Policy Debate: Are Americans Overworked?
Issues and Background
In the last twenty years the amount of time Americans have spent at their jobs has risen steadily. Each year, the change is small,
amounting to about nine hours, or slightly more than one additional day of work. In any given year, such a small increment has
probably been imperceptible, but the accumulated increase over two decades is substantial....
The rise of worktime was unexpected. For nearly a hundred years, hours had been declining. When this decline abruptly ended
in the late 1940s, it marked the beginning of a new era in worktime. But the change was barely noticed. Equally surprising, but
also hardly recognized, has been the deviation from Western Europe. After progressing in tandem for nearly a century, the
United States veered off into a trajectory of declining leisure, while in Europe work has been disappearing. Forty years later,
the differences are large. U.S. manufacturing employees currently work 320 more hours -- the equivalent of over two months --
than their counterparts in West Germany or France.
~Juliet Schor, The Overworked American
Medieval peasants worked less than we do. On average, as former Labor
Secretary Robert Reich points out, we Americans work 350 hours...
nearly one year more every five years. Thus, during an average work life,
western Europeans enjoy nearly ten years more free time than do their
American counterparts. In Europe, workers receive an average six weeks of
paid vacation. We Americans average two. And 26% of us got no vacation at
all last year.
~John De Graaf
Contrary to the popular “time famine” perception, evidence from the historical record
of time diary surveys shows that there has been a slight increase in women’s leisure
time over the period from the 1960s to the 1990s, while for men there has been only a
slight decrease.
~Oriel Sullivan and Johnathan Gershuny
The publication of The Overworked American by Juliet Schor touched off a debate concerning
the length of U.S. workweeks. Schor found that in the U.S.:
- workers have experienced longer workweeks in recent years,
- rising labor force participation rates for married females has significantly increased the total
amount of working time provided by a typical household,
- workers receive less vacation time than in most other industrialized countries.
While individuals generally report that they feel increasing time pressure, an analysis of time-use diaries find that
the mix of leisure and work time has not changed substantially in recent decades. Robinson and Bostrom (1994) and Sundstrom (1999)
suggest that there may be an upward bias in reports of hours worked by those workers who report relatively long workweeks.
Jacobs (1998), however, provides evidence suggesting that the Robinson and Bostrom results may be the result of random
errors in the data and not the result of biased reporting of hours worked.
Even if their work hours have not increased substantially, U.S. workers tend to work more weeks per year than workers
in all other industrialized economies. While two weeks of paid vacation is common in many U.S. occupations, most Europeans
receive at least twice as much paid vacation time each year.
In some industries, overtime use has increased in recent years. To understand this phenomenon, it will be helpful to examine why firms may elect
to use increased overtime. When firms desire to change their level of labor use, they may do this by:
- changing the number of workers,
- changing the length of the average workweek, or
- changing both employment and the length of the average workweek.
If the only costs of labor are hourly wage costs and if worker productivity remained constant throughout the workday,
firms would be indifferent among all combinations of hours worked and employment levels that provide the same total
quantity of labor hours. In reality, however, this choice is not that simple:
- The marginal product of labor that results from an additional hour worked is likely to decline as the length of the
workweek increases beyond some point (workers get tired as they work more hours).
- The Fair Labor Standards Act requires that overtime pay of "time and a half" be paid when the workweek exceeds 40 hours.
- Some costs (known as "quasi-fixed costs") vary with the number of workers hired, but not with hours worked. These costs
include hiring and training costs as well as the costs of some expensive fringe benefits such as health insurance.
A firm minimizes its costs if it selects a mix of inputs at which the marginal physical product resulting from the last
dollar spent on any input equals the marginal physical product resulting from the last dollar spent on any other input. In
selecting the mix of hours worked and employment levels, firms must compare the additional output resulting from a given increase
in spending on overtime hours against the additional output resulting from the same additional spending on new hires.
During an economic expansion, cost-minimizing firms will use a longer workweek if the cost of expanding output in this manner
is lower than the cost of expanding output by hiring new workers.
Firms are expected to attempt to increase the length of the average workweek if the quasi-fixed costs associated with adding
new employees rises. This may be the result of increases in hiring costs, training costs, or employee health insurance costs.
