We’ve Been Measuring the Economy All Wrong

Questionable theoretical assumptions drive economic models to rubber-stamp disastrous policy changes.

Illustration of a pressure guage attached to a quarter
Illustration by The Atlantic. Source: Getty.

The Trump administration made some bold claims about the 2017 Tax Cuts and Jobs Act, which slashed the corporate-tax rate. Larry Kudlow, the head of the former president’s National Economic Council, said it would boost GDP so much that it would “virtually” pay for itself. Steven Mnuchin, the Treasury secretary, went further, saying the tax cut would “in fact create additional revenue.”

This was fantastical nonsense; tax cuts rarely if ever pay for themselves. But the Congressional Budget Office gave the Trump team’s sales pitch one important boost: The agency ran the bill through its model and concluded that it would have a positive, if muted, effect on long-term growth. Most Republicans were going to support Trump’s bill no matter what, but now they could do so with a straight face.

The CBO’s relatively sober prediction was wrong, a group of respected economists is charging. The Budget Office, along with other forecasters, assumes that when corporations get a tax cut, they take some of that money and reinvest it in their business, boosting growth and productivity. That assumption ignores a central dynamic of today’s economy: Many sectors are dominated by a small number of huge companies.

Much has been written over the past few years about the rise of corporate concentration. The Biden administration has made fighting monopolies one of its key economic-policy priorities. Yet none of the leading economic models takes consolidation into account. This is more than a purely technical concern. The risk is that conventional modeling is misinforming policy makers, Republican and Democratic alike, on how to structure policies that affect everyone. Governments may be relying on models that are too stuck in theory, too slow-changing, and too simplistic to be truly helpful.

“The existing models do not work well, particularly in the moments when it really counts,” such as a financial crisis, Joseph Stiglitz, the former chief economist for the World Bank, told me. “The underlying economics, the assumptions that go into the economics, are very badly flawed.”

Stiglitz is an adviser to American University’s new Institute for Macroeconomic & Policy Analysis, an initiative to update those assumptions and improve those forecasting tools. Today, IMPA is launching a new economic-forecasting model that researchers believe better captures how the economy works and, by extension, how policy changes will really play out.

Competitive economies and monopolized economies behave very differently, economists have found. When there’s plenty of competition, corporations tend to plow money into their business, investing more in research and development or raising worker pay to attract talent. But if a company doesn’t face much real competition, the pressure is off—it can simply pass the extra money along to shareholders. A model that ignores this distinction might be suitable for a tidy theoretical economy, but it won’t match the messy one we live in.

In the inaugural paper using IMPA’s model, the economists Lídia Brun, Ignacio González, and Juan Montecino conclude that the Trump tax bill was “harmful to the economy”—it slowed down growth and amped up inequality. Slashing the corporate-tax rate from 35 percent to 21 percent did not boost workers’ wages by thousands of dollars a year, as Trump appointees had predicted. Nor will it boost GDP in the long term. The IMPA model finds instead that cutting the corporate-tax rate “reduced the funds used for productive investment” by shunting money into investor payouts. What’s more, it suggests that raising taxes on business monopolies might stimulate growth by lowering those firms’ stock-market returns and thus spurring investors to pour money into more dynamic businesses.

The relationship between corporate-tax rates and business investment is a fiercely debated topic among economists, and some forecasters I spoke with didn’t think that rising corporate concentration was causing them to overestimate the economy’s growth. Mark Zandi, the chief economist at Moody’s Analytics, who did not work on the IMPA project, told me that concentration shows up indirectly in other variables commonly included in big forecasting models.

That said, several forecasters agreed that conventional models rely on questionable, even laughable, assumptions. Some models, for instance, assume that the country has perfectly competitive labor markets in which workers have total freedom to switch jobs and are paid precisely what they’re worth to a company’s bottom line. And the models generally don’t account for the ways in which having markets dominated by so few competitors—Google with web search, Amazon with online shopping—might skew profits, investments, and wages. “The assumption that there’s no market power is just wrong,” Stiglitz told me. “It’s so obvious. In many sectors of our economy, we don’t have anything that approaches that level of competition.” He cited the tech sector, drug stores, even dog food.

Creating the new IMPA model to account for monopoly power in the United States was a three-part process, Montecino and González told me. The researchers first constructed a complex mathematical model capturing a variety of factors that contribute to growth. They then tested it against historical data, seeing how well it would predict the 2015 economy using financial numbers from 2012, for instance. Finally, they let other economists critique it and review its predictions.

One of those economists was Kimberly Clausing, a tax expert at UCLA School of Law. She said that she appreciated the attempt to make a model that accounted for enormous companies’ outsize power. “Things look different when things get hyperconcentrated,” she told me. “Look at the economy of the 1970s, when there was less concentration. The labor share of income was higher. Investment was higher. So many of the main macro variables performed differently.”

Many of the forecasters I spoke with mentioned that forecasting is just plain hard—and something economists have not gotten much better at in recent decades. Chris Varvares of S&P Global Market Intelligence, a forecaster with decades of experience who is not affiliated with IMPA, noted that the economy is enormous and complex. Impossible-to-predict events, such as the coronavirus pandemic and the war in Ukraine, happen all the time. Even the world’s most influential economists often disagree on what causes what. “There’s not always good data,” he told me. “That’s just something we have to live with.”

That said, IMPA’s economists stressed that rising corporate concentration has profoundly changed our economy over the past several decades. It seems past time for it to also change how we model the economy.