Last week’s new economics data were a mixed bag. But on balance I’d have to say I’m more discouraged than when the week began.
July auto sales weren’t so bad, edging up a bit (on a seasonally unadjusted basis) from both the previous month as well as July of 2009. The latter is not a trivial accomplishment, since the beginning of the cash for clunkers stimulus had been one factor in the sales a year ago. For the last 5 months we’ve held about 15% above the values of the previous year, but nevertheless still 30% below the corresponding 5 months of 2006.
The ISM manufacturing PMI index slid for the fifth consecutive month, standing at 55.5 for July. But as I keep reminding people, the consistent string of observations above 50 means that a plurality of managers have been reporting that each month this year was better than the previous month. To give myself some reassurance about that, I looked at a regression of quarterly real GDP growth (at an annual rate) on the quarter’s average PMI reading from 1948:Q1 to 2010:Q2. The regression has an R2 of 0.43, and says that we might usually expect a PMI of 55 to be associated with about 4% real GDP growth. In other words, if everything else in the economy were doing as well as the last ISM manufacturing report suggests, our Little Econ Watcher would be smiling away.
But the biggest worry remains the 131,000 fewer Americans working in July. Even if you add back in the 143,000 lost temporary Census jobs, employment was only up by 12,000 jobs, very far below what is needed to keep up with growing population, let alone to make any progress in bringing the millions of unemployed Americans back to work. Menzie takes some comfort in the fact that hours worked are still going up, and that an important factor in the drag– the loss in state and local government employment– is something that policy can address. But I would still describe the lack of progress made on employment at this point as pretty discouraging.
The Aruoba-Diebold-Scotti Business Conditions Index, which does not use hours, and does not adjust the monthly employment changes for the effects of temporary Census hiring, is certainly discouraged. In fact, the combined effect of last week’s increase in new claims for unemployment insurance plus the weak employment report pushed the ADS dangerously close to the -0.8 value that historically would often mean that a new recession could be starting. The most recent inference of that series, however, can be heavily influenced by the latest data, and it would be a mistake to make too much out of the most very recent values.
But it would also be a mistake to say that everything is going just fine.