In my first post about Arne Kalleberg’s Good Jobs, Bad Jobs, I described the author’s general framework for understanding the increasing polarization of work. He considers many factors: the social and economic forces that have transformed the economy (globalization, new technologies), the composition of the labor force (by education, race, gender, etc.), the mediating role of other institutions (government, financial institutions, unions), and the organization of work itself (“high-road” and “low-road” employment strategies). Now I want to take a closer look at work polarization, starting with changes in the distribution of occupations.
Kalleberg ranks broad occupational groups according to measures of job quality. He confirms that occupations in the middle range have lost workers, while occupations at the high and low ends have added workers. For example, a lot of the losses have involved semiskilled machine operators and administrative support workers, where new technologies have reduced the need for labor. My Dad’s first job after college was as a statistical clerk, armed with only a slide rule and an adding machine. When I taught statistics a generation later, my students and I had computers to do the busywork. When I later became an independent financial planner, my desktop computer provided all the administrative support I needed.
Remember the old IBM slogan, “Machines should work; people should think”? Many optimistic social scientists of the 1960s and 70s expected technological change to improve the quality of work, eliminating the drudgery and expanding the creativity. Workers could take the benefits of their high productivity as higher wages and/or more time off the job. Part of this vision has come true. High-end occupations–managerial, professional and technical–have expanded. But they haven’t expanded enough to employ all the workers displaced by technological obsolescence or outsourcing. In the long run, new technologies may create as many jobs as they destroy; in the short run, destroying a good job may be easier than creating one. Capitalists are job destroyers as much as job creators, as the current political debate over Bain Capital illustrates. Getting the same work done using fewer American workers can be easy with new technologies and a global supply of labor. Doing something new and paying someone a good wage to do it is more challenging. It may involve taking more risk, making an investment in worker training, and finding a market for a new good or service. Much of the work worth doing has social–not just individual–benefits, so the demand may not be there without some financial commitment from the taxpayers.
The phenomenal growth of low-wage service jobs results partly from the growth in the number of workers who are available for such jobs. This in turn results from a combination of labor-force characteristics (so many workers who lack the skills required for higher-level work) and work organization (too many companies adopting “low-road,” cost-cutting approaches to labor instead of “high-road” investments in human capital). These factors are reinforced by the demand for cheap services in a society with so many low-income households. Of the ten occupations with the largest projected job growth from 2006 to 2016, seven are low-wage sales or service jobs (such as retail salespersons, food preparation and service workers, home health care aides, and janitors).
Kalleberg’s findings on wages are consistent with those of other researchers: More workers at the high and low ends and fewer in the middle. Between 1973 and 2009, real (inflation-adjusted) wages for workers in the 95th percentile have increased from $39 to $55 per hour for men, and from $24 to $41 per hour for women. At the median (50th percentile), wages have increased slightly for women ($11 to $14) but have stagnated for men (around $18). In the 20th percentile, wages have also increased a little for women ($8 to $9), but have fallen for men ($12 to $10). And at the bottom, the minimum wage of $7.25 has not been adjusted for inflation, and so it has declined in real value since its peak in the 1960s.
Wage disparities have increased the correlation between educational level and income. The economic advantage of a college education has increased, although the cost and the debt one incurs has too. The United States ranks relatively high on average education and other measures of skill, but it also ranks high on the proportion of workers in very low-paying jobs. Other countries have had more success in maintaining decent wages. Kalleberg says,
Institutions matter for wage-setting: inequality tended to be greater in liberal market economies such as the United States, Britain, and Canada, which have relatively weak unions and decentralized patterns of wage-setting, whereas inequality was relatively low in countries such as France and Germany, which have relatively centralized wage-setting mechanisms and stronger unions.
This more sociological view contrasts with the economic theory of skill-based technical change (SBTC), which attributes wage disparities primarily to individual worker qualifications, without much regard for institutional variables.
Kalleberg also finds increasing inequality in the availability of benefits, especially health insurance and defined-benefit pension plans. The United States is unusual in its reliance on employers to provide such benefits at their discretion, rather than making them rights of citizenship. As stable employment relations and the social contract between management and labor have broken down, this system provides fewer reliable benefits, especially for workers at the low end. The proportion of workers covered by defined-benefit pension plans has been cut in half since 1980. Defined contribution plans like 401(k)s shift the risk of poor investment performance to the individual worker.
One particularly discouraging aspect of job polarization is the effect it has had on racial inequality. Whatever progress we have made in combating racial discrimination has been counteracted by the economic forces and decisions that have tended to polarize the labor force. Occupational segregation by race declined a bit in the 1970s, but stopped declining after that. Many of the jobs that had helped other ethnic groups move into the middle class–especially manufacturing jobs with modest skill requirements but good wages–are no longer available, and what good jobs there are have higher educational requirements. To make matters worse, minorities (and women) who do have college educations still don’t obtain their proportional share of the good jobs, suggesting that discrimination remains a problem at the high end of the occupational spectrum.