Chris Gash
The United States has become better known for inventing things than making them. That goes for everything from Apple’s iPhones to Levi’s jeans. Low-cost foreign labor is an obvious reason factories have migrated abroad, but economists have been fretting that for more than a decade, US manufacturing productivity—the efficiency of production, or how much input is needed to produce a given amount of output—has been in decline.
However, the measures traditionally used to gauge manufacturing productivity have dramatically understated its performance, suggests research by the Federal Reserve Bank of Philadelphia’s Enghin Atalay and Nicole Kimmel, University of Chicago’s Ali Hortaçsu, and Chicago Booth’s Chad Syverson. The researchers base their findings on the premise that in rapidly changing industries, the traditional yardsticks—producer and import price indexes—fail to fully capture quality improvements and therefore understate productivity growth. Instead, consumer prices, which more thoroughly adjust for quality changes but aren’t normally used to gauge manufacturing productivity, may be a better measure.
The size of the manufacturing sector means that its productivity performance carries major implications. Concerns over the apparent stagnation of the sector’s productivity are heightened by the fact that manufacturing is generally regarded as a hub of innovation. In contrast to its 10 percent share of aggregate employment, it accounts for more than two-thirds of research-and-development spending and corporate patents.
For decades, productivity in US manufacturing advanced far faster than productivity in the economy as a whole. Between 2009 and 2023, however, there was a slight decline in the sector’s total factor productivity, which is a broad, Bureau of Labor Statistics’ measure of output per unit of labor and capital input. This dip looks even worse in light of the 0.8 percent TFP growth per year during this time for the overall economy.
The researchers find that most of the stagnation can be explained by productivity changes in one area: computer and electronic product manufacturing. Although many products in this category, including well-known consumer brands such as the iPhone, are made in China, many others—including various semiconductors and medical electronics—are made in the US. In the analysis, that industry alone explains nearly all of the productivity growth since 1987 and deceleration since 2009, according to the study.
Productivity mismeasured
And the fact that most of the sector’s change in productivity growth is in this area—the manufacturing of products with fast-changing technologies—offers a clue to understanding what’s going on, Syverson says. The researchers argue that the slowdown can be largely explained by properly accounting for quality improvements.
Their underlying idea is that in order to measure real output and productivity, you have to account for quality. Consider a Toyota Corolla, which has Japanese ownership but is made in the US: The 2025 model, reflecting 50 years of technological advancement, is indisputably higher quality than the 1975 model. If a new Corolla sold in 2025 for the same amount it did in 1975 (adjusted for inflation), the factory that produced it would be considered more productive because it produced a superior car at the same price.
The researchers compared three measures of inflation: producer prices from the Bureau of Economic Analysis’s industry gross output deflators, import prices from the Bureau of Labor Statistic’s Import/Export Price Indexes, and consumer prices from the BEA’s Personal Consumption Expenditures Price Index.
Using the PCE Price Index, the researchers find that inflation in the computer and electronic product subsector is “substantially less” than what the corresponding producer and import prices suggest, evidence that when it comes to understanding manufacturing productivity, taking consumer prices into account may provide useful information. “This pattern is consistent with too little quality improvements being incorporated into producer and import price indices,” they write.
The difference, which exists only in manufacturing, is concentrated in durable goods, and especially among products—such as computers, phones, tablets, and the like—experiencing the greatest quality adjustments. Overall, durable goods prices rose 2.6 percentage points less each year when measured using consumer prices rather than producer prices or import prices, the researchers write.
The assertion that consumer prices more comprehensively measure quality improvements—and therefore growth in real output and productivity—than do producer prices indicates that conventional national account data likely understate the real output growth of durable goods manufacturing. The researchers estimate that between 1997 and 2023, the underestimate of annual TFP growth was 1.6 percentage points for durable goods and 0.5 percentage points for nondurable goods, yielding a 0.8 percentage point average annual growth rate for manufacturing as a whole.
Correcting for the undercounting of quality improvements indicates that manufacturing’s TFP has continued to grow since the late 2000s, albeit at a slower pace. While the study’s conclusions apply most prominently to the information and communications technology sector (including computers and electronics products), they could have implications for the entire US manufacturing sector, especially if other kinds of products start experiencing similar jumps in quality.
Enghin Atalay, Ali Hortaçsu, Nicole Kimmel, and Chad Syverson, “Why Is Manufacturing Productivity Growth So Low?” Working paper, September 2025.
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.