Faux “solutions” such as universal basic income, a tax on “robots,” or regulations that shackle innovation, are not only unnecessary, but would also be harmful, slowing income growth and keeping workers out of the labor market.
The findings, interpretations and conclusions are those of the authors and do not necessarily reflect the views of the European Investment Bank
>> You can download the essay here
Over the last decade productivity has declined in most advanced economies and Europe’s performance has been especially disappointing: since the last financial crisis labor productivity in the 28 EU member states grew just 0.7 annually. At this rate, it will take a century for Europe’s per-capita incomes to double. No wonder there is political unrest across the continent.
Thankfully, a “next production revolution”, enabled in part by artificial intelligence (AI), is emerging. This could boost the growth in Europe’s productivity, wage and gross domestic product (GDP) perhaps to as soon as five to ten years from now, but fully capturing the benefits of this new production revolution will require European policy makers and the European public to embrace, rather than resist its rapid emergence and the transformation of most industries in the continent.
At present developed countries appear to be in the midst of a period of relative stagnation: the existing information and communication technology system has reached to the top of the “S curve” (the S curve describes the shape of the technology lifecycle: at the bottom of the curve progress is slow, as the curve steepens progress speeds considerably, until reaching the top of the curve, when it slows again).
A decade or two ago, rapid improvements in operating systems, computer chips, broadband speeds, and smartphones mattered a lot. People and companies rushed to buy new computers when the latest Intel processor and Microsoft operating system came out, in the process scrapping perfectly good computers. But today, these and related technologies are not only improving more slowly (Moore’s Law, the prediction that computing power would double every 18 to 24 months, has slowed), but are already so good that scrapping existing equipment in favor of new ones is a less compelling proposition: computers, smartphones and broadband speeds are “good enough” for the majority of tasks.
This maturity, more than any other factor, likely explains the slowdown over the last decade in both capital investment and productivity in European economies. [2]
Economist Robert Gordon, perhaps the most widely cited pessimist, argues that advanced economies have picked virtually all the “low-hanging fruit” and future growth will stagnate. But Gordon and other pessimists do not fully appreciate the potential of the new technologies to improve in price and quality and therefore transform industries. As one example, Gordon dismissed the productivity potential of autonomous cars, failing to understand that the reduction in accidents and the decrease in traffic jams would generate an estimated $1 trillion in annual savings to the U.S. economy. [11]
Conversely, the techno-utopians, such as World Economic Forum leader Klaus Schwab, see the next production revolution as qualitatively different as past transformations and believe that the technology is advancing at an exponential rate. He writes, “We stand on the brink of a technological revolution that will fundamentally alter the way we live, work, and relate to one another. In its scale, scope, and complexity, the transformation will be unlike anything humankind has experienced before”. [12] Schwab and other pundits tell us that powered by artificial intelligence, fully autonomous vehicles, humanoid robots and other breakthroughs, change will come at rates that will make the Industrial Revolution look like a period of stability.
But such utopians base their predictions on over-optimistic assumptions. One is that computing power will advance at past rates. But, as mentioned, Moore’s law has already slowed and it is doubtful it will continue ad infinitum. [13] As Sanjay Banerjee, Director of the Microelectronics Research Center at University of Texas at Austin puts it, “no exponential is forever.” [14]
Second, there is simply no reason to believe that this coming technology wave will be any different in pace and magnitude than past waves. Each past wave led to improved technology in a few key areas (e.g., steam engines, railroads, steel, electricity, chemical processing, information technology, etc.) and these were subsequently used in other sectors. But none completely transformed all industries or processes.
Even if the share of jobs impacted by technological change is likely to be modest, the impact on individual workers can be challenging. While the past wave of automation had larger impacts on middle-wage jobs, both in services and manufacturing, it looks likely that the next wave will have significantly larger impacts on lower-wage and lower-skill occupations. Indeed, the correlation between average wage of an occupation and risk of automation in the United States is negative and quite large (-0.59 for the Oxford University estimates and -0.52 for the ITIF estimates). The correlation with average years of schooling for each occupation and risk of automation is also negative and large (-0.64 for Oxford, -0.51 for ITIF). And when using ITIF data, the occupations that have the highest risk of being automated have the lowest median wage ($32,380), the occupations with the next highest risk have the second lowest median wage ($34,990), and so on. The White House Council of Economic Advisors also used the Oxford data and found that 83 percent of jobs making less than $20 per hour would come under pressure from automation, as compared to 31 percent of jobs making between $20 and $40 per hour and just 4 percent of jobs making above $40 per hour. [26] The OECD also estimated that 44 percent of American workers with less than a high school degree hold jobs made up of highly automatable tasks while 1 percent of people with a bachelor’s degree or higher hold such jobs. [27] There is no reason to expect different effects in Europe, given the similarities of the economies and technologies to be used.
