More evidence that technology and increased worker productivity, not trade, were behind US factory job losses

From Mark Perry.

"Here’s the abstract from the NBER research paper “Recent Manufacturing Employment Growth: The Exception That Proves the Rule” by Robert Z. Lawrence of Harvard’s JFK School of Government:

This paper challenges two widely held views: first that trade performance has been the primary reason for the declining share of manufacturing employment in the United States and other industrial economies, and second that recent productivity growth in manufacturing has actually been quite rapid but is not accurately measured. The paper shows that for many decades, relatively faster productivity growth interacting with unresponsive demand has been the dominant force behind the declining share of employment in manufacturing in the United States and other industrial economies. It also shows that since 2010, however, the relationship has been reversed and slower productivity growth in manufacturing has been associated with more robust performance in manufacturing employment. These contrasting experiences suggest a tradeoff between the ability of the manufacturing sector to contribute to productivity growth and its ability to provide employment opportunities.
While some blame measurement errors for the recently recorded slowdown in manufacturing productivity growth, spending patterns in the United States and elsewhere suggest that the productivity slowdown is real and that thus far fears about robots and other technological advances in manufacturing displacing large numbers of jobs appear misplaced.

The chart above displays one of Professor Lawrence’s most interesting findings (to me), which is described below:

The chart above displays one of Professor Lawrence’s most interesting findings (to me), which is described below:

The rate at which the US manufacturing employment share has fallen over the past three decades as the economy globalized further—about 0.4 percentage points per year—is remarkably similar to the rate at which it fell in the 1960s and 1970s. This similarity suggests that the forces that operated in the early years continue to dominate. As shown in the chart above, when a trend line fitted to manufacturing employment data from 1960 to 1980 is projected out to 30 years, it precisely predicts the share of manufacturing employment in 2010. Without knowing about China’s rise, the North American Free Trade Agreement (NAFTA), or the formation of the World Trade Organization (WTO), a forecaster in 1980 would have been able to accurately predict the number of workers employed in manufacturing in 2010 without almost any error!
Donald Trump has blamed “offshoring and trade” for declines in manufacturing employment, singling out what he considers are flawed trade agreements as the major driver of these developments. While there is evidence that the US trade deficit has contributed to some of the losses of manufacturing jobs since 2000, far more powerful even during this period has been the impact of faster productivity growth interacting with unresponsive demand for goods. For example, 985,000 US manufacturing jobs estimated to have been lost due to Chinese imports between 1999 and 2011 (Acemoglu et al. 2016) represent less than a fifth of the total loss of more than 5 million US manufacturing jobs over the same period.

And what about the significant break in the trend starting in about 2010 for factory jobs as a share of total employment? Has the US become more internationally competitive due to lower energy prices, higher wages in China, technological innovations like 3-D printing and/or a rebound from the Great Recession? None of those reasons are convincing, according to Professor Lawrence:

As can be seen in the chart above, since about 2010, the long-run trend line no longer fits the data. While still declining, manufacturing’s share in US employment since 2010 has fallen much less than its long-run trend. Indeed, the small drop in the share of just 0.29 percentage points over the six-year period between 2010 and 2016 is less than the 0.40 percentage point annual average that was typical over the prior five decades. Recent manufacturing employment behavior has also been radically different than it was in the early 2000s. In the economic recovery after the 2001 recession, manufacturing employment did not increase at all, but, after reaching a trough in 2010, it had increased by 1.6 million jobs by 2016.
Instead, slow productivity growth appears to be the source of the relative strength in manufacturing employment growth over the post-2010 period. Since 2010 the rapid productivity growth in manufacturing has ground to a halt: Between 2010 and 2016, output per full-time employee in  manufacturing declined by 2.2 percent. This pace is far below the manufacturing labor productivity growth of 4.3 percent achieved annually between 2000 and 2010.

MP: Regardless of the reason for the recent departure in the long-term trend since 2010 of factory jobs as a share of total employment, it’s important to note the consistent and gradual downward trend in manufacturing employment that took place during the entire 50 year period between 1960 and 2010. As Professor Lawrence demonstrates, the claim by Trump and his fellow protectionists/scarcityists/mercantilists/nationalists that China, offshoring, and international trade are to blame for the loss of more than 7 million US factory jobs since the peak in 1975 is largely incorrect and exaggerated. Rather, the real “culprit” behind the loss of manufacturing employment in America, at least until recently, has been the same force that has eliminated millions of US agricultural jobs: technology, innovation, and increased worker productivity."