The origins of this research came out of personal experience renovating homes and then putting them on the market and wondering what were the dynamics at play when an agent lists a home for sale: What are the agents’ incentives, what kinds of services do they offer, how do they get paid. This led to the question of misalignment of interests between the sales agent and the homeowner. The agent only gets a small percentage of the sales price ‘on the margin’ i.e. the last few thousand dollars of sale price earn the agent a minimal additional fee, and so do they really work for their clients to get the best price possible.
The theory is that the dynamic really changes when there is an offer on the table. At that point the agent has earned 98% of the commission they are going to earn. If they recommend to a seller to hold out and wait for another two, three weeks for a better offer, and consequently have to incur time and money expenses showing the home, advertising etc., the amount of incremental commission is not worth it to the agent – even if the additional sales price is worth it to the seller.
Recognizing that agents also sell their own houses, the study set out to test this theory by looking at how agents perform when they sell their own houses versus how they do when they sell on behalf of clients. This data is readily available because agents are required to report when they are selling their own home as a mandated disclosure. The insight this perspective brings is akin to seeing what physicians a doctor takes her own family to, or what does a car mechanic do when they work on their own car.
In short, the study gives an opportunity to see what the expert does when they serve themselves, versus when they are hired to do it for a client.
The way that agents are currently compensated creates a misalignment between the agent’s incentives and the home seller’s incentives. When a home is sold, there is a commission that is paid to the agent of 5-6% that is split with half of that going to the seller’s agent and half going to the buyer’s agent, and then the agent has to split again with their brokerage which differs from company to company, and from agent to agent, and in all cases reduces the share of the commission to the individual agent. If we assume that the broker split is 50%, then the share that the listing agent gets somewhere between 1.25% and 1.5% of the total sale price.
Now you may think that this is what you want; that as the sale price goes up for the home, the agent is compensated more because they get a percentage of the sale price as their commission. So you might think that the motivations of the agent and of the seller are aligned, but it is the magnitude of the incentives on the margins that creates the misalignment.
SAVINGS ARE SIGNFICANT
Let’s say you get an offer of $637,000 for a home. At that point the agent has earned around $17,500 in commission. To get an additional 4% for the sale of the home would add over $25,000 to the sale price that would go straight into the seller’s pocket, but would earn the agent only an additional $350. That $350 is only 2% more commission for the agent and in nominal dollars. This is just not worth the extra effort working for two or three weeks more, doing more showings and open houses, and continuing to advertise the property. The agent would rather recommend to the seller that they accept the offer, take their commission, and move on to the next deal – but the seller would clearly be better off with the extra $25,000. The problem is, therefore, that the commission rate returns on the margin such a small amount to the agent that gives them such a low incentive to proceed relative go the home seller.
The researchers looked at this typical case and compared it to what happens when an agent sells his or her own house. In this case it is the agent that is getting the lower offer and has the option whether or not to take it or to wait for the higher offer.
When agents sell their own house they keep the house on the market for about 10% longer and they end up selling it for almost 4% more than an otherwise identical house when representing a client.
These findings are amplified where the information available to the seller is less. For example, in an area where all houses are very similar, say in a tract home where the neighbors’ houses are almost identical, sellers see what nearby homes are sold for and are better informed about the value of their own home so they are going to be less inclined to accept a lower price. However, where homes are very different in an area, it is harder for a home owner to predict the value of their house and so they are more dependent on the agent’s advice and likely more inclined to take an early offer that may be lower than they otherwise could get. This is exactly what the study found; that where homes are less similar, agents sold their own homes for proportionally more than when they sold their own homes in areas where homes were very similar.
So what is going on where you have this uniformity of commission structures nationwide that has persevered despite the pressures of a competitive economy? There seems to be something ‘special’ about needing two agents in the transaction; one on the buy side and one on the sell side. This two-agent structure makes the model ‘stickier’ than has been seen in other industries, such as the travel agent industry, or the stockbroker industry where there is only one agent involved.
In the travel industry, once technology allowed for efficient disintermediation of the agent and it became possible to book a flight or a hotel directly, there was nothing to stop consumers from going straight to the airline or hotel. Similarly, with the stock broker; as soon as the consumer could buy stocks directly using their computer or telephone bypassing the agent, the agent role became redundant.