Several studies have confirmed that the use of overtime during an expansion rises by a larger amount for those occupations
in which quasi-fixed costs are higher.
While there is some question concerning whether full-time workers work more or fewer hours, there is no doubt that combined
hours worked by both spouses in married households has increased. Rising labor force participation rates for married females
has resulted in a substantial increase in the quantity of labor supplied by a typical married household. It is possible that
the popular perception of a time crunch may be the result of the more difficult time pressures facing households in which both
spouses work.
It is clear that American workers are working substantially more in the paid labor market than workers in other industrialized
nations. Much of the debate focuses on whether this is the result of voluntary worker choice or whether this is a decision
imposed on workers by their employers.
Primary Resources and Data
- Monthly Labor Review
http://stats.bls.gov/opub/mlr/mlrhome.htm
The Monthly Labor Review provides a good source of relatively nontechnical studies concerning
hours worked and other characteristics of the labor force.
- Bureau of Labor Statistics
http://www.bls.gov/
The Bureau of Labor Statistics provides an extensive collection of U.S. labor force statistics.
- International Labor Organization, "Working longer,
working better?"
http://www.ilo.org/public/english/bureau/inf/magazine/31/work.htm
This online document summarizes an International Labor Organization study that finds that average hours worked by U.S. workers
is the highest among industrialized countries. U.S. productivity is also the highest among industrialized economies.
- International Labor Organization, "Key Indicators of the Labor Market"
http://www.ilo.org/public/english/employment/strat/kilm/index.htm
The Key Indicators of the Labor Market study compares hours, productivity, wages, and other labor market indicators
across countries. While the full report and data is not available online, this website contains tables of statistics on several
labor statistics.
- Phillip L. Rones, Randy E. Ilg, and Jennifer M. Gardner, "Trends in hours of work since the mid 1970s"
http://stats.bls.gov/opub/mlr/1997/04/art1full.pdf
Phillip L. Rones, Randy E. Ilg, and Jennifer M. Gardner, in this April 1997 Monthly Labor Review article, examine trends in hours worked from
the mid-1970s through the early-1990s. They argue that that the popular perception of substantial increases in the length of
the workweek is inaccurate. After adjusting for changes in the age distribution, Rones, Ilg, and Gardner find that, between 1976
and 1993, the average work week increased by 0.2 hours for males and 1.0 hours for females. They note, though, that the
proportion of the population working 49+ hours per week has increased as has the proportion of women engaged in year-round work.
Thus, while the average work week has not changed substantially for women, the number of hours worked per year has increased due
to an increased in average annual weeks worked. (The Adobe Acrobat viewer plugin is required to
view this document. You may download this viewer by clicking
here.)
- John P. Robinson and Ann Bostrom, "The overestimated workweek? What time diary measures suggest"
http://stats.bls.gov/opub/mlr/1994/08/art2full.pdf
In this August 1994 Monthly Labor Review article, John P. Robinson and Ann Bostrom discuss differences in the estimated
length of the workweek between survey data and time-diary studies. They note that estimates of the length of the workweek
based on employer surveys are unreliable since they do not take into account self employment and second jobs. Robinson and Bostrom
argue that survey questions that ask respondents to recall hours worked in the previous week are subject to potential biases
and are subject to a variety of interpretations concerning what constitutes work time. Time-diary data indicates that
workers tend to overestimate the number of hours worked when asked for the length of their workweek. The magnitude of the
reporting error is larger for workers who report working more than 40 hours per week. Workers who report working 60 or more hours
per week on surveys are found to work approximately 53 hours when time diaries are used. (The Adobe Acrobat viewer plugin is required to
view this document. You may download this viewer by clicking
here.)
- Linda L. Stinson, "Measuring how people spend their time: a time-use survey design"
http://stats.bls.gov/opub/mlr/1999/08/art3full.pdf
In this August 1999 Monthly Labor Review article, Linda L. Stinson discusses the problems associated with constructing
a time-use survey. She notes that one of the problems with time-use surveys is the treatment of simultaneous activities, such
as time spent monitoring children or cleaning the house while also watching television. Stinson discusses the evolution of
and current practices in time-use studies. (The Adobe Acrobat viewer plugin is required to
view this document. You may download this viewer by clicking
here.)
