The New Division of Labor: How Computers Are Creating the Next Job Market

by Frank S. Levy, Richard J. Murnane

As the current recession ends, many workers will not be returning to the jobs they once held--those jobs are gone. In The New Division of Labor , Frank Levy and Richard Murnane show how computers are changing the employment landscape and how the right kinds of education can ease the transition to the new job market.


The book tells stories of people at work--a high-end financial advisor, a customer service representative, a pair of successful chefs, a cardiologist, an automotive mechanic, the author Victor Hugo, floor traders in a London financial exchange. The authors merge these stories with insights from cognitive science, computer science, and economics to show how computers are enhancing productivity in many jobs even as they eliminate other jobs--both directly and by sending work offshore. At greatest risk are jobs that can be expressed in programmable rules--blue collar, clerical, and similar work that requires moderate skills and used to pay middle-class wages. The loss of these jobs leaves a growing division between those who can and cannot earn a good living in the computerized economy. Left unchecked, the division threatens the nation's democratic institutions.


The nation's challenge is to recognize this division and to prepare the population for the high-wage/high-skilled jobs that are rapidly growing in number--jobs involving extensive problem solving and interpersonal communication. Using detailed examples--a second grade classroom, an IBM managerial training program, Cisco Networking Academies--the authors describe how these skills can be taught and how our adjustment to the computerized workplace can begin in earnest.

Genres: Economics, Education, Nonfiction, Business, Technology

192 pages, Paperback

First published April 12, 2004

Reviews

[Tony WANG · Rating 3 out of 5]

This book was first published in the 2004, thus, reading it in 2020 and beyond is rather futile and renders minuscule insight. Many of the predictions and recommendations were common sense i.e. the authors iterated numerous times that “expert thinking and complex communication” are two key skills employers value. Needless to say those traits are highly desirable since the turn of the century due to the advent of machine learning and artificial intelligence.

In hindsight, i.e. right now, it is common sense in this day and age that those kind of “skills” are literally a must for ALL graduate roles in any field. I do, whoever, agree that a firm grasp in reading, writing and mathematics are important and is especially essential in the future. And that being a critical lifelong learner and continuously upgrading oneself is the key in any job markets of the future.

[Richard · Rating 3 out of 5]

This raised some interesting ideas; however, it is a little dated. While the authors most likely intended to write a book that would stand the test of time, a few of the predictions have turned out to be wrong. The primary example would be self-driving cars. The authors hold driving up as a complex task that would be difficult to teach a computer. It is certainly a complex task; however, self-driving cars are now a reality.

There is still some value in this book; however, I would recommend only reading the first and last chapters.

[Mehmet · Rating 4 out of 5]