This is harder when there are two agents involved, one on each of the buy and the sell sides of the transaction. In this case while you may have a model that evolves that disintermediates the sales side of the equation, or sharply reduces the cost in some way to the home seller, you still have a buy side agent to contend with. This buy side agent may be reacting to the new model and resisting the changes that they are seeing to their industry by steering their clients away from homes that are put on the market in a way that is unfavorable to their own interests i.e. the reduced commission or full disintermediation model where sellers don’t need a listing agent at all.
Evidence of this is found in other research conducted by the same authors that compared the sales prices of agents who use a flat-fee model with those of agents who use the full commission model. The flat fee agents were able to sell for as much as the full commission agent but it took longer to achieve this result. This indicates steering by buyer agents away from the listings of flat fee agents. That said, this result also shows that although the sales price was not higher, the net return to the home seller was higher because they did not have to pay the full commission of the selling agent – they just paid the lower flat fee.
While steering clients away from a home because an agent does not like the listing agent’s business model is unethical, possibly illegal, its illicit implementation does serve to enforce the existing full commission model on the industry in general and may help explain why the real estate industry has these commission rates nationwide that vary very little.
YOUR (BUYER) AGENT IS NOT FREE
One factor holding up the disruption of the industry is this notion that buyers have that the agent representing them is free: They are not. While the seller does, technically, pay both buy and sell side agents, the seller is using buyer money to pay them. Buyers assign excessive authority to their agent because contractually it is the seller who is paying the buyer agent and so, technically, not a cost to the buyer. However, this is not how the transaction works economically and it is indeed the buyer who must find a larger down payment, and a larger loan to consummate the transaction in order to provide enough funds to pay the buyer agent also.
Syverson and Levitt found that the magnitude of the sales price difference between the commissioned agent selling their own property and the commissioned agent selling a client’s property became smaller over time. It was more pronounced during the early periods of the study’s data set, the early 1990’s, and less so in the later period, the early 2000’s. This indicates an erosion of the commissioned agent’s exclusive ownership of market data and an increasing sophistication on the part home selling public. As the availability of home sales data has become ever greater, the information gap between the agent and the home seller has reduced, and the consumer is now more alert to the value of their own home and less likely to take an offer simply because an agent insists it is likely their best option. Whether this phenomenon has reached the point at which the fee based agents’ rise in the market is imminent remains to be seen.
Professor Siegmann's research examined the role of the real estate agent in the house prices transaction, and compared the for-sale-by-fee agent, the new generation of agent that has risen since the advent of the internet, with the performance of the traditional commissioned agent. In the Netherlands, where the study was conducted, commission rates for an agent selling a home is between 1.5% to 2%. This compares with an average of around 5.5% in the United States.
In fact, commission rates in the United States are among some of the highest in the world at an average of around 5.5%. Unlike the United States, most countries have seen significant drops in real estate commissions in the last 15 years or so. The lowest in the world are in Northern Europe at around 1.7%, and the highest is in Mexico at 7.5%. For those countries that have seen drops in commission rates, the average decline has been 34%. That there has been no significant change in the US gives pause for thought: Why is there such pricing uniformity among ostensibly competing agents?
‘Before the internet everyone needed an agent to buy a house and an agent to sell the house’. Agents maintained exclusive access to the multiple listing service and so were the gatekeepers for sellers to a market of buyers, and to buyers who wanted to see what was available to buy.
However, since the advent of the internet, buyers and sellers now have full access to detailed information about all the houses that are available on the market, and yet sellers are still paying, on average 5.5% to agents who are erstwhile intermediaries. This begs the question that, if information is freely available to everyone, ‘what is the use of an agent?’ The former role of the agent was to match buyers and sellers by providing access to a proprietary data network to each side of the transaction, but now the network itself provides direct access for buyers with seller and vice versa.
FOR SALE BY FEE (FSBF)
The for-sale-by-fee agents in this study involved the Seller in showing their house rather than the agent on the principle that the owner is, presumably, the best person to show the house because they know the house better than anyone else. This model of for-sale-by-fee agent emerged once the availability of information became more widespread on the internet. The for-sale-by-fee agent can offer considerably lower fees than the commissioned agent because, by taking the showings from the agent services, a time consuming component of the agent task is eliminated. In addition, relative to the homeowner, the agent is not as good at showing the home because, simply stated, they do not know as much as the agent.
