing agent manifested in the co-broke is easily observable to all selling agents. Of course, listing agents could offer a lower total commission rate without lowering the co-broke. Even in this case, listing agents who want to attract more customers by offering lower commission rates will need to advertise their offer to potential customers. Making offers known to potential customers will also reveal them to selling agent.²⁰
• Home buyers do not have direct knowledge of all available listings. If home buyers had perfect knowledge of all available listings, it would be difficult for selling agents to steer buyers from considering the listings of price cutting listing agents. Buyers do not have perfect knowledge. In particular, buyers do not have access to the MLS database which contains listings.²¹ Rather, selling agents act as intermediaries between buyers and MLS information. In fact, in many ways, a key role of selling agents is filtering the vast amount of MLS information in a way that is targeted to the buyer's specific preferences. With the ability to filter, comes the ability to hide. Due to the Internet, buyers today have access to far more listing information than they had in the past.²² However, even in cases where buyers have ability to learn about listings selling agents are attempting to hide, buyers may still be at an informational disadvantage. For example, a buyer may bring a listing to a selling agent's attention only to be told that the selling agent is not showing that property since the listing agent is breaking the law.

Under these conditions, it is easy for selling agents to identify a listing agent who lowers her commission rate. Further, since the listing agent is relying on selling agents to help market her listing, selling agents can reduce the listing agent's potential pool of buyers by steering buyers. Finally, because they have access to only less than complete listing information, buyers will not know they are being steered.

3.B. AGENTS MAY HAVE A UNILATERAL INCENTIVE TO STEER

Steering does not require coordinated behavior among selling agents.²³ Steering may simply be an individual selling agent's optimal strategy. As described above, listing agents are required to post their co-broke offer in the MLS database. Selling agents can easily compare the different co-broke offers across listings.²⁴ Other things being equal, selling agents have the unilateral incentive to steer buyers towards listings with higher co-brokes and away from listings with the lower co-brokes since they will generate greater income by doing so.²⁵

If selling agents cared only about co-brokes, one might think they would have no incentive to steer if all co-broke offers were equivalent. However, virtually all selling agents also act as listing agents for other transactions. As a result, selling agents are not only motivated by the co-broke. Ultimately, they are motivated by both co-brokes and the commission they can earn in the capacity of listing agent. As a result, a price-cutting listing agent cannot necessarily prevent steering by simply reducing their total commission but keeping their co-broke offer comparable to other listings.

Consider the case of two identical listings with equal co-broke offers of 3%, but one of the listings carries a total commission rate of 4.5% and the other a total commission rate of 6%. For these two particular listings, a selling agent is indifferent between which home she sells. However, if she equally markets both listings to her buyers (i.e., does not steer), she validates the fact that a home seller only needs to pay 4.5%, rather than 6%, to get equal marketing intensity. In doing so, she will have a difficult time, tomorrow, in her role as a listing agent, trying to justify to a home seller why it is worth paying any more than 4.5%. In short, because selling agents are concerned about their future commissions as listing agents, they may still have a unilateral incentive to steer buyers away from a price- cutting listing agent's listing even when co-brokes are equal.²⁶

The distinction between the role of current co-brokes versus future commissions in the capacity of listing agent creates a potential asymmetry among agents. Virtually all selling agents, regardless of market share, share the same incentive to steer toward listings with higher co-brokes, other things being equal. However, for a given co-broke, agents with more future listings have a greater incentive to steer away from current listings with lower total commissions.²⁷ Agents, or the brokers they work for, with more future listings simply have more to lose from a future reduction in the level of total commissions. As a result, agents with the greater market share have a greater overall incentive to steer. Incidentally, agents or brokers with greater market share also possess the greatest ability to steer since they can influence a larger portion of the potential pool of buyers.²⁸

The asymmetry among agents also suggests that steering creates a greater disincentive to lower co-brokes than total commissions. All agents share the same unilateral incentive to steer away from listings with lower co-brokes. However, given a co-broke comparable to other available listings, agents with larger future market shares will have a greater incentive to steer away from listings with lower overall commissions. Agents with few future listings will have very little incentive to steer buyers away from listings so long as the co-broke is the same. For these agents, the payoff to marketing listings that other (high market share) selling agents might be less interested in marketing can easily outweigh the loss resulting from lower commission fees in the future. When entry is fairly easy and there are many already licensed, but inactive, agents waiting on the sidelines, there are likely many agents of this type. In sum, other things being equal, we would expect price-cutting listing agents to reduce their total commission before lowering their co-broke offer.²⁹

3.C. IMPLICATIONS OF STEERING

We have explained how the possibility of steering creates a disincentive to compete on price (i.e., lowering commission fees). However, we are not suggesting that the possibility of steering will somehow eliminate price competition.³⁰ Even in the presence of steering, we expect agents will continue to compete on price. The possibility of steering simply implies that price-cutters do so with the risk of degrading their service. Listing agents can still win clients adopting a low price, low quality strategy. Our point is that price competition among agents will be less intense than it would be in the absence of steering–in a world where a lower price did not result in lower quality. It follows that a reduction in the ability to steer or the effectiveness of steering will lead to more intense price competition.³¹

The relationship between the ability to steer or the effectiveness of steering and intensity of price competition suggests a number of testable implications. For example, since steering is more difficult when buyers have more complete access to available listings, commission fees should decline as buyers obtain greater access. Since agents with greater market share have a greater incentive to steer and a greater ability to steer, commission fees should decline as markets become less concentrated. Similarly, we expect agents with greater market share to steer more than agents with less share.

We leave it to future work to test the implications of steering, but offer an anecdote from 1927 when a state association official noted, "It has not always been easy, especially in some of the smaller towns, to raise old established rates, but if the brokers will stand by their guns, it will unquestionably work out in the end."³² It stands to reason that steering buyers in smaller towns will be more difficult than in cities since in a small town buyers will likely be aware of all available listings even without the assistance of a selling agent. Since steering is more difficult in small towns, price competition should be more intense. Therefore, the quote is not inconsistent with the predicted effects of steering.

