Median Price Not Telling the Whole Truth

When July’s housing stats came out last week, the most confusing piece of data was that despite skyrocketing local inventory and tightening lending across the nation, the median price still jumped up 2.3% from June, bouncing back into double-digit YOY territory at a 10.6% increase since July 2006.

While the local press has their own theories—apparently based on whimsical fantasies and proud pink ponies—I’ve been doing some actual investigating into the more detailed reports from the NWMLS and have come up with a theory of my own.

The shortcomings of the median price as an indicator of actual price changes have been discussed here before, but typically in a hypothetical sense. For instance, we know that if low-end home buyers stop buying homes, while middle and high-end buyers keep buying, the median will increase. However, in the past we have not been able to observe this happening in King County via the data we have available to us. As you are about to see, I believe that is no longer the case.

The last time we investigated this topic we looked at actual price breakdowns. This data is not generally available to the public though, so I came up with a reasonable alternative: regional breakdowns within King County. This data is publicly available in the KCBreakout pdf files that are released by the NWMLS every month. Furthermore, it serves more or less the same purpose as price breakdowns, since the county can be split into three general regions:

  • low end: South County (areas 100-130 & 300-360)
  • mid range: Seattle / North County (areas 140, 380-390, & 700-800)
  • high end: Eastside (areas 500-600)

The areas’ approximate June 2006 median price was:

  • South County: $360,000
  • Seattle / North County: $480,000
  • Eastside: $640,000

So, if we were to see (for example) a drop in sales in South King County coupled with an increase in sales on the Eastside while the median value of homes sold in each of those areas remained unchanged, the net effect on the county-wide median would be an increase. Let’s see what has actually happened in the King County single-family house market since January of last year:

NWMLS King County Sales Breakdown 01.2006-07.2007
thin dashed line: 6 month rolling average | Click to enlarge

From January of last year through about January of this year, the Eastside region averaged around 30% of the total closed sales, Seattle / North County averaged roughly 33%, and South County come in at about 37%. However, look what started to happen in February. The Eastside held pretty steady, but the percentage of sales in Seattle vs. South County began to converge, and essentially swap. This was followed up last month with a relatively large jump in sales on the Eastside, from 29.8% of the county-wide total in June to 33.5% in July.

In fact, I believe that almost all of last month’s increase in the county-wide median can be attributed to this spike in sales on the Eastside. To get a better idea of the magnitude of this change, take a look at this graph, which shows the monthly deviation from the six-month rolling average for each area:

In July, the Eastside had a 15% larger portion of the total closed sales in the county than it had averaged over the January-June period, while the South County came in 11% lower. Seattle held mostly steady in July (just 1% lower), having had its large spike in April (up 11%), also primarily at the expense of sales in the South County.

What happens when the sales breakdown goes from 30-35-35 to 33-32-35 in one month, shifting sales from the least expensive area to the most expensive? Ta-da, you get a higher county-wide median.

So how much of the increase in King County’s SFH median can be explained by shifting sales patterns, rather than actual rising home prices? Before February of this year, I would have said “none.” But with Seattle’s share of the total homes sold in the county showing a steady increase, South County sales experiencing a steady decrease, and Eastside sales taking a sudden spike in July, I am now much more inclined to say “quite a bit.”

I don’t think this explains away all of the median price increase over the last few months, but I do think it accounts for a good portion of it, especially last month. In theory, if sales distributions are indeed skewing the median, we should see the YOY change in the Case-Shiller Index begin to diverge from the YOY change in the median, since the Case-Shiller method avoids this particular shortcoming. I’ll keep you posted when July Case-Shiller data is released.

Update: Guess what? My prediction about the July Case-Shiller data was dead-on. MoM change was just 0.2%, YoY was 6.86%. That’s quite a bit different from the median’s 10.6% YoY increase.

0.00 avg. rating (0% score) - 0 votes

About The Tim

Tim Ellis is the founder of Seattle Bubble. His background in engineering and computer / internet technology, a fondness of data-based analysis of problems, and an addiction to spreadsheets all influence his perspective on the Seattle-area real estate market. Tim also hosts the weekly improv comedy sci-fi podcast Dispatches from the Multiverse.


