“The health of local economies today will affect their ability to adapt and thrive in the automation age.
The US labor market looks markedly different today than it did two decades ago. It has been reshaped by dramatic events like the Great Recession but also by a quieter ongoing evolution in the mix and location of jobs. In the decade ahead, the next wave of automation technologies may accelerate the pace of change. Millions of jobs could be phased out even as new ones are created. More broadly, the day-to-day nature of work could change for nearly everyone as intelligent machines become fixtures in the American workplace.
Until recently, most research on the potential effects of automation, including our own, has focused on the national-level effects. Our previous work ran multiple scenarios regarding the pace and extent of adoption. In the midpoint case, our modeling shows some jobs being phased out but sufficient numbers being added at the same time to produce net positive job growth for the United States as a whole through 2030.
The day-to-day nature of work could change for nearly everyone as intelligent machines become fixtures in the American workplace.
But the national results contain a wide spectrum of outcomes. A new report from the McKinsey Global Institute, The future of work in America: People and places, today and tomorrow (PDF–4.41MB), analyzes more than 3,000 US counties and 315 cities and finds they are on sharply different paths. Automation is not happening in a vacuum, and the health of local economies today will affect their ability to adapt and thrive in the face of the changes that lie ahead.
The trends outlined in this report could widen existing disparities between high-growth cities and struggling rural areas, and between high-wage workers and everyone else. But this is not a foregone conclusion. The United States can improve outcomes nationwide by connecting displaced workers with new opportunities, equipping people with the skills they need to succeed, revitalizing distressed areas, and supporting workers in transition. Returning to more inclusive growth will require the combined energy and ingenuity of business leaders, policy makers, educators, and nonprofits across the country.
Local economies have been on diverging trajectories for years
Cities and counties across the United States are entering this period of technological and labor market change from different starting points. We used a mathematical clustering method to categorize all US cities and counties into 13 archetypes based on their economic health, business dynamism, industry mix, labor force demographics, and other characteristics (download the full list of locations in each segment). This approach reveals that the differences between local economies across the country are more nuanced than a simple rural-urban divide or regional variations. Our 13 archetypes can be grouped into five segments with common patterns:
• Urban core. Twenty-five megacities and high-growth hubs account for roughly 30 percent of the US population and are the nation’s most dynamic places. The high-growth industries of high tech, media, healthcare, real estate, and finance make up a large share of these local economies. These cities have higher incomes, faster employment growth since the Great Recession, high net migration, and younger and more educated workforces than the rest of the country—but also high levels of income inequality. Many are experiencing congestion and affordable housing shortages.
Differing outlooks for local economies
Partner Susan Lund explains why the impact of automation will play out differently depending on where you live.
• Urban periphery. These 271 counties are the extended suburbs of US cities. Home to 16 percent of the US population, they also have seen strong net migration, attracting people moving out of cities in search of more space. In most of these counties, a large share of the population works in nearby urban areas. Healthcare, retail, logistics, and local services are large parts of these local economies.
• Niche cities. These 56 much smaller towns and cities, home to 6 percent of the US population, have found success by building on unique features. In college-centric towns, a major research university dominates the local economy. Silver cities, many of which are in Florida, are fast-growing retirement destinations. Small powerhouses, such as Bend, OR, and Provo, UT, have built economic clusters around technology and other industries; they have the fastest economic growth rates and second-highest rate of net migration across our archetypes. All niche cities are attracting both workers and companies with a low cost of living and a high quality of life.
• Mixed middle. Almost one-quarter of the nation’s population is found in these 180 stable cities (such as Cincinnati and St. Louis), smaller independent economies (such as Lancaster, PA, and Winston-Salem, NC), and the manufacturing hubs that we call “America’s makers” (such as Rockford, IL, and Oshkosh, WI). Neither thriving nor in distress, these places have slower economic and job growth, higher unemployment, and workforces with slightly lower educational attainment than those in urban core cities. Some of America’s makers are on an upward trajectory, while others are in decline.
