Here’s an op ed I wrote for The Age in response to the shock and horror of a scandal that’s brewing down here in Victoria. Real estate agents are (gasp) underquoting house prices at auctions.
Ask a real estate agent what price they expect a house theyre auctioning to sell for, chances are his indicative price will be a massive underestimate. For him the more the merrier at the auction and who knows, you might get your hopes up and end up joining the bidding.
The practice of underquoting has been outlawed in Victoria since February 2004 but, as The Age reported on its front page last Wednesday, its still rampant.
Still, it beats me why we invoked the heavy hand of the state (yet again) to protect the naïve from the minor inconvenience of turning up to an auction only to be outbid. (This is while were promised a 25% reduction in red tape).
Even here regulation has its own unique costs and unintended consequences. Consumer Affairs Brochures informing us that these transgressions have now been outlawed no doubt led some over-serious souls to have greater confidence in indicative auction prices.
Although we know that markets cant work without well informed buyers, and that sellers have a range of incentives to conceal the truth from them, governments are still back in the stone age when it comes to addressing the problem.
Firstly theres St Augustines regulation a worthy goal thats not enforced. Lord make me virtuous, but not yet. Then theres St. John Nepomucenes regulation. St John is the patron saint of floods victims, but well come back to him.
When governments are in something must be done mode, a flood of mandatory disclosure regulation is never far away. Whether it does any good it cant do much harm, can it?
But disclosing information is not the same as informing. Sometimes it can be the opposite. Consumers can only absorb so much after which more information doesnt inform, it obfiscates. Thus by demanding excessive disclosure in the form of client risk assessments and forty page product disclosure statements Financial Services Reform bamboozles investors and helps financial planners dress up their own commission-driven selling of investment products as independent financial advice, complete with government validated Financial Services Licenses.
That reminds me about St John Nepomucene. He might be up there protecting us all as best he can from the deluge. His own fate? He drowned. But I digress.
Governments have been too focused on the morality plays that drive our tabloid media. Thats not to say that we shouldnt redress clear wrongs. Indeed prohibitions on misleading conduct provide a basic foundation for a well informed market at least if they are enforced!
But theres a better more positive way we should try alongside minimum standards of disclosure. The way consumers deal with their abundant ignorance is by boning up not on the product (you can never know enough) but rather on the sellers reputation.
I dont need to be a computer expert or to audit the hardware and software on all the computers on offer to know that Apple products are unusually easy to use. Reputation, not first hand knowledge, is my guide.
But what if Im choosing a real estate agent, an investment advisor, a hospital or a school?
Often Ill have no direct experience. My grapevine might consist of the singular of just a handful of people. And chances are that none of them has the expertise or information to be confident of their assessment. Maybe the real estate agent just got lucky or the obstetrician was unlucky.
We can do much better. Not by punitive action against rogues though that mightnt hurt but by generating good information and getting it out there.
What if real estate agents were required to record on a website the price they told the vendor they could achieve? The website would be publicly inaccessible until after the sale. But everyone could then see how accurate the quote was.
Wouldnt that make it easier to choose an estate agent? You could just ask them what price they thought they could achieve, and then compare bids always alongside the accuracy of their past bids.
If you formalize this process you end up with what Ive called the Gruen Tender, which I invented while traipsing with my cancer riddled father from one oncologist to another. Each service provider predicts how well theyll perform before the event and has their performance tracked against their prognosis. A real estate agency would predict the price it can get for your home, while a hospital clinic (say an obstetric delivery ward) would predict the chances of a trouble free delivery.
Over time the accuracy of these prognoses would build up enabling us correct future bids for the optimism or pessimism of previous prognoses. The result? Just as a simple auction ensures that a sale goes to the person willing to pay the most, so the highest (adjusted) bid in a properly run Gruen Tender would identify the bidder most likely to do the best job!
Should we rush out and regulate to make it so? I doubt it at least at the outset. But governments can do a lot to encourage good practice to emerge from the competitive dynamics of the industry. (Note to Kevin Rudd and Ted Baillieu so can Oppositions)
Alan Fels provides an example as chairman of the ACCC. Despite raised eyebrows, he launched AAMIs customer charter a set of promises to customers about service standards that are audited and publicly reported upon annually. The charter has been expanded and AAMI now has many imitators.
The market should produce estate agents wholl subject themselves to similar disciplines not out of altruism, but to attract customers by demonstrating their quality. But right now things are too cosy. A bit of government activism in calling for such practices and public support when they emerge could get the ball rolling.