  1. 1
    Chris says:

    Interesting post, and a good theory.
    But see if you can work in some more pink ponies to the analysis.

  2. 2
    Joel says:

    Excellent post. This is why I come to blogs for information; more analysis, fewer quotes from “experts” and “industry insiders”.

  3. 3
    Matthew says:

    Very interesting Tim, good work.

  4. 4
    deejayoh says:

    That’s an interesting way to test the theory Tim. I think it holds water. Do these areas represent 100% of KingCo? If so, I wonder if you constructed a “synthetic” median – by applying the % of total from the graphs above to the median for each market, and then adding them together – whether it would track what is happening with median for KingCo? That might be another way to test it – see how high the correlation is.

  5. 5
    Grivetti says:

    hahaha! Sadly Tim you’ve left out Ballard! See, as it sits now, someone accidentally threw up a “For Sale” sign in next door to the Sands Showgirls (a block up from Meshugy’s lil’ Uptight Oasis) for Townhomes that hadn’t even been built yet!!! What a howler!!!

    Yes, each one was going to be priced around 600K, but sitll, yuppies and telecommuters with dish-sized WiFi connections camped out in Gortex(TM) tents for the eventually first ‘bid’ on this little piece of paradise…. its true! Like Angelina Jolie, some were even discussing letting their biological clocks tick out, so they could scrape a little more child-free profit from their Microsoft contractor jobs before adopting from diversely culturally lands afar (a good notion considering Ballard’s about 99.9% suburban white) so they could afford the neg-am, no-docs.

    Too bad, sadly the prankster was apprehended and burned at the stake by over-zealous, hyper-leveraged first time buyers willing to get a piece of the Norwegian dream here in B-town. Yep, its true, no new glossy 600K townhomes… that’s right Meshugy, its going to be a Ballard Wal*Mart!!! I know, I know, but think of all the diversity it’ll bring!

  6. 6
    j says:

    newspapers wonder why their readership continues to dwindle. it’s pieces like this vs. the trash that they put up for july numbers that basically sums up the reason MSM is headed down.

  7. 7
    Eric says:

    Perhaps the reason higher end homes are selling quicker right now is one simple reason: interest rates. Jumbo Loans are going up faster than the Space Shuttle right now. A lot of high end buyers being patient may have just decided to lock in now before it was too late.. smart move on their part (relatively speaking).

  8. 8
    TJ_98370 says:

    Excellent post Tim. I think you have proved the point that changes in median price statistics can be attributed to changes in price data mix (distribution) as much as actual overall market trends.

    It’s been said many times before and in many different ways:

  9. 9
    JohnnyBigSpenda says:

    Your theory is pretty solid man. I’ve always been suspect of the ‘median home price’ as a meaure of property value increases. What is the true measure that should be used to understand the price trends? Is there a publicly available measure that compares historical prices of a particular home to the recent sale price?

  10. 10
    The Tim says:

    What is the true measure that should be used to understand the price trends? Is there a publicly available measure that compares historical prices of a particular home to the recent sale price?

    That’s the question of the hour, Johnny. The Case-Shiller Home Price Index (see related posts) does a better job of measuring actual home prices than the median does, although it is not without its own shortcomings.

    I’m still working with the numbers, trying out things like what Deejayoh suggested above, to hopefully come up with an additional metric we can look at for Seattle. I’ll keep you all posted.

    Thanks for the positive feedback, everyone.

  11. 11


    I’d add to support your theory on the artificial price increases in the Seattle area is meaningless, re: the “long-term” picture.

    How many of these $300-600K units were just “100% equity flipped properties” with a little extra cash to cause the increase from older baby boomers?

    Remember John Lennon’s “Imagine” song? Imagine when all the 100% equity property owners retire soon to cash in and are left with paultry new world order $45K average Seattle household income first time buyers to support their retirements.

    Horrifying thought isn’t it.

