Here’s the Secret Ingredient in Economic Growth

https://www.nytimes.com/2022/04/27/opinion/total-factor-productivity.html

Version 0 of 1.

The secret ingredient in economic growth is what economists call total factor productivity. You can always increase output by putting more people or more machines to work, but that consumes real resources. Total factor productivity is the growth you get without having to add labor or capital. Think of it as a new blueprint or a smarter business model that puts existing resources to better use.

Thomas Philippon, an economist at New York University who wrote a 2019 book against corporate concentration, “The Great Reversal: How America Gave Up on Free Markets,” argues in a new paper that his fellow economists have been looking at total factor productivity all wrong. I interviewed him this week.

The standard complaint among economists is that the rate of growth of total factor productivity has been falling, and since it’s a key ingredient of economic growth, that’s bad for living standards in the United States and around the world.

But Philippon says the very notion of measuring a rate of growth for total factor productivity is mistaken. Growth rates apply to phenomena that are multiplicative, such as the spread of the coronavirus in an unprotected population, he says. Each person who gets infected spreads the virus to a certain number of others, and the infected population rapidly grows.

Total factor productivity doesn’t work like that, he says. Instead, he says, there is a certain increase in know-how each year, and the size of that increase is more or less fixed. To simplify, imagine there have been three inventions per year for decades — the overall number of inventions keeps growing, but not at an accelerating pace.

In math terms, Philippon is arguing that the growth in total factor productivity is linear, not exponential. To make his case, he invites you to look at the data.

This chart is reproduced from his paper. The squiggly line is the actual increase in total factor productivity since 1947. The line with a constant slope represents no change in the amount of know-how that’s added each year. Notice how tightly it fits the data.

The line that curves up and away from reality is exponential growth, also known as percentage growth. Philippon chose the curve to fit the historical data for the first half of the period in question, namely 1947 to 1983. It’s a very poor fit for the second half of the period, 1983 to 2019.

Total factor productivity growth “is not exponential,” Philippon writes in the paper. “New ideas add to our stock of knowledge; they do not multiply it.”

If this is true, it’s a big deal, because if know-how is merely additive, it will over time become increasingly small as a percentage of a growing economy. As my Opinion colleague Paul Krugman wrote in his book “The Age of Diminished Expectations”: “Productivity isn’t everything, but in the long run it is almost everything. A country’s ability to improve its standard of living over time depends almost entirely on its ability to raise its output per worker.”

Philippon isn’t arguing, unrealistically, that the rate of invention has been flat since prehistoric times. He spots three “structural breaks” in the data — times when the rate of invention shifted upward. The first was between 1650 and 1700, the second around 1830 and the third around 1930.

Among economists, adding structural breaks to your model sets off alarm bells. It can be an ad hoc way of massaging the model to match reality. But Philippon says the three breaks he found all have plausible explanations: In each case, the step changes seemed to have been caused by the widespread adoption of world-changing general-purpose technologies: advances in agriculture and cottage industries in the first transition; steam power and iron-making in the second; and electrification and the internal combustion engine in the third.

Philippon warmed to the subject during my interview with him. While we spoke he came up with the idea that the structural breaks were like shifting the gears on a car. “Maybe think of it as shift to a higher gear? Let’s try that metaphor,” he said. The changes around 1930 were so big that they were like shifting twice, into third and then into fourth, he said. “The question for the future is whether A.I. is going to be the fifth gear.”

Philippon earned degrees in economics and physics in his native France before getting his doctorate from the Massachusetts Institute of Technology in 2003. He has advised governments and central banks in the United States, Hong Kong and France. “The Great Reversal,” his 2019 book, argued that a new era of oligopoly was costing the average American household about $5,000 a year in higher prices.

David Weil, a Brown University economist, told me that he doesn’t entirely buy Philippon’s argument that total factor productivity growth can only be linear: “If I invent something it makes us 5 percent more productive. Next year when you invent something, you would be sitting down in a better lab thanks to my invention.” And Weil said that while Philippon’s linear model may fit the data well, there’s no reason to expect the rate of invention to be fixed. “Technological progress depends on both the difficulty of the scientific terrain and the amount of R & D that gets done,” both of which vary, he said.

Gregory Mankiw, a Harvard economist, called Philippon’s approach “clever” in an email, without fully endorsing it. “It suggests that we should not expect a return to the growth rates we experienced in the past — a suggestion that Robert Gordon and others have been making in other ways,” he wrote. (Gordon, of Northwestern University, has argued that the information-technology revolution isn’t as transformative as past industrial revolutions.)

Mankiw is right that plenty of other economists have puzzled over why total factor productivity has not grown more and what to do about it. The Productivity Institute, a think tank funded by and based in the United Kingdom, is devoted pretty much full time to those questions. When Covid hit in 2020 and people switched to working and shopping from home, some economists expected a big jump in total factor productivity as a result of the more efficient business model. It didn’t show up in the data.

Imagine a factory in which each generation of robots builds the next, better generation of robots. In that case, advances would build on advances and total factor productivity growth really would be exponential, Philippon said. But that’s a rare exception, he said: “Anyone who has spent time with a lawyer and still thinks growth in total factor productivity is exponential — we aren’t living on the same planet.”

If you’re an average homeowner who owns a home for the national average of seven years, your mortgage will account for only about 30 percent of your total cost of homeownership, according to a study by two economists at Fannie Mae, which packages mortgages for resale to investors. About half of the total cost of ownership for the average homeowner is other ongoing costs, including taxes and utilities, according to the study, by Jaclene Begley, an economist, and Mark Palim, vice president and deputy chief economist. The killer? About 20 percent of total costs are incurred in two narrow windows at the beginning and the end of ownership: expenses related to buying and selling.

“The climate is changing in ways unfavorable to life, except for microbes, jellyfish and fungi.”

— Edward O. Wilson, “Half-Earth: Our Planet’s Fight for Life” (2016)

Have feedback? Send a note to coy-newsletter@nytimes.com.