Likewise governments should ask loudly why share brokers and investment advisors dont keep auditable sample portfolios of investments they recommend to demonstrate the superiority of their investment acumen. Just think of the opportunities for those in the industry who broke ranks and started doing so. (If theyre any good that is!)
And in medicine? Well governments are the major buyer of clinical health services. Why not use Gruen Tenders to award the work? That way wed find out which hospitals provided the best service in different areas. Theyd be the ones that were winning all the jobs they could handle.
So lets get government out of the stone age and into that information age we keep hearing so much about. Lets escape the deluge of mandatory disclosure regulation. And lets forget about St Augustine and do it now!
Perhaps not, the tender would tell you who is most likely to predict their own performance best – the guy doing the worst job may actually have a very accurate prediction ratio, or in the case of an oncologist, be taking only patients with a very poor prognosis. Information wants to be free, no question, but accuracy of prediction is not a good indicator of ultimate performance.
David – a bid in a Gruen tender involves two pieces of information – a prognosis and a correction to that prognosis based on the accuracy of past prognoses. So the corrected bid is not a measure of accuracy, it’s a measure of how well you can be expected to do the job.
I should have linked to the previous posts on this on Troppo and will now do so. They’re here and here.
Nicholas Gruen wrote:
Sorry, my poor interpretation jumping ahead. A question though, in the case of health care providers, would this tend to bias towards those with lots of predictions rather than potentially good providers with little experience? Surely you’d be better off running a bigger analysis that removed local effects (i.e. smoothed out those providers who continually get landed with terminal patients for example).
The example I’m thinking of here is animal related to do with phenotypic genetics. To identify the best rams, geneticists use BLUPS (best linear unbiased prediction) which try to remove environmental effects from comparing genetic performance.
For sheep this is easy – farmer A might have a ram that continually throws big fat lambs on his property, but he also might have good rainfall and supplements his stock feed (so the performance of the ram is not 100% related to the performance of his offspring, but there is a definite link). That ram may well be just an average performer when his semen is used in another flock. Obviously it’s a lot more complicated than that (genetics must also take into account the contribution of that sires parents etc).
As an example, the best real estate agent might not be the obvious choice based on accuracy of prediction and best local outcome – he might just be working in an area that is very easy (i.e. only sells flats from one estate and has no competition). Similarly, the oncologist with the best outcomes may be working in an area where they see only lots of easily treated cases – their statistics will look stellar and their predictions very accurate but might ultimately be misleading when presented with your particular case.
The real problem is “linkage” – you need to be able to cancel out the environmental effects (usually done in animals by using other rams in your flock or sending your ram out to other flocks to be used). I don’t know whether this can be done with oncologists though. I’m not sure, but I could guess that a simplistic analysis of performance based the terms of the Gruen Tender could end up being very misleading in some cases, a classic case of information becoming noise. For simple comparisons it may well turn out to be tolerably reliable. Best wishes for your dad.
In real estate their is a real imbalance in information available to agents and to the punters.
In the case of Real Estate issues there is one step which would flood the market with accurate information easily, simply and cheaply. In fact most of this information already exists and is available to real estate agents (for a fee).
There should be a free searchable online database of house prices over the lat 50 years. It would be simple to search it by street or suburb and use it todo desk comparisons and also to conduct kerbside inspections. The information on final prices is already collected by the stamp duty process. The simple addition of quoted range would cost very little in fact agents could be compelled to provide it a very small imposition as they will already have the info for their own use and for the advertisments.
If the information was freely available on the net then people could do their own research or contract for a small fee those businesses that would inevitably pop up offering to do the hard yards in comparisons and checking. A parallel WIKI, edited for defamation etc, could supply all the additional knowledge needed.
I think the rules banning “dummy” bidding are uneccessary and unenforcable. In my view dummy bidding is part of the game of testing the market.
I don’t see why it can’t and shouldn’t be similar to truth in advertising laws – if an agent is going to advertise a price, it should be a price that they would accept in the absence of a better offer. No house should be passed in at or above the price advertised. No auction should start at or above the price advertised.
It’s simple dishonesty. And dishonesty with a real cost – people paying for inspections, to investigate the property before purchasing; all of this costs money. Money that’s lost because of blatant dishonesty by the real-estate agents. And that must have an effect.
I’m surprised that the free-marketeers out there would argue that dishonesty in advertising is a vital part of the market.
