Since the onset of the global financial crisis in 2007, the US economy has been expanding steadily but at a slower pace than before the crisis. Had US output grown at its pre-recession rate, its level would be about 10 percent greater than it is today (figure 1).1 The important and widening output gap relative to the pre-recession trend suggests that the economic downturn is not only cyclical but also structural. The US economy is like a car that has left the highway and is now running close to a lower speed limit on a rural interstate.
Figure 1 US output: Actual level compared to extrapolated prerecession trend
Sources: Bureau of Economic Analysis and author's calculations.
What factors are driving this deviation from the pre-recession trend? Output is composed of employment (the number of employed people) and productivity (the output per worker). The visual impression is quite striking: The US economy suffers from a clear productivity slowdown (figure 2), while the employment level has more than recovered its pre-recession trend (figure 3). But long term factors are hurting the employment outlook.
Figure 2 US productivity: Actual level compared to extrapolated prerecession trend
Sources: Bureau of Economic Analysis, Bureau of Labor Statistics, and author's calculations.
Figure 3 US employment: Actual level compared to extrapolated prerecession trend
Sources: Bureau of Labor Statistics and author's calculations.
What can explain the productivity slowdown? The slowdown predates the Great Recession, suggesting that the recession itself is not the major cause. The US productivity decline since the mid-2000s seems instead to derive from the waning growth in information and communications technology, compared to the exceptional growth in the late 1990s to early 2000s.
Historians and economists are debating whether lower productivity growth will be temporary or longer-lived. Robert Gordon, professor at Northwestern University, has championed the view that future innovations will not be as important as some previous ones—or as he puts it, that "all the low-hanging fruit have been picked." On the other hand, Erik Brynjolfsson, professor at the MIT Sloan School of Management, challenges this pessimistic view and argues that innovations take time to show up in statistics. Overall, the future of productivity remains highly uncertain.
As for the labor market, the most notable trend has been the continued and steep decline in the unemployment rate (figure 4) in response to the early fiscal stimulus and extended aggressive monetary policy. As the unemployment rate is now approaching its "natural level," the Federal Reserve has decided to tighten monetary policy—and is expected to tighten even more this year. Gains from any further drop in the unemployment rate are thus limited.
Figure 4 US unemployment rate and nonaccelerating inflation rate of unemployment (NAIRU)
Sources: Congressional Budget Office, Bureau of Labor Statistics, and author's calculations.
While the labor market is on the mend, looking solely at the falling unemployment rate overstates the strength of the recovery. Other indicators of labor market health suggest that there is more work to be done. The recession may have played a role in the drop in the participation rate, but that trend clearly preceded the crisis: The labor force participation rate for men has fallen downward for nearly 60 years, while the growth in women's labor force participation has stalled since 2000. The overall participation rate has been increasing since the end of 2014, but has not yet recovered to its pre-recession level (figure 5). Although there is room for the rate to move upward as the economy strengthens, long-term forces will continue to exert downward pressure on labor force participation.
Figure 5 US labor force participation rate
Sources: Organization for Economic Cooperation and Development, Bureau of Labor Statistics, and author's calculations.
Finally, and perhaps more worrying, the working age population has been increasing at a significantly slower pace than before the crisis (figure 6). It is hard to understand why the recession would have played a role here. This phenomenon is the natural outcome of the aging "baby boomer" generation. In fact, baby boomers have been progressively leaving the working population as they retire, a trend that will likely continue until 2030. The dramatic changes in age demographics are likely to slow economic growth, strain retirement systems, and jeopardize fiscal sustainability.
Figure 6 US working age population compared to extrapolated prerecession trend
Sources: Organization for Economic Cooperation and Development, and author's calculations.
What can be done? First, the objective is to boost productivity growth. So-called "structural reforms," which include labor market, product market, and tax system reforms, can help. Given their negative effect in the short run, they may have a greater impact if accompanied by appropriate transitory fiscal stimulus. However, the determinants of productivity are still not well understood and the final effects of those reforms on productivity and output remain highly uncertain.
Second, improving labor force participation is essential to tackling inequality and ensuring broad-based growth. Criminal justice reforms, by encouraging employers to hire formerly incarcerated individuals, could help mitigate the negative effects of mass incarceration on labor force participation. Family-friendly policies (such as paid family leave, flexible scheduling, or affordable high-quality childcare), which can help parents to better balance paid work and family, could also play a major role in jump-starting the growth engine by encouraging women's participation to the labor force.
Finally, the aging population poses a major challenge for the US and other advanced economies. Immigration and pension system reforms (such as welcoming more young skilled workers, or raising the retirement age) could alleviate shortages in the working age population in the short run while preserving fiscal sustainability in the long run. If implemented wisely, those reforms could help the US economy re-enter the highway. But there is still a long and uncertain way to the interchange.
1. For all variables, the trend is estimated over a 6-year window. The 8 quarters preceding the start of the recession are excluded from the computation of the trend. The extrapolation line starts at the mid-point of the estimation window. Different specifications give similar results.