, , , , , ,

I’m chagrined to admit that, in the rush to blog in a timely way about economic data, I’ve lost sight of a big orange flag that needs to be waved regarding the U.S. Government’s manufacturing production statistics.

These data deserve special importance, in my opinion, both because of industry’s special importance in creating a genuinely durable American prosperity, and because they represent a gauge of manufacturing’s health that should be relatively uncontroversial (unlike, for example, manufacturing employment, which generates heated debate over whether it’s mainly been falling for worrisome reasons, like declining competitiveness, or ultimately heartening reasons, like productivity growth).

But that doesn’t mean that the manufacturing output figures put out each month by the Federal Reserve are problem free. In recent years, researchers have marshaled impressive evidence that these inflation-adjusted numbers have been considerably overstated by flaws in measuring rapid price declines in information technology hardware. More specifically, because the prices of surging imported inputs (parts and components) in these sectors supposedly are falling much faster than government statisticians can track, the inflation-adjusted production of the American made content of these goods has been growing much slower than Washington has reported. According to the most authoritative investigation, this overcount amounted to as much as 20 percent between 1997 and 2007.

Without commenting on whether this conclusion is fully or partly justified, it’s still revealing to show how profoundly production in IT hardware has affected real manufacturing output since the Great Recession struck.

In this exercise, I’ll use July, 2007 as the baseline – not the recession’s official start that December – because that was the month high tech hardware goods production peaked. From that point, through the end of the recession in June, 2009, inflation-adjusted manufacturing output overall plunged by 20.16 percent. But strip out computers, computer parts, semiconductors, related devices, and telecommunications gear, and the real manufacturing output drop was somewhat greater – 21.48 percent.

The gap between the two has been wider since the recovery began. From June, 2009 through last month, real manufacturing production has grown by 27.41 percent. But if IT hardware is removed, this figure falls all the way to 22.43 percent. That’s not as big a difference as that between manufacturing output and the (so far) booming automotive sector (27.41 percent versus 20.42 percent), which I described last Friday. But it’s significant nonetheless.

Another way to portray the importance of high tech hardware to domestic manufacturing’s fortunes: U.S.-based industry is 1.73 percent larger as of last month in real terms than it was in July, 2007 – the pre-recession peak for inflation-adjusted IT production. But strip out that IT output, and domestic manufacturing’s real production is down by 3.87 percent during this period.

In July, the Commerce Department, which generates most of the U.S. Government’s raw economic data, announced it would create a “data czar” (my term for its planned “Chief Data Officer”). Everyone concerned with the fortunes of domestic industry should hope that Secretary Penny Pritzker reveals her choice sooner rather than later, and starts getting the position filled. Without better data, too much of American economic policymaking – and analysis inside and outside government – will continue to be based on flying if not blind, then seriously vision impaired.