COMMISSION AGENT – A CASE WHERE YOU DON’T GET WHAT YOU PAY FOR
The researchers’ thesis was that the for sale by fee broker to underperform the high paid commission broker. But what they discovered was that the for-sale-by-fee broker not only sold for a higher price but did it slightly quicker than the commissioned agent.
‘This is really remarkable because it means that they are cheaper and better’. For-sale-by-fee agents do better than commissioned agents irrespective of whether it is a cheap house or an expensive house, it does not depend on whether a home is in a rural area or in the city, does not matter if it is for houses that typically take a long time to sell or a short time to sell. Having a for-sale-by-fee agent, where you do the showings to buyers yourself, will give you a better result – you will sell for more, in less time, and it will cost you considerably less to do it.
Simple Conclusion: For-sale-by-fee brokers, when you do the showings yourself, do better than traditional commissioned brokers. Period.
WHY DO FSBF AGENTS OUTPERFORM COMMISSIONED?
There are two possible reasons. One is that the homeowner knows much more about their house than an agent who may be juggling multiple homes at once. The agent cannot know the details of a house as well as the seller does, and this kind of personal information about a home is of value to a buyer. This information could include things like when it was renovated, which contractor did the job, how is the neighborhood and other similarly intimate insights into the home.
The other reason is that inherent in the commissioned agent’s fee structure is the disincentive to prolong a sale to squeeze out a higher price. If an offer comes in that is lower than asking price it is in the commissioned agent’s best interest to recommend the sale because the incremental commission earned from rejecting such an offer and seeking another, higher priced, buyer is not worth the effort. This is a similar theme as we have seen evidenced in other research. If the open houses and showings are managed by the agent but left to the seller to conduct, that additional layer of time consuming labor is removed enhancing the for-sale-by-fee agent’s willingness to prolong a sale in pursuit of a better result for the seller.
Siegmann and his co-authors pondered whether these results were measuring something that they could not see; something else other than the variables they were considering. What they conclude is that there are ‘smart sellers of houses’ who are very comfortable with showing people around and making sure that their house gets sold. It also shows that the added value of professional agent is not that high; there are enough people who can do it themselves.
FSBF IN THE US MARKET
There are many studies, some covered already in this podcast series, and some coming up, that find similar results here in the US. The overall picture is a puzzle; why is it, other than by tacit collusion, that agents in the US all charge the same commissions and yet claim to compete. This is especially true in the millennial internet age where it is very easy to put pictures up online to show your house and provide direct access to buyers who can search freely and without the need of an agent. Perhaps it is because the agent realizes that as soon as they start competing on price, their business is over.
As a buyer ‘you do want to talk to the seller. Why would you not want to talk to the seller? This seems to be a story told by the agents who say “well, but that’s not good, we are professional sellers”… If I sell something that is really valuable, I would like to tell the buyer what I know about it.’
What is being seen in the European market is that agents are beginning to offer more creative ways to advertise their services and to structure their fees. There are those who charge a flat fee, and who offer a modular service, like photography upgrades, or staging advice. In other words, the market is evolving such that sellers are paying for the actual services they are getting and not just for someone to be in the middle.
It is only a matter of time before these models start to dominate the landscape here in the U.S.
Attitudes to marijuana have changed dramatically over the last 50 years. In the 1950’s only 12% of people interviewed approved of marijuana, but in recent Gallop polls that number has increased to over 60%. Now there is a majority of adults in the U.S. support the legalization of marijuana, and as opinions have shifted over the last couple of generations so, slowly, have regulations especially on the State level. As of time of publication, there are 28 states plus Washington D.C., that have legalized medical marijuana, although on the federal level marijuana is still classified as a Schedule 1 drug and deemed illegal to possess or consume.
As medical marijuana has become legal, some states have begun to legalize recreational marijuana. Currently (as of time of publication) there are eight states plus Washington D.C. that have legalized recreational marijuana, and the first two to do this were Colorado and D.C.
