The productivity/pay gap and phony debates

from Dean Baker

New York Times columnist Peter Coy did a piece yesterday questioning the existence of a gap between productivity growth and the typical worker’s pay. This gap was established decades ago by my friends and former colleagues at the Economic Policy Institute (EPI). The fact that it is now being questioned says a lot about economics and even more about politics in this country.

First, let me be clear, my purpose is not at all to beat up on Coy. I’ve known him for many years and consider him a very good reporter/columnist who tries to get things right. The fact that he could be caught up in this mud-throwing effort speaks volumes about the influence of well-funded conservative think tanks and their ability to push their agenda even when it has no basis in reality.

The Productivity-Pay Gap, What Is at Issue?

When my former boss at EPI, Larry Mishel first began to write about the gap between productivity and pay, more than three decades ago, he was always very clear: the gap was between the productivity growth and the pay of the median or typical worker.

He and various colleagues at EPI, including Jared Bernstein, John Schmitt, Josh Bivens, Heidi Shierholtz, Sylvia Allegretto, and Elise Gould, were always pointing out that the gap was not between productivity and average pay, the source of the gap was within the wage distribution. In other words, money was going from workers at the middle and the bottom to workers at the top, like CEOs and Wall Street types, not from wages to profits.

For this reason, it was a bit jarring to read that there is little gap between productivity and average wages as some sort of revelation. We all knew this for several decades. (This is why some of us are less than convinced about the stories of excess monopoly power being the big issue in the economy. That would better fit a story of the money going to profits, not high-end workers.)

Anyhow, the other aspects of the ostensible controversy have long been dealt with by EPI and others. (I have written on the issue a couple of times over the years.) We typically use the Consumer Price Index to deflate wages while we use the GDP deflator to calculate productivity. The latter generally shows a lower rate of inflation than the former, which could explain some of the gap. To get around this issue, it’s necessary to use a common deflator, which I have done, as has EPI in its more recent analyses. It barely changes the story.

There is also an issue that an increasing share of compensation is going to non-wage benefits like employer-provided health care insurance. Actually, this was mostly a story in the 1980s and 1990s, not in this century. In any case, we deal with this by assuming that the median worker gets the same share of their compensation in non-wage benefits as the average worker.

Coy also raises the bizarre issue that EPI looks at growth in economy-wide productivity, rather than just the non-farm business sector. The issue, which was apparently raised by some of the conservative economists he consulted, is that this includes housing, which involves very little labor and therefore implicitly has high productivity.

It’s not clear what point they were attempting to make, but as a practical matter their argument goes the wrong way for their point. EPI is comparing the growth in productivity with the growth in real wages. The growth in economywide productivity has generally been slower than the growth of productivity in the non-farm business sector over the last half-century. Using this measure for productivity reduces the gap between productivity and wages.

In short, none of the issues raised by Coy is at all new. They have been well-known for decades and thoroughly addressed by EPI. The big issue is that they are now raised again in the country’s leading newspaper to muddy the waters around a long-established fact. It’s a bit like someone digging up old and long-refuted criticisms of Johannes Kepler’s theory of a sun-centered solar system.

Distracting from the Real Issue

As EPI and others have always contended, the question is why we had a quarter-century-long period of rapid growth (1947-73), where the gains were broadly shared, to a period of slower growth where a disproportionate share of the gains go to high-end workers. Coy notes comments from the conservative economists that this just reflects the greater productivity of high-end workers.

But that just begs the question. Productivity is determined by how we structure the economy. It is not an intrinsic feature of individual workers.

Starting with my favorite example, government-granted patent and copyright monopolies, and the fact that we have made them longer and stronger over the last half-century, have a huge impact on the “productivity” of the workers best situated to benefit from them. To take my poster child, how rich would Bill Gates be in a world where the government did not threaten to arrest anyone who copied Microsoft’s software without his permission? He would likely still be working for a living in that situation. (Okay, he might be collecting his Social Security now.)

There are millions of workers who collect much larger paychecks than would be the case if we had different rules on patents, copyrights, and other forms of intellectual property (IP). These rules are very much a policy choice, not an inherent feature of the economy. And just to connect the dots, higher prices for items like prescription drugs, medical equipment, software, computers, etc. mean lower real wages for everyone not benefitting from the IP rules.

A second way that we have structured the economy to redistribute income upward is to foster the growth of a bloated financial sector. In the mid-seventies, the narrow financial sector (securities and commodities trading and investment banking) was about 0.4 percent of GDP. Today it is around 2.2 percent ($600 billion a year). Many of the country’s biggest fortunes are earned in the financial sector, and the major banks and private equity funds have plenty of people pocketing millions or tens of millions annually.

Here too the laws have an enormous impact. If we had financial transactions taxes, similar to the sales taxes most of us pay on food, clothes, and other items, we would have much less of our money going from our pockets to the Wall Street crew. Similarly, if we did not structure bankruptcy law to be so user-friendly, private equity companies would have much less fun taking over corporations and stripping their assets.

We also have structured our rules of corporate governance so that the CEOs and other top executives can rip off the companies they work for. One of the most striking items that got attention in the UAW strike last fall, was how much less money the top executives of the major car makers in Europe and Japan get paid than the CEOs of the Big Three here.

These are companies that are every bit as large and often more profitable than the Big Three. Yet their top executives get paid far less. In the extreme case, the CEO of Honda got less than one-tenth the pay of the CEO of GM.

This is due to a corrupt corporate governance structure where boards of directors see themselves as allies of top management, not agents of shareholders looking to rein in excessive compensation. The best story here is Elon Musk’s effort to get $55 billion out of Tesla, almost as much as the aid package for Ukraine.

Anyhow, according to advanced economic theory, more pay for top executives means less money for everyone else. And just to be clear, the excessive pay goes beyond the CEO. If the CEO gets $25 million, the other top execs likely get $10 million to $15 million, and the next tier can get paychecks in the single-digit millions.

Then we get the selective protectionism commonly known as “free trade.” This is selective protectionism because our trade policy has been structured to place manufacturing workers in direct competition with low-paid workers in the developing world. This has the predicted and actual effect of pushing down their wages, largely eliminating the manufacturing wage premium. Since manufacturing had historically been a source of good-paying jobs for workers without college degrees, this lowered the pay of less-educated workers more generally.

We could have applied free trade policies to professions that employed highly educated workers, like doctors and dentists. These professionals here earn two or three times as much as their counterparts in other wealthy countries. All the arguments for gains from trade would apply big-time to plans to make it easier for foreign-trained professionals to work here, but such proposals are nowhere on our policy agenda.

On the other side, full employment policies disproportionately benefit workers at the bottom end of the wage ladder. Many of us have long argued for these policies and we clearly see the impact with the strong economy of the last few years. Unions also make a huge difference in the pay of their members. And, if the minimum wage had kept pace with productivity growth, as it did between 1938 and 1968, it would be over $26 an hour today.

To be clear, when workers were getting the much higher minimum wage in 1968, relative to the economy’s average productivity, they must have been at least as productive as their pay, or they would not have been hired. (The unemployment rate was under 4.0 percent in 1968.) In short, productivity is far from some inherent characteristic of individual workers, it is the result of how we structure the economy.

Anyhow, this is a long detour from the debate over the productivity-pay gap, but it is where the discussion needs to take place. (Yes, this is the topic of my book Rigged [it’s free].) The right-wing economists challenging EPI’s analysis are just blowing smoke and trying to hide the real issues. We need to talk about how we have structured the economy to increase inequality within the wage distribution, and not pretend that this problem does not exist.