Photo: The White House/Pete Souz
The economic data releases in the United States have been all over the place this year.At the beginning of the year, warm weather gave a boost to certain economic data points that usually slow down in the winter.
Many warned that weakness in the numbers would show up in the spring because business investment normally reserved for that time of year had already taken place through the warm winter. The data did indeed turn down.
More recently, we’ve seen some big upward seasonal adjustments to key data like the July jobs and yesterday’s retail sales numbers.
This has some sceptics crying foul because July tends to be a month of downward revisions.
BNP Paribas economist Julia Coronado says that even though July is usually a month of downward seasonal adjustments, what we’re seeing now is no conspiracy:
Conspiracy rumours flew after the number as the strength in July resulted from an unusually strong upward seasonal adjustment. Total retail sales fell 0.9% on a non-seasonally adjusted basis, while core sales fell 0.8%. The seasonal adjustment pushed growth in core sales up 1.6pp as compared to a seasonal adjustment of 0.25% last July, and a downward adjustment of 1.6pp in 2010.
Indeed, the seasonal factor has generally been negative in July for the past 10 years. We saw a similar phenomenon of a large July upward seasonal adjustment in restaurant hiring in nonfarm payrolls. What seems to be occurring is that the seasonal factors were abnormally large in the negative direction earlier in the year to compensate for the impact of the unusually warm winter, i.e., spending and hiring occurred earlier than usual because of the warm winter.
Since seasonal factors have to cancel each other out over the year, there has to be a period of abnormally large, positive adjustments in the summer. This is no conspiracy; it is simply a reflection of shifting seasons and economic behaviour, but it does complicate reading the tea leaves.
Business Insider Emails & Alerts
Site highlights each day to your inbox.