While this occupational differential will have some negative impacts, overall it is likely to be extremely positive. If low skill jobs are more likely to be automated it will mean that individuals with lower incomes are more at risk of displacement. And given their more limited resources (finances, social networks, and skills) making successful transitions to new employment can be more difficult.
At the same time, however, automating more lower-wage jobs will mean fewer of these jobs. Because the firms employing lower-wage workers in these increasingly automated occupations will be able to lower prices of their goods or services, consumers will have more purchasing power. That spending will create jobs at all wage levels. The net result will be an occupational shift to middle- and higher-wage jobs. This will be an unalloyed plus for many workers now stuck in low-wage occupations where it is difficult for employers to raise wages because of low productivity levels. But it is incumbent upon policy makers to enact policies and programs to more effectively help these workers successfully make employment transitions.
In addition, many workers in low-wage jobs have more skills than they need for their current job (the college grad waiting tables). This suggests that some workers in low-wage jobs have enough skills to move into higher paid, moderately skilled jobs. [28] In fact, a European Commission study estimated that 40 percent of EU workers are overqualified for the jobs they hold. [29] Some are in these occupations due to choice; but in other cases it is because there are not enough jobs in Europe that require a college education. These workers should have an easier time transitioning to newly created middle-wage jobs.
The big challenge
If Europe is to avoid an even greater populist, neo-Luddite backlash against the next production revolution, policy makers will need to take greater and more effective steps to help regions and individuals at risk from technology disruption.
One place to start is with better help for lagging regions. Some workers who lose their jobs from new technologies can and will move to regions where employment growth is stronger, but not all workers are willing or able to do so. As such, smart policies and programs to spur growth in lagging regions can help minimize social disruption from the next production revolution.
But the biggest challenge will be to help individual workers make successful transitions. European policymakers should embrace the concept of “flexicurity,” as Scandinavian nations have, which commits not to ensure that workers will never get laid off, but to minimize the number of workers at risk; and then, for those who are laid off, provide support so they can make successful and expeditious transitions. Policies limiting lay-offs will only postpone the inevitable. Likewise, providing laid-off workers with very generous and long-term benefits will not only help ensure higher unemployment rates, but also lead to more workers being out of the labor market for long periods of time, hurting the very workers the benefits are intended to help. For the longer a worker is out of the labor force, the harder it is for him or her to re-enter. Rather the goal should be finding a balance between being overprotecting and too severe.
To do that, policy makers should adopt the operational models of some of the world’s best-in-class programs, such as Singapore’s SkillsFuture program. The lessons from Singapore are fourfold. First, government policy needs to make a major commitment to skill development and workforce transition. Second, such efforts need to be closely linked to employers and markets, including through training vouchers and credits. Germany has done an excellent job in this regard with its longstanding and widespread employer-supported apprenticeship system. Third, such efforts need to be much more flexible and take full advantage of advanced information technology tools. Finally, incremental changes in existing institutional arrangements will not be enough. If policy makers are to respond effectively to the challenges of a more turbulent labor market, they will need to drive significant institutional reform, particularly in the high school and higher education sectors; provide more support for institutions focused on technical training; and provide skills valued by employers.
European nations may want to focus on several areas. The first is to enable more workers to obtain “better” skills and other competencies so that if they are dislocated by technology they will be better positioned to make a successful transition. One key is to shift the education system, particularly at the high school and post-secondary levels, towards an increased focus on teaching both “21st century skills” such as teamwork, analytical skills, critical thinking and more technical skills.
New skills required
As Manuel Trajtenberg writes in a study that addresses the next production revolution, the skills employers desire are seldom taught in school. Employers want workers with strong analytical, creative, and adaptive capabilities, but few secondary or collegiate schools impart these competencies. [30] Moreover, schools appear to be teaching technical subjects such as computer science and statistics poorly when compared to the needs of the next economy. [31] Thus, reforms such as high school career academies; [32] project-based learning; reducing the rigidity of state high school graduation course distribution and graduation requirements; and a focus on increased adoption of workforce-focused classes, including business, statistics, and engineering, would all help future workers have a stronger base of skills with which to manage a more turbulent workforce. In addition, more should be done to encourage and support corporate partnerships with new kinds of high schools. For example, IBM has worked to develop P-TECH (Pathways in Technology, Early College High School) in New York City, which runs from grade 9 to grade 14, and works to give students marketable skills in information technology.