3.D. DIRECT EVIDENCE OF STEERING

Numerous commentators and participants in the residential real estate industry acknowledge steering behavior. A main motivation for the FTC's 1983 investigation was "complaints from sources within brokerage industry claiming harassment and boycotting of [agents] who charge lower than ‘customary' commission rates..."³³ As part of its investigation, the FTC conducted a survey of 154 "alternative" [agents] in 1979. The FTC defined alternative agents as "those who charge and promote a commission rate or fee that is at least 2 percent lower than the fee prevailing in their geographic area, or who offer and promote services that differ significantly from those generally offered in their geographic area." Alternative agents which provided MLS access tended to be "full-service brokers, offering to consumers the same package of services as the traditional brokers." In other words, the FTC's alternative brokers were full-service brokers that undercut their rivals' prices. Of the alternative agents providing MLS access, 91% reported that they either frequently or occasionally experienced "refusals by other brokers to show homes listed by [their] business," with 50% reporting this as a frequent problem. Similar percentages reported threats or disparagements of their business to clients and prospective clients.³⁴ The FTC report also found that these same alternative agents succeeded in selling only 62% of their listings compared to 88% for all agents, and that only 29% of their sales were cooperative sales compared to 66% for all agents.³⁵

To the best of our knowledge, more recent empirical evidence of steering does not exist. Nevertheless, anecdotal evidence abounds. Numerous non-traditional agents have described instances where their listings were subjected to steering, or how they encourage sellers to offer a co-broke comparable to the "prevailing rate." The Government Accountability Office reported in a 2005 study that, "Some discount full-service and discount limited-service brokerage firms we interviewed said that other brokers had refused to show homes listed by discounters."³⁶

A number of comments submitted by non-traditional agents to the joint DOJ/FTC Workshop on "Competition Policy and the Real Estate Industry" also described instances of steering.³⁷ One non- traditional broker wrote, "I have had agents tell me personally they have been instructed by their broker not to show our properties."³⁸ Another wrote:

Another agent, who has initiated several lawsuits against real estate associations and MLSs, wrote:

Finally, a panelist at the Workshop submitted:

There exists academic research considering the impact of non-traditional agents on listings' final sales price, time-on-market and the likelihood they sell at all. While these papers do not purport to examine steering behavior, we can use their results as indirect evidence on the economic significance of steering. Namely, if steering degrades the service of discounting agents, we expect that, other things being equal, their listings will sell for a lower price, stay on the market longer, or be more likely not to sell at all.

Ford and Rutherford (2003) consider the impact of limited service agents on sales price and time-on-market.⁴² Rutherford and Yavas (2005) consider the impact of full-service discount agents on sales price, time-on-market, and the likelihood of whether a listing will actually be sold. In each study, it should be true that the group of agents under consideration are likely to charge lower prices than their counterparts. The basic finding in these two papers is that limited service listings and discounted listings sell for marginally lower prices, stay on the market a little bit longer than non-limited service and non- discounted services, and are slightly less likely to be sold at all. While these results are consistent with steering, they also suggest steering did not have a large impact on the quality of service offered by discounting agents.⁴³

3.E. THE ROLE OF THE INTERNET

Home buyers' lack of complete information about all available listings is one of the key factors that makes steering possible.⁴⁴ As buyers' information about available listings becomes more complete, selling agents' ability to steer is reduced, and, as a result, so is the disincentive to compete on commission fees.⁴⁵ In particular, we suspect the recent explosion of listing information available via the Internet would pose a very serious threat to the success of steering strategies.

4. WEAKENING THE LINK BETWEEN COMMISSION RATE REDUCTION AND SERVICE DEGRADATION: REBATES, NON-DISCRIMINATORY MLS LISTING POLICIES, AND MLS-ONLY LISTINGS

We explained above why greater price competition will result when there exist ways to cut price without service degradation. In this section, we explain how rebates, non-discriminatory MLS listing policies, and MLS-Only listings can break the link between service degradation and price cuts.

4.A. REBATES

In most states, listing and selling agents are permitted to rebate a portion of their commission fees to home buyers and sellers. The immediate impact of a rebate is to reduce directly the overall commission fee paid to listing and selling agents. In this sense, rebates are no different from simply charging a lower commission fee, and one should ask why the ability to rebate can mitigate the disincentive to compete through lower commission rates.

The ability to rebate should not change listing agents' incentives at all. Offering a rebate to attract home sellers is not appreciably different from directly offering a lower commission fee — an option already available to listing agents. As a result, for listing agents the same disincentive to lower their commission fee described above applies to rebates.⁴⁶

In contrast, rebates clearly alter selling agents' ability and incentive to lower their commission fee. Without rebates, selling agents have no control over the commission fee paid by the home buyer and seller.⁴⁷ Rebates give selling agents the ability to control the commission fee. As a result, with rebates, selling agents can compete for additional business by lowering their commission fee. While it might appear that giving selling agents the ability to compete on price may be offset by a disincentive to use this ability, this is not the case.

Home buyers primarily desire to find the house that best matches their preferences at the lowest net sales price possible. As a result, a selling agent might attempt to attract more customers by offering home buyers a higher rebate than her rivals; other things being equal, a higher rebate decreases the net sales price paid by the home buyer. Unlike the case with listing agents offering lower commission fees, a selling agent can offer a higher rebate while keeping other things equal. Namely, offering a higher rebate than her rivals will not change a selling agent's ability to help a buyer find the house that best matches their preferences at the lowest net sales price possible.⁴⁸

Selling agents still require the cooperation of listing agents to bring transactions to closing, and listing agents may have an incentive not to cooperate with selling agents that offer rebates by steering listings away from the these selling agents' buyers.⁴⁹ However, listing agents will not have the ability to steer. The selling agent can easily see all available listings in the MLS, and when she finds the best match for her buyer, she will contact the listing agent directly.⁵⁰ The listing agent has the fiduciary duty to convey all offers to the home seller.⁵¹ Even in the event that the listing agent attempts to steer the listing away from a buyer, the selling agent will already know where the desired property for sale is located and can notify the home seller without going through the listing agent if necessary (and the listing agent knows this).⁵²

In sum, rebates provide a means for cutting price that mitigates quality degradation. First, rebates allow selling agents the opportunity to directly compete on price. Second, a price cut by a selling agent is not as susceptible to steering as a price cut by a listing agent. As a result, rebates are a particularly useful vehicle for increasing price competition.⁵³ A rebate ban would prevent selling agents from competing on price. Moreover, a rebate ban would eliminate a type of price cut that is not susceptible to steering. Currently, in eleven states agents are banned from offering rebates.⁵⁴

4.B. NON-DISCRIMINATORY MLS LISTING POLICIES

Buyers are more likely to learn about a listing if it appears in a publicly viewable database.⁵⁵ In particular, there is no way for selling agents to hide publicly viewable listings from buyers. Once a buyer is aware of listing, steering that buyer away from the listing will be more difficult than if the buyer never learned about the listing. As a result, a listing that appears in a publicly viewable database is less susceptible to steering. Non-discriminatory MLS listing policies assure that lower commission listings will remain publicly viewable.

