The pace of productivity growth and misallocation in the United States
Productivity is a term bandied about by academics, policymakers, and pundits alike, yet most of us aren’t sure what it means—even though we know it when we see it. In common day usage, the term evokes images of workers diligently typing away at desks or laboring hard on the factory floor, as well as annual performance evaluations by the boss. These everyday views of productivity are accurate yet at best ephemeral and at worse confusing—especially when trying to sort out how productivity relates to overall economic growth in the United States.
But for economists focused on economic growth, productivity is the Holy Grail. It’s the source of long-run economic growth and steady wage gains for workers. And that’s why, in the wake of the Great Recession, there are fears that productivity’s best days are behind us. Has productivity growth declined? And if so, what’s the reason? Answering these questions requires a clear understanding of what we mean when we talk about productivity.
First, there is what economists call total factor productivity. This measurement of productivity originated in spirit, if not name, from the research of Nobel Laureate Robert Solow on long-run economic growth. Solow, now Professor Emeritus at the Massachusetts Institute of Technology (and a member of our steering committee), was trying to figure out how much increased savings in an economy could increase the pace of long-run economic growth. He found that savings don’t do much at all.
Solow parceled out the sources of growth to increases in the labor force (population growth), increases in capital (savings), and a residual term that captured all the unexplained part of growth. His calculations showed that this residual was the main driver of long-run economic growth. Faster population growth or higher savings might boost growth in the short term, but then per-capita growth would return its old rate. Only increases in Solow’s residual seem to increase the pace of long-run economic growth.
This residual is often called “technology,” but it’s also known as total factor productivity. The concept started out explaining output growth for countries, but it can be used to describe the performance of smaller economies such as cities or even individual firms. Total factor productivity tries to figure out how well a country, city, or firm combines capital and labor to create output. Whether changes in this efficiency of output come from technology, management practices, culture, or some other factor is a matter that economists fiercely debate. To put this back in the context of the Solow model, understanding what increases TFP growth is the same as understanding what’ll boost output per capita growth.
But, if you try to look up U.S. government statistics on productivity, you’ll likely find data on labor productivity. Labor productivity is what it sounds like—a measure of how productive labor is. More specifically, it shows how much output is created for every hour of work done by a laborer. Take the amount of output produced, divide it by the number of worker hours it took to create the output, and the resulting number is your labor productivity. Importantly, though, labor productivity also is derived from total factor productivity. Take the growth in total factor productivity, add the increase in the capital workers can use from investments and account for changes in the kinds of workers in the labor force (including increases in educational levels and demographic changes), and you have the increase in labor productivity.
A debate about the source of long-run economic growth and wage growth would be important in any context. But it is especially relevant today given the slowdown in productivity growth over the past decade compared to growth levels seen before a burst in productivity growth in the late 1990s and early 2000s. Since 2003, labor productivity has slowed considerably from its pace during that period and compared to earlier periods in the postwar era. According to data compiled by John Fernald, an economist at the Federal Reserve Bank of San Francisco, trends in labor productivity look something like this:
- 1948 to 1973—grew at an annual average of 3.3 percent
- 1973 to 1995—growth declined to an annual average of 1.48 percent
- 1995 to 2003—growth bounced back to an annual average of 3.38 percent
- 2003 to 2007—growth slowed again to an annual average of 1.57 percent
- 2007 to 2013—growth increased only slightly to an annual average of 1.83 percent
Why these wide swings in labor productivity? A look at total factor productivity growth provides some clues. Here are the trends in total factor productivity over the same periods:
- 1948 to 1973—grew at an annual average of 2.15 percent
- 1973 to 1995—growth declined to annual average rate of 0.47 percent
- 1995 to 2003—growth improved to an annual average of 1.81 percent
- 2003 to 2007—growth declined to an annual average of 0.71 percent
- 2007 to 2013—growth improved marginally to an annual average 0.75 percent
Other contributors to labor productivity mostly increased over these periods. Due in part to rising education levels, growth in what Fernald calls “labor quality” went from 0.27 percent in 1948-1973 to 0.43 percent in 1973-1995. Labor quality growth ticked down just a bit during the 1995 to 2003 period to 0.40 percent and fell quite a bit more, to 0.24 percent, during the 2003 to 2007 period. It has since jumped up to 0.59 percent in the 2007 to 2013 period.
Similarly, more capital was a contributor to the increase in labor productivity growth, with the growth rate of capital deepening rising from 0.57 percent in the 1973 to 1995 period to 1.17 percent annually from 1995 to 2003. It’s fallen since then, averaging 0.61 percent during the 2003 to 2007 period and falling again to 0.49 percent during 2007 to 2013. (See Figure 1.)
As these data show, we really can’t blame the slowdown in labor productivity specifically on declines in capital accumulation, investment rates, or a deterioration in labor quality. According to Fernald’s decomposition, something else happened that caused a decline in the underlying pace of total factor productivity. What exactly could be the cause of this decline?
A recent working paper by Stanford University economist Charles Jones is helpful for thinking about this conundrum. Jones splits up total factor productivity into two factors. The first is the contribution of technology or knowledge to the pace of economic growth. Call it the stock of human knowledge. The second factor is far more nebulous. Jones calls it “M.” He’s very upfront about the fact that “M” could very well could stand for “measure of our ignorance” about the sources of growth. Jones does venture that given new research on productivity, that “M” might also be called “misallocation.”
Misallocation stories about productivity are less about the need to spur new innovations and more about understanding that many firms and individuals are already quite productive or have the ability to be more productivity if put in the right situation Jones highlights one paper that shows how women and people of color’s entrance into higher-skill occupations provided a significant boost to economic growth. That’s a matter of the misallocation of society’s workforce. The same phenomenon is evident in another study that shows the potential gains from allowing more workers to live in high-productivity cities.
There’s a similar story when it comes to firms. Research from the Organisation for Economic Co-operation and Development shows that there are firms operating among its 34 developed and rapidly development member nations that have quite high total factor productivity growth. The problem is that the insights and innovations from those firms haven’t spread to the rest of the firms.
What’s broken that transmission mechanism from these vanguard firms to the rest of the population? The decline in the rate of new business formation could be a likely candidate. If fewer innovative start-up firms are being created, then it makes sense that it would take longer for new ideas about how to run companies to diffuse throughout these economies. In the United States, the rate at which new firms enter the economy has been on the decline since the late 1970s.
But what explains the slowdown in the start-up rate? That isn’t clear, but some research by economists Ian Hathaway of Ennsyte Economics and Robert E. Litan at The Brookings Institution indicates that the slowdown in population growth and an increase in business consolidation could be the culprits. One of the reasons the number of public companies is on the decline in the United States is because of increasing mergers and acquisitions activity.
Since the late 1970s, overall productivity gains haven’t translated into broadly shared gains for the entire workforce as compensation inequality has increased. And since 2000, the gains from productivity across the U.S. economy are accumulating more and more toward the owners of capital instead of compensation for workers, as the labor share of income has been on the decline. Possible links between less equitable economic growth and declining productivity growth need to be explored because the pace of productivity growth sets the bound for how much standards of living can rise.