Two weeks ago today, online retailer Amazon announced Amazon Go, a brick-and-mortar grocery store concept without checkout lines. Shoppers simply scan their smartphones when they enter the store, and the company's detection technology tracks items that have been removed from the shelves, charging a customer's Amazon.com account when they leave.
It's the most recent signal amid the increasing noise about artificial intelligence and its impact on work. Other announcements include:
• Fast food chains such as McDonald's, Hardee's, and Carl's Jr. disclosing their plans to eliminate cashiers in their stores
• Wal-Mart trying out drones in its warehouses to check its inventory
• Amazon.com experimenting with drone delivery to consumers
• Alphabet's subsidiary Waymo providing a glimpse of its self-driving minivans
• Billionaire Vinod Khosla asserting that 80% of jobs in an IT department could be replaced by AI-type systems
• IBM's Watson recommending treatment plans that matched actual suggestions from oncologists in 90% of cases
One thing is clear: the automation onslaught is already underway.
An Argument Against The Hysteria
Technology columnist Christopher Mims argued last week that fears over artificial intelligence are overblown. "A long trail of empirical evidence shows that the increased productivity brought about by automation and invention ultimately leads to more wealth, cheaper goods, increased consumer spending power and ultimately, more jobs," he writes in the Wall Street Journal.
Last year The Hamilton Project, an economic policy initiative at the Brookings Institution, published a paper entitled, "The Future of Work in the Machine Age," in which the authors put a finer point on Mims' argument:
There is a consensus that, historically, technological progress has created winners and losers, but over the long run, new technology has tended to create more jobs than it has destroyed, while increasing society’s productivity and wealth. For example, between 1900 and 2000, the proportion of the U.S. workforce in agriculture fell from 41 percent to 2 percent, yet agricultural output rose dramatically and there was no long-term increase in the unemployment rate, even as a greater proportion of the population participated in the labor force (Autor 2014b). The children and grandchildren of the workers who might have tilled farmland in 1900 are now computer programmers, radiology technicians, and pilots—jobs created by technologies that were unknown in 1900. And nearly everyone is economically better off.
The coming technological developments will transform the economy in groundbreaking, unimaginable ways, at an arguably unprecedented speed and magnitude. Optimists maintain that robotics and machine learning are likely to substitute specific tasks that free workers to focus on new ones in the same occupation. But pure substitution will happen too, and the World Economic Forum projects a net loss of five million jobs between 2015-2020 despite the cushion of upskilling, redeployment, and productivity enhancement.
Simply put, this economic revolution may not unfold like prior ones.
Even if the results are, over the long run, good for everyone, something must be done about worker displacement in the interim.
Time For A New Educational Movement
Mims writes that when farm automation threatened the U.S. agricultural economy, the country adapted to the changing circumstances:
In response, they created the “high school movement,” which required everyone to stay in school until age 16. It was hugely expensive, both because of the new schools and teachers, but also because these young people could no longer work on the farm. But it better prepared workers for 20th century factory jobs and fueled the explosion in college attendance after World War II."
Broad-based education was key to reducing the dislocation from technological innovation.
Is it possible, then, to spark a new educational movement to address the unique challenges of the fourth industrial revolution?
Despite high unemployment during the Obama administration, many jobs have remained open for significant periods of time. In 2011, with the U.S. unemployment rate at 9%, a survey of 2,000 U.S. companies found that 600 had positions open for more than six months that they could not fill. A skills gap has persistently plagued the economy, and this phenomenon will worsen in the wake of further disruption.
"More than half of all jobs (54%) in the US today are middle-skill jobs that require more than a high school diploma, but not a four-year degree," the National Skills Coalition reports. "Yet only 44% of workers are trained to the middle-skill level."
The supply of skilled workers has not been keeping up with demand. In April, U.S. Chamber of Commerce President and CEO Thomas J. Donohue penned an op-ed in which he acknowledged the role employers must play in bridging this gap and filling the 5.6 million jobs. The fields most affected by the gap are advanced manufacturing, health care, engineering, and computer science.
Education, in all its guises, is not a panacea. Some people won't be suited for these future jobs, no matter the amount of education they receive. We need to be honest with ourselves that structural unemployment and sustained under-employment are possible outcomes.
Nevertheless, the choices we make today about how to educate the population, including workers whose skills are becoming obsolete, will determine the degree to which our country flourishes and prosperity is shared in the new economy.
The Economic Resiliency Of College Towns
Recent history shows that education can offset losses in a dynamic economy. The trend of offshoring to China spurred the manufacturing downturn of the 1990s, decimating blue collar jobs. Yet some areas have been surprisingly resilient, and many are home to a major university, the Wall Street Journal reports.
Auburn University, in addition to providing employment and training to the local workforce, has attracted new businesses. In 2014 GE selected its new Auburn plant to be the first to use 3-D printing to make products at scale, believing the university could supply an educated workforce and collaborate on research projects.
In another example, after Mattel moved its production from Kentucky to Mexico in 2002, the affected county's unemployment rate reached 9.3%. Soon thereafter, Pella, a custom window maker, moved into the empty plant and cozied up to nearby Murray State, counting on the university for graduates, influence over public schools, and help with software and manufacturing issues. Since then the county's unemployment rate has dropped to below 4%.
