Total factor productivity (TFP) measures the portion of economic growth that cannot be explained by increased inputs such as capital, labour and other factors of production. Paul Krugman, a Nobel-prizewinning economist, once described it as "inspiration", not "perspiration". There are "at least 1,001 ways" to improve TFP, according to Arnold Harberger of the University of California, Los Angeles—one of his favourite examples being a clothing boss who got 20% more out of his seamstresses by playing background music in his factory. TFP is often associated with technology and efficiency.
The most widely cited cross-country estimates of TFP come from the Penn World Table, begun by economists at the University of Pennsylvania in the 1970s and now overseen by Robert Feenstra of the University of California, Davis, Robert Inklaar and Marcel Timmer of the University of Groningen. The project calculates comparable estimates of GDP and factors of production, holding prices constant across time and space. By one estimate, two-thirds of empirical work on growth across countries draws on it.
Xi Jinping has called a rise in TFP the "hallmark" of China's "new productive forces". China's workforce has fallen by over 20m since 2016, and its rapid rate of capital accumulation will become harder to sustain as its population ages and its saving rate declines.
The tenth edition of the Penn World Table (released 2021) reported that China's TFP actually shrank between 2009 and 2019, ranking 83rd out of 118 countries. Hal Brands of Johns Hopkins University and Michael Beckley of Tufts University used similar calculations in their book "Danger Zone", which coined the term "Peak China" and likened the country to the Soviet Union.
The latest edition, released in October 2025, paints a different picture. It reports that China's TFP grew by 2.3% a year from 2009-19, ranking sixth in the world; over the most recent decade (to 2023) it ranked third. The transformation reflects a change in methodology: the new table adopted China's official GDP figures, whereas earlier editions relied on alternative estimates by Harry Wu of Peking University, which produced lower growth rates. Robert Inklaar argues that departing from official numbers confused researchers, and that singling out China's data was unwarranted when other emerging economies also raise doubts. Critics note that Mr Wu's physical-output measures of industrial production become harder to apply as Chinese industry grows more sophisticated, and that his assumption of 0% growth in output per worker in services such as education, finance and health care—though in line with international experience—may understate reality.
Barker's Proof: Proofreading is more effective after publication.