While the standard pros-and-cons arguments have not changed – opponents claiming that marijuana is a gateway drug, proponents arguing that removing it from illicit trade will reduce crime – no-one has really conducted any empirical research yet on the impact of legalization simply because, until recently, there has been no data to analyze.
Effect on Home Prices
The current study adds to the debate by addressing a very specific question: What happens to home values (single family residences) in the neighborhood of a store that converts from being a medical marijuana store to one that is permitted to sell for recreational purposes. What makes this paper’s results so important is that in examining the impact on single family homes values, what the study does is to flatten out the positive and negative effects of having a recreational marijuana store nearby, and to examine the net effect on home prices. In other words, home prices changes likely capitalize the overall impact of bringing a store to the neighborhood by accounting for the impact of both negative and positive effects.
Denver, Colorado provides a good location to examine because, one, it was one of the earliest to legalize recreational marijuana so has the longest history for this that can be studied, and, two, Denver provides a rich source of publicly available data. For other states and cities considering legalizing recreational marijuana, the results of the current study provide a finding that might be of use in making legislative decisions about whether to approve.
The study’s objective, therefore, was to study whether or not there is an impact on a neighborhood’s home values, and, if there is, to what extent is there an impact…
Marijuana in Colorado
Background. In Colorado in 2000 the state legalized medical marijuana but kept the industry very small to limit early growth. This changed a few years later and the state started to relax legislation allowing for the industry to start to flourish, and by 2012 it had become so widely accepted that when Amendment 64 to legalize recreational marijuana in the state’s constitution was put on the ballot, it passed. This allowed for the beginning of recreational sales to start in the beginning of 2014, a little over a year later.
As the state developed a process for managing the expansion it became clearer where these retail marijuana stores were going to be located. By the end of 2013 precise locations that would get licenses to sell recreational marijuana had been identified, and those locations were to be a subset of existing medical marijuana stores.
In December 2013, when the list of approved sales locations was released, the researchers looked at these locations and drew circles starting at a 1/10th mile radius and moving outwards at incrementally greater distances. A tenth of a mile is about a city block – so, at this distance, the impact on a single location would encompass approximately a four-block area (one block in each direction).
House Prices Shoot Up 8%
What the study found was there is a benefit on home values on a very localized basis. Homes within a one tenth mile radius of a newly designated recreational marijuana store go up roughly 8% more relative to homes situated further away; homes situated further away experience neither a negative nor a positive effect.
Why Do Prices Go Up?
From a statistical and methodological perspective, the result shows a strong causal effect: Convert a medical marijuana facility to a recreational one, and home prices within a block radius will shoot up relative to homes situated further away.
What the study does not do is explain why that happens. Why do you think it might happen? Is it:
The researchers had anticipated a negative effect going in to the study, so were surprised results that so confidently show a positive effect on home values. As the debate advances in other states as to whether to legalize recreational marijuana, the results of this study should point to at least one positive impact where, in all likelihood, there is a tendency to erroneously predict a negative one. For those ‘not-in-my-back-yarders’, understanding that their home values could go up substantially might help in taking a different view on the possible impact.
What Value Does the Real Estate Agent Bring?
The question that Jonathan Meer and his co-author Douglas Bernheim set out to answer was, what value does the broker add to a residential real estate transaction? One of the challenges in researching this is that the listing services that an agent provides are generally bundled so it is difficult to separate them out to be able to analyze them independently of each other. These services include listing the property on the MLS, taking photographs, staging advice, listing prices, paperwork, showings, handling other brokers and such like. It is a large bundle of services, but what is strange, when you think about it, is the percentage payment model for the services provided. Why is the value of all these bundled services, that might only total a few hundred or at most a couple thousand dollars, result in a commission cost that could be ten times or more as much as the actual cost of the of the services themselves? Not only is this somewhat strange from a financial perspective, but it sets up a classic case of what is known as the principal-agent problem.*
CLICK HERE to read the Shownotes.
*The principal agent problem occurs when one person, the principal, hires another, the agent, to act on their behalf in some manner, but where there is imperfect monitoring of the agent’s performance. This creates a dilemma whereby the agents are motivated to act in their own best interests, which are contrary to those of their principals.
Real Estate Agency Unpacked
The classic example of this is the auto-mechanic who knows a lot more than you do, and where you have to trust them to do the work properly and advise you accordingly, but where the incentives for both agent and principal might not be aligned.