College towns are remarkably responsive to industry ebbs and flows, but it's simplistic to say that the presence of a university is sufficient to produce economic revival. Merced, CA and Binghamton, NY -- both home to universities -- are still struggling to recover from the downturn.
And automation is expected to cut deeper into the economy than globalization did.
Government Investment In Workforce Development
Some say that the government should do more to retrain displaced workers.
In 2008, Georgetown University Professor Harry J. Holzer testified to Congress that the federal government should play a more active role in addressing this problem by investing more heavily in workforce development.
The US spends a tiny fraction of its GDP on this issue, a far lower share "than what is spent on active labor-market policy by almost any other industrialized nation," Holzer says.
In his testimony, he summarizes the lessons learned from the evaluations of workforce development programs: "The most promising models of workforce development today involve partnerships among industry and employer groups, community colleges, state and local agencies (including workforce boards), community groups, and intermediary organizations that can bring them all together."
Holzer believes that a greater investment will prevent some of the adverse social and economic consequences of high unemployment.
The federal government's involvement in worker displacement dates back to the Great Depressionwhen Congress passed the Wagner-Peyser Act, creating a nationwide system of employment offices to match workers to jobs.
This investment was expanded in the 1960s and 1970s and today nine federal agencies spend $18 billion to administer 47 programs.
As early as the 1990s the Government Accountability Office issued a series of reports that raised questions about the efficiency and effectiveness of the federally funded employment and training system. They concluded that a structural overhaul and consolidation of these programs was needed.
In 1998, the Workforce Investment Act reorganized federal employment and training programs. A new feature of WIA was the creation of career centers across the nation to co-locate job search, counseling, and training services and, ideally, transform the fragmented employment and training system into a coherent, one-stop system. Unfortunately these programs continue to be disparate and duplicative.
More alarmingly, the impact of these programs -- and taxpayers' dollars -- are hard to pin down. At the time only five of the programs had conducted detailed impact studies. "The five impact studies generally found that the effects of participation were not consistent across programs, with only some demonstrating positive impacts that tended to be small, inconclusive, or restricted to short-term impacts," the GAO found.
In 2012, the WSJ reported that of people who completed a dislocated-worker training program, 78% said they found a job within three months, but only 38% said they landed a job in the same field as their training (a Labor Department spokesperson said that the success rate is closer to 50% when those who didn't respond to the survey are factored out). No efforts are made after three months to monitor whether those trained are working in the fields they selected.
Concerns over fragmentation and accountability led to the 2014 Workforce Innovation and Opportunity Act, which consolidated various programs into a single funding stream and requires them to record and report on how many people get new jobs as a result of program participation.
I'm skeptical that additional governmental measures will rid a sprawling administrative apparatus of its waste and redundancy. At a minimum, these programs enable politicians to say that they're doing something to address the problem. The preference for talking points during re-election season creates considerable resistance to change.
Navigating The Emerging Learning Ecosystem
The impact of these programs are marginal for another reason: few people use them.
In 2010, when the unemployment rate hovered around 10% and the number of unemployed neared 15 million, the Labor Department stated that 6.7 million people used its WIA services. When counting only those who used staff-assisted services, however, this number plunges to 1.1 million and comes down even further to more like 130,000 when constrained to those who took advantage of a federal training program.
That's right: 130,000 out of 15 million.
A new ecosystem of market solutions is emerging, offering alternatives to federally funded programs. It exists, for the most part, outside of traditional institutions and encompasses services like headhunting, career counseling, life coaching, executive recruiting, resume writing, outplacement, temp agencies, career fairs, networking organizations, community colleges, online courses (e.g., MOOCs), and on and on.
Despite the availability of training, workers struggle to select programs that yield a high return in the job market. The Hamilton Project identified the following five reasons why:
• Lack of awareness of offerings, especially certificate and career-oriented programs that have a high ROI
• Poor visibility into variation in financial returns across programs
• Difficulty assessing their personal readiness for available training programs
• Lack of access to effective career counseling and guidance
• Inability to compare the quality of programs
Perhaps more people would take advantage of the abundant learning opportunities if information about them were consumable, comparable, and flat-out correct.
The Learning Platform Of The Future
Unless we act boldly, technology disruption will further widen economic inequality and tear apart the social fabric that remains. Now is the time to lay the groundwork for a resilient economy able to withstand the shocks coming our way. Education is essential to that future, but a new approach is called for.
We need to offer our youth multiple pathways and reemphasize vocational training as a viable -- even celebrated -- option for students.
But this movement must extend well beyond secondary and tertiary education. We need a system that enables adults to pursue continual, iterative learning that is personalized to their needs and available on-demand.
To be effective, it must start with a tight partnership between the public and private sectors. Unlike the Beltway, where government officials are removed from the crushing realities of market competition, business will need to ensure that training is grounded in real-world practicalities.
Somehow, real-time data on labor supply and demand should be made widely accessible, like electricity, to stakeholders engaged in solving this puzzle. Then qualified coaches can use it and other tools to direct job-seekers to fields where the future is bright.
If gaps exist between their skills and job requirements, workers can browse a marketplace of learning opportunities, one that is provider-agnostic. When they finish their supplemental learning, staffing experts assist in placing them at employers.
Finally, when more shocks come -- and they will -- workers can fluidly reengage the system to navigate their next transition.