The economic recovery is now four years old, and job growth remains a big problem. Since the recession technically ended in June 2009, American businesses have expanded their workforces at an average annual rate of 1.4 percent, creating some 6.1 million new jobs. The good news? We’re creating new jobs at twice the rate seen in the first four years of the previous expansion. The bad news? These gains are much smaller than those seen in the early years of the expansions of the 1980s and 1990s. Economists have been wondering whether the slow pace of jobs growth is a feature of this particular cycle, or whether American businesses have lost some of their storied capacity for generating new jobs. The answer, of course, is that both factors are at work. And while there’s growing optimism that the pace of job creation will pick up, there’s reason to be skeptical. In the coming decade, new robotic technologies could further aggravate the problem.

History tells us that we should be doing better than we are. Job creation tends to bounce back more sharply after a deep recession than after a shallow run. But not this time. In the four years, since the Great Recession of 2008–09 ended, private employment has risen at an annual rate of 1.4 percent. That’s 61 percent below the rate of job creation following the deep downturn of 1981–82, after which jobs rose at an annual rate of 3.6 percent. A similar disparity is evident between job creation in the first four years of the two most recent expansions that followed more moderate recessions. The 0.7 percent annual rate of job growth over the first four years of the 2002–07 expansion was 68 percent less than the 2.2 percent annual job gains seen in the first four years of the 1991–2000 expansion. Something has clearly changed.

One obvious change is that every successive expansion since the 1980s has seen progressively lower rates of economic growth, especially in the early years. U.S. GDP grew by an average of 5 percent per year in the first four years of the 1980s expansion; by 3.4 percent per year in the first four years of the 1990s expansion; by 3 percent per year in the early years of the 2002–07 expansion; and by just 2.3 percent per year in the current expansion.

Keynesians have insisted that the slower economy should explain much of the recent slowdown in job gains. We can calculate roughly how much of the slowdown in job gains can be traced to the slower economy by adjusting the rates of job creation for the rates of overall growth. Had the economy grown as rapidly over the past four years as it did in the first four years of the 1980s expansion, we could have seen 3 percent annual job gain, much higher than the 1.4 percent average gains we’ve seen. Since jobs actually grew in the early 1980s by an average of 3.6 percent per year, as much as 80 percent of the current slowdown in job creation may simply reflect slower economic growth— in other words, it's cyclical.

As for the other 20 percent, well, structural changes may be to blame. Put another way, U.S. businesses now respond to economic growth by creating fewer jobs than they used to. Technological advances, of course, are one of the driving forces at play here. The countless applications of information technologies (IT) across every industry and economic activity have created considerable wealth, but they also displace more jobs than they create. The U.S. manufacturing workforce, which contracted nearly 28 percent over the last two decades, fell from 16,480,000 positions in 1992 to 11,951,000 in 2012. All of these job losses can be accounted for by workers with high school diplomas or less, whose number in manufacturing declined by more than 40 percent. The picture is different for workers with the skills to operate in an IT-dense workplace: over the same 20 years, manufacturing jobs held by college graduates increased by 2.4 percent, and the number with graduate degrees jumped 44 percent.

The latest threat to jobs, according to many technologists, is coming from robotics, the application of information technologies to new forms of kinetic hardware. Today businesses worldwide employ some 1.4 million industrial robots, mainly in automobile and electronics assembly. Those numbers appear to be rising quickly. For example, FOXCONN, the Taiwan-based giant that assembles 40 percent of the world’s consumer electronics—and employs 1.2 million workers around the world—has announced plans to purchase 1 million new robots over the next three years.

A new report from the Atlantic Council catalogs the growing number of large-scale, public-private R&D programs under way. The U.S. effort is led by DARPA, NASA, and firms such as Raytheon and iRobot, with grants from the NSF National Robotics Initiative playing a venture-capital role. In Japan, the FANUC Corporation and the Ministry of Economy, Trade and Industry have taken the lead. In Korea, the Ministry of the Knowledge Economy is working with LG and Samsung. And in Europe, the European Network of Robotic Research is collaborating with companies such as Philips and the ABB Group.

No one can predict the direction or dimensions of robotics a decade from now. Nevertheless, the next generation of the technology will be able to draw on important recent developments, such as the first open source Robot Operating System as well as advances that allow robots to retrieve and manipulate objects outside the structured environment of an assembly line. In the past year, for example, Willow Garage released a new personal robot that can fold laundry and pour beer, the French firm Robotsoft showcased robots that monitor elderly patients, Italian and Swedish firms offered robotic landscapers, a Japanese company unveiled its new robot teachers, and South Koreans developed robots to assist firefighters and provide basic child care. The first large-scale application of the technology may well involve transportation. Drone technology could force early retirement on thousands of pilots, and future variations of Google’s driverless car could displace tens of thousands of teamsters, cabbies, and bus drivers.

Like prior waves of technological innovation, these new modes of operating are likely to make the global economy more efficient and enhance many people's quality of life. But as machines continue to displace humans in a range of fields, they may exacerbate our structural problems with jobs growth.