At the same time, nations can do more to encourage employers to expand workforce training efforts. This can include wider use of portable skills credentialing; supporting sector-wide training and development plans, as Singapore has done; establishing an “Investors in People” program modeled on the UK’s effort to offer annual awards to employers who do the best job of investing in their workforce; supporting industry-led skills alliances; promoting greater use of apprenticeship programs, as Germany has done; and increasing use of portable training accounts, such as those established in France. [33]
European nations could also productively cooperate on how to better use technology to facilitate online skill assessment, career navigation, training and workforce placement. Many government-run websites are now limited in their offer. Governments should consider partnering with the private sector to improve their digital offer. For instance, in the United States the Markle Foundation’s Skillful Initiative, funded in part through Microsoft Philanthropies, has partnered with LinkedIn to help Colorado workers identify training for in-demand occupations. [34]
Flexicurity, a guiding principle?
Finally, the concept of flexicurity needs to be more than a commitment; it should evolve into active labor market policy. It needs to be a guiding principle by which European nations orient themselves toward technological change. Increasingly, many in Europe appear to view technologically driven employment loss as so disruptive to the individuals affected that society should attempt to slow the pace of change to a more “human” pace, or at minimum, not do anything to accelerate it. Bill Gates speaks for many in Europe (and the United States) when he says, “At a time when people are saying that the arrival of that robot is a net loss because of displacement, you ought to be willing to raise the tax level and even slow down the speed of that adoption somewhat to figure out, ‘OK, what about the communities where this has a particularly big impact? Which transition programs have worked and what type of funding do those require?’" [35]
Embracing flexicurity as an overarching guiding principle means rejecting these notions and acknowledging that technology-based productivity growth, some of which may lead to job displacement, is fundamentally a progressive force, without which wages and living standards will grow more slowly. To be more open to technological innovation, we must not apply the “precautionary principle”, but rather accept the hypothetical risk posed by technology. Imposing restrictive regulations on technologies based on speculative fears would only slow their development and limit their benefits. Countries should instead embrace the innovation principle, which says that policymakers should address risks as they arise, or allow market forces to address them, and not hold back progress because of hypothetical concerns. [36]
If European nations work together in a spirit of embracing the new production revolution, and ensure that the benefits are widely shared, they can look forward to a more prosperous economic future.
Artificial Intelligence will change the way we work. The EIB’s loans help SMEs and start-ups in Europe and beyond keep pace with the new emerging technologies. Read more about the EIB’s commitment to the digital innovation.
The findings, interpretations and conclusions are those of the authors and do not necessarily reflect the views of the European Investment Bank.
Download the essay
Find out more about Robert D. Atkinson.
© European Investment Bank 2018
Photos: © Getty Images. All rights reserved
Notes
[1] Robert. D. Atkinson, "The Past and Future of America’s Economy: Long Waves of Innovation that Power Cycles of Growth" , (Edward Elgar, 2006).
[2] Robert D. Atkinson, “Think Like an Enterprise: Why Nations Need Comprehensive Productivity Strategies”, (Information Technology and Innovation Foundation, May 2016).
[3] Daniel Castro and Joshua New, “The Promise of Artificial Intelligence”, (Center for Data Innovation, October 2016).
[4] Irving Waldawksy-Berger, “‘Soft’ Artificial Intelligence Is Suddenly Everywhere”, (The Wall Street Journal, January 16, 2016).
[5] Ibidem.
[6] Robert D. Atkinson, “It’s Going to Kill Us!’ and Other Myths about the Future of Artificial Intelligence”, (Information Technology and Innovation Foundation, June 2016).
[7] Daniel Bentley, “Why Ford Won’t Rush an Autonomous Car to Market", (Fortune, December 6, 2017).
[8] Rodney Brooks, “Robots, AI, and Other Stuff”, (Rodney Brooks Blog, January 1, 2018).
[9] Ibidem.
[10] Carlota Perez, "Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages", (Edward Elgar Pub: 2003).
[11] Robert D. Atkinson, “The Coming Transportation Revolution”, (Information Technology and Innovation Foundation, October 2014).
