An economist got lost in the woods during a blizzard. He found a tiny hut with firewood and started a blaze to warm up. The hut was so small his head stuck out the window while the fire was charring his boots. But he felt fine since, on average, the temperature was about right.
That is not the only case where an average misled someone. Just read sundry pundits who now predict when the recession will be over or when the stock market will go up again, based on historical averages.
I just heard one explain on radio that since the economy apparently entered a recession in July and since recessions average 10 months in length, the economy will be growing again in April 2009. Don’t rely on such simplistic reasoning.
Averages afford useful insights in some situations. If you know the heights for 2,000 college students you not only can calculate their average height, but also the likelihood that one of your students will be over seven feet tall or under five.
Give my neighbor the meteorologist 80 years of weather data and he can compute not only average rainfall, but the probability of floods or drought on our farm.
Such predictions of height or rainfall involve calculations made from a large number of observations, hundreds of students and decades of weather readings. Moreover, one reasonably can assume that student height and rainfall follow some sort of “normal distribution,” the familiar bell-shaped curve centered on the average.
These assumptions don’t apply to the 10 recessions that have been officially identified since World War II. Yes, they average 10 months of declining output for our economy before growth returns. But drawing inferences from 10 events is very different from 100 or 1,000 events.
Moreover, it is not prudent to assume that these 10 recessions were somehow randomly drawn from a “normal population.” Rainfall and human growth result from well-defined phenomena in biology and physics.
It is not clear that all recessions stem from uniform causes. Indeed, history tends to show the opposite. The circumstances leading to one recession usually are very different from those leading to another. So don’t assume they all play out in the same way over the same time.
Here are some reasons current circumstances differ from the previous 10 recessions:
- None was preceded by a bubble in housing values as large as now.
- In none did as many major financial institutions fail as now, nor did the Federal Reserve and U.S. Treasury intervene as massively as now.
- In none had household debt increased by as large a factor over the preceding decade as now.
- In none did new and complex financial instruments like credit default swaps or collateralized debt obligations loom as large on financial institutions’ balance sheets.
- In none did foreigners hold as large a proportion of our national debt.
This is not to say we face another Great Depression. It is rather that we face true uncertainty, in Rumsfeldian terms an “unknown unknown.” Simple averages of past experience may be worse than useless.
© 2008 Edward Lotterman
Chanarambie Consulting, Inc.