ECONOMIC ANALYSIS: Don’t make too much of July economic data

There are dates on the calendar that make some of us tremble. The Ides of March was a bad one, as I recall, for a certain Roman emperor long ago. Stock market traders know and fear those triplewitching days when futures and options contracts expire.

But for those of us who track the regional economies around the state, it’s really a whole month that makes us sweat. It’s the month of July, thanks to the screwy data we receive for the period. There are more oddball, inexplicable and exaggerated movements in employment, hours and earnings data for the month of July than at any time of the year.

It’s normally a down month for most parts of the state, and this year proved no different. Compared with month-ago payrolls, employment declined in every major city statewide except Columbus, which managed to eke out a 0.2-percent increase. Anderson, Gary and Bloomington saw the largest declines in the month, while the state as a whole lost 1.2 percent of payroll employment.

That sounds dramatic, but it’s really not. This is the time of year for shutdowns and layoffs in countless businesses, especially in manufacturing.

It’s also a time when most K-12 teachers don’t show up on public school payrolls. In short, the downturn is due more to the season of the year, rather than the underlying fundamentals of the state economy.

That’s why economists and statisticians frequently employ the process of seasonal adjustment, to filter out the movements in data that stem from the calendar rather than the economy. On a seasonally adjusted basis, employment growth was essentially flat across the state, with payrolls declining by a minuscule 2,000 jobs statewide.

But it is not just the seasonal ups and downs that can make July so unusual. It’s the special events, usually involving manufacturing, that can make the data do mighty strange things.

For example, if you look at which major cities grew fastest in July, one stands tall over the rest-Kokomo. Comparing July 2005 employment against the same month in 2004-which removes the seasonal effect-Kokomo registered a whopping 10.3-percent job increase.

That huge gain was due to an even more astronomical 34.2-percent increase in payrolls in the metro area’s manufacturing sector.

But I doubt if anyone in Kokomo will even notice. That’s because the supposed growth is nothing more than a statistical quirk, brought on by the temporary shutdown of some major facilities in 2004 that didn’t occur again this year. Compared with a figure that was unusually low, this year’s employment looks very high.

It’s weird, but it’s also real. Those furloughed workers in 2004 showed up as unemployed in July, helping the city’s jobless rate zoom beyond 12 percent for the month, only to fall back to normal in August. And those same “jobless” workers filed for compensation from unemployment insurance, sending claims skyrocketing up to 3-1/2 times their June level, again falling back to Earth in August.

To my way of thinking, that’s an abuse of the unemployment-insurance system, but that’s a whole different issue. For economy-watchers across the state, it serves as an apt warning: Beware the month of July.