Discriminatory MLS listing policies prevent certain types of listings from appearing in publicly viewable databases of listings derived from the MLS (e.g., realtor.com). For example, some MLS policies prevent listings that are not "exclusive right to sell" listings from appearing in publicly viewable databases.⁵⁶ Yet, it is precisely those listings that include lower commission fees or allow home sellers to undertake a greater share of marketing effort that are most likely to be listed as something other than "exclusive right to sell." As a result, these policies effectively prevent many lower commission listings from appearing in publicly viewable databases.

Any discriminatory MLS policy that automatically prohibits lower commission listings from appearing in publicly viewable databases creates a direct link between service degradation and price cuts. Such MLS policies enhance selling agents' ability to steer buyers away from lower commission listings. Non-discriminatory MLS listing policies mitigate the degree of quality degradation that would otherwise result.

4.C. MLS-ONLY LISTINGS

Home sellers have always had the option to act as their own listing "agent." Listings sold in this way are commonly referred to as "for sale by owner" or "FSBO" (pronounced "fizz-bo"). By substituting their own effort for that of a listing agent, FSBO sellers avoid paying any listing agent commission.⁵⁷ However, because FSBO sellers still need to find potential buyers, selling agents still play a important role in the success, or lack thereof, of the FSBO option.⁵⁸

When a home seller chooses not to enlist the services of a listing agent, their listing will not appear in the MLS.⁵⁹ As a result, selling agents cannot learn of FSBO listings via the MLS. Similarly, potential buyers cannot view FSBO listings in publicly accessible databases derived from the MLS. Selling agents who choose to steer in response to the commission reduction will be able to do so relatively easily.⁶⁰ While the FSBO option offers a home seller an extreme commission reduction, there is substantial risk of degrading the potential buyer pool. In other words, with the FSBO option there is a direct link between a commission rate reduction and the quality of selling agent services a home seller can obtain.

Some listing agents offer to list a home seller's home in the MLS and provide no additional services for a flat fee. This type of offer is commonly known as an "MLS-only" listing and, depending on the geographic location, is typically priced around $400.⁶¹ The MLS-only option largely preserves the FSBO approach to selling, but for a nominal fee places the listing in the MLS.⁶² Placement in the MLS exposes the listing to all selling agent members of the MLS, and makes the listing visible in publicly viewable databases derived from the MLS. As a result, a MLS-only listing is less susceptible to steering than a FSBO listing. An MLS-only listing offers a comparable price cut to the FSBO option without the same amount of service degradation. Therefore, the MLS-only option weakens the link between a lower commission and service degradation.

In some states, listing agents are subjected to minimum service requirements.⁶³ Minimum service requirements prevent listing agents from offering an MLS-only option. In most cases, an agent who wants to provide an MLS-only option will also be required to provide delivery of offers and counteroffers, to assist in the developing or negotiating offers and counteroffers, and to answer any questions regarding offers and counteroffers.⁶⁴ These additional services will increase agents' costs and, as a result, limit the degree to which agents can charge lower commissions for an MLS-only type service.

In sum, while MLS-only listings offer a means to break the link between a lower commission and service degradation, a minimum service requirement limits the amount by which commissions may be lowered. A higher price for the MLS-only option will make it a less attractive option.

4.D. REBATE BANS, DISCRIMINATORY MLS POLICIES, AND MINIMUM SERVICE REQUIREMENTS

Rebates, non-discriminatory MLS policies, and MLS-Only listings all provide a means severing the link between lower commission rates and service degradation. In each case, mitigating quality degradation results from frustrating agents' ability to steer. In this light, rebate bans, discriminatory MLS policies, and minimum service requirements can all be viewed as policies which accommodate steering behavior and chill price cutting behavior. Preventing these measures can increase price competition for brokerage services.

5. WHY MITIGATING DISINCENTIVE TO COMPETE ON PRICE MATTERS

Regardless of the specific reason for any disincentive to compete on price, in this section, we explain why mitigating any disincentives to compete on price is important for economic efficiency. In the textbook model of perfect competition, any price exceeding the minimum of long-run average cost attracts entrants who increase supply until price is reduced to the perfectly competitive level. The reduction of price to the minimum of long-run average cost generates two important results. First, the amount of production and consumption is efficient (i.e., consumer and producer surplus are maximized).⁶⁵ Second, production is achieved at the minimum possible cost.⁶⁶

If price persists above the perfectly competitive level, these two important results will no longer obtain. First, price will remain above the minimum of long-run average cost leading to an inefficient level of production and consumption. This first effect results in what is commonly known as "dead- weight loss." Second, entrants and existing firms may expend resources to capture the available rents (i.e., profits) resulting from the persistently supra-competitive price. Such rent-seeking behavior has the effect of turning rents into costs, and implies that production is no longer achieved at the minimum possible cost.⁶⁷ Dead-weight loss coupled with rent-seeking costs comprise the overall social cost of the disincentive to compete on price. The magnitude of these two components depends on the particular characteristics of the market under consideration.

Consumer demand for residential real estate brokerage services is likely to be relatively inelastic.⁶⁸ As a result, a supra-competitive price will not generate much dead-weight loss. The primary effect of a supra-competitive price will be the transfer of consumer surplus from consumers to brokerage firms. Therefore, mitigating any disincentive to compete on price will not eliminate much dead-weight loss since there is likely very little to begin with.⁶⁹

On the other hand, mitigating the disincentive to compete on price for residential real estate brokerage services could eliminate substantial social costs associated with rent-seeking. A number of academic authors have written about why the residential real estate brokerage services industry is likely to suffer from rent-seeking behavior. These authors make the basic point that supra-competitive prices combined with easy agent entry will likely lead to higher average cost per transaction as more and more agents enter and compete for a fixed number of transactions via non-price means.⁷⁰ In equilibrium, non-price competition will completely dissipate any agent rents resulting from supra-competitive prices.⁷¹ In this case, mitigating the disincentive to compete on price can have a big effect. Namely, greater price competition will return rents to consumers in the form of consumer surplus and prevent rents from being transformed into costs.

