June 14, 2017

21st Century Automation: Policy Responses



Previously, we discussed the historical tensions inherent in the First Industrial Revolution, and how the rapid pace of technological change contributed to emerging social tensions. In the short-term, the growing pains were significant, as more and more people began moving to cities and experienced firsthand the costs associated with the advent of a more industrialized economy. However, there were also long-term costs that resulted from the inaction of early 19th century policymakers—specifically, the rise of ideologies that eroded people’s trust and confidence in liberal institutions. The Niskanen Center’s Will Wilkinson expressed this sentiment well in a post-election blog on the need to revitalize liberalism:

Liberal norms and institutions are under constant corrosive pressure from natural, deep-seated illiberal tendencies that we’ve only recently managed to suppress and/or harness at all. These latent atavistic instincts cannot be effectively neutralized in general or in advance because they constantly find expression in novel, unpredictably powerful guises as our culture, economy, and technology evolves.

As our technology evolves, it is inevitable we will continue to encounter “corrosive pressure” from “illiberal tendencies” that speak to people’s fears, rather than their aspirations. While we cannot possibly prepare for all future scenarios that might degrade our institutions, there are a number of issues that we can begin addressing that may yet help ensure we avoid the same pitfalls experienced during the First Industrial Revolution.

The Issues the Robots Can’t Solve

There are a number of tangential policy areas that will have an impact on how quickly and easily people adapt to the changing technological landscape. Some, like ensuring a robust social safety net and education reform, have been focal points of attention. Others, like the cost of housing, have largely flown under the radar. All of them, however, are important to the broader conversation of automation and the future of work.

Reforming Housing Policies

Impediments to economic growth aren’t solely tethered to dwindling returns from productivity gains. For example, some believe that restrictions on the supply of housing in dense urban centers where total factor productivity is high (in particular, Silicon Valley) has led to significantly lowered aggregate economic growth. Urban enclaves on the coasts are a significant source of economic activity. By some estimates, U.S. cities with populations exceeding 150,000 contribute to almost 85 percent of GDP. These are the areas with the most potential for high-income earnings, but high costs of living can have a significant deterrence effect for citizens seeking to capitalize on those opportunities. As Greg Ferenstein recently noted:

In one of the most productive cities in America, San Francisco, average rent has rocketed past $3,500 a month, mostly because anti-skyscraper residents have made it illegal to build apartments in half of the city. Getting a permit to build a tall apartment complex can take upwards of 10 years because neighborhood groups have broad regulatory authority to delay construction.

As a result, talented engineers are fleeing the city, and their dreams for creating the next Facebook or Google are going with them. Not everyone needs to live in San Francisco, but it’s much easier to build high-growth companies in places with a dense concentration of talent. The fewer people who can afford to live in big cities, the less innovative America will be.

If fewer people can afford to take advantage of the network effects in major metropolitan areas, their ability to take advantage of better paying jobs is at risk. Every lost opportunity for an individual is also a potential loss for society, with fewer people participating in, and contributing to, the innovation economy. Diminished potential for innovation could stymie economic growth, while exacerbating social tensions. Much of the modern housing issue in American cities echoes those of early 19th century Britain. As Robert C. Allen wrote in a 2007 Oxford University working paper:

As British cities expanded, growing labour demanded bid up the price of housing and land, and much of the income gain was transferred to urban landowners. Faced with a rising cost of housing, workers responded by reducing their consumption: the result was overcrowding. The were limits as to how far this process could be pushed, and those limitations meant that rising rents translated into a rising share of income spent on housing.

According to a report published by the California’s Legislative Analyst’s Office (LAO), a major roadblock to meeting housing demands in California’s most sought-after coastal neighborhoods (Los Angeles, Oakland, San Diego, San Francisco, San Jose, and Santa Ana–Anaheim) has to do with resistance from existing residents. Fear of negative impacts on their home values combined with perceived negative complications related to increased development contribute to hostile attitudes towards developing land for residential purposes—the “Not In My Backyard,” or NIMBY, effect.

So-called “growth control” policies also play a role in limiting the development of housing in California’s coastal communities. According to the same LAO report, two-thirds of California cities have growth control policies. Policies such as limiting the number of homes built in a given year or limiting the height of buildings, place a physical limitation on the ability of developers to meet the demand for housing in these highly sought after regions.

Policy reform in this space will be complicated due to the hyperlocal nature of housing policy. The power to reform restrictive zoning laws is often in the hands of those who directly benefit from manipulating the housing supply through such mechanisms. Implementing housing policy reforms at the state level may decouple zoning regulations from local politics aimed at restricting housing supply.

New Approaches to Education

Regardless of how the future of work unfolds, investing in America’s talent today will be essential to reaping the economic benefits of the world of tomorrow. Pursuing policies that incentivize STEM education programs in primary, secondary, and postsecondary institutions will be essential in maintaining an innovative workforce that will be equipped to deal with rapid technological change. Perhaps more importantly, however, a reassessment of the role of certified education in society is needed. That means focusing on different means of accrediting the workforce, outside of the traditional four year university.

