Robots are evoking some deep economic anxiety these days. They’re routinely mastering human tasks -- driving cars, trading securities, diagnosing diseases - - that not long ago appeared permanently beyond their capabilities. And as automated technology advances at an exponential rate, more and more jobs, in more and more fields, will be done by intelligent machines in the very near future.
This transition will involve some scary trade-offs. Growth and productivity will probably accelerate, and low-cost, high-quality goods will probably proliferate. But many workers will find their skills obsolete and their ability to compete diminished. Unemployment could be exacerbated. Wage stagnation for the middle class could persist or worsen. And inequality seems likely to widen.
For all that, we remain optimistic. Throughout history, technology -- from the steam engine to electricity to the computer -- has upended old ways of doing business and created useful and edifying new fields of human endeavor. This long cycle of creative destruction suggests, however, that the robot revolution will be a time of significant turmoil. And that the more we prepare for it now, the better off we’ll be.
As the digital economy churns through old industries, workers will need to become increasingly creative and open to change, and governments will need to grow more nimble in encouraging innovation and cushioning the blow for those left behind. As Erik Brynjolfsson and Andrew McAfee argue in their book “Race Against the Machines,” the guiding principle for revamping public policy as this revolution unfolds is flexibility.
That means, first, preserving labor-market flexibility by fighting the urge to inhibit job losses in industries where robots will inevitably displace humans. A smarter way for the government to encourage hiring is to start shifting from taxes on employment, like the payroll tax, and toward taxes on consumption and pollution, like carbon. Investing in education, as always, will be critical. Expanded worker-training and placement programs would help ease the burden of changing jobs and better align workers’ skills with what employers need. And reforms that help make health-insurance policies more portable would be a boon to workers looking to try new fields.
Second, the government needs to do a better job of encouraging innovation. It may seem paradoxical that one way to mitigate the stresses of surging innovation is yet more innovation. But the key to the next economy will be people who can think of new fields, new applications and new businesses that robots can complement. Entrepreneurs will create whole new categories of jobs -- many of which we haven’t imagined yet -- by combining human ingenuity with the efficiency and precision of robot labor. Entrepreneurialism will thus become an increasingly essential skill.
Congress can help by resisting the urge to regulate new industries overzealously, keeping taxes on small businesses low and making it easier to start new companies. It should also reform the immigration system by expanding H-1B visas for temporary workers, ensuring that those who graduate from U.S. universities with advanced degrees in science, technology and math automatically qualify for green cards, and by creating a “startup visa” to encourage foreign entrepreneurs who have attracted investment or revenue in the U.S. to stay and create jobs.
Perhaps the best investment the government can make to encourage innovation for the robot economy is in research and development. That requires increased funding for government research centers, such as the National Science Foundation. More important, it means spurring private investment by making the R&D tax credit, which businesses use to help fund research projects, permanent. The credit has been perpetually at risk of expiring since its inception in 1981. Making it permanent would give companies an incentive to plan long-term research and would yield social and economic benefits that almost certainly exceed its cost to taxpayers.
Third, the outdated American patent system needs to be further overhauled to crack down on patent trolls while ensuring it remains accessible to independent inventors. The U.S. Patent and Trade Office should require more clarity from applicants for software patents, reject more infringement claims based on bogus patents, shorten patent terms and seek more input from outside experts. Such changes should encourage innovation rather than litigation.
Finally, robots will probably worsen inequality by shifting income from workers toward owners of capital. This process, known in economics as “capital-biased technological change,” is already under way: After holding roughly steady for decades, the share of income that accrues to labor declined markedly in the past 10 years, partly due to increased automation. One solution, as we’ve argued before, is to start taxing investment income at the same rate as wages, while eliminating or significantly reducing the corporate-income tax.
Another option is strengthening the social safety net, preferably in ways that increase the returns to low-paid work. The earned-income tax credit is one such program, and ought to be expanded. But orthodox redistribution alone will ultimately be insufficient, and if carried too far could be counterproductive.
Here’s a more radical idea. Fresh thinking is needed on how to give low-income workers a stake of their own in the nation’s stock of capital while shielding them from undue financial risk. Done right, an equity-based component in Social Security benefits would be a step in the right direction. Along the same lines, some have suggested that the children of poor families should be granted a modest portfolio of stocks at birth, with lockup provisions. The point is that we need to fundamentally rethink how the government addresses inequality in the new economy.
History suggests that as creative destruction works its ruthless magic, the efficiencies robots yield will eventually lead to new jobs, new discoveries, new ways of enjoying life. They may free us to do ever more satisfying work. But as Brynjolfsson and McAfee warn, “There is no economic law that says that everyone, or even most people, automatically benefit from technological progress.”
That’s one more reason to start preparing now. The robot revolution may be liberating, but it won’t be bloodless.
To contact the Bloomberg View editorial board: firstname.lastname@example.org.