In a recent article, Hsieh and Moretti attempted to empirically test whether a lack of price competition has led to wasteful rent dissipation through excessive agent entry.⁷² Their basic hypothesis is that, "if commission rates do not vary and barriers to entry are low, ... there will be more market entry and more social waste in cities with higher housing prices."⁷³ The testable implications of their hypothesis are threefold: first, "the fraction of [agents] in a city (relative to the population) should be increasing in the price of housing;" second, "the productivity of an average [agent] should be lower in a city with higher housing prices;" and third, "[agents] should be indifferent across cities: their real wage should be uncorrelated with the price of housing."⁷⁴ Using data from 1980 and 1990, the authors cannot reject any of these testable implications. The authors conclude that "higher commission fees in more expensive cities are dissipated by excessive entry of [agents]." Finally, they estimate that excessive agent entry led to between $1.1 billion and $8.2 billion (depending on assumptions) in social loses across 282 U.S. cities in 1990.

There exists other evidence supporting the hypothesis that relatively easy agent entry will dissipate available rents. For example, NAR commissioned a study by Arthur D. Little in the late 1970's. Arthur D. Little reported, "Salesperson transactions rates in the residential real estate brokerage industry (houses listed or sold per unit time per full-time equivalent salesperson) have been decreasing in recent years."⁷⁵ According to the FTC's 1983 report, "reasons for declining productivity [in the Arthur D. Little report] include a tendency for firms to compete by adding staff or increasing services to clients rather than reducing price."⁷⁶

The FTC report also provides interesting historical evidence from California of excessive entry in response to available rents. In 1936, shortly after the Great Depression, the California market was booming. "With commissions fixed, the average price of homes and lots increasing, and the number of transactions increasing, increasing entry had become a problem. The Chairman of the California Association's Brokers' Division observed that there was a ‘Flood of New Licensed Agents–The increase in licensed brokers is all out of proportion to the increase in business up to the present time.' The Chairman called for the better ‘protection' for brokers."⁷⁷ Despite attempts to choke off entry with tougher licensing laws, the entry "problem" continued. In 1958, the Real Estate Research Program at UCLA undertook a study for the California Association of Realtors regarding the Association's proposed increase in the recommended commission rate from 5 percent to 6 percent. According to the FTC report:

More recently, Geoffrey D. Lewis of Re/Max International, Inc. made a similar assessment. Lewis testified to Congress that, "The lack of increase in agent income despite rising housing prices is due to the large increase in the number of agents entering the industry. ...These agents are being drawn in by the increase in housing prices. ... As a result of the increase in the number of agents searching for transactions, the average number of transactions per agent is decreasing."⁷⁹ At the same time, there is evidence that agent income has been decreasing. According to NAR's President elect, Pat Vredevoogd, "Income of Realtors working as sales agents, who make up two thirds of the membership of NAR, [decreased] from $41,600 (2002) to $38,300 (2004)."⁸⁰ A NAR economist reported, "That's not surprising. The number of new agents entering the market in the past couple of years has outpaced the home sales growth and even the home price growth. So, given the fact that the Realtor membership has increased far more than actual home sales, it's not surprising that the median income has fallen."⁸¹

Data on the number of NAR members, number of home transactions and commissions are consistent with these assertions. Over the past 30 years, one can see a clear decrease in agent productivity when commission fees increase. Consider the Chart in the Appendix which displays inflation adjusted commission fees per transaction alongside the number of transaction sides per NAR member.⁸² Commission fee increases in the late 1970's, late 1980's and early 2000's are all accompanied by sharp declines in the number of transaction sides per NAR member.

The combination of higher commission fees, declining agent productivity, and lower agent income has been described as "The Tragedy of the Commission" by Hsieh.⁸³ Consumers are clearly worse off when commission fees increase, but the real tragedy is that the agents themselves are no better off.⁸⁴ Not surprisingly, NAR (and other interested parties) has attempted to co-opt this "tragedy" as evidence of fierce competition. However, NAR's claim of "intense competition" mistakes rent seeking waste (i.e., competition for rents) for allocative and technical efficiency (i.e., true competition). Without a doubt, agents are competing fiercely. But, if commission fees remain elevated, agents are simply wasting resources by scrambling for a bigger, though elusive, share of the pie.

Finally, it is important to note that rent seeking entry on the part of agents does not imply that all rents from higher commission fees will be transformed into costs. There are many other industry participants with claims on commission fees. As explained by Vredevoogd, "real estate licensees work as independent contractors who are individually responsible for their business and administrative expenses, which typically include: state license fees, MLS membership fees, continuing education expenses, professional association dues, ... [and] franchise fees."⁸⁵ Of course, brokerages also take their cut. To the extent entry among these entities (i.e., professional associations, MLSs, etc.) is more difficult, rents accruing to them are less subject to rent seeking waste.⁸⁶ While that is "good" in the sense that it implies a portion of the rents will less likely be converted to costs, it also provides motivation for these entities to maintain the status quo generating the wasteful rent seeking behavior in the first place.⁸⁷

In summary, supra-competitive commission fees will create both allocative and technical inefficiencies. While both reduce total welfare, we suspect that the latter is more significant. More specifically, there exists considerable evidence that persistently high commission fees generate wasteful rent seeking behavior. Any factor creating a disincentive to compete on price, contributes to the wasteful behavior. By supporting, or strengthening, the disincentive to reduce commission rates, rebate bans, discriminatory MLS listing policies, and minimum service requirements exacerbate technical inefficiency (i.e., raise the costs) of providing residential real estate services and decrease consumer welfare.

APPENDIX

[D]

Sources: Number of transactions from HUD, "U.S. Housing Market Conditions," 4th Quarter, 2006. Number of National Association of Realtors members from www.realtor.org. Total commissions from Department of Commerce, National Income and Product Accounts.

FOOTNOTES

* U.S. Department of Justice, Economic Analysis Group. The views expressed are not those of the U.S. Department of Justice. I greatly appreciate numerous helpful comments from Ken Heyer and W. Robert Majure. Any remaining errors are my own.

1. We realize "commission fee" appears redundant. However, the distinction between commission rates and fees is often confused. Consider, for example, recent testimony before Congress, "... full-service commission rates have been trending down over the past decade. They have gone from 6.1 percent in 1991 to the current average rate of 5.1 percent. That is not a one percent decrease, it is a 16% decrease. How many other industries have experienced that level of decrease in pricing over the same period?" (emphasis added). Testimony of Geoffrey D. Lewis on behalf of Re/Max International, Inc., before the Subcommittee on Housing and Community Opportunity, Committee of Financial Services, U.S. House of Representatives, July 25, 2006. Obviously, a decrease in the commission rate does not imply a decrease in "pricing," i.e., the commission fee. To avoid any confusion, we always qualify commission with either fee or rate. Ultimately, it is the fee that consumers care about.