  12. 12
    deejayoh says:

    Tim –
    You probably already plan to do this, but I was thinking that if the number you get from what I suggested tracks the actual median – then you want to take the same set of median prices for those areas and weight them at a constant level over time (probably proportionate to the number of SFH in each area). That should give you an estimate of the “true” price trend that normalizes away mix shift. could be pretty cool.

  13. 13
    Joel says:

    Microsoft hiring slows but still grows in state

    Microsoft’s 7.1 percent local growth rate “might be considered a slowdown for them, but that’s still pretty darn quick growth,” Gonzalez said.

    Hmmm, sounds familiar.

  14. 14
    Joel says:

    That was supposed to be a link. Lets try it again: Microsoft hiring slows but still grows in state

  15. 15
    Dean says:

    Complete breakdown of july eastside stats are on PI blog.

  16. 16
    melonleftcoast says:

    Great analysis! Thank you! I was beginning to wonder when Seattle would (finally) arrive at the party.

    From what I’ve seen in Boston (lived there from 2000- 2006) and now the SF Bay Area, I’d say you are spot on. It took a while to register the price decreases in Boston because of this.

    I think if individual properties were evaluated to determine the price decrease/increase, this may (or may not?) illustrate your point a bit better. For example, compare recent sales prices (or even asking prices) for houses in each of your locations (South County, Seattle, North County, Eastside) to comparable houses from the past year. It would be more work to compile that data and do the evaluations, but I think it may give a better picture of what is going on.

    Another point I’d like to add, and maybe this could be added to your analysis, although it might be difficult to get all the info., is the effect that good school districts have on keeping the home prices up.

    From my observations (not data reviewed), in both the Boston area and on the SF Peninsula, sales are relatively brisk for communities with the combination of excellent proximity to jobs and excellent schools. Even now, when adjacent, but less desirable communities have really started to have price drops and large decreases in sales, it is not unusual for houses with excellent locations and excellent school districts to still go for asking price or above.

    Thanks again for all your data crunching!

  17. 17
    patient says:

    Those crazy numbers from the PI blog must be from a similar trend on the Eastside as Tim describes for county as a whole. I.e it’s the higher end homes that sell and not a run-up in value of homes. +40% in Kirkland and +50% in West Bellevue in one year…I don’t think so.

  18. 18
    patient says:

    Sorry I got the West Bellevue number wrong. It’s +34% increase in median price and not +50%. Still, it’s crazy numbers that can’t possible indicate the apreciation of homes. It’s got to be a change of the type of homes that were sold.

  19. 19
    rose-colored-coolaid says:

    melonleftcoast, I bet people who buy into nice school districts or near jobs actually intend to live in the house they purchased. Maybe even for a period of time > 3 years.

    It seems like a mindset of buying a house to live in might minimize the number of ‘motivated’ sellers, which would reduce supply and better support current prices.

  20. 20
    TJ_98370 says:

    A good question to ask is how are the recently tightened lending standards going to effect future median sales price statistics?

    The affluent will be the least affected by lending restrictions. They aren’t bothered by such trivial details as 20% down requirements. Won’t that fact further skew median sales prices upward as cash-poor, credit damaged would-be buyers drop out of the middle and lower end of the market because they cannot secure a mortgage?

  21. 21
    melonleftcoast says:


    definitely a possible scenario. for friends of mine that have school-aged kids and like their current schools, the last thing they want to do is move their kids. so, more stability for those districts, right?

    however, how many families stretched to buy homes in those excellent school districts with ARMs, and now won’t be able to afford their loan reset, and are not able to refi their house because they lowered their credit score by charging furniture for their new-to-them house to their credit card?

    if they are in a highly desirable area (i.e. excellent school districts), at least they have a better chance of avoiding foreclosure by selling their house.

    it sounds like you are saying their have been a lot of speculators/flippers in Seattle area that have driven up prices? i also saw that happen in Boston, with everyone and their plumber brother-in-law buying two-families and flipping them as condos for quite a tidy profit. it really drove the two-family market off a cliff, to the point that the only people buying two-family houses were the flippers, because they cost way too much to buy and rent out one of the units (fairly low rents in Boston compared to buying). Now, people are having trouble selling their two-families because they are still priced too high, but most of the flippers have left the market.

    it will be interesting (for me :) ) to continue to compare Boston’s RE decline to the SF Bay RE decline to the Seattle RE (potential) decline.