• Low-growth and rural areas. This group, which includes 54 trailing cities and more than 2,000 rural counties, is home to one-quarter of the US population. Many trailing cities, such as Flint, MI, and Bridgeport, CT, are former industrial towns with declining economies. Rural counties encompass somewhat better-performing places (Americana) and struggling areas (distressed Americana). In these segments, populations are older, unemployment is higher, and educational attainment is lower than the national average. Things are somewhat brighter in the 192 rural outlier counties that have found some success with tourism or mining and energy.
What will the future of work look like in your community?
A new report maps local labor markets today and weighs the impact of automation on people and places.
The economic performance of these segments has been diverging for decades, and that trend accelerated after the Great Recession. While all areas of the country lost employment during the downturn, job growth since then has been a tale of two Americas. Just 25 cities (megacities and high-growth hubs, plus their urban peripheries) have accounted for more than two-thirds of job growth in the last decade (Exhibit 1). By contrast, trailing cities have had virtually no job growth for a decade—and the counties of Americana and distressed Americana have 360,000 fewer jobs in 2017 than they did in 2007.
Population growth has also tilted toward urban America. High-growth hubs, small powerhouses, and silver cities have grown by more than 10 percent since 2007, and most urban peripheries are also growing. Residents have been moving out of megacities, stable cities, America’s makers, and trailing cities. Immigration has more than offset domestic population losses in megacities and stable cities, but populations in rural Americana counties grew by less than 1 percent—and distressed Americana is shrinking.
Growing economic divergence might have been expected to prompt more people to move from distressed areas to thriving job markets. Yet geographic mobility in the United States has eroded to historically low levels. While 6.1 percent of Americans moved between counties or states in 1990, only 3.6 percent did so in 2017. Furthermore, when people in rural segments and less vibrant cities do move, it is usually to places with a similar profile rather than to megacities or high-growth hubs (Exhibit 2). Differentials in the cost of living, ties with family and friends, and a growing cultural divide all partially explain these patterns, but more research is needed to understand these patterns.
Automation will not be felt evenly across places or occupational categories
Previous MGI research has found that less than 5 percent of occupations can be automated in their entirety, but within 60 percent of jobs, at least 30 percent of activities could be automated by adapting currently demonstrated technologies.1 What lies ahead is not a sudden robot takeover but a period of ongoing, and perhaps accelerated, change in how work is organized and the mix of jobs in the economy. Even as some jobs decline, the US economy will continue to create others—and technologies themselves will give rise to new occupations. All workers will need to adapt as machines take over routine and some physical tasks and as demand grows for work involving socioemotional, creative, technological, and higher cognitive skills.
We modeled scenarios with varying timelines for the widespread adoption of automation technologies in the American workplace and base our research on the midpoint adoption scenario. Our model shows some local economies experiencing more disruption than others. At the high end of the displacement spectrum are 512 counties, home to 20.3 million people, where more than 25 percent of workers could be displaced. The vast majority (429 counties) are rural areas in the Americana and distressed Americana segments. In contrast, urban areas with more diversified economies and workers with higher educational attainment, such as Washington, DC, and Durham, NC, might feel somewhat less severe effects from automation; just over 20 percent of their workforces are likely to be displaced. These differences are explained by each county’s and city’s current industry and occupation mix as well as wages.
The coming wave of automation will affect some of the largest occupational categories in the US economy.
The coming wave of automation will affect some of the largest occupational categories in the US economy, such as office support, food service, production work, and customer service and retail sales (Exhibit 3). Nearly 40 percent of US jobs are currently in occupational categories that could shrink between now and 2030. A common thread among shrinking roles is that they involve many routine or physical tasks. Because these roles are distributed across the country, no community will be immune from automation-related displacement.
