Economic data doesn’t jibe with economic reality

By James McCusker
For the Bellingham Business Journal

Some years ago Alan Greenspan, then chairman of the Federal Reserve, admitted that he wasn’t satisfied with the clarity of the economic picture drawn by the available data. To get a better handle on what was actually happening in the economy, he was relying more heavily on anecdotal information he obtained by talking with people across the country.

Today’s economists are faced with a similar problem. It isn’t that the economic data is of poor quality. If anything, its quality is probably as high as it has ever been — higher, in fact, in many cases.

What has happened, though, is that the data does not seem to match the reality of what we see, hear, and talk about in our daily lives. This creates a lingering uneasiness, for it limits our ability to recognize good news, or bad, in the economic data. Few people find joy in the low unemployment rate, for example, because in their minds it no longer reflects the kind of economic reality it once did.

On Wall Street, of course, they enjoy a different sense of reality. The importance of a number doesn’t stem from its accuracy but from how “other people” are likely to react to it. It’s a human version of the thought process governing the actions of a steer that isn’t afraid of thunder but knows it is likely to spook the herd.

The Bureau of Labor Statistics (BLS) published its Employment Situation Report for October and it was quickly credited for the stock market’s gains last Friday, and later blamed for the stock market’s decline the following Monday. This report obviously has powers.

The Wall Street explanation for the upward and downward movement is that the employment report “number” — the 271,000 new jobs that were filled during the past month — encouraged optimism, but after investors actually read the report later their mood changed.

In one important respect, the attitude-changing power of the employment report was less about reading than about remembering. It is difficult for investors to reflect or remember much of anything in the fast-moving environment of today’s financial markets, and a report can portray a very different picture from its headline number. That is especially true of the monthly jobs report, which suffers from a split personality problem of long standing.

The BLS obtains the data for its employment situation report in two ways: the establishment survey and the household survey. Both are sample-based and they are independently carried out— so independent, in fact, that they can show opposite results in a particular month. That kind of direct contradiction isn’t common, but appears often enough to remind us that we are dealing with sample data from two distinct and different sources.

In general, the household survey data tends to be more volatile than the establishment survey. On the other hand, the household survey provides a part of the employment picture that is unavailable from employers’ payrolls.

Volatility of the data, however, is endemic to both surveys. The result is that month-to-month changes are often difficult to interpret. The BLS notes that, “An over-the-month employment change of about 100,000 is statistically significant in the establishment survey, while the threshold for a statistically significant change in the household survey is about 500,000.”

The “number” that set off the Wall Street optimism last Friday was the 271,000 new jobs added to payrolls in October. That was higher than the experts had predicted and sounded like the economy was rebounding. Remembering volatility and looking at the three-month average of 137,000, though, prompts less exuberance.

A look at the Household Survey side of the report also draws out less exuberance. There, the number of new part-time jobs was greater than the number of new full-time positions — suggesting that the rebound might turn out to be temporary. This data also fits into our anecdotal evidence of Internet retailers hiring especially large numbers of workers early for the holiday season due to the scarcity of logistics workers to handle order fulfillment.

The possibility that the improved jobs picture is temporary is very troublesome to the Federal Reserve, whose policy decision to raise interest rates hinges on improvements in the jobs environment. Almost all Wall Street economists interviewed last Friday believed that the latest jobs report meant an interest rate increase in December. The higher rate would start affecting the economy, though, just as the temporary, seasonal workers were being laid off — imperfect monetary policy timing at best.

Our economy, our workplaces, and our labor force are all going through significant structural changes as we search for ways to compete successfully in a global economy. It shouldn’t surprise anyone that our economic forecasting models are unconvincing — making our policy decision-makers uneasy. They should be.

James McCusker is a Bothell economist, educator and consultant. He also writes a column for the monthly Herald Business Journal.

Related Stories