Introduction
Increases in income inequality in the U.S. over the past quarter century have been well documented (Murphy and Welch, 1992; Karoly, 1992; Freeman, 1997; Levy and Murnane, 1992; Katz and Autor, 1999). There have been three main facts to which everyone agrees. Income and wage inequality increased in the 1980s, stabilized in the late 1980s and early 1990s, then it began to increase until the late 1990s when it once again stabilized (Freeman, 1997; Lee, 1999). Generally, the workers who fared the worst in these changes were those who did not finish high school. They saw their wages relative to college graduates slip by at least 30% (Freeman, 1997, Lee, 1999; Mishel, et. al., 2001). Finally, women generally saw their situation improve relative to men over the period (Karoly, 1992; Freeman, 1997). From the data, it appears as if low skilled men suffered the brunt of these changes (Lee, 1999).
There has been a lively theoretical and empirical debate over the causes of these changes (for some review articles, Topel, 1997; Fortin and Lemieux, 1997). Some have concluded that most of the change was caused by the increase in demand for skilled labor caused by technological change (Katz and Murphy, 1987; Bresnahan, et. al, 1999; Krueger, 1993). Others have focused attention on institutional factors, such as the decline in unions and the lack of increase in the minimum wage (Lee, 1999; Freeman, 1997; Card, 1992). Still others have tried to examine how the continuing shift from manufacturing to services and the increased exposure to world markets has helped skilled workers and hurt unskilled workers (Freeman, 1997; Bluestone and Harrison, 1986). Finally, some have focused on how immigration patterns have depressed the wages of 3 low skill workers (Borjas, 1998). This debate turns very much on how one measures these factors and their effects.
A related debate concerns how work and jobs have changed in the past 25 years. Many observers argue that during the 1980s, the employment relation in the U.S. began to change for all workers (for example, Osterman, 1999; Harrison, 1999; Gordon, 2000; Pfeffer and Baran, 1988; Blair and Kochan, 2000). Firms began to redefine who their core workers were. They began to downsize, outsource, and employ more contract workers. This made workers generally more insecure, and as we will show, dissatisfied with work. This paper will review the literature on this subject and try to link these changes to shifts in income inequality.
We will provide descriptive evidence consistent with the view that work changed as income became more unequally distributed. The literature shows very clearly that not only did workers on the bottom of the skill distribution fare poorly in terms of losing ground on wages, they also had less safe working conditions, found themselves working less regular shifts, had fewer benefits such as pensions and health care, and lower job security and job satisfaction. In essence, the increases in wage inequality were accompanied by a growing insecuritization of work for those at the bottom. The evidence is somewhat different for those at the top. While they experienced more insecurity at work as well, they also benefited from the changes in employment relations. Their benefits remained more stable. For those whose incomes went up the most, there was an increase in job satisfaction and an increased sense of efficacy at work. Those with the highest incomes also had increased hours of work which they appear to mostly enjoy. Work has become more intense for all. But, those at the top have had more opportunities to enjoy work while those at the bottom have seen their work lives grow more insecure.
Our review has the following structure. First, we consider more carefully the argument about what has changed in the employment relations of various groups of workers in the past 25 years. Then, we consider the evidence that tries and measures those changes. We will explicitly try and link these changes to changes in income inequality wherever possible. Finally, we will consider what research is implied by our review.