These losses will not necessarily manifest as sudden mass unemployment. Many occupations are likely to shrink through attrition and reduced hiring. This has already been occurring in office support roles, for instance. Offices once populated by armies of administrative assistants, research librarians, and payroll and data clerks now run with leaner support teams and more digital tools. Administrative assistants, bill collectors, and bookkeepers lost a combined 226,000 jobs from 2012 to 2017.
Even as some occupations decline, the US economy should continue to grow and create new jobs in the years to 2030. But the occupational mix of the economy is evolving and could do so at an even faster pace in the decade ahead. While employment in categories such as office support and food service may decline, our scenario suggests strong job growth in healthcare, STEM occupations, creative fields, and business services. Growth and displacement may occur even within the same occupational category. In customer service and retail sales, for example, counter attendants and rental clerks may decline, but more workers could be added to help customers in stores or to staff distribution centers.
Despite new occupations and overall job growth, one worrisome trend could continue: the hollowing out of middle-wage jobs. Our analysis suggests that by 2030, they could decline as a share of national employment by 3.4 percentage points. Our model shows employment in low-wage jobs declining by 0.4 percentage point, while employment in the highest-wage jobs grows by 3.8 percentage points. But the growth of high-wage opportunities can be realized only if workers can obtain the necessary education and skills. Forging career pathways to help people move up and finding sources of future middle-wage jobs will be essential to sustaining the US middle class.
In the decade ahead, local economies could continue to diverge
Workforce transitions will play out differently in local communities across the United States (Exhibit 4). Our findings suggest that net job growth through 2030 may be concentrated in relatively few urban areas, while wide swaths of the country see little employment growth or even lose jobs.
The 25 megacities and high-growth hubs, plus their peripheries, may account for about 60 percent of net job growth by 2030, although they have just 44 percent of the population. Individual standouts like Phoenix and Austin have diverse economies and high concentrations of the tech and business services that may boost job creation. But even the most thriving cities will need to connect marginalized populations with better opportunities.
Some niche cities are also well positioned. Small powerhouses could enjoy 15 percent employment growth on average by 2030, fueled in many places by technology businesses. Silver cities are riding a wave of growth as the retirement-age population swells. Employment in this segment could grow by 15 percent as seniors drive demand for healthcare and other services—and as more of them continue working past traditional retirement age. College-centric towns may see 11 percent employment growth over the next decade; they can build on their well-educated talent pools.
25 megacities and high-growth hubs, plus their peripheries, may account for about 60 percent of net job growth by 2030.
On the other end of the spectrum, the decade ahead could be a rocky one for rural America (interactive). Low-growth and rural areas as a group account for 20 percent of jobs today but could drive as little as 3 percent of job growth through 2030. Our model indicates anemic 1 percent employment growth over the entirety of the next decade in the more than 1,100 rural Americana counties. Rural outlier counties should continue to sustain growth through natural resources and tourism, although they may manage job growth of only 3 percent. The picture is worst for the roughly 970 distressed Americana counties that are entering the decade in poor economic health. Our model suggests that these areas could experience net job loss, with their employment bases shrinking by 3 percent.
The mixed middle cities are positioned for modest jobs gains. Some could manage to accelerate growth, but in a period of change and churn, others could slip into decline. Many stable cities and independent economies have relatively educated workforces and could become attractive regional outposts for corporations looking to expand into lower-cost locations. America’s makers may see mixed results; they will need clear strategies to shift to advanced manufacturing and rebuild local supply chains.
Less educated workers are most likely to be displaced, while the youngest and oldest workers face unique challenges
Understanding who holds the occupations with the highest automation potential today is an important first step for designing targeted interventions and training programs (Exhibit 5). We find that individuals with a high school degree or less are four times more likely to be in a highly automatable role than individuals with a bachelor’s degree or higher—and as much as 14 times more vulnerable than someone with a graduate degree.