- Jerry A. Jacobs, "Measuring time at work: are self-reports accurate?"
http://stats.bls.gov/opub/mlr/1998/12/art3full.pdf
Jerry A. Jacobs, in this December 1998 Monthly Labor Review article, raises questions concerning the bias in self-reported
hours that was found in the study by Robinson and Bostrom. He presents evidence suggesting that the apparent overestimate of
self-reported hours worked for individuals who report long work weeks is the result of random measurement error. Jacobs
finds that the correlation between reported work hours are and measured work hours is very high. He
argues that there is no convincing evidence of biased reporting of hours worked. (The Adobe Acrobat viewer plugin is required to
view this document. You may download this viewer by clicking
here.)
- Daniel Hecker, "How hours of work affect occupational earnings"
http://stats.bls.gov/opub/mlr/1998/10/art2full.pdf
Daniel Hecker analyzes patterns of work hours and wages by occupation in this October 1998 Monthly Labor Review article.
He observes that the use of extended workweeks vary substantially by occupation. In most occupations, hourly wages increase with
the length of the workweek, but in some occupations, hourly wages are lower for those who work longer hours. (The Adobe
Acrobat viewer plugin is required to view this document. You may download this viewer by clicking
here.)
- William A. Sundstrom, "The Overworked American or the Overestimated Workweek? Trend and Bias in Recent Estimates of Weekly Work Hours in the United States"
http://business.scu.edu/economics/faculty/wsundstrom/pdf/overworked.pdf
In this November 1999 working paper, William A. Sundstrom analyzes the upward bias in Current Population Survey estimates
of the length of the average workweek. Using time-diary surveys and and results from the Bureau of Labor Statistics
Current Employment Statistics survey, he finds that the bias in the estimate of the workweek has increased between 1960 and 1995.
Sundstrom's results suggest that workers with higher levels of education tend to overestimate the length of their workweek by a larger amount.
He argues that rising educational attainment is a primary cause of the growth in the bias in workweek estimates.
(The Adobe Acrobat viewer plugin is required to view this document. You may download this viewer by clicking
here.)
- Cornell University, "The Cornell Couples and Careers Study"
http://www.human.cornell.edu/che/BLCC/.../couplesandcareers.pdf
This document provides information about the Couples and Careers Study conducted by Cornell
University. This study collected information on a sample of dual-earner households. This study
finds that workers feel constrained by career considerations to work more than their desired number
of hours. Evidence is also presented that indicates that men still spend more time in paid market
labor than their working wives. (The Adobe Acrobat viewer plugin is required to view this document.
You may download this viewer by clicking
here.)
- U.S. Department of Labor, Bureau of Labor Statistics, "American Time Use Survey"
http://www.bls.gov/tus/
This web site provides information about the American Time Use Survey developed by the Bureau of Labor Statistics. This
survey is designed to measure the allocation of time spent in a wide range of activities. Links to related resources
are provided on this site.
- International Association for Time Use Research
http://www.iatur.org/
The web site for the International Association of Time Use Research contains information about and links to a wide variety
of web sites that provide information about international time use research.
- Centre for Time Use Research, "Multinational Time Use Study (MTUS)"
http://www.timeuse.org/mtus/
The Multinational Time Use Study is an attempt to provide an international comparison of time use. This project, developed
in coordination with several national and multinational statistical agencies, has attempted to harmonize data collection
methods for time use studies and to manage and disseminate the data from these studies.
Different Perspectives in the Debate
- Juliet Schor, "Overworked American: The Unexpected Decline of Leisure"
http://www.amazon.com/exec/obidos/ASIN/046505434X/002-0964804-8950469
This Amazon.com web page contains a link to an excerpt from the book that helped focus the debate over whether
Americans are overworked.
- Kristin Roberts and Peter Rupert, "The Myth of the Overworked American"
http://www.clevelandfed.org/Research/Com95/0115.pdf
In this January 15, 1995 Federal Reserve Bank of Cleaveland Economic Commentary, Kristin Roberts
and Peter Rupert examine trends in hours worked using data from the Panel Study of Income Dynamics
for the years 1976 to 1988. They note that total hours of work did not change much during this
time period. Roberts and Rupert note that the major change in time use was a shift from home work
to market work for married women.
