Not so many years ago, there was a real fear that all jobs would be overtaken by computers. Now we know that is bunk. Technology has indeed changed the face of work, but not in the ways many people once predicted. In a provocative new book, economists Frank Levy and Richard J. Murnane discuss the kinds of jobs that are likely to endure and the kinds that will fall by the wayside—and the resulting impact on society. Their book, The New Division of Labor: How Computers Are Creating the Next Job Market, asks and answers four questions:
Levy and Murnane recently joined forces for an e-mail interview with HBS Working Knowledge. Levy is the Daniel Rose Professor of Urban Economics at the Massachusetts Institute of Technology; and Murnane is Juliana W. and William Foss Thompson Professor of Education and Society at Harvard University.
Martha Lagace: In The New Division of Labor, both of you make the very interesting point that during the Industrial Revolution, technology brought low-skilled workers to the fore and displaced higher-skilled craftspeople such as weavers and clock makers. Something different is happening now. How do you see work evolving over the next couple of decades?
Frank Levy and Richard J. Murnane: Strange as it sounds, computerized work creates both high-skilled and low-skilled jobs. With a few exceptions, it is the "middling skilled jobs" that are most at risk. To understand this pattern, begin with the fact that computers excel at jobs that can be described as a series of logical rules. Think about the airline kiosks that handle self-service check-in:
This "rules-based" repetitive work occurs most frequently in clerical jobs—particularly back office work—and in assembly line work. These jobs are also vulnerable from a second direction because the ability to describe a job in rules makes it easier to move the jobs to a lower wage country with minimal misunderstanding.
Conversely, three main types of work cannot be described in rules:
1. Identifying and solving new problems (if the problem is new, there is no rules-based solution to program).
2. Engaging in complex communication—verbal and non-verbal—with other people in jobs like leading, negotiating, teaching, and selling.
3. Many "simple" physical tasks that are central to janitorial work, waiting on tables, and other service work. (For example, entering an unfamiliar room and making sense of what you see is trivial for a human but extremely difficult to program.)
You can see these effects at work in Figure 1 below, which shows how occupations have evolved over the last thirty years. Low-skilled service work—for example janitors and security guards—is growing in importance, as are sales jobs and high-skilled technical and professional jobs. The relative declines come in manufacturing and clerical jobs—rules-based jobs. Computers are not the only factor in this hollowing out, but they play an important role.
Advances in computerization in the coming years will hollow out the occupational distribution even more, leaving a smaller and smaller percentage of the U.S. labor force employed in manufacturing and clerical jobs. Workers with the skills to do the growing number of managerial, technical, and sales jobs will prosper. Those without the requisite skills will be forced to compete for service jobs, the number of which is growing, but which do not pay well because almost all workers can do these jobs.
Q: Should anyone today feel secure about his or her job skills and prospects? What should someone do to better ensure that he or she will have well-paying work for the future?
A: The age of computerized work began sometime in the late 1960s. Since that time, U.S. employment has grown by 40 million jobs (+75 percent). So the challenge of computerized work is not to compete for a declining number of jobs, but rather to respond to the changing mix of jobs.
Least secure are people who have made a living carrying out the same task repetitively. While many clerical and production jobs fit this description, some higher-end jobs fit the description as well—floor traders in securities exchanges whose jobs are being taken by computerized trading networks.
Most secure are the people who know how to learn new material rapidly and how to communicate effectively with other people. It also helps, of course, to do work that must be done here—teaching a fourth grade class or detecting U.S. marketing trends rather than solving physics problems that might be solved equally well in China.
Q: And what about children? Given the conditions of typical school systems (at least in the U.S., though your book has implications worldwide), what can our readers who are parents do to help their school-age children develop necessary problem-solving and complex communication skills?
A: Parents should keep four ideas in mind when thinking about their children's education.
1. Everything starts with reading, math, and writing skills. A child needs strong reading and mathematics to learn particular subjects deeply enough to develop effective problem solving. A child needs to write well to communicate effectively, an increasingly important component of virtually all well-paying work.
|The challenge of computerized work is not to compete for a declining number of jobs, but rather to respond to the changing mix of jobs.|
2. The traditional subjects should be taught in a manner that emphasizes relationships as well as facts—relationships that allow a child to transfer knowledge to new problems. Mathematics should focus both on the right answer and on the proper structuring of the problem. History should focus both on facts (e.g., the date of the battle of the Spanish Armada) and structures that explain why facts are important (e.g., the relationship between the battle of the Spanish Armada and the colonization of the United States).
3. Check the tests your child takes. The kind of learning we describe is measured through English language tests that require students to write compositions, and mathematics tests that require students to figure out an effective problem-solving strategy and describe this strategy in writing. For better or worse, teachers teach to the test and so if your child only takes multiple choice tests, working on the structure of relationships, problem-solving skills, and writing are likely to get short shrift in the classroom.
4. Your child shouldn't be spending much time on developing computer skills per se. Computers should be a natural part of teaching and learning subjects like science, social studies, and the three Rs. Employers now assume that applicants know how to use computers. What they care about is whether applicants can use computers to carry out effectively the problem-solving and complex communication tasks that technological advances have made more important.
Q: How optimistic or pessimistic are you that individuals and societies will be able to meet the challenges the "new division of labor" represents? What political and social consequences can you imagine if (or when) people are left behind?
A: In the long run, the U.S. economy can be very flexible: 120 years ago, half of the population still worked on farms. So over the long run, we are optimistic that better education and training will prepare most of the workforce to do meaningful work in the computerized world. But education takes time and change is happening fast. Forty years ago, John Kennedy could say, "The rising tide lifts all the boats" and be correct. He governed in a lucky economic time when technology and trade did not favor one group of workers over others.
While politicians still invoke Kennedy's language, it no longer applies. Technology and trade are still engines of economic growth but the engines now favor educated and skilled workers. Less educated workers are paying a big price so the nation as a whole can advance. The short run challenge is to work out a better safety net—long-run training, assistance in areas like health insurance—to get through this period.
Q: What are you working on next?
A: After fifteen years of collaboration—sometimes crazy, mostly fruitful—we have decided to work on different topics. Murnane is writing a book showing how new statistical methods developed in the social sciences can improve educational research. Levy is studying how technological advances and outsourcing are changing the work of radiologists in the United States. We live a mile apart and plan to continue our Sunday walks together that have led to most of our good ideas.