To fully understand reported market statistics one must understand a few statistical terms, how they are computed, and why they don't necessarily tell the whole story.
The three most common statistics used are: mean (or average), median, and mode. The most commonly misunderstood is the median while the mode is seldom reported. It's easiest to understand these terms using some numbers as an example. Assume during one month a home sold at each of these prices:
100,000 100,000 200,000 300,000 400,000 600,000
Mean (or average) = 283,333; median = 250,000; mode = 100,000; total sales = 6
There are six numbers in the series; therefore the median is midway between the two centered numbers - (200,000 + 300,000) divided by 2 = 250,000. What's important is that 50% of the numbers are below the median and 50% are above the median. This is the definition of the median.
The mode is the number in the series that occurs most often. In this case there are two home sales at 100,000.
When you read a headline such as "Home prices in the month increased 10%", look closely to see exactly what increased - the mean or the median. In either case, the statistic is most representative of the population or group of homes sold rather than actual increasing (or decreasing) prices. Look at the series of numbers again with two changes underlined.
100,000 100,000 100,000 200,000 300,000 400,000 600,000 700,000
Mean (or average) = 312,500; median = 250,000; mode = 100,000; total sales = 8
What happened to the price of homes just looking at these statistics? What's most important that one can infer?
Based on the data and computed statistics for the second group of homes we cannot conclude anything about a change in prices or home values although the mean did increase by about 10%. Actual "price increases" can only be done by comparing the sale of the same or like homes from one period to another.
For this example what is most important, in my view, is that the sales volume increased by 33%. People are buying, that ultimately will result in decreasing inventory followed by price increases. The average price increased only because of two additional sales one was a very high-priced home, comparatively, and one was a low-priced home each at the extremes.
The distribution of homes that sold, in this example, is skewed to the lower priced homes. From this you might infer that it may be the increased availability of lower priced homes that is driving the sales volume upward. Perhaps price reductions are making more homes available in the lower priced segment. Does this means that prices are going down? It very well could mean that.
There are huge issues with any RE reporting. The fact that mean or median home prices are dropping like a rock doesn't necessary mean that prices are falling (although it probably means falling prices at a much lower rate than reported). It could simply mean that what is selling is skewed toward the lower end of the market. Especially given the use of the first time buyer credit, it stands to reason that entry level homes are going to be selling at a faster pace and therefore contributing much more to the mean/median.
But then, that's pretty much what you said...
Great blog post. It is difficult for people in the industry to even understand this, let alone the general public. And, this is something that should really be understood so that when people are looking at true statistics they can understand how to use them appropriately.
This has been issue of mine for a while. With the first time homebuyer credit, we have seen so many houses sold in the lower price ranges that someone could misconstrue that prices are falling if they are looking at the wrong data.
HI Bob, Good Post
And Julia...nice comment. We are seeing the same thing w the tax credit sales.
Verlyn