I found this book fascinating. At the same time, it raised some interesting questions as well. The book is about the effects of automation/computerization on the job market: which occupations will be declining, which jobs will grow, and how those jobs will be changed by computers? What skills will be required to succeed in the growing jobs? And how can the education system foster those skills in our children and youth? That is quite some ground to cover, but the authors do a fantastic job. The authors cover the period 1969-2003 to answer those questions.
The authors' (both of whom are economists) main idea can be summarized as follows: computerization is changing the way we work. It destroys some jobs, while it creates others. On the net, it creates more jobs than it destroys. There is a hollowing out: jobs that are provide lower-middle and middle-class earnings are declining. Jobs that are either low-paying, low-skill (service jobs that cannot be routinized and automated such as food service, janitors) or high-end, high-paying (management, professional services, administrators, nurses etc.) will continue to grow. These patterns are already visible in the data covering 1969-2003. The proportion of clerical (e.g. mortgage approval) and blue-collar manufacturing jobs (such as windshield installment) have declined from more than 50% to less than 40%. While some people may think that manufacturing jobs eroded because of offshoring, the two explanations are not mutually exclusive. In fact, offshoring of jobs such as call centers and manufacturing) and computerization are related. For jobs to be offshored, their content have to be scripted and converted into a set of rules (e.g. if the customer says this, then answer that way). The very conversion paves the way for programming and subsequent automation of the work (witness automated customer help menus).
The high-end jobs continue to grow because they cannot be easily programmed. Part of the reason is that they defy simple rules. They involve interpersonal communications, reading people's body language, or solving complex, novel problems. Another reason is that people know more than they can tell. This is called tacit knowledge. If knowledge is tacit, it is not possible to write programs that tell the computer what to do in every instance.
These jobs will require expert thinking and complex communication skills. The authors use several examples of each. For example, a car mechanic may have to be an expert thinker as the computer codes do not cover all eventual car problems. When faced with a novel problem, the expert thinker relies on tactile or auditory sensory information (e.g. the way a car sounds or feels) to solve it. These are not easy to program at all. Similarly, a financial advisor and a manager both have to communicate and/or persuade the other party. To do so, they need to evaluate and interpret the other party's voice, inflection, body language, etc. These are extremely complex behaviors and signals that currently computers cannot reliably imitate.
All this means that we need to teach expert thinking and complex communication skills to the future workforce. In the last part of the book, the authors discuss how this can best be achieved. They go through corporate training events and standards-based education as potential ways of acquiring these skills.
All in all, this was an eye-opening book for me; one that shows how simplistic predictions, negative (computers will eat all our jobs) and positive (computers will be our best friends and will result almost purely in job creation) are both wrong. It shows how the job market has actually changed, and how it will likely change going forward. It also puts forward a theoretically sound and well-supported explanation of why this is the case, and what kinds of skills our youth and children will need in order to succeed in the workplace of the future.
My biggest concern with the book is that it may be a bit dated. Though it was fascinating to read Herbert Simon's predictions from 1960 about how computers will change the job structure in corporations came to pass, at times I wondered, is this still true? For instance, the authors made heavy use of an example of a truck driver making a left turn as an instance of non-routine physical movement, and how difficult it is to program that since there are so many sensory inputs. While my reading of the news is that computer scientists are still struggling with autonomous vehicles, I also had the feeling that they are advancing fast. Which made me wonder whether some of the predictions still hold. I have high confidence that the general contours of their predictions will hold true, but the specifics may be quite different. It would be nice to see how far we have come, and whether we are still on the same track that the authors predicted. While a paper by David Autor (a colleaegue of the authors with whom they published a QJE article in 2003 that forms the basis of chapter 3 of the book) updates some of the results, one still wonders how the authors feel about the potential of machine learning as opposed to simple routine programming, and whether they would still hold on to their predictions.
Nevertheless, a clearly exposited, thoughtful book about the effect of automation on the future job landscape. Highly recommended for sensible people who are skeptical of all the hyperbolic stuff out there.

[Ashley Marc · Rating 3 out of 5]

Just as others noted, it is a bit outdated now...
For instance a lot of the skills and jobs that they had presented as immune to automation are getting there now. I reference this book in my own book at several areas precisely to highlight that and take some lessons away from it.

I would recommend this for people who really interested in the topic as a whole to have on their shelf. But if you are to read just one book on it, go for something more recent.

[Jeff · Rating 4 out of 5]

The Division of Labor is a great articulation of the need for an expansion of the global goals of public education. Levy and Murnane argue that computers have altered our economic world, but that it does not lead to mass unemployment. Rather it leads to a hollowing out of the middle of the job market, because computers are better at running algorithms or rules-based tasks. So the jobs that we loose are the blue-collar and middle class jobs that do rules-based tasks. However, what computers can't do is solve complex problems and communicate ideas is subtle complex ways. So what people will continue to offer to the labor market is expert thinking and complex communication. However, for everyone to have access to the new jobs, then we have to teach our children the expert thinking and complex communication skills. If we are serious about improving education, we need to take these ideas seriously and think about how we prepare all children to think through complex problems and communicate with teams about these ideas.

[Jysoo · Rating 4 out of 5]

This is another book on the issue of “men vs. machine”. Compared with the “Race Against the Machine” which discusses more theoretical framework of the issue, this book is more practical with narrower focus. The authors start with the question “what kind of tasks do computer perform better than humans?” --- rule based logic. On the other hand, humans are good at expert thinking and complex communication. The authors discuss their implications on the job market, with very interesting statistics. They also present some examples of how to train people with skills necessary for their future. I truly enjoy the book.

[Tim · Rating 4 out of 5]

If you want to understand the large forces shaping our dynamic labor market's evoilution, this is one of the best books you will find. Insightful but slightly on the scholarly side. It's been a while since I read this, and I recall that their policy prescriptions seemed oriented to typical government-knows-best solutions.

[Shirley · Rating 4 out of 5]

Nothing new, but nice to read through the conclusions and especially see data w.r.t. impact of technology jobs (which skills more in demand, which less in demand) and what this means for our K-12 education system. Heavier on the business cases than anticipated.

[Johnny Go · Rating 5 out of 5]

I heard Richard Murnane speak in Harvard. Brilliant guy. This book helps put into perspective the impact of technology on jobs and education.