On news late last week that the latest federal data on unemployment had North Carolina’s jobless rate falling below the national average for the first time in four years, state business leaders and politicians were ecstatic.
They would have every right to be enthusiastic — if the trend proves to be a lasting one. But judging by past experience, it is important not to jump to conclusions about these kinds of statistics, and to interpret them carefully.
The latest report puts North Carolina‘s unemployment rate at 5.2 percent in March, down from 6 percent in February and the lowest rate since April 2001. The national jobless rate actually edged up a bit in March, to 5.7 percent. North Carolina’s improvement did not seem to be limited to any one sector, but was to be found broadly throughout the labor market.
The first thing to keep in mind about these data is that they come from the federal government’s household survey, the smaller of the two federal surveys used to track employment trends. The larger one is conducted of business establishments, and is usually cited when analysts are trying to track the performance of industry sectors. It is also the survey that Democratic politicians have been citing over the past year to argue that the national economy under President Bush has shed more than 2 million jobs. Because it is limited to business respondents already known to federal agencies, the establishment survey probably fails to pick up job creation in the early part of a recovery because it ignores start-up companies and self-employed entrepreneurs.
On the other hand, the household survey has a smaller sample size and may bear an opposite, optimistic bias. Some folks who tell pollsters than they are employed are 1) embarrassed about being without a job, 2) employed part-time but say otherwise, or 3) new enrollees in the latest multi-level marketing scam. Discrepancies between the two surveys can have additional causes, as economist Bruce Bartlett explained in this helpful column, such as the double-counting of positions and differences in definitions and timing.
Both sets of data ought to be followed over time, with the assumption that the truth probably lies somewhere in between. In March, the establishment survey apparently showed another net loss of jobs in North Carolina, just over 2,000 positions, rather than the kind of significant gain that would reflect a widening and deepening recovery.
Secondly, both series of survey data are revised constantly. I’ve found this out the hard way — even rates going back several months can be different tomorrow than what you saw yesterday because the Bureau of Labor Statistics has updated them based on new information or after detecting an error. This is not a problem, actually, in the long run. It’s a good thing. No one should pretend that economic statistics are scriptural. Unfortunately, however, it means that preliminary data for the just-completed month can turn out to be off, sometimes far off, of what actually occurred.
I’m quite willing to believe that North Carolina is, like the rest of the country, in the midst of a robust economic recovery. It would be about time. But before anyone uncorks the bottle in celebration, I would recommend letting its contents age for at least another month or two.
Hood is president of the John Locke Foundation and publisher of Carolina Journal.