In a word: jobs. For most of us, the job determines our standard of living, receives much of our energy, and defines who we are. (Maybe that's not how it should be, but it's often the way it is…)

For the past half century, computers have brought change. The content of jobs has changed. So have pay rates and locations.

A compelling, fact-based explanation of these trends is available in a book by Frank Levy and Richard J. Murnane: The New Division of Labor: How Computers Are Creating the Next Job Market. I dipped in with interest when I saw a prominent acknowledgement to David Autor, a former teaching assistant of mine who is now a professor of economics at MIT.

The book examined employment data and case studies to address several key questions:

1. What do computers do better than people, and vice versa? In essence, computers are best at the lower half of a spectrum of increasingly complex IF=>THEN information processing. That is, computers are much better at following explicit rules (arithmetic) and, in many cases, at decoding and following implicit rules (mortgage underwriting). On the other hand, people are still best at recognizing patterns generated by implicit rules (detecting cancer from mammograms), and, especially, at finding patterns where the rules can't be inferred (or can't yet be inferred, as in innovative problem-solving or complex visual processing such as a truck driver turning left against traffic).

So, computers are best at programmable routines (no surprise there). People, on the other hand, are best at explaining complex things to other people (sales and professional services jobs) and at the creative problem-solving required for innovation. As computers work their way throughout the global economy, complex communications and innovation are becoming more important, not less.

2. What is this doing to the division of labor? Basically, computing has substituted for routine physical and cognitive work. Between 1969 and 1999, blue collar and administrative support workers dropped from 56 percent to 39 percent of the workforce. Absorbing this decrease, all other work categories gained (service workers, sales, technicians, professionals, and managers). With routine work handled by computers, or by workers at a distance over a network, the wage advantage for more creative/educated people has risen dramatically. From 1973 to 2000, while the real wages of high school graduates dropped, those of college graduates rose. College grads started with a 40% premium over high school salaries and ended up earning 90% more than those who stopped at high school. The bottom has been falling out for those without much education, while the ceiling has been dramatically raised for those at the upper end.

3. How is this different from the past? The original industrial revolution displaced relatively higher income craft workers in favor of those moving off the land to factories and to eventual life in the middle class. Through much of the 20th century, technology-based productivity impacts fell rather evenly across rich and poor so Kennedy could credibly claim that "a rising tide lifts all boats."

With computer-based productivity, however, the burden is falling much more heavily on low income workers. The gap between rich and poor grows. Social mobility and cohesion suffers, at least within the U.S. if not globally.

What should we be doing? Two goals stand out:

1- Educate people for needed, high-paying jobs that computers can't readily displace. These are jobs involving complex interpersonal communications and innovation: an increasingly important part of sales, management, teaching, design, and hundreds of other occupations. Education for these occupations must focus on how to learn and re-learn, not just on facts. The book makes a strong argument in favor of serious testing to measure progress and control the process on an individualized basis. Technology creates the need for education and can also be a big part of meeting the need. Reforming education, on a lifetime basis, with technology -- is thus perhaps our highest priority 'cross-boundary' IT investment, at least according to Levy and Murnane.

2- Provide a better safety net and support for social cohesion. While IT-based productivity should, in the long run, lift all boats, the politics of transition are dangerous. Success is not guaranteed. We need wise leaders.

* * * * *

When it's "back to basics," is preparing people for good jobs the top priority? Should reforms in education and the social safety net be at the top of the list for those taking the cross-boundary IT agenda seriously?

What's your take on these issues?

You can find the book here: http://pup.princeton.edu/titles/7704.html