  22. 22
    waitingforseattletocool says:

    To see how the median data might be skewed, take a look at a snapshot of current activity on Mercer Island.

    The median price of homes currenly listed is $1.649M.

    The median price of homes currently subject to inspection (STI) is $3.42M.

    There are 10 homes STI out of 152 total listings in all price categories.

    There are zero homes STI out of 40 homes listed under $1M.

    There are 2 homes STI out of 96 homes listed under $2M.

    There are 8 homes STI out of 55 homes listed over $2M.

  23. 23
    carlislematthew says:

    “Area 560, Kirkland experienced a huge 40.41% increase in Median price over last year! Inventory was up 21.19% and PENDING sales from last year were up 5.13%”

    Anyone out there live in Kirkland? I find it VERY VERY hard to believe that the median value of Kirkland homes went up 40% in a year. A 40% increase in the median price of sold houses, sure!

    This surely is the biggest argument for “a change in the mix of sales” that I’ve ever seen. I mean honestly, just look at the numbers! 40%? Come on! No analysis required there…

  24. 24
    carlislematthew says:

    “The affluent will be the least affected by lending restrictions. They aren’t bothered by such trivial details as 20% down requirements. Won’t that fact further skew median sales prices upward as cash-poor, credit damaged would-be buyers drop out of the middle and lower end of the market because they cannot secure a mortgage?”

    I think you’re right. At some point, however, you have to consider that if you’re affluent you’re *probably* selling a house in order to buy a nicer one. There really aren’t that many people around right now that are suddenly affluent and don’t already own a house.

    In that situation you have two choices. You can either make your purchase contingent, which is a nail in the coffin of the purchase if the sale is taking longer (because it’s a lower end home). Or, you can just buy that new house and then sell the old one. A lot of people have been in that situation in other parts of the country and were too confident in the market and the sale of their old place.

    I know someone down in San Diego who was in that situation and it was very painful for them. Things went from “Oh, it’ll sell – it’s a great house” to “I just want to get rid of that damn house” pretty quickly. It took 3 or 4 (not sure) price decreases and nine months. Several buyers fell through due to financing, and this was BEFORE all the recent turmoil and further tightening of lending.

    This is the very definition of “motivated seller” and it is *those* sellers that will drive down prices, especially now they’ve seen what has happened in other parts of the country.

  25. 25
    melonleftcoast says:


    i also have some friends that bought before selling their old house. one couple bought a ~$1M house in a very affluent suburban town near Boston (one of the best school districts in the state). they tried to sell their downtown Boston condo, but put it on the market with a ridiculous price (their realtor “bought” their listing). It sat. And sat. And sat. They paid two mortgages for seven months. Finally, it sold for more than $150K below the original list price. Ouch! I think they can afford the “loss”, though. I hope!

    I also have friends in Seattle that bought a house on Mercer Island before they sold their house in Seattle (moved because the schools in their Seattle neighborhood were not good). I haven’t spoken with them since July, but their Seattle house had been on the market since April, and two deals had already fallen through due to buyers not being able to get financing. I hope they found a buyer before this current craziness started!

  26. 26

    […] Median Price Not Telling the Whole Truth […]

  27. 27

    […] SFH summary: August 2007 Active Listings: up 43% YOY Pending Sales: down 26% YOY Median Closed Price*: $477,345, up 9.7% […]

  28. 28

    […] Median Price Not Telling the Whole Truth, August […]

  29. 29

    […] September 2007 Active Listings: up 40.21% YOY Pending Sales: down 32.14% YOY Median Closed Price*: $450,000 – up 5.88% […]

  30. 30

    […] October 2007 Active Listings: up 36.76% YOY Pending Sales: down 29.88% YOY Median Closed Price*: $443,950 – up 0.90% […]

  31. 31

    […] price for a specific home. There are a few reasons for this. First, as has been discussed here a couple of times before, the median price can easily be (and has been) skewed by a change in the […]

