I recently read a very interesting paper, The Growing Importance of Social Skills in the Labor Market, by Harvard professor David Deming. Deming’s paper shows that over the past several decades, labor markets have been increasingly rewarding social skills, that is, interpersonal skills that facilitate interactions and communications with others. He presents evidence that since 1980, social-skill intensive occupations have enjoyed most of the employment growth across the whole wage spectrum, and that employment and wage growth have been particularly strong in jobs that require both high cognitive and high social skills.
Deming’s paper builds on the work of MIT economist David Autor, a leading authority on the impact of technological change on the US labor market. A few weeks ago I wrote about a paper he published this past summer on the history and future of workplace automation. In that paper, Autor summarized the sharp polarization of job opportunities that’s taken place in the US over the past two decades, noting that job opportunities have significantly expanded in both high-skill and low-skill occupations, while they have significantly contracted for mid-skill jobs.
The reason for this polarization is that automation has been most successful when applied to mid-skill routine tasks, that is, tasks or processes that follow precise, well understood procedures that can be well described by a set of rules. The occupations most susceptible to automation have included blue-collar physical activities such as manufacturing and other forms of production, as well as white-collar, information-based activities like accounting, record keeping, and many kinds of administrative tasks. As a result, these occupations have experienced the biggest declines in employment opportunities and earnings.
On the other hand, non-routine tasks are much harder to describe by a set of rules that a machine can follow. These include manual jobs in fast-food restaurants, janitorial services and health-care aides that require relatively low skills and education, as well as high-skill jobs that involve expert problem solving and complex communications requiring strong cognitive skills and a college education.
Deming’s paper sheds light on a mystery that’s been puzzling economists for the past few years. Since 2000, the growth of high-skill jobs has markedly slowed down. What accounts for the unexpected slowdown in the growth of these jobs?
Could it be that advances in technology, software, robotics and AI are now taking the potential for automation to a whole new level, including tasks requiring intelligence and cognitive capabilities that not long ago were viewed as the exclusive domain of humans? Perhaps automation is encroaching upward, beginning to substitute for professional, technical, and managerial occupations. Perhaps, the job polarization Autor wrote about only 5 years ago has turned into a downward ramp, with automation now affecting all jobs, routine and non-routine alike.
But, as Autor points out, this explanation does not fit well with the patterns of IT investment since 2000. The automation of mid-skill blue and white collar occupations in the 1980s and 1990s was accompanied by heavy capital investments in IT and manufacturing. There has been no similar surge in IT investments since 2000. “Instead,… information processing equipment and software investment was only 3.5 percent of GDP, a level last seen in 1995 at the outset of the dot-com era.”
Deming’s research sheds a strong light on this mystery. He analyzed changes in the task content of work using data from the Occupational Information Network (O*NET), which is administered by the US Department of Labor. He then focused on the changes in intensity in four key kinds of tasks:
- routine - degree of automation and of repetition of the same physical activity;
- nonroutine analytical - requirements for mathematical reasoning;
- service - degree of service orientation and of assisting and caring for others; and
- social skill - extent to which job requires coordination, negotiation, persuasion and social perceptiveness.
Deming’s overriding finding is that difficult-to-automate jobs are increasingly those requiring social skills. He presents evidence for three importance facts about the US labor market:
- “Employment growth in social skill-intensive occupations has occurred throughout the wage distribution, not just in management and other top-paying jobs.”
- “Since 1980, employment and wage growth has been particularly strong in occupations with high cognitive and social skill requirements. In contrast, employment has fallen in occupations with high math but low social skill requirements, suggesting that cognitive skills are increasingly a necessary but not sufficient condition for obtaining a high-paying job.”
- Routine occupations tend to require relatively low social skills, which further explains their employment and earnings declines over the past few decades.
Why are social skills so important in our fast-changing, technology-intensive economy? Let’s discuss a few potential answers to this important question.
Computers have made huge advances in automating many physical and cognitive human tasks, especially those tasks that can be well described by a set of rules and thus captured in software. But, as Professor Autor argued in a 2014 paper, despite continuing advances in AI and robotics, the “challenges to substituting machines for workers in tasks requiring flexibility, judgment, and common sense remain immense.”
Central to his argument is the concept of tacit knowledge. Explicit knowledge is formal, codified, and can be readily explained to people, captured in software and executed by a machine. Tacit knowledge, on the other hand, is the kind of knowledge that we’re often not aware we have, and is therefore difficult to transfer to another person, let alone to a machine. Generally, this kind of knowledge is best transmitted through personal interactions and practical experiences. Everyday examples include speaking a language, riding a bike, and easily recognizing many different objects and animals.
As Deming writes, “computers are still very poor at simulating human interaction. Reading the minds of others and reacting is an unconscious process, and skill in social settings has evolved in humans over thousands of years. Human interaction in the workplace involves team production, with workers playing off of each other’s strengths and adapting flexibly to changing circumstances. Such nonroutine interaction is at the heart of the human advantage over machines.” Social skills are increasingly vital in our work, precisely because they are the most intrinsically human and least susceptible to automation.
As work is becoming more team-based, you’d expect that workers with strong social skills who are more able to work well with others will be more valuable. Good teamwork increases productivity through comparative advantage, that is, the notion that different members of a team should specialize in those tasks that they’re best at. Workers with good social skills should be better able to play off each other’s strengths, quickly learn which tasks they are each best at, and flexibly adapt to changing circumstances.
Deming developed a mathematical model of workplace team production to formally test these ideas. The paper describes the model and its various findings in great detail. These include:
- People with higher social skills earn more money, even after controlling for cognitive skills and other factors affecting wages;
- Those with both strong cognitive and social skills command the highest wages, and their earnings have continued to grow since 2000; and
- Workers with high social skills tend to gravitate toward, social-skill intensive occupations, where they command higher earnings.
These findings are not surprising. The digital revolution has had a profound impact on all our social interactions, - giving us the ability to communicate through a variety of means and devices, easily share all kinds of information, and collaborate on just about any endeavor. Digital technologies, and the platforms and tools they’ve enabled, have radically amplified our social skills, transforming the way we interact and work with each other. It stands to reason that social skills are more valuable today than at any other time in history.