Justin Fox: One way to ease the U.S. debt crisis? Productivity

In a May report on alternative scenarios for the long-term U.S. budget outlook, the Congressional Budget Office estimated the impact of productivity growth that was faster or slower than the 1% annual average in its baseline forecast. It made a big difference.

These forecasts were completed before the passage of the Republican-backed budget bill, which by the CBO’s reckoning markedly worsened the debt outlook. But the basic point still holds – higher-than-expected productivity growth could turn what is starting to look like an out-of-control debt problem into a much more manageable one, with the debt-to-gross-domestic-product ratio flattening rather than continuing to rise.

That’s “could,” not “would.” Deutsche Bank economists Jim Reid, Henry Allen and Raj Bhattacharyya, who took note of these CBO projections a few weeks ago, warned that higher productivity growth “may simply encourage policymakers to continue with fiscal giveaways – so debt could still rise.” Still, it would definitely give policymakers in Washington more options if total factor productivity growth averaged 1.5% a year over the next three decades instead of 1%.

How likely is that? Well, it’s a pace that the U.S. has certainly attained in the past but has failed to meet for most of the past half century. For most of the past two decades it hasn’t even reached 1%.

I converted the quarterly TFP growth statistics from the Federal Reserve Bank of San Francisco into rolling five-year figures because it can be awfully hard to see the trend in quarterly or annual data. Another approach is to divide economic history into what you might call productivity eras, which Northwestern University economics professor emeritus Robert Gordon did in the 2010 paper from which these statistics were taken:

Total factor productivity and multifactor productivity are basically same thing; the latter just happened to be the term Gordon used in that paper. Elsewhere, Gordon has called TFP “the best available measure of innovation and technical change.” Labor productivity is simply output divided by hours worked and can thus be reliably increased by replacing people with machines. TFP or MFP attempts to factor in the cost of the machines by dividing output by a weighted average of labor and capital input (with labor weighted at 0.7 and capital at 0.3).

Measuring capital input requires some judgment calls, different TFP measures cover different parts of the economy, and even the way of measuring inflation (which all the growth numbers cited here are adjusted for) can have an impact. But the basic picture of huge productivity gains in the middle part of the 20th century, a sharp slowdown in the 1970s and 1980s, and a revival from the late 1990s until the Great Recession shows up in all the different productivity statistics I’ve seen. Another slump in the 2010s gave way amid the dislocations of the COVID-19 pandemic to somewhat stronger growth, although that seems to have faded.

Great hope is attached to the potential of generative artificial intelligence to make knowledge workers more productive, with Goldman Sachs predicting a 1.5 percentage point increase in annual productivity growth and McKinsey Global Institute a range of 0.1 to 0.6 percentage points annually just from generative AI, and 0.5 to 3.4 coupled with other technologies. Then again, 2024 economics Nobel winner Daron Acemoglu of the Massachusetts Institute of Technology crunched the numbers from those and other recent studies and concluded that a more realistic forecast for the next decade would be a 0.05 percentage point annual TFP increase, although greater gains could come later. It took so long for the spread of computers to boost TFP growth that there was much talk in the 1980s and early 1990s of a “productivity paradox,” but the gains finally did come. AI could be similar.

Change like this usually doesn’t come smoothly or easily, though. If today’s chatbots make the leap to artificial general intelligence and start taking over all of humans’ cognitive tasks, it will certainly increase productivity, but there may be some less welcome side effects. More prosaically, the period with the fastest productivity growth in U.S. history began with a Great Depression that doubled the federal debt-to-GDP ratio in less than five years. Higher productivity growth could help ease U.S. fiscal troubles, but it probably won’t solve them.

Justin Fox is a Bloomberg Opinion columnist covering business, economics and other topics involving charts. A former editorial director of the Harvard Business Review, he is author of “The Myth of the Rational Market.”