In The Overworked American: The Unexpected Decline of Leisure (1992, Basic Books), Juliet Schor (a Harvard-based economist) examined government survey data and found that between 1969-87 annual hours worked increased by 163 hours (i.e., one month) for the typical American worker. This 163-hour rise in hours worked reversed a century-long downward trend in time spent in paid labor.
The causes of this decline in leisure are found in the decline of unions (which fought for reduced work days earlier in this century), and a consumer-oriented society in which workers take on additional overtime to pay for a material lifestyle fed by abundant credit. In addition, employers seeking to increase revenues can reduce their labor costs by asking their employees to work overtime rather than hire additional workers. Automation and downsizing increases the workload on remaining workers after each round of cost-cutting. Schor recommends European-style job sharing and shorter work weeks to reduce workloads, and increased use of flextime.
Appearing at a time when Japanese writers portrayed Americans as lazy, inefficient, and reluctant to compete, The Overworked American was an immediate hit in the media, and struck a responsive chord among researchers, activists, and human resource professionals concerned with maintaining a balance between work and family obligations. But, critics wondered how Americans could be overworked when the labor force is increasingly composed of temps, contingent workers, and involuntary part-timers. Others challenged Schor's data and her diagnosis of overwork.
To begin, scholars note that Schor's relied on data that asked workers to estimate how many hours they worked in the past week, a question subject to large recall errors. An alternative approach is to ask survey respondents to keep detailed diaries of time use. In an August 1994 paper in the Monthly Labor Review ("The overestimated workweek?"), John Robinson and Ann Bostrom found that respondents recalled longer work weeks that what was actually shown in their time diaries. Moreover, government data shows that the average work week declined from 38.2 hours in 1972 to 34.5 hours by 1996 (Council of Economic Advisors, Economic Report of the President, 1996, Table B-43). Only an enormous increase in moonlighting could reconcile Schor's data on increased annual hours worked with government statistics showing shorter work weeks (which seems implausible given that fewer than 10 percent of workers hold more than one job).
The reader should note that Schor defined work as the sum of time spent in paid labor plus the amount of time performing essential tasks in the home (such as cooking, cleaning, and child care). In this light, men's time in paid labor actually declined between 1969-87 while their household work increased. Meanwhile women's market time increased over this period while their home work decreased. These gender-based changes in paid employment and household work resulted in the net 163-hour increase in the work year noted above. Understanding gender differences in market and household work is key to reconciling the government's data showing shorter work weeks with Schor's overwork thesis.
Using longitudinal data from the Panel Study of Income Dynamics, Barry Bluestone and Stephen Rose (in "Overworked and Underemployed: Unraveling and Economic Enigma," The American Prospect, March, 1997) examined the length of annual hours in market work for individuals ages 25-54 (the authors did not examine time spent in household work). Consistent with Schor's data, they found that market work declined from 1967-82, but increased by 66 hours since 1982 (which is less than the 163-hour increase reported by Schor). For individuals, then, there is only a slight hint of overwork since 1982.
When men and women are considered together in families, however, the trend in annual hours worked slopes upward from 2,850 total hours of employment in 1967 to nearly 3,500 hours in 1988. Moreover, Bluestone and Rose found a distinct pattern of loading up on overtime hours in the years before a experiencing a spell of unemployment. Men were more likely to engage in this pattern of "feast before famine," which resulted in their wives working more hours. In other words, job instability among fathers resulted in mothers working harder than before.
Bluestone and Rose's analysis shows that Americans are both overworked and underemployed. Other family members are working harder to compensate for job insecurity of breadwinners. Moreover, when the education of family members is taken into account, those with college degrees are the most overworked, but their real incomes grew by one-third since 1967. For those who have failed to complete college, however, working harder is a response to stagnant real family incomes. Therefore, policies which seek to reduce the length of work weeks will only be successful if they are coupled with policies which increase job security and improve living standards.
