Talking about work

Monthly Review, July-August, 1997 by Doug Henwood

It's quite reminiscent of the buzz around globalization. Ancient features of capitalism, job-displacing technical innovation, and the lust to cross national borders, are posited as recent transformations. This leads to a dramatic misspecification of the enemy: instead of capital and its lust for profit, innovation and cosmopolitanism themselves are demonized. Demonizing these things, instead of talking about their radical transformation, is one of the worst consequences of the rejection of Marx.

Bright Future

The optimistic line is known to anyone who picks up a copy of Wired, or even just watches TV. Technology is changing everything. All the old fixities - big industry, big government, big things - have become endlessly fluid. Hierarchy is gone in the new team-oriented world; innovation is ceaseless, bubbling up from the bottom. Self-reliance and interconnectivity together make the future, and a libertarian spirit pervades it all. Education and skill are the keys to the new world, not ownership or connection.

This mode comes in several flavors. The right wing one, exemplified by George Gilder and the "digerati" of Wired, is deeply laissez-faire. But there's a liberal one as well, embodied by Robert Reich and even Bill Clinton. A few years ago, when he was Secretary of Labor, Reich said that "the most rapidly growing job categories are knowledge-intensive; I've called them 'symbolic analysts.' Why are they growing so quickly? Why are they paying so well? Because technology is generating all sorts of new possibilities.... The problem is that many people don't have the right skills." (The "problem" Reich is talking about is presumably poverty and social decay.) Clinton believes and says very similar things.

Reich, it seems, didn't pay much attention to the projections of an agency he once supervised, the Bureau of Labor Statistics (BLS), nor does Clinton apparently. On the next page the BLS's projections of fastest-growing occupations between 1994 and 2005; they bear no relation to Reich's fantasies, or those filling the pages of Wired. These thirty job categories alone account for one-third of total employment today, and for over half the next decade's projected growth. Of the top thirty, those that look like symbolic analysts account for 7 percent of employment now, and 13 percent of projected growth. Most look quite mundane. Engineers, computer professionals, and associated technicians together now account for about 3 percent of total employment and under 7 percent of projected growth. It's hard to see from this how "the problem is that many people don't have the right skills."

Just what do people mean by skills anyway? If the U.S. ruling class were seriously worried about illiteracy and innumeracy, they'd spend more money on education. While it may be that we'd all be better off if workers were better trained, there's no sign that actual employers are demanding such a workforce now.

One problem with analyzing skill is that it's a difficult concept to define and measure. Economists typically use a mix of education and experience to describe the skill of workers, but these are only inexact proxies. At the big picture level, the rising "education premium" - the earnings advantage enjoyed by those with advanced degrees - is often cited as proof of the rising demand for skilled workers. But comparisons of U.S. regions show that the higher the unemployment rate, the greater the education premium. Since average unemployment rates have drifted upwards since the golden age - they averaged 4.6 percent in the 1950s and 1960s, and 6.8 percent since 1980 - it may be that the rising education premium is just a sign of a slack labor market.

More direct attempts to measure employers' wants offer no support for the skills thesis. To measure the changes in skills requirements in the United States, David Howell and Edward Wolff linked changes in employment in 64 industrial and 264 occupational categories from 1960-1985 to their descriptions in the BLS's Dictionary of Occupational Titles. They found that low-skill service industry work grew more quickly than high-skilled - and, perversely, the more rapidly growing service industries "required higher skills but paid lower wages than the low-growth service industries." The picture was brighter in manufacturing, but it's a shrinking field. For nonsupervisory workers as a whole, the most rapid growth "was in the highest skill segment but in the lowest wage segment."

If the "problem" that Reich was talking about was social polarization, then there's not much evidence that skills or their lack have much to do with it. Most of the increase in inequality over the last twenty-five years has come within demographic groups - people grouped by occupation, age, sex, and schooling - and not between them. A better place to look for explanations of polarization would be the erosion of the working class's bargaining power because of union busting, deregulation, and capital mobility.

Unpacking "Skills"