2. Real commission fees per home transaction computed using CPI-deflated residential "brokers' commissions on sale of structures" from U.S. Department of Commerce, National Income Accounts, divided by number of residential transactions from U.S. Department of Housing and Urban Development, "U.S. Housing Market Conditions," 4^th Quarter 2006.

3. Number of licensed agents estimated based on number of National Association of Realtor members. See, www.realtor.org/libweb.nsf/pages/fg003. Number of transactions based on U.S. Department of Housing and Urban Development, “U.S. Housing Market Conditions,” 4^th Quarter 2006.

4. If the premise is that there is little price competition, arguing that rebates – a form of price competition – will generate more price competition begs the question. What makes rebates different? Similarly, why should entry by limited service agents lead to more price competition when there is ample evidence of substantial full service agent entry. What makes limited service agents different?

5. Note, we are not arguing that commission rates are uniform as others commonly argue. Indeed, the evidence we have seen argues against commission rate uniformity. For example, see "The Price of Residential Real Estate Brokerage Services: A Review of The Evidence, Such As It Is," John C. Weicher, Real Estate Law Journal, vol. 35, 2006.

6. By law, commission rates are negotiable. Agents work for brokers, and brokers often influence the listing contract commission rate. We ignore this extra level of complication throughout this paper and simply refer to "agents."

7. hungryagents.com provides a technology for home sellers to execute an actual auction (though auction participation is limited to member agents). It has been alleged that many home sellers do not realize that commission fees are negotiable. We are skeptical of such claims. However, it might be the case that many home sellers are reluctant to negotiate commission rate for reasons we describe below.

8. For a good description of some of the key points of divergence, see "Property Rights and the Pricing of Real Estate Brokerage," Randall Bartlett, Journal of Industrial Economics, 1981.

9. For an empirical test of the agency problem see, "Market Distortions When Agents Are Better Informed: The Value of Information in Real Estate," S. Levitt and C. Syverson, NBER Working Paper, 2005. Levitt and Syverson find that when agents sell their own homes (i.e., agents incentives are perfectly aligned with home owner's incentives), their homes sell for about 4% more and stay on the market about 10 days longer, after controlling for a wide range of housing characteristics. In other words, the authors find evidence that home sellers are induced by agents to sell too quickly and at a price that is too low.

10. Of course, there are likely other factors that influence commission rates. We choose to focus on the principal-agent problem and steering because of their prominence.

11. Costly effort refers to any costly input that increases the net sales price or decreases the time on market. Examples include: open houses, advertisements and time spent searching for buyers. Costly effort is distinguished from a listing agent's (quasi) fixed costs of selling a listing, i.e., the cost of listing the home on the MLS, researching comparable properties, dealing with paper work, etc.

12. More complex listing contracts might provide another way to generate more effort. Describing the design of such contracts is beyond the scope of this paper.

13. For a mathematical model that generates this result, see "Downward Rigidity in Real Estate Commissions," R. Santore and D. Bruce, University of Tennessee Working Paper, 2001.

14. Discouraging buyers from purchasing can take many forms. It might be something as subtle as an agent showing a buyer two listings, but emphasizing the attractive features of one and the unattractive features of the other, or telling a buyer a particular listing is not open for showing at a time that is convenient for the buyer. Discouragement can take more extreme forms such as telling a buyer that the listing agent for the desired property is unethical or breaking the law (sometimes referred to as "spooking"), or simply by refusing to show certain agents' listings to buyers ("boycotting"). Note: The terms steering, spooking and boycotting may be used slightly differently among different industry commentators.

15. Previous authors have described this potential effect of steering. "Kickbacks, Specialization, Price Fixing, and Efficiency in Residential Real Estate Markets," Bruce Owen, Stanford Law Review, 1977 ( p. 949, "...MLS members simply may be less anxious to sell homes listed by brokers who charge less than the fixed rate.") "Competition and Efficiency in Transacting: The Case of Residential Real Estate Brokerage," John H. Crocket, AREUEA Journal, 1982 (p. 218, "A brokerage firm that solicits listings by offering lower commissions may find itself disadvantaged since agents from other firms may prefer to concentrate their efforts where higher commissions are available. This tendency may be reinforced by boycotts or other discriminatory practices.") "The Residential Real Estate Brokerage Industry," Federal Trade Commission Staff Report, 1983, hereafter, "FTC Report" (p. 40, "One result of cooperating brokers' apparent tendency to steer buyers away from the listings of discount brokers is that discount brokers may be at a substantial disadvantage in marketing their listings." and, p. 55, "Prices do not drift, we hypothesize, because price-cutters will lose the crucially valuable cooperation of their fellow brokers.") "Real Estate Brokerage: Factors That May Affect Price Competition," Government Accountability Office, 2005, hereafter, "GAO 2005" (p. 13, "A discount broker may advertise a lower commission rate to attract listings, but the broker's success in selling those homes, and in attracting additional listings in the future, depends in part on other brokers' willingness to cooperate (by showing the homes to prospective buyers) the sale of those listings.") "The Residential Real Estate Brokerage Industry: What Would More Vigorous Competition Look Like?" Lawrence White, New York University, Law & Economics Research Paper Series, 2005 ( p. 6, "A high-fee-upholding agent who has a potential buyer may threaten to or actually boycott the listings of a price-cutting seller's agent.") "Bringing More Competition to Real Estate Brokerage," Robert Hahn, Robert Litan and Jesse Gurman, AEI-Brookings Joint Center for Regulatory Studies, 2005 (p. 10, "Other agents can avoid bringing buyers to houses listed by agents who offer lower commissions or operate under an alternative business model.") While these authors have described the potential link between price cutting and steering by rivals, they have only provided a partial explanation for the existence of strategy. Below, we provide a more thorough explanation.

16. This is just another dimension of the homeowner's principal-agent problem described in the previous section.

17. The causal link between price (i.e., commission fee) and quality (i.e., time on market and gross sales price) for brokerage services is a clear departure from the textbook version of price competition where price can be altered independent of product quality. More generally, the resulting disincentive to compete on price is not uncommon in theoretical models. Well accepted models describe a variety of circumstances (e.g., differentiated products, capacity constraints, retaliation) in which marginal cost pricing will not obtain even under ostensibly competitive conditions. The pivotal feature of these models is that undercutting rivals' prices does not result in enough new business to make a further price cut profitable. When it is difficult to win additional customers by cutting price, the incentive to do so is mitigated. A unique feature of the instant case is that the quality of an agent's service is substantially controlled by his rivals.