[12] Klaus Schwab, “The Fourth Industrial Revolution: what it means, how to respond”, (World Economic Forum, January 14, 2016).
[13] Robert D. Atkinson, “50 years of Moore's law, but for how much longer?”, (The Hill, April 16, 2015).
[14] “Are Advancements in Computing Over? The Future of Moore’s Law”, (ITIF Event, November 21, 2013).
[15] Robert D. Atkinson, “False Alarmism: Technological Disruption and the U.S. Labor Market, 1850–2015”, (Information Technology and Innovation Foundation, May 2017).
[16] Boston University School of Public Health, Behavioral Change Models, “Diffusion of Innovation Theory”, accessed January 1, 2018.
[17] Three percent growth per year leads to a doubling in 27 years.
[18] Robert D. Atkinson, “False Alarmism: Technological Disruption and the U.S. Labor Market, 1850–2015”, (Information Technology and Innovation Foundation, May 2017).
[19] Organisation for Economic Co-operation and Development (OECD), "Technology, Productivity and Job Creation: Best Policy Practices", (Paris: OECD, 1998), 9, accessed March 7, 2016.
[20] Carl Benedikt Frey and Michael A. Osbourne, “The Future of Employment: How Susceptible Are Jobs to Computerisation?”, (Oxford Martin School, University of Oxford, Oxford, September 17, 2013).
[21] Ben Miller, “Automation Not So Automatic”, (The Innovation Files, September 20, 2013).
[22] OECD, “Automation and Independent Work in a Digital Economy”, (May 2016).
[23] James Manyika, Susan Lund, Michael Chui, Jacques Bughin, Jonathan Woetzel, Parul Batra, Ryan Ko, and Saurabh Sanghvi, “Jobs Lost, Jobs Gained: Workforce Transitions in a Time of Automation”, (McKinsey Global Institute, December 2017).
[24] Michael Chui, James Manyika, and Mehdi Miremadi, “Four Fundamentals of Workplace Automation”, (McKinsey & Company: November 2015).
[25] OECD calculations based on the Survey of Adult Skills (PIAAC) (2012) and Arntz, M. T. Gregory and U. Zierahn (2016), “The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis”, (OECD Social, Employment and Migration Working Papers, No. 189, OECD Publishing, Paris).
Note: Data for the United Kingdom corresponds to England and Northern Ireland. Data for Belgium corresponds to the Flemish Community.
[26] Executive Office of the President, “Artificial Intelligence, Automation, and the Economy”, accessed January 5, 2018.
[27] Melanie Arntz, Terry Gregory, Ulrich Zierahn, “The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis”, OECD Library, OECD Social, Employment and Migration Working Papers, May 2016.
[28] One recent study found that over one-third of U.S. college graduates are overeducated in terms of the jobs they have, with similar numbers for EU nations. See Research working paper by Brian Clark and Arnaud Maurel from Duke University and Clément Joubert from University of North Carolina at Chapel Hill titled “The Career Prospects of Overeducated Americans” using data from the National Longitudinal Survey of Youth 1979 and Current Population Survey to look at overeducation’s effects on employment and wages over time. To analyze these effects, the researchers tracked almost 5,000 college graduates for 12 years after they entered the workforce. Their study shows that over one-third of college graduates are working in what the researchers call “overeducated employment”; https://www.bls.gov/opub/mlr/2013/article/clayton.htm; https://www.bls.gov/osmr/abstract/ec/ec060110.htm.
[29] European Commission, Skills Panorama, “Skill Underutilization across Countries in 2014”.
[30] Manuel Trajtenberg, “AI as the next GPT: A Political-Economy Perspective”, (NBER Working Paper No. 24245, January 2018).
[31] Only 7.7 percent of U.S. high school students pass a statistics class, compared to approximately 87 percent for geometry.
[32] Betsy Brand, “High School Career Academies: A 40-Year Proven Model for Improving College and Career Readiness”, (American Youth Policy Forum, November 2009).
[33] Ministère du Travail, Mon Compte Formation, accessed January 7, 2018.
[34] LinkedIn, “Training Finder, Denver Colorado”, accessed January 7, 2018.
[35] Kevin J. Delaney, “The robot that takes your job should pay taxes, says Bill Gates”, (Quartz, February 17, 2017).
[36] Daniel Castro, “Digital Decision-Making: The Building Blocks of Machine Learning and Artificial Intelligence”, (Information Technology and Innovation Foundation, December 2017).
