Housing stuff, continued...

This was going to be a part of the above post, but it grew to be far too long. So, here it is:

Meanwhile, in another story from yesterday, the Sun puts on it's analytical shoes and cuts it up in an anecdotal piece -- disguised as "statistics" -- comparing housing prices and income. The story tells of several families who make decent money but still can't afford to buy anything anywhere (well, anywhere that matters...you can still get a nice house in Salisbury...but why?). There's even a fancy graphic meant to show how out of whack housing prices and incomes are.

Of course, none of this is really new. Most people, whether they're trying to buy a home or not, probably know that it's a tough market to get into, especially for first time buyers. I don't want to fault the Sun for just keeping this on people's radar and more importantly for showing that it's "regular" people -- not "those" people -- who need affordable housing.

What I do fault the Sun for, however, is bad statistics. Unless they've just mislabeled their graphic, which you can see here, the entire analysis is based on averages, which are just about useless when it comes to comparing things like home values and incomes. As most of you probably know, averages can be highly skewed by extreme data points -- for instance, a family that earns $10 million a year or lives in a $10 million house will inflate the averages for their community.

With no limit on the high end and a very clear limit on the bottom end (zero), average income statistics for a place will never paint a truly accurate picture; they will always be too high. Which is why you almost always here things in terms of median -- or the middle point of the distribution; that is, 50 percent of people earn less than median income and 50 percent earn more. Median is much less response to outliers.

To see the bad statistics in action, go look at the chart and compare Allegany and Garrett Counties. Allegany has a higher average income, but the average home price in Garrett is three times that of Allegany. Why? My guess is this has something to do with the expensive houses around Deep Creek Lake.

Now, take a look at the 2000 Census data for each county. Median house value for Allegany was $71,100 and for Garrett it was $86,400. Suddenly, things make a little more sense.

As I said, I'm glad the Sun decided to do the "research" and write the story. I just wish they knew what they were doing.