Here's another one of my tortured mathematical analyses to explain why it just doesn't matter what happens with median.
Many of us have pointed out a likely explanation for the continued increase in median price (despite sky high inventory, reduced sales) is that as the market has slowed down, it is the low price sales that are disappearing, skewing the median up. In other words, it is is a change in the shape of the distribution not a upward shift of the curve that is causing this phenomenon
If you watch this data, you probably know that Seattle's average price is always above median price - since there are a greater number houses that sell on the low end, the "mid point" is below the mathematical average. So for this argument to be true, the average price should be moving further away (higher) from the median price. If it is the low price sales that go away, then even if prices are stagnant - the median price will move up because there are fewer sales on the left end of the curve.
Take a look at the charts below. I compared YTD median prices for each MLS area to YTD average prices for both 2006 and 2007. When you compare average to median across areas, for 2006 the average:median ratio was 1.156:1. For 2007 it is 1.192:1. They are definitely moving apart - the spread has increase 3%.
Not conclusive, but certainly lends support to the argument that has been put forth.