  32. 32

    […] summary: November 2007 Active Listings: up 41% YOY Pending Sales: down 26% YOY Median Closed Price*: $435,000 – 0% CHANGE […]

  33. 33

    […] 2008 Active Listings: up 56% YOY (new record) Pending Sales: down 31% YOY Median Closed Price*: $435,000 – up 1.28% […]

  34. 34

    […] Active Listings: up 61% YOY (yet another new record) Pending Sales: down 32% YOY Median Closed Price*: $429,900 – down 0.01% YOY (basically […]

  35. 35

    […] I think it’s a good time to revisit the median price discussion that I first brought up last August. […]

  36. 36

    […] March 2008 Active Listings: up 57% YOY Pending Sales: down 37% YOY (new record) Median Closed Price*: $439,900 – down 3.3% […]

  37. 37

    […] SFH summary: April 2008 Active Listings: up 49% YOY Pending Sales: down 28% YOY Median Closed Price*: $448,500 – down 3.6% […]

  38. 38

    […] May 2008 Active Listings: up 42% YOY Pending Sales: down 39% YOY (new record) Median Closed Price*: $440,000 – down 6.2% […]

  39. 39

    […] May 2008 Active Listings: up 42% YOY Pending Sales: down 39% YOY (new record) Median Closed Price*: $440,000 – down 6.2% YOY (new […]

  40. 40

    […] May 2008 Active Listings: up 42% YOY Pending Sales: down 39% YOY (new record) Median Closed Price*: $440,000 – down 6.2% YOY (new […]

  41. 41

    […] SFH summary: June 2008 Active Listings: up 25% YOY Pending Sales: down 27% YOY Median Closed Price*: $449,700 – down 4.3% […]

  42. 42

    […] SFH summary: July 2008 Active Listings: up 24% YOY Pending Sales: down 24% YOY Median Closed Price*: $445,000 – down 7.5% […]

  43. 43

    […] SFH summary: August 2008 Active Listings: up 16% YOY Pending Sales: down 16% YOY Median Closed Price*: $423,950 – down 11.2% […]

  44. 44

    […] SFH summary: September 2008 Active Listings: up 5% YOY Pending Sales: up 15% YOY Median Closed Price*: $415,000 – down 7.8% […]

  45. 45

    […] SFH summary: October 2008 Active Listings: up 3% YOY Pending Sales: down 22% YOY Median Closed Price*: $392,000 – down 11.7% […]

  46. 46

    […] summary: November 2008 Active Listings: up 2% YOY Pending Sales: down 21% YOY Median Closed Price*: $395,000 – down 9.2% […]

  47. 47

    […] YOY Pending Sales: down 5% YOY Closed Sales: down 35% YOY Months of Supply: 7.8 Median Closed Price*: $382,500 – down 12.1% […]

  48. 48

    […] summary: December 2008 Active Listings: up 6% YOY Pending Sales: down 16% YOY Median Closed Price*: $403,500 – down 7.2% […]

  49. 49

    […] YOY Pending Sales: down 21% YOY Closed Sales: down 42% YOY Months of Supply: 7.5 Median Closed Price*: $375,000 – down 12.8% […]

  50. 50

    […] YOY Pending Sales: down 2% YOY Closed Sales: down 36% YOY Months of Supply: 5.7 Median Closed Price*: $363,850 – down 17.3% […]

  51. 51

    […] YOY) Pending Sales: up 15% YOY Months of Supply: 4.5 (-20.2% MOM, -26.8% YOY) Median Closed Price*: $380,000 – down 14.6% […]

  52. 52
  53. 53

    […] been nearly two years since we originally detailed how changes in the sales mix affects the median price, and with King County’s SFH median bumping up over $30,000 in the last three months, now […]

  54. 54

    […] price is not the same as rising prices. For a refresher course, hit these three posts:August 2007: Median Price Not Telling the Whole TruthJuly 2009: Median Price Still Being Distorted by Geographic Shifts in SalesMarch 2010: Declines in […]

Leave a Reply

Use your email address to sign up with Gravatar for a custom avatar.
Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Please read the rules before posting a comment.