Brad Delong has posed an excellent question. With today's release of quarterly GDP growth at 5.7%, we'd expect unemployment to go down per Okun's law. Unemployment figures, however, haven't gone down about 2% as predicted. Instead of 10% unemployment, it should be 8%. What's up?
First and foremost, I believe the relationship of Okun's law to still hold true. Calling it a law is a misnomer. It's more of a highly correlated relationship, but Okun's law sounds much better than Okun's Highly Correlated Relationship between GDP and unemployment. The lack of the predicted 2% drop in unemployment still falls well within the the range of the linear relationship. A little variance never hurt anybody.
The lack of corresponding drop between GDP and Unemployment doesn't mean Okun's Law is invalid, especially when it's still within the variance. For the same reason that the best quarterbacks still throw interceptions, we can still see figures like these. We shouldn't worry too much about such flukes unless they're systematic enough to create a pattern, or they're such complete weirdos that defy all expectations. The current GDP/unemployment relationship would not fit in either category.
Give the economy a few more quarters of growth and it would be hard for unemployment to resist falling.



