In recent years, the rate of productivity growth, both in the US and in every major economy worldwide, has slowed. At the same time, and particularly in the US, we’ve seen an increase in income inequality, with the top 1 percent seeing increased income while compensation for median-wage workers has been close to flat for decades. Are these two trends related? Or are there other factors at play?
This was the subject of several presentations at a conference I attended at the Petersen Institute for International Economics.
Since I have recently heard a number of economists debate the implications of artificial intelligence and automation on productivity, wages, and employment, I was curious whether or not presenters at the Petersen Institute would portray technology-related changes in the workplace as driving income inequality.
At the conference, a paper given by Former Treasury Secretary Lawrence Summers and Anna Stansbury demonstrated that in general, productivity improvements still lead to growth in median income, and suggested that progress in technology is not depressing income. Instead, Summers and Stansbury suggest that other factors may be responsible for the recent productivity slowdown. And in another presentation, former Chairman of the Council of Economic Advisors Jason Furman (top) pointed to the creation of fewer firms, lower mobility, an increasing concentration of wealth, and monopolies as more important factors in flat compensation.
The point of the conference was to examine what might happen if productivity continues to stay low, and participants discussed how such a reality would affect debt sustainability and tax policy, noting that the impact in these areas depends mostly on what happens with interest rates and inflation. There was some debate as to whether or not productivity growth really drives real interest rates, though there was a consensus that productivity growth does lead to improved living standards over time.
Based on what I am hearing at most technology conferences, there is a belief that we’re seeing faster technological change than ever before, which is increasing disruption in the workplace and also driving income inequality. But based on the economic statistics and what I hear at economics-oriented conferences, I wonder if the problem is actually that we’re seeing less technological change in most of our organizations than what we were accustomed to in the past, and that has resulted in lower productivity growth.
Is Reduced Dynamism and Competition Causing Lower Productivity Growth and Increased Inequality?
Furman, also a professor at Harvard, and Peter Orszag, of Lazard and former Director of the Office of Management and Budget, shared research that sought to determine if the productivity slowdown and the increase in inequality share a common cause.
Furman said that between 1948 and 1973 productivity increased at 2.8 percent per year, but that since 1973, this has dropped to 1.87 percent. Between 1948 and 1973, 90 percent of the population saw an increase in their share of income, while the top 1 percent of earners saw their share drop; since 1973, that trend has reversed, which has lead to increasing inequality.
Furman said the traditional explanation has been that skills-biased technology change leads to inequality, but he argued that reduced dynamism and reduced competition were the common cause behind both the productivity slowdown and the increase in inequality.
For evidence of reduced dynamism in the economy, Furman pointed to the creation of fewer new firms in the economy, and much less hiring by “young firms,” or firms less than five years old. He also discussed research that shows the rate of both job creation and job destruction is in fact declining, and that there is less migration of people, presumably previously driven by economic opportunities. Much of this runs counter to the prevailing narrative that technology is causing rapid change in the job market. (See my earlier stories from the recent Techonomy and Fortune Brainstorm conferences.)
Regarding reduced competition, Furman noted that we’ve lately seen an increase in the rate of return on capital, even though business investment has trended down. Meanwhile, concentration has increased in most sectors of the economy.
Furman listed several possible explanations for this: We could be seeing more natural monopolies, particularly with network externalities favoring the big tech companies. We seem to be having less antitrust enforcement, with the agencies not objecting to smaller mergers in particular. Common ownership has grown, due to the growth of mutual funds and similar instruments. Land use restrictions and occupational licensing may be contributing to lower mobility. Furman said we are seeing more differences in productivity and inequality across firms but less within them, as most of the benefits of productivity are going to the highest-performing firms. In the end, Furman said that it comes down to policy decisions, and he said we have an opportunity to make both improving productivity and equality part of the economic agenda by reducing the barriers that people and businesses face.
Productivity and Pay: Is the Link Broken?
Former Treasury Secretary Lawrence Summers, currently of Harvard University, and Anna Stansbury, also of Harvard, presented a paper looking at the link between productivity and pay.
Summers talked about studies which show that real wages and productivity used to track together, but since 1973, that behavior has changed. But since 1973, though productivity has risen—at a slower rate than previously—the wages of median workers have been relatively flat.
Summers wonders if that means that raising productivity growth no longer raises the average American’s income, or whether the decrease is the result of other changes that have occurred since 1973, including the reduction in labor bargaining points, or competition from other places.
Taking a look at the statistics represented visually, Summers said, productivity and compensation seem to track together, though compensation growth has been slower, and it looks likes the two are linked, despite fluctuations in productivity growth versus wage growth.
Stansbury went into greater detail, and showed that in times of higher productivity growth, the typical American worker has seen higher pay growth, this being the case for both the median worker as well as production/nonsupervisory workers’ (as defined by the Bureau of Labor Statistics) compensation. Summers and Stansbury estimate that a 1 percent increase in productivity growth is associated with two-thirds to 1 percent higher median pay growth, and half to two-thirds of a percent higher pay growth for production/nonsupervisory workers.
Looking at the numbers, Stansbury said, the gap between productivity and wages increased less during productivity booms than during productivity slowdowns, but she said they saw “no evidence productivity growth is causing stagnation.”
Summers pointed out that if the ratio of compensation between the mean and median worker was the same in 2015 as it was in 1973, median compensation would have been around 32 percent higher. Based on the numbers, he said that if the rate of productivity growth since 1973 had been the same as it was between 1948-1973, mean compensation would have been 59-76 percent higher, and median compensation would have been 65-68 percent higher. In other words, he said, “success in increasing productivity growth is likely to translate into wage growth.”
Summers said this work has made him more skeptical of technology-based explanations for increased inequality. The paper shows inequality tended to rise faster during the productivity slowdowns of 1973-1996 and 2003-2015 than during the productivity booms of 1948-1973 and 1996-2003.
Summers wasn’t sure about Furman’s hypothesis on monopoly power and dynamism, and said that while his ideas were broadly consistent with their findings, the hypothesis better explained the falling labor share of the economy than the share of relative wages between the mean and median workers. He said the general tendency to outsource would be expected to create more inequality without monopoly power, and said he thought most of the changes in concentration were not due to mergers, but rather to organic growth in firms such as Facebook and Google.
Reacting to these presentations, Jaana Remes, an economist and partner at the McKinsey Global Institute, agreed that there was evidence that productivity and pay were “delinked.”
But Remes noted that manufacturing has contributed two-thirds to the decline in labor’s share of US GDP, and while there are many possible factors—such as the declining power of unions, automation, offshoring, and outsourcing—she said it isn’t obvious what the connection to wages is. She said low growth in wages reduces the incentive to invest in automation.
Regarding Furman’s paper, Remes said she saw no evidence that rising corporate concentration has contributed to the productivity growth slowdown. She noted that there has been much higher concentration in the automotive parts industry since 2004, but that that industry has seen significant productivity improvement. Similarly, she said the rise of large-scale retail stores—and more recently e-commerce—has led to both more concentration and more productivity.
Remes said both papers should improve our understanding of the what’s going on here, but added that “our job is far from done.” In particular, she pointed to the “digital transformation” that is happening to the economy, and said we have a long way to go before we understand it.
Source: http://bit.ly/2ADgLvU