CONTACT US with Your Comments
The same applies to the real estate agent who does not capture a meaningfully significant extra amount of commission for significant amount of additional effort they must put in to extracting the best price for a client. For example, working to get an additional $10,000 from a home sale, might be of significant value to the homeowner, but to the sales agent, the amount of additional commission is too small to warrant the extra effort it would require in getting it. The incentives are not aligned between the principal, who would like the extra $10,000, and the agent, from whom getting the extra money for the principal, is not worth the extra effort it would take.
The objective of the study was to unpack the broker expertise services – showing the home, advising on pricing, negotiations – from the value of listing the home on the MLS, and the paperwork services which can be readily calibrated in terms of their actual costs. The Stanford University faculty staff market provided a ‘usefully unusual’ market in which the unpacking of services could be studied, and served as the foundation upon which the analysis was conducted. The university has an office called ‘The Faculty Staff Housing Office’ which acts as a multiple listing service (MLS) because the university retains ownership of the land and limits who is eligible to purchase homes. So many of the real estate agent’s functions are subsumed by this office. The office lists the homes for sale and provides all necessary paperwork to consummate a transaction in the 800 or so homes on the Stanford campus – the equivalent of about 40 blocks in a typical metropolitan area.
Using an Agent Reduced Sales Price by 6-7%
In the late 1990’s, and though not required, there was a sudden uptick in the number of home sellers in this neighborhood using a broker to sell their home, going from none in some years to up to 60% of sales by the mid-2000’s. This came about because of some aggressive marketing on the part of local agents to use their services.
The study allowed for differences in house characteristics, size, number of bedrooms, features etc. They were also able to identify those homes that were sold multiple times over the course of the 30 years of the study and to compare those that sold with an agent against those that sold without an agent.
Finding: The same home sold with the aid of a real estate agent sold for 6% - 7% less than when sold without the aid of an agent.
This provides ‘evidence of very, very strong agency costs, that is the real estate agents’ incentives are aligned differently. They would like to sell the home relatively quickly and if they sell the home two weeks earlier for ten thousand dollars less that means that they are essentially giving up $200 in order to put in maybe 10-20 fewer hours of work into the sale of the house which when you work out the hourly rate of that is not a crazy decision to be making.’
Listed at Lower Prices, Homes Sold Faster
The researchers also discovered that homes listed by agents were much more likely to sell significantly quicker when listed by an agent, which reinforced the idea that agents were motivated to sell for a lower price in a shorter time. These findings were very similar to another famous study by Chad Syverson and Steve Levitt (of Freakanomics fame), where they looked at the Chicago market and compared sales by an agent of client homes, and compared them to sales by that same agent selling his or her own home. They found that when selling their own home, agents take longer to sell, and sell for more. [AG: I will be covering this study in a future episode]. Syverson and Levitt took this to be evidence of the agent-principal problem and it supports the findings of the Stanford study.
The Stanford study also noticed that homes listed with an agent typically listed at a lower initial asking price than those homes listed without, further pointing to the agent-principal conflict where the agent just wants to sell the home for a good price quickly, but not necessarily for the maximum price and to take the time doing so.
While this study is restricted to a unique real estate market, it is nevertheless of a decent size, being a reasonable equivalent in scale to a city neighborhood, and has been replicated in some regards to far larger, more generic real estate markets by other studies – such as the Syverson and Levitt study in Chicago.
Certainly, the internet has begun to make some of the listing services increasingly obsolete. It is important that the individual decides independently whether the services of an agent are in their best financial interests, and it is important that in making this decision that the homeowner is aware of the principal-agent conflict and that the agent is likely operating from a different set of incentives.
The pressure to unbundle services is likely to become more prevalent: buying an MLS listing, or photography for a home as separate and distinct services versus paying 5-6% of the sales price of a home for those same services for example.
The market forces that are driving this unbundling of services could drive the real estate agent to obsolescence the same way as it did, most notably, with the travel agent.
As the generation that is becoming more comfortable with using technology to do pretty much everything in their lives comes on stream to start buying and selling homes, they are also likely to be increasingly uncomfortable with the current bundled services model that comes with a high commission base pricing structure.