« High Oil Prices and the Return of “Resource Nationalism" | Main | "The End of the Age of Milton Friedman" »

Saturday, April 12, 2008

Consumer Sentiment versus Job Growth

Paul Krugman notes the large decline in consumer sentiment recently:

Depressed consumers, by Paul Krugman:

University of Michigan Index of Consumer Sentiment

Surpassing the father

Wow. The latest read on consumer confidence is stunningly bad. We’re probably in the early stages of recession — but consumer sentiment is already worse than it ever got under Bush I.

Something’s happening here. What it is ain’t exactly clear. But my guess is that the way the Bush boom crested without ever leaving most people better off than they were in 2000, and the sense that there’s something fundamentally rotten with the system, have produced a wave of disillusionment and revulsion that goes beyond the real economic effects so far.

Here's the relationship between the year over year growth in job hires (data here) and consumer sentiment since 2001 (a shorter time period than in the first graph). It's not a perfect fit, but the two series move somewhat in unison after mid 2002.

The steady decline in the rate of growth of job hires beginning in 2004, and the negative growth in hires at the end of the sample are also notable. Relative to the previous year, hires are currently falling at around 2%:


Some notes: First, the graph shows year over year growth rates. Monthly growth rates for hires look quite a bit different, there is far too much noise to see any pattern (the growth rates shown in the graph are equivalent to the average monthly growth rate over the past year). Second, I'm not sure this relationship is causal. If it was, I would think that the growth in job hires causes confidence rather than the other way around, though I can think of stories of how a revealed lack of confidence about the future would cause firms to cut back on hiring today. I suspect both of these series reflect the general state of the economy and that is what is driving the co-movement instead of one of the two causing the other. Third, I wish I had a longer series for hires so I could determine if the lack of correlation at the beginning of the sample is an aberration, or a common feature over the extended, earlier sample. But given the available data, and it does appear that the two series have moved together over the last few years. That's expected as consumer confidence ought to reflect economic conditions.

Update: Reading Floyd Norris and Paul Krugman gave me an idea about how to extend the analysis using a different series. This is employment growth for all workers aged 25-54 versus consumer sentiment (both are three month moving averages):

Employment Growth (ages 25-54) and Consumer Sentiment

Update: The graph above shows monthly employment growth rates, but I meant to show year over year rates. Using year over year growth rates gives:


The fit is good at times, but there are notable departures, e.g. the late 1990s.

    Posted by on Saturday, April 12, 2008 at 10:18 AM in Economics, Unemployment | Permalink  TrackBack (0)  Comments (13)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference Consumer Sentiment versus Job Growth:


    Feed You can follow this conversation by subscribing to the comment feed for this post.