Bailey’s article cites an excellent report by Michael Mandel and Bret Swanson. It correctly identifies the potential of the information technology and digital sectors to radically transform physical industries in the coming years. They argue that it is necessary “to upgrade our education and workforce development systems to dramatically expand the number of Americans who can help create, and thrive in, the digitally-enabled economy.” Reducing the cost of higher education while “boosting collaboration between higher education and industry leaders and improving the relevance of curricula” are also cited as important public policy goals. Promoting more skills-based certification and training programs, especially through apprenticeships and trade schools, are necessary to ensuring workers are better prepared (and at a cheaper cost) for the emerging jobs in the digital economy.

Consider Germany, where according to The Wall Street Journal, “roughly half of high-school graduates opt for high-octane apprenticeships rather than college degrees.” Through a system of collaboration between employers, educators, and the federal government, students in Germany are afforded the option of developing in-demand skills via an apprenticeship arrangement with an employer. Apprentices participate in a dual-training program where they split their time between on-site training and in-classroom instruction. Funding for this scheme is low-cost to the state. Federal agencies provide public funds for the development and promotion of apprenticeship schemes, whereas the majority of the costs involved with training and educating an apprentice is the responsibility of the employer.

Heeding these calls can help fundamentally alter the future of work for future generations. Of more immediate concern, however, is the largest growing sector of the active labor market: older workers. According to a report produced by the New America Foundation in partnership with Bloomberg, one quarter of the workforce will be 55 or older by 2024.

Investments in opportunities for lifelong-learning programs will be necessary to enable those most at risk of being displaced by automation, thus allowing the older generation to participate in a new economy shaped by technological advancement. Incentivizing lifelong education for this at-risk population through income tax credits or grants will be necessary to inspire and motivate those who might otherwise be reluctant, unwilling, or unable to adapt to the changing nature of work.

Traditional education policy has focused on delivering resources for young students to obtain an equitable education regardless of where they live or their economic background. This same line of thinking will need to be adapted to provide an equitable education regardless of age. Providing resources for local community colleges and universities to offer new and innovative curricula, such as online MOOCs or apprenticeship learning, will provide diverse and affordable pathways for older workers to pursue new opportunities.

Social Safety Net Provisions

Although, the United States has fairly robust social welfare programs, they can certainly be improved. A report published by the Executive Office of the White House points out that domestic spending on active labor market programs amounts to just 0.1 percent of GDP. The Organization of Economic Co-Operation and Development (OECD) average, by contrast, is 0.6 percent. Some tout the need for retraining programs or incentivizing employers to invest in their workforce through a Worker Training Tax Credit.

While training or retraining our workforce to be productive in an economy powered by automation is a possible solution, there is potential to leave out segments of the population that are educationally disadvantaged. It is important to consider policies that will enable this segment of the population to transition into new work or to better plan for leaving the workforce.

In an interview with the MIT Technology Review, economic historian Joel Mokyr notes that “in the modern capitalist system your occupation is your identity.” Policies should help enable Americans to maintain their ability to function as productive members of society during their transition from low-skill labor to new jobs powered by automation. This can be achieved by improving the social insurance system.

One such proposal, recently floated by Rep. Ro Khanna, is to expand the Earned Income Tax Credit (EITC). The Niskanen Center’s Samuel Hammond has written in support of Rep. Khanna’s proposal, arguing that:

major disruptions to work on the horizon, with more and more routine jobs being automated by robots and artificial intelligence. The result will not be a lack of jobs, but rather a challenging period of transition to new jobs that leverage uniquely human capabilities like caring and emotional intelligence.

As Hammond has previously noted, expanding the EITC, in contrast to worker retraining programs like Trade Adjustment Assistance, can be a powerful incentive mechanism for both workers and firms to “aggressively seek each other out.” Mitigating the potential negative effects of automation on America’s labor force will require a thoughtful combination of expanding or augmenting current social nets while being open to policies that will benefit older workers participating in the workforce.

Conclusion

Robots probably aren’t going to eat all the jobs; at least not anytime soon.

Nonetheless, it is incumbent upon policymakers and technology analysts to start grappling with the potential outcomes of a more automated society. Arguing for doing nothing simply because the gains from automation could outweigh the potential costs ignores the inherent uncertainty of future events. If the labor displacement effects from automation are far greater than we anticipate, the unraveling of institutional trust could lead to unintended consequences that actually forestall future progress, while producing greater market uncertainty.

A dynamic economy that embraces innovation is, on net, a good thing. We shouldn’t forestall a future of limitless possibilities for the contentment of the present. But that isn’t an excuse for ignoring the many practical hurdles that exist between the present and that future. When Bailey and others argue that their opponents’ “lack of imagination blinds them to how people will use … technology to conjure millions of occupations now undreamt of,” it fails to acknowledge those roadblocks that currently exist, and which may be erected in the future.

None of this is to suggest we should back off innovation and digitization of the economy. Quite the opposite: if anything, we should aim for more innovation and digitization of the economy. What we cannot do, however, is simply hope for the best. We need to focus on how best to mitigate the risks associated with automation of the economy, and soon.