As the major (and not-so-major) media continue to hyperventilate about consumer confidence, nobody seems to be asking if this really is a leading indicator. Turns out, it's not much of one at all. That doesn't mean it's meaningless, it's not just much of a predictor of consumer spending.
The St. Louis Fed makes public economic data that would make your head spin, just before it nodded off to sleep. While they don't have the Conference Board's numbers, they do have the University of Michigan index of consumer sentiment, a reasonable proxy. They also have an inflation-indexed history of personal spending. I compared the two over the history of the Consumer Sentiment index, from January of 1978. That period covers just about every conceivable economic condition except a depression, including a major runup in gas prices during 1978-80.
Since the Index is constrained between 0 and about 100, rather than compare actual spending, which rises almost continuously, I compared the Sentiment Index to the change in personal spending, both monthly and annual. I compared them contemporaneously, and both leading and lagging the Index by 1, 3, 6, and 12 months.
The best correlation was a contemporaneous comparison between the Sentiment Index and the year-over-year change in spending. Here's a chart, with the Sentiment Index on the left, and the spending rescaled to overlay:
This is about a .70 correlation, meaning that just about half of the annual change in spending can be explained by consumer sentiment. And we still haven't established a cause-and-effect here, either. (The correlations to monthly changes were awful, under .20.)
But at the risk of burying the lead, compare the correlations when I lead the Sentiment Index (i.e., when I use it as a leading indicator) with when Iag it:
Consumer sentiment is a terrible leading indicator, but a tolerably good follower. What that says to me is that people's perceptions of the economy, and even of their own finances, are more a reflection of what they've been reading than of what they've been saving.
Applying this to our current situation, we shouldn't be surprised if personal spending doesn't rise much this month compared to last year, but we certainly shouldn't be projecting Christmas sales on the basis of this number.