David,
Let me take your examples. The real estate agent who’s got lucky in the past. He’s in a good area and finds it easy to predict and then to achieve high prices.
But you invite them to assess their prospects in selling yourhome. Not some statistical artifact. They have an incentive to give you an accurate prediction of their own likely performance (because otherwise they will find their bids being downgraded in competition with other agents because of a history of overoptimism).
So their past isn’t relevant except for its value in showing how accurate their future predictions can be expected to be. And their predictions relate to your situation nothing else.
You write.
Presuming the oncologist wants to keep his credibility (and this is now a measurable entity) they want to give you the best chance they can (to win the job), so long as it’s consistent with retaining their credibility. So they may have had easy jobs previously, but they’ve got to assess their chances with your cancer, not the jobs in the past.
The beauty of this system is precisely that it never gives anyone an incentive to turn someone away as a bad risk – it just gives them an incentive to put in a pessimistic bid – which is very worthwhile information to be generating.
The remaining concerns relate to statistical representativeness it seems to me. One can of course argue that the data one has at any given time isn’t enough to give a statistically robust measure of confidence in which case one doesn’t use the system at that point or one uses it with the appropriate caveats.
Another way of thinking of the Gruen Tender is that it’s a risk rated way of allocating work, but the risk rating isn’t read off some table which is how I think risk rating is now done. The service provider gets the opportunity to risk rate every single job if they wish – or of course to just default to some more aggregated standard table of risks.
Lefty – if you have free online access to the prices of similar houses sold in the last year in the same street or suburb then it doesn’t matter what the agent says or where the auction start price is – in my view all auction start prices are at zero or at the reserve.
Auctions work because both buyer and seller think they will get an advantage from the process – the buyer thinks they will get it below market price and hopes for rain and a bad turnout of cash strapped punters, the seller hopes for a bunch of cashed up idiots and sunshine – the agent wins no matter what.
If sellers should be required to make public their reserve then buyer should also perhaps disclose their cut off point.
I’ve bought and sold real estate by both auction and the fixed price method. Neither one is better, although I do think that in Melbourne sellers get a bad deal from agents on auctions, having to pay all costs, most of which are corporate branding exercises by the agent. And agents don’t disclose the kick backs (“volume rebates”) they get from the suburban glossy ad rags for adverts which the seller (ultimately the buyer) has payed for.
My understanding of the UK system is that they generally don’t have auctions but fixed price and gazumping is very common and I think legal.
Not all markets are rising markets. (anecdote warning) A long time ago a house I was considering buying went DOWN 2% each of the 3 weeks I was making up my mind so that by the time I bought it was 5% cheaper without any negotiating on my part. Two years before that when there were no auctions around, houses were advertised in the weekday morning and sold / gone by the afternoon with no chance to even make an offer or inspect unless it was done that morning.
In real estate, I think the best law change that could be made would be for all auctions to start at the reserve price. This way the seller is guaranteed to get a price they are happy to accept, and the bidders know that they are actually in the game from the start.
It seems utterly absurd to me that an auction can go up thousands of dollars from the starting bid and still be not high enough for the seller to accept.
The Gruen Tender reminds me of the old Imperial Chinese regulation that doctors must display one lantern outside their house for each patient that died under their care in the past year.
FX – agents do disclose the commission they get from what you rightly call their branding exercises – I think they’re required to by regulation. But my guess is that most people don’t ask for their commission back (I do). And good as it is, data on past sales (which incidentally I think is available in the ACT via the http://www.allhomes.com.au website) still leaves plenty of wiggle room for quality of building, aspect etc etc etc.
There’s another thing going on here which I didn’t raise in my article – because it wasn’t really about real estate agents (they were just the hook) and also because of limitations of space. There’s lots of irrationality (and some dishonesty) flying around in this market on behalf of both buyers and sellers and one can see agents as just rationally responding to that.
The market for real estate agents could work much better if clients required agents to put their money where their mouth is – and paid success fees. They don’t because standard practice amongst agents is to make success fees an add on, rather than a substitute for the basic agency fee.
Why does this persist?
I guess because a lot of clients are not thiking in an economicly rational way about the incentives, because there are cultural inhibitions on people getting too bloshie in their negotiations – they’re trying to get along as well as get a good deal. Further it’s hard to know how hard to bargain because if you pay less the agent can often put you at the bottom of their list and favour other properties (on which they qualify for more commission) ahead of yours.