18. While the direction of the disincentive is clear, the magnitude of the disincentive will depend on a number of (difficult to quantify) factors. Among others, these include: how much steering takes place (I.e., If steering were perfect, in the sense that a commission fee reduction resulted in no buyers coming to view the listing, there would be no incentive to cut the commission fee since no home seller would ever take the offer.), the degree of heterogeneity among home buyers (I.e., If all home buyers are identical, then a larger pool of buyers is relatively unimportant — anything less than nearly complete steering would be ineffective.), and home buyer discount factors (I.e., If home buyers don't care much about how quickly their homes sell, steering will not be as effective.). Quantifying the size of the disincentive is beyond the scope of this paper in which we simply argue that the disincentive exists, and that rebates, limited service agents, and non-discriminatory MLS listing policies mitigate the disincentive.

19. FTC Report, 1983 (p. 37, 92% of agent assisted transactions are listed on MLS, and "...approximately 66 percent [of sales] involve more than one broker."). According to the National Association of Realtors, "In many markets across the country, over 50% of real estate sales are cooperative sales." See, http://www.realtor.com/basics/allabout/realtors/why.asp?poe=realtor.

20. Secret price cuts after a contract is signed will not affect the initial competition to win listings.

21. In a comment provided for the "Competition Policy and the Real Estate Industry" workshop (hereafter, "DOJ/FTC Workshop") held by the Federal Trade Commission and the Department of Justice, October 25, 2005, Barry stated, "Buyers are never aware they are being steered. The buyer agent makes a selection of homes to show, and since the public sources of homes never show the commission offered, buyers are never aware when their agents select out the homes with lower priced commission offerings." Workshop public comment 19 at 57. See, http://www.usdoj.gov/atr/public/workshops/reworkshop.htm.

22. We discuss in more detail below how public access to listings can limit agents' ability to steer.

23. Note, the "cooperative marketing" described above has nothing to do with "coordination" in an antitrust sense. It merely indicates that more than one agent is often involved in real estate transactions. Unfortunately, loose language can generate confusion about whether steering is a unilateral strategy or the result of coordination. For example, the FTC Report, p. 36, "Individual brokers, we hypothesize, police the system by withholding cooperation in selling listings which carry a lower than customary ‘split' or commission. In doing so, they engage in both typical profit-maximizing (refusing to sell their services for less than the going rate) and also prevent a collective lowering of commissions generally." Of course, just because steering is unilaterally optimal does not somehow imply that there exists no coordinated behavior. Our point is simply that steering in no way requires coordinated behavior. It is beyond the scope of this paper to discuss whether there exists some coordinated effort to maintain commission fees. N.b., if there were some sort of coordination on price, agents' unilateral incentive to steer in response to price-cutting rivals would provide a powerful mechanism for sustaining coordination. This may help explain the apparent success of explicit price-fixing behavior throughout much of the last century. For more detail on explicit price-fixing behavior see FTC Report, p. 195.

24. In practice, comparing the potential commission fee from one listing to the next is more difficult than simply comparing the relevant co-brokes. For example, selling agents cannot foretell the final sales price, how long it will take to close a transaction, how much cost they will entail to close a transaction, or the likelihood their buyer gets the home. In other words, other things are rarely equal. Selling agents' inability to perfectly compare potential commissions across listings generates some protection from steering for price cutting listing agents. Nevertheless, we presume selling agents will do their best to generate higher commissions.

25. Previous authors have noted this unilateral incentive. Supra, fn. 15. FTC Report, p. 75, "As a matter of self-interest, brokers may tend to steer buyers toward the homes that involve a better commission split just as merchants may tend to promote the sale of those items which will bring them the largest returns." GAO 2005, p. 13, "When choosing among comparable homes for sale, brokers have a greater incentive — all else being equal — to first show prospective buyers homes that offer other brokers the prevailing commission rate than homes that offer a lower rate." Note, we are not alleging a breach of fiduciary duty though it certainly may result.

26. Whether the incentive to steer exists depends on whether the future (discounted) benefit of a higher commission rate outweighs the present cost of forgoing a commission at a lower rate.

27. Of course, the co-broke need not be identical across listings for this to be true.

28. We note here that due to heterogeneous incentives to steer among agents, a potentially interesting topic for future research is the possibility of "coordinated steering." To give one example, suppose there is a broker with large market share, and many smaller brokers in a market. According to the above, the smaller brokers might have limited incentive to steer so long as co-broke is comparable. The largest broker, however, will have a strong incentive to steer even with a comparable co- broke. In response to a price-cutting entrant, the largest broker might make it known that he is boycotting the entrant's listing, and anyone else who doesn't also boycott will be boycotted by the largest broker as well. Since the MLS publishes the name of the selling agent, the largest broker will know which of the smaller brokers breaks the boycott, and with a large market share, the punishment doled out by the largest broker could be severe.

29. Many flat-fee, MLS-only listing offers encourage their clients to offer the prevailing MLS co-broke in order not to discourage selling agents from showing their listings. See, for example, http://flatfeelisting.com/mls.htm or http://www.valuemls.com/faq.htm.. If the listing agent has to absorb the entire commission reduction, she will likely not be willing to reduce the total commission by as much as otherwise.

30. In the extreme case where steering implied no buyers could observe properties with lower commission fees, price competition would cease. Such an extreme case is nothing more than an abstract hypothetical.

31. Consider the analogy of a differentiated product market. Just because price-cost margins are positive, does not imply that differentiated product rivals do not compete on price. Indeed, they do. Moreover, as the degree of differentiation decreases, price competition becomes more intense (i.e., price approaches marginal cost).

32. FTC Report at 195.

33. FTC Report at 1.

34. In one case a letter bearing the local board of realtors' letterhead was sent to clients who had listed their homes with an alternative broker. The letter read, "As a homeowner who has listed their home for sale with _____, please be advised that numerous Realtors, associated with the MLS, have chosen not to show or sell your home, due to the unethical conduct of _____ towards their fellow Realtors." FTC Report, p. 156. Today, such threats are referred to as "spooking." See, http://openmlsinstitute.org/en/cms/?112, "Owners often hold their own open houses when listed with a discount realty firm. Unethical agents attempt to spook these owners by telling them agents will steer buyers away from the home."

35. FTC Report at 154. While these statistics are consistent with steering, other factors could generate similar results. For example, agents charging lower commissions may exert less effort in selling homes.

