NAR revising 2007-2011home sales figures

Logo of the National Association of Realtors.
Image via Wikipedia

There has been a lot of talk about the National Association of Realtors having to revise their numbers downwards.

I wanted to jump on this story but I am afraid to speak without understanding what is going on here.   Here is Larry Yun’s, Chief Economist at NAR, explanation from CNN’s article:

Yun said the database NAR uses to track existing home sales, the Multiple Listing Service (MLS), has led the real estate agency to over-count existing home sales for several reasons.

The MLS database only includes home sales listed by realtors, and excludes homes listed by owners, providing a very narrow view of the market. And because more people are using realtors to list their homes instead of selling them independently, realtor-listed sales numbers have become artificially inflated, said Yun.

In addition, some of the assumptions NAR used in calculating its data have become outdated, since they were based on 2000 Census data.

From my perspective, this a legitimate reason the data could be wrong.  The MLS stats don’t record all the transactions, just those that have a Realtor involved.  With this dramatic of a shift in the housing market, very few are selling their home on their own now and a higher percentage are using Realtors – which would change their formula.  I question the reason why they were using 2000 Census Data, but I know there are big issues going on the credibility of the 2010 US Census.  So either Census data numbers anyone uses are suspect.  (CalculatedRisk has an article on that topic today, and they has been covering the issues that have been surfacing with the 2010 US Census for a while now.)

On a side note, you may remember an earlier post I had about the population of the Twin Cities.  The data from the Met Council showed we were losing population, while the 2010 Census and ACS surveys showed we were increasing population.  CalculatedRisk covered these problems as well. (Updated post on ACS by Calculated Risk)

This story will develop and we will be receiving the revised numbers soon.  There are enough people on the band wagon throwing out accusations, so I will only make this short opinion.  My opinion right now is that this is an error, not a manipulation.  Time will tell if my opinion is accurate.


How this impacts the information I have been posting:

The numbers I have been using for the National stats are now in question.  We will find out next week by how much when they release their revised figures.  Based on Larry Yun’s statement, I don’t think we will be seeing earth shattering differences.

The local stats I believe are accurate.  The primary source I have used for the local stats has been from the Minneapolis Area Association of Realtors and these are pure MLS stats, there is no re-calculation with Census Data.  The only thing to keep in mind on these stats is the number of transactions that were For Sale By Owner in 2004, 2005 at the peak was more than it is today.

Going forward:

Even with the error in the data, it is still the most relevant source of the data.

I will continue to watch this story and keep you posted.

Going forward, we will have to question the Data I am sharing a little more.  If you find information suspect, please post a comment about your perspective and we can all discuss it and become a little wiser.



UPDATE 12/16/2011:   NAR will be releasing the revised data Dec 21st, along with the methodology.

A. The numerical computation sheet will be available after the official release of new figures at 10 a.m., December 21, 2011.


Enhanced by Zemanta

Download my free property search app



Realtor. Twin Cities Area specializing in New Construction, Development and Investment Properties.


  1. says

    Craig, Very thorough explanation, and I agree with most of your thesis. However, there’s more to it just than just the for sale by owner sales; there was also double counting for certain types of sales; investor, builders, etc. It’s these types of sales that they should of gotten right, and frankly many of us have been questioning for a long time. As you and I have discussed before, if outside firms can get it right, why is it that the NAR is struggling, and why have they ignored complaints for so long. The bottom line is that if they are looked to as the authority for “figures and stats,” then this puts a nick in their credibility. I’ve been a member of the NAR for over 23 years, and rely heavily on their numbers, so all of “us” need to hold their feet to the fire and make sure they get it right.

    • says

      Great insight. I don’t have an argument at this point, just waiting until the data comes in to make a decision. It appears there is a lot of mis-information early on in stories like this and I don’t want to jump to any conclusions just yet.

      Exactly how was this Double Counting taking place? I have heard that, but have never understood or seen proof of “double counting”. If an investor purchases a home, fixes it up, turns around and re-sells it in 3 months – that is 2 sales not 1 according to the MLS. How is that double counting? Are we saying that only “owner occupant” sales should be counted? And as far as outside firms getting it right, how do we know that? Wasn’t CoreLogic recently in trouble for overstating their financials? How can we trust them either?

      • says

        The best example is when a developer couldn’t get his condo building going, he needed sales. So he offered investors (straw buyers) to purchase one or several units, with the option of redeeming the properties if they changed their mind (I was offered to take part in a couple of these, but I refused). A (straw) buyer changes their mind and the builder buys it back. He then sells to another investor with the same opportunity, but this time the investor actually sells it right away and makes money. Then the development starts selling units like crazy and that buyer sells it again within the same year. So that is four sales within one year, and yes I seen it happen. Matter of fact I’ve seen it where the builder/developer purchased their own units and used them as comparable sales.

        When you compare all of those sales (whether they are technically legal sales or not) against sales from prior years when this didn’t happen you get “fuzzy math.” It looks much better than it really is. Do a little research on condo sales in Florida where this was rampant. In the end you have a lot of (straw/investor) buyers with units that won’t sell because they’re over-priced and the market collapses (or known as a bubble) because you run out of buyers.

        So the NAR uses some calculation to estimate all of the non-MLS sales (for sale by owner, investor) and adds them to their own data. What you have is a bunch of non-arm’s length transactions making up the market that don’t meet the definition of “fair market value.” The NAR uses the census numbers to correct there estimate of how many actual buyers there were. What happened was that the NAR’s number of sales wasn’t matching up with the real world, and frankly I believe they knew or had an inclination. They wanted to wait until the census numbers came out to verify, and low and behold they were off.

        They were given a lot of grief about this, but they refused to address it, and that’s where the rub comes in. They defended their stance and said everyone else was wrong. I wish I kept all of the articles about this but I didn’t, so take it for what it’s worth. It’s no accident that this is all coming out during a slow news cycle before the holidays. They’ve had the census data for a while. I realize that CoreLogic, Census, Case/Shiller, all could have numbers that don’t make sense, and quite frankly should be questioned and made to come clean, but the NAR refused to come clean until 5 YEARS later.

        • says

          I have been reading about this too, so I agree with you that this issue has been out there for a while.

          I still am having trouble seeing fault directly on NAR here, if anything it shows how flawed the US Census data is. If there is fault for NAR, it may be that they knew the Census data is flawed and chose to use it anyhow. Shouldn’t we more upset with the US Census data? (along the line of what CalculatedRisk has been writing about for months now.)

          As far as Straw Buyers, that was a serious problem in 2003, 2004, 2005, and maybe some in 2006 – but I don’t think this will be much of a factor on the revising.

          I would really like to know how they are calculating these figures with the Census data – that to me would clarify and give me a better understanding.