There’s also a whole host of psychological tricks agents use both on buyers and sellers. People fall for them – it’s often hard not to – we’re human. I recall in one sale one of the agent’s consultants stage wispered to another one (audible to me but not the buyer) that this buyer we were negotiating with after the public part of the auction was the only buyer. I knew it was a lie, but it still managed to put doubt in my mind.
Likewise if you try to negotiate a success fee with them they say “What price do you want to get” and then offer to make the success fee start counting from there. That’s a psychological tactic. I say that I want a much lower base commission with the success fee cutting in earlier – but that’s out of their comprehension. Can’t do.
I toy with the idea of trying to start a real estate agency that was built from better micro-economic principles, but we’d have to lie to buyers (and maybe even sellers) to perform as well as the existing agents (They wouldn’t lie if it didn’t work!)
And that doesn’t seem like much fun to me.
I don’t know if it goes back 50 years or not but here’s one for the Canberra region, Wollongong, South Coast, Illawarra and Southern Highlands areas:
http://http://allhomes.com.au/c/ah?
(Click on the ‘past sales data’, enter the town or address and down the left side there will be an title called ‘sales history’.
Sorry about the link stuff-up (a preview would be nice :-) ).
ron- that allhomes.com.au site is fairly impressive. I havent got time to check it but does it also search back info for places NOT advertised with allhomes?
I liked the set out and land value history rates and past 15 year sale history. I’d assume, from my observations, that ACT housing might be relatively uniform in type?
I wonder if there are less
whingescomplaints about the system in ACT.I don’t know for sure, FXH. It’s used by local real estate agents in those areas, one of whom sent me the link when I was trying to track down a possibly fraudulent property transfer. (It was transferred but was not fraudulent – legally, anyway.)
nic – It’s my information that Agencies get another end of year only rebate on ads related to overall volume placed.
I agree that here is a lot of irrationality in home buying. To be fair it’s generally some one buying a home, not just an investment, and the emotional issues of home run deep in most humans. Enough to fight wars over at times. Property isn’t like say cars- it stays put – so that comparison is not all that easy to the average person.(although I’d maintain its easier than people think)
At least with cars you can compare prices and drive cars of friends etc before you hit the sellers and the Red Book, or is it Green Book, gives a fair idea of prices.{this is an interesting site http://www.carbroker.com.au/car-dealer-tricks-tactics-scams.php
Given the above and the fact that it will be most peoples largest purchase ever I’d suggest that at least a light touch of the “dead hand” is justified. The extent to which the dead hand should reach is the issue.
One of the true things agents say is that position matters. People will pay a premium for a known street name, although two or three streets back essentially ofers the same living conditions.
That is certainly true and speaking to the few estate agents that will slum it with me I hear many true tales of sellers who have unrealistic views of what their house is worth. As an example if you have ever tried to look at something in the Trading Post you’ll mostly find that the price asked second hand is higher than the cheapest new price around. The concept of sunk costs isn’t very evident. “tell’em they’re dreamin'”
ebay has sorted some of this out but equally many are struck by auction fever and pay over the price. Recently I got a new item off ebay that is available from the seller for $600 anytime (I rang up and asked) the buy history on ebay shows prices up to $1100 for the item. aaah to be able to price all products like airline seats.
FX – yup the allhomes site appeared to list all sales in the ACT (through the dataset at the titles office which is where the data comes from in the first place) – and they add other data to it obviously based on listings.
Or at least they did when I last bought a house a few years back.
cheers,
Christopher
Since I believe actual prices paid at auction are the truest reflection of current market value, it really pisses me off that local newspapers of Rural Press – some monopoly – appear reluctant to report auction sales.
I have quizzed them on this with non-commital responses which makes me fairly convinced that they are doing the real estate companies bidding. That and perhaps that they are too lazy to attend an auction to find out. Either way it is a disservice to readers and at worse could be said to be a conspiracy between them and realty advertisers.
I like the idea of going on-line, but it is restrictive for many. One of my arguments to the local rag was that I would chase an auction price by retrospectively checking stamp duty paid to the NSW office of State Revenue. Like hitting them with a feather I agree.
I’d like to see some discussion of the “winner’s curse” principle as it applies to errors in providers’ estimates. The optimism factor doesn’t reflect the average bias in the estimates; it reflects the average bias in the winning estimates (a.k.a. “bids”). For example, it all providers were unbiased in their estimates, but made randomly distributed errors around the unbiased mean, then all providers would seem to have a bias toward optimism as a result of the buyers’ non-random choice of which bids to accept.