36. GAO 2005 at 13.

37. See, http://www.usdoj.gov/atr/public/workshops/reworkshop.htm.

38. Durham, Public Comment 15 at 1.

39. Moore, Public Comment 215 at 1.

40. Barry, Public Comment 19 at 57.

41. Farmer, DOJ/FTC Workshop Tr. at 74.

42. "Impact of TRELA 1995, Limited Service Representation, and Agent Certification in Texas on the Marketing and Pricing of Single-Family Residential Real Estate," A Technical Report delivered to the Real Estate Center at Texas A&M University, J. Ford and R. C. Rutherford, June, 2003.

43. The authors did not intend to address the efficacy of steering. As a result, they did not control for a number of potentially confounding factors. For example, it might be the case that sellers choosing to use a limited service or discount agents do not require as much marketing from selling agents to get a good price for their house. In other words, it is possible that some groups of sellers are relatively immune to the degrading effects of steering, and it is precisely these sellers that are most likely to choose a service that may otherwise be susceptible to steering. For example, a recent NAR survey suggests that many for-sale-by-owner sellers sell their homes to someone they know. For many of these sellers, the buyer was likely predetermined and steering would not have been a factor since the seller was not relying on selling agents to procure a buyer. Also, both studies rely on data from a single MLS in the Dallas/Ft. Worth around 2001-2002. It is possible that steering has a bigger effect in other areas or at different times (e.g., not in a sellers' market).

44. See section 3, "Steering" above.

45. "The 4 ½% Solution," Money, October 1, 2004 ("With 70% of home shoppers now surfing their local MLS on the Web, any agent who tries to boycott a discounter ... risks having to explain to clients why they weren't shown, say, that little colonial on Maple Lane that so perfect online. As [one discount agent] puts it, ‘Buyers are smart. If their agent won't show them a property they want to see, they're going to call the listing agent.'")

46. Rebates may enable listing agents to lower commission fees without detection in a way that does not seem possible without rebates. Namely, a listing agent can pay rebates at closing outside of the settlement. While such a transfer from a listing agent to a home seller increases the consumer surplus of the home seller, it should have no effect on commission fee competition between listing agents to win additional customers since it takes place after the competition for listings occurs. Of course, a listing agent might develop a reputation for rebating or even advertise the fact that they will rebate. In that case, rebating would directly affect price competition. However, it would also fall prey to the exact same type of steering described above. The founder of an online referral service offering buyers and sellers rebates noted the importance of secrecy in a recent article, "[Rebating] has to be done quietly... The confidentiality of this makes it work." See, "‘Secret Agents' Quietly Offer Real Estate Rebates," Glenn Roberts Jr., Inman News, March 7, 2006. Even if rebating does not directly affect commission fee competition, it is an important vehicle to maintain because it provides a way for listing agents to transfer surplus to consumers.

47. As described above, the commission rate is negotiated by the listing agent and the home seller.

48. Here we are focusing on the fact that a lower selling agent commission fee will not change the amount of cooperation from rival agents. It may be true that a rebating selling agent will have a weaker incentive to exert costly effort in finding a suitable home for his buyer. We do not expect this to be an important effect since it is often the case that rebate offers are made conditional on buyers undertaking more search effort themselves. In other words, rebate offers can be thought of as a partial solution to the principal-agent problem.

49. Our focus above was on selling agents steering buyers away from sellers. Listing agents may also attempt to "steer" listings away from certain buyers. For example, listing agents might refuse to show their listing to an interested buyer. White (2005), see note 15 above, and GAO 2005 both mention this possibility. Listing agents may have an incentive to do so because listing agents today, are selling agents tomorrow. By cooperating with a rebating selling agent today, a listing agent will be supporting more aggressively pricing rivals they will have to compete with tomorrow. Of course, as discussed above in the context of selling agents, for many listing agents the potential future (discounted) loss may easily be outweighed by the immediate gain. Aside from the potential future loss, a listing agent will seemingly have a preference for buyer's represented by a rebating selling agent. The listing agent will get her agreed upon commission rate regardless of what the selling agent does. Moreover, a buyer who knows he is getting a rebate should be willing to offer a higher purchase price than a buyer who is not getting rebate, and a higher sales price implies a bigger payoff for the listing agent. In short, apart from the future potential loss, it seems listing agents may have a disincentive to steer their listings away from buyers getting discounts. As a result, evidence of listing agent steering would indicate that some agents consider the potential future loss to be substantial. See footnote 53 for an example of listing agent steering.

50. Note that in the converse case where a selling agent attempts to hide listings from potential buyers (i.e., steer buyers away), the listing agent has no direct way of knowing the potential buyer even exists. For example, listing agents cannot see an inventory of buyers in the MLS database.

51. Note that in the converse case when a selling agent attempts to steer a buyer from a listing, the fiduciary duty of the selling agent to the buyer is far less clear. In most cases, the selling agent in a transaction acts as the sub-agent of the listing agent. In that case, the selling agent has a fiduciary duty to the seller, not the buyer. Additionally, selling agents are in no way forced to accept a listing agent's offer of sub-agency. In this light, a selling agent who steers a buyer is simply turning down an offer to be a sub-agent.

52. In some instances it may be difficult for the selling agent to contact the home seller directly. For example, the home owner may no longer live in the home for sale. A recent news article described what agents from Redfin, an online brokerage featuring rebates, do when faced with listing agents' attempts to steer: "Redfin agents contact the sellers and let them know that their agent will not show the house. When they cannot find a phone number, they send a registered letter. When sellers have moved, they track them down through the relocation service that moved them." See, "The Last Stand of the 6-Percenters?" Damon Darlin, New York Times, September 3, 2006. Also note that MLS databases often list the property owner's name, and sometimes their telephone number.

53. Since rebates are commonly paid outside of settlement, it is extremely difficult to obtain any data on the prevalence of rebates and, as a result, (net) commission fees. This is particularly unfortunate since evidence showing insignificant amounts of rebating would tend to reject steering as an explanation for limited price competition. Alternatively, evidence of limited commission rate (i.e., the listing contract commission rate) competition coupled with significant rebating would be consistent with steering behavior.

54. Alaska, Oregon, North Dakota, Kansas, Oklahoma, Missouri, Louisiana, Mississippi, Alabama, New Jersey and Tennessee.

55. A recent survey by National Association of Realtors found that 80 percent of home buyers used the Internet during their home search in 2006, and 24 percent of recent home buyers first located the home that they purchased on the Internet. See, 2006 National Association of Realtors Profile of Home Buyers and Sellers.