Because some racial minorities have lower educational attainment, we find they are more vulnerable to being displaced by automation. Hispanic workers, for instance, are overrepresented in food service roles and have the highest rate of potential displacement among all minority groups, at 25.5 percent (7.4 million individuals). For African Americans, the potential displacement rate is 23.1 percent (4.6 million individuals). White workers have a potential displacement rate of 22.4 percent, and Asian-American workers have the lowest rate at 21.7 percent.
Automation will affect workers across age brackets, but both the youngest and oldest segments of the labor force face unique risks. Young people will need new career paths to build skills and gain a foothold into the working world. Tens of millions of Americans can think back to their first jobs in retail or food service—roles that gave them valuable soft skills and experience that propelled them on their way. But these are the very roles that automation could phase out. Roughly 14.7 million workers under age 34 could be displaced by automation; almost half of them are in roles with high separation rates, so employers lack incentives to retrain and redeploy them. It will be important to create a wider variety of pathways from high school to work, perhaps through apprenticeship.
Automation will affect workers across age brackets, but both the youngest and oldest segments of the labor force face unique risks.
On the opposite side of the generational divide, some 11.5 million US workers over the age of 50 could be displaced by automation. Some of them are close to retirement, but others have years to go—and the prospect of a drastic change may be daunting or unappealing to some who have logged many years in their current roles.
Many of the specific jobs most at risk from automation skew heavily toward one gender or the other. Men, for example, make up the majority of drivers and assembly line workers, while administrative assistants and bookkeepers are predominantly female. Overall, women represent 47 percent of the displaced workers in our midpoint automation scenario, while men are 53 percent. Based on the current gender share of occupations, our modeling suggests that women could capture 58 percent of net job growth through 2030, although the gender balance in occupations can and does change over time. Much of this is due to women’s heavy representation in health professions and personal care work—and some of these roles are low-paying. Improving the representation of women in the tech sector is a priority; today they hold only 26 percent of computing jobs in the United States.
Local business leaders, policy makers, and educators will need to work together to chart a new course
The next decade will bring every community new challenges—but also new opportunities to boost innovation, productivity, and inclusive growth. Even in the nation’s most prosperous cities, large populations are already struggling to find a place in the new economy and keep up with the rising cost of living. But in general, cities are more diversified and have more resources and investment flows on which to draw. Reinvention will be a harder task for trailing cities, some manufacturing towns, and rural counties that never bounced back from the Great Recession. The good news is that there is a growing tool kit of potential solutions, and many promising pilots are under way. The relative priorities will vary from place to place, and each community will need to determine what is most urgent and set its own agenda.
Connecting workers with new opportunities
A central challenge in the automation age will be connecting millions of displaced workers to new, growing jobs. Some may need to change jobs within the same company, and employers would provide the necessary training in these situations. But many workers may need to switch employers or make even bigger moves to different occupations in new locations. For these workers, governments and other stakeholders can help to make local labor markets more fluid and easier to navigate.
A central challenge in the automation age will be connecting millions of displaced workers to new, growing jobs.
In a more technology-driven world, job-matching efforts can be aided by a range of new digital tools and should run on easily accessible digital platforms. New online tools can assess an individual’s skills, suggest appropriate career choices, and clarify which jobs are in demand and the credentials needed to obtain them. Many efforts are under way to centralize and standardize information on skills, job postings, and credentials.
Geography itself can be a barrier to connecting to new opportunities, given the declines in Americans’ mobility. It is sometimes suggested that people should simply leave distressed places and move to where the jobs are. But this greatly oversimplifies the weight of this decision for individuals who may have deep personal and family ties to the places where they live, as well as economic barriers to leaving. Addressing the affordable housing shortage in the fastest-growing urban areas would enable people who do want to move for better opportunities to do so (and would create demand in the construction sector at the same time). Because there is a national benefit to improving labor market fluidity, policy makers might consider providing relocation assistance or tax credits.
