A Dismal Future: The Projected Decline of U.S. Manufacturing

There has been considerable ink spilt about the relative health and prospect of U.S. manufacturing, with the majority saying that all is well. Most of this is wrong because many pundits focus on total output rather than manufacturing value-added as a share of GDP, and virtually all ignore the massive overstatement of output in official U.S. statistics that occurs because of how computer output is measured.

When looking at real (inflation-adjusted) U.S. manufacturing value-added as a share of GDP for 18 of 19 U.S. manufacturing industries (leaving out NAIC 334: computers and electronics), the picture has been and continues to look bleak. Manufacturing value-added as a share of GDP declined from around 12 percent in 2006 to around 10 percent in the second quarter of 2021. If this annual rate of decline continues for the next 30 years, U.S. manufacturing output as a share of GDP will be just 6.9 percent. Given that U.S. national power, both militarily and economically, is projected not as shares but as absolutes, a steadily shrinking share of manufacturing means a reduced ability to project power.

A note on computers. As ITIF has noted, because the Bureau of Economic Analysis (BEA) vastly overstates output growth in the computer and electronics components industry (NAICS code 334), overall manufacturing growth is significantly overstated. According to the BEA, output in this one industry, which accounted for fewer than 11 percent of manufacturing jobs in 2000, grew 417 percent in the 2000s. This amounted to more than 100 percent of U.S. manufacturing output in the 2000s, because the other 18 sectors saw collective declines. But this surge in computer and electronics output doesn’t reflect an actual increase in the number of computers the United States is producing. In fact, U.S. companies have been producing fewer computers as manufacturing has shifted offshore. Rather, this massive growth in output and productivity is simply a result of how BEA measures quality adjustments caused by computer speeds increasing according to “Moore’s Law.”

It’s time for Washington to take U.S. manufacturing decline, especially in advanced sectors, more seriously. This means ignoring those who say that manufacturing doesn’t matter, or that those seeking to grow manufacturing are engaged in nostalgia, or that a national manufacturing policy violates the sacred tenants of free-market religion. It is time for policymakers to ignore the idealogues and embrace manufacturing pragmatism. One place to start is to pass the U.S. Innovation and Competition Act and Sen. Chris Coons’s (D-DE) Industrial Finance Corporation Act and other provisions, as laid out by my ITIF colleague Stephen Ezell. Absent these steps, we can expect continued decline in U.S. manufacturing.