56. For an example involving the board of realtors in Austin, TX see, www.ftc.gov/opa/2006/07/austinboard.shtm.

57. In most cases, FSBO sellers will still offer compensation to any selling agent who procures a buyer.

58. According to the National Association of Realtor's 2005 Profile of Home Buyers and Sellers, 65% of home buyers first learned the home they purchased was for sale via a real estate agent (36%) or the Internet (29%)–the vast majority of internet searches focused on real estate agent websites, 15% first learned from a yard sign, 6% from an acquaintance, and 5% from print newspaper advertisement.

59. Only member agents of the local MLS are permitted to enter listings in the MLS.

60. Indeed, selling agents who rely on the MLS for listing leads may not even become aware that certain FSBO listings are available.

61. See, for example, www.flatfeemlslisting.com and www.forsalebyowner.com. In virtually all cases, home sellers choosing the MLS-only option will need to offer compensation to any selling agent procuring a buyer.

62. Consider the following example. In 2006, the average price for an existing home in the U.S. was $268,300 (HUD, U.S. Housing Market Conditions, fourth quarter, 2006). Suppose the selling agent in a typical transaction is paid 3% of the sales price, or $8,049. A FSBO seller would pay nothing to a listing agent, but most likely still offer to compensate the selling agent $8,049. Suppose an MLS-only seller would pay $500 to a listing agent, and $8,049 to the selling agent for a total of $8,549. In return for the extra $500, or 6% more money, the MLS-only seller gets his listing listed in the MLS. Since MLS- only fees do not typically increase with the price of the home, the percent increase will be even smaller for homes selling above the average price. Of course, whether an extra $500 is worth it or not depends on how much service degradation is eliminated by including the listing in the MLS. Unfortunately, we have no way to measure this.

63. Currently, Alabama, Idaho, Illinois, Indiana, Iowa, Texas, and Utah have minimum service requirements. Further, Kentucky, Michigan, Mississippi, and New Mexico recently have considered but not adopted minimum service requirements.

64. The law in Missouri is fairly representative and states that all agents who enter into an exclusive brokerage agreement "...shall provide at a minimum, the following services: (1) Accepting delivery of and presenting to the client or customer offers and counteroffers to buy, sell, or lease the client's or customer's property or the property the client or customer seeks to purchase or lease; (2) Assisting the client or customer in developing, communicating, negotiating, and presenting offers, counteroffers, and notices that relate to the offers and the counteroffers until a lease or purchase agreement is signed and all contingencies are satisfied or waived; and (3) Answering the client's or customer's questions relating to the offers, counteroffers, notices, and contingencies. See,Mo. Rev. Statutes § 339.780(7)(1)-(3).

65. In economics, this result is referred to as "allocative efficiency."

66. In economics, this result is referred to as "technical efficiency."

67. See, "Price and Non-Price Competition," George Stigler, Journal of Political Economy, Vol. 76, No. 1 (1968); "The Social Costs of Monopoly," Richard A. Posner, Journal of Political Economy, Vol. 83, No. 4 (1975); "Quality Variation When Prices Are Regulated," Lawrence J. White, Bell Journal of Economics and Management Science, Vol. 3, No. 2 (1972); and The Theory of Industrial Organization, Jean Tirole, 1989.

68. The main alternative to agent services is for home sellers to sell on their own. Most people lack the time or expertise to sell their home. Further, access to the MLS is conditional on hiring an agent. Commission fees are also a relatively small percentage of the overall value of a transaction. In particular, higher commissions are typically accompanied by higher property values, so there may be an offsetting income effect.

69. Of course, if one is concerned with maximizing consumer welfare, then mitigating the disincentive to compete on price will have a favorable effect. From an economic perspective, maximizing total (i.e., consumer and producer) welfare is more important than maximizing consumer welfare.

70. See, "Competition and Efficiency in Transacting: The Case of Residential Real Estate Brokerage," J. H. Crocket, Journal of the American Real Estate and Urban Economics Association, 1982; "The Welfare Effects of Non-Price Competition Among Real Estate Brokers," T. J. Micelli, Journal of the American Real Estate and Urban Economics Association, 1992; "Real Estate Brokers, Nonprice Competition and the Housing Market," G. K. Turnbull, Real Estate Economics, 1996; and "Agency and Brokerage of Real Assets in Competitive Equilibrium," J. T. Williams, The Review of Financial Studies, 1998.

71. This is the extreme outcome. It need not be the case that rents are completely dissipated. Further, while non-price competition will generate additional costs, it is possible for these additional costs to be offset by any consumer surplus that is created by non-price changes in service. For example, airline passengers surely enjoyed the side effects of the "sandwich wars" during the period of airfare regulation. Of course, if the additional consumer surplus were greater than the additional cost, such services would have already been provided.

72. See, "Can Free Entry Be Inefficient? Fixed Commissions and Social Waste in the Real Estate Industry," C. Hsieh and E. Moretti, Journal of Political Economy, 2003.

73. Ibid., 1093.

74. Ibid., 1097.

75. Vincent Giuliano, et al., The Challenge of Success, report of an independent study commissioned by the National Association of Realtors, Cambridge, MA: Arthur D. Little, Inc., 1979, at 71-72.

76. FTC Report at 62.

77. FTC Report at 89.

78. FTC Report at 90.

79. Lewis Testimony, July 25, 2006.

80. Testimony of Pat Vredevoogd, NAR, before the Subcommittee on Housing and Community Opportunity, Committee of Financial Services, U.S. House of Representatives, July 25, 2006.

81. Yun, DOJ/FTC Workshop, Transcript at 225.

82. Each transaction has two "sides:" a listing agent side, and a selling agent side.

83. "The Tragedy of the Commission," C. Hsieh, powerpoint presentation, FTC/DOJ Workshop on Competition Policy and the Real Estate Industry, October 25, 2005.

84. Hsieh does not mention the possibility that if agents' costs have decreased, a steady (or declining) income may leave them better off.

85. Vredevoogd testimony at 12.

86. The same is likely true for "star" agents who likely enjoy considerable market power. So, the "tragedy of the commission" may run more like the "comedy of the commission" for some participants. Future research should consider the extent to which membership and licensing fees have increased in recent years.

87. Rents accruing to professional associations, MLSs and other fee collectors may also create their own market distortions and wasteful expenditure of resources. For example, MLSs might wastefully expend resources trying to protect their profitable arrangement by, say, designing rules to prevent limited service agents equal MLS exposure; or NAR might wastefully expend resources lobbying Congress to keep banks out of the real estate industry.