Retraining workers and providing lifelong learning
Workforce skills have been a growing concern in the United States for many years. Now technology demands new and higher-level skills, including more critical thinking, creativity, and socioemotional skills. The skills needed in fast-growing STEM roles, in particular, are continuously evolving. The old model of front-loading education early in life needs to give way to lifelong learning. Training and education can no longer end when workers are in their twenties and carry them through the decades.
The old model of front-loading education early in life needs to give way to lifelong learning.
Employers will be the natural providers of training and continuous learning opportunities for many workers. But millions who need to switch employers or change occupations will need training options outside the workplace. All levels of government, nonprofits, education providers, and industry associations can play a role here. Midcareer workers need to continue paying their bills while they train for the next chapter in their careers; they require short, flexible courses that follow the boot camp model, teaching new skills in weeks or months rather than years.
The challenge ahead is to scale up the most successful programs. Using data to track employment outcomes will be essential so that funding can be channeled into what works and individuals can make more informed choices about their own training and careers. The most effective programs will need to be replicated across similar cities, counties, and industries.
Creating tailored economic development strategies to boost job creation
Every community, from the most dynamic to the most distressed, faces economic development issues that need to be solved at the local and regional level. For megacities and high-growth hubs, the priorities may be connecting disadvantaged populations with new opportunities, adding affordable housing, and improving transportation. The communities in the mixed middle segment need to accelerate economic growth and focus on entrepreneurship and skills development.
For rural counties, the road is tougher. Many of these places do not have the vibrancy, economic activity, or inflows of investment or people to create new jobs. No amount of workforce retraining can solve the bigger challenge of lack of economic activity. Individual companies can help to ease this strain by considering whether there is a business case for establishing operations in more affordable parts of the country that need the investment.
Turning around places that have lost their economic dynamism is a multiyear journey, but it is possible.
Turning around places that have lost their economic dynamism is a multiyear journey, but it is possible. Each community will have to take inventory of its assets, such as available industrial space, natural attractions, local universities, and specialized workforce skills. That data can form the basis of an economic development plan built around a growth engine industry that can create jobs and spillover effects. The next step is attracting investment, which does not have to come from within the United States. Subsidies and tax incentives can be part of the tool kit, but they need to be backed by a rigorous business case.
The growing acceptance of remote working models could be a positive trend for creating jobs in rural counties, whether full-time work-at-home employee roles or contract work. But it will take a push to continue building out fast, affordable broadband in the regions that still need service. The Rural Innovation Initiative, recently launched in nine communities nationwide, is building outposts for workers in the downtowns of rural cities, aiming to spur professional collaboration and nurture tech talent across the country.
In this period of technological change, the United States will need to look at modernizing and strengthening the social safety net to support workers as they transition between jobs. Some of the people most likely to be affected are already living paycheck-to-paycheck. For them, even a short period of disruption could provoke tremendous stress.
Support can take many forms: longer and more flexible income support programs during periods of unemployment, relocation assistance, training grants, and earned income tax credits. Portable benefits—tied to the worker rather than the employer—could offer stability to people who need to move between opportunities and geographies as well as covering the millions of Americans who are self-employed or independent workers.
Wages and purchasing power are real concerns. Although a tighter labor market may increase wage growth in the short term, it will take sustained growth to counter the trend of wage stagnation, which dates to the 1980s. Policy makers and employers alike cannot ignore the implications if a large share of the population is falling behind.
The United States does not have to let opportunity concentrate in a limited number of places, some of which are straining at the seams, while others wither. Policy choices, along with increased public and private investment in people and in the places that need it, can create more inclusive growth. Companies can make a difference, too, in recognizing that talent, space, and untapped potential are available all over the country. It is possible to turn this period of technological change into an occasion to create more rewarding jobs and build better learning systems and career pathways that serve more Americans. The challenge is not fighting against technology but preparing US workers to succeed alongside it.”
By Susan Lund, James Manyika, Liz Hilton Segel, André Dua, Bryan Hancock, Scott Rutherford, and Brent Macon
More information in www.mckinsey.com