- Oriel Sullivan and Jonathan Gershuny, "Cross-National Changes in Time Use: Some Sociological (Hi)stories Re-Examined"
http://www.iser.essex.ac.uk/pubs/workpaps/pdf/2001-01.pdf
In this May 30, 2000 working paper, Oriel Sullivan and Jonathan Gershuny examine trends in time use using cross-national data
for the period from the 1960s to the 1990s. They find that the division between work and leisure time has been relatively
stable over this period. In particular, they find that leisure time for women has risen slightly during this period
while male leisure time has decreased slightly. These results cast serious doubt on the arguments for "overworked" Americans.
A number of possible explanations for the popular perception of increasing time pressure are examined.
(Another interesting result of this study is a finding that, even though fertility has declined, parents spent more time in
child care activities in the 1990s than in earlier decades.)
- The Economist, "The Land of Leisure"
http://www.economist.com/displaystory.cfm?story_id=5476124
This February 2, 2006 article summarizes a recent study by Mark Aguiar and Erik Hurst (listed below) that examines
trends in hours worked in the U.S. It is observed that during the period from 1965 to 2003, hours worked
declined for U.S. citizens at all educational levels. Hours devoted to leisure activities have risen, by
4-8 hours a week for a typical individual during this period. While hours worked in the labor market have
increased, particularly for woman, time spent in household work have declined by more than enough to offset
this increase in market labor supply.
- Mark Aguiar and Erik Hurst, "Measuring Trends in Leisure: The Allocation of Time Over Five Decades"
http://www.bos.frb.org/economic/wp/wp2006/wp0602.pdf
Mark Aguiar and Erik Hurst examine time-use surveys for the years 1965 to 2003. They observe that
males have increased time spent in leisure activities by 6-8 hours per week during this interval
(primarily as a result of a reduction in work hours) while females have increased weekly time spent in
leisure activities by 4-8 hours (primarily due to a reduction in time spent in household work). The
range of estimates for the increase in leisure time is due to the use of alternative definitions
of leisure time activities. (The Adobe Acrobat viewer plugin is required to view this document.
You may download this viewer by clicking here.)
- PBS NewsHour, "Overworked America?"
http://www.pbs.org/newshour/bb/business/july-dec99/overwork_9-6.html
This transcript of a September 6, 1998 NewsHour contains an analysis of whether Americans are overworked. Researchers who argue
for and against this proposition present their cases in this discussion. Walter Olson, Kathleen Gerson, Geoffrey Godbey, and
Paula Rayman participate in this discussion.
- Ron L. Hetrick, "Analyzing the recent upward surge in overtime hours"
http://stats.bls.gov/opub/mlr/2000/02/art3full.pdf
Ron L. Hetrick, in this February 2000 Monthly Labor Review article, finds that the economic expansion of the 1990s
resulted was accompanied by a larger than usual expansion in overtime hours in manufacturing industries. He observes that
the average weekly overtime for manufacturing workers increase by 1.6 hours per week between March 1991 and early 1997. While
an increase in both overtime and employment is common as the economy recovers from a recession, the portion of the increase in
labor use that was attributable to a longer workweek was larger than usual. The increase in the use of overtime is
largest in those industries that rely on a highly skilled labor force. (The Adobe Acrobat viewer plugin is required to
view this document. You may download this viewer by clicking
here.)
- Barry Bluestone and Stephen Rose, "Overworked and Underemployed: Unraveling an Economic Enigma"
http://www.prospect.org/print/V8/31/bluestone-b.html
Barry Bluestone and Stephen Rose examine whether Americans are overworked in this article appearing in the March/April 1997
issue of The American Prospect. They note that Juliet Schor found that workers were overworked, yet other researchers had found
that the average work week was declining and part-time work rising. Bluestone and Rose suggest that both of these observations may be
the result of growing job instability. This instability causes workers to work longer hours when these are available and reduced
hours in other periods due to layoffs or shortened workweeks offered by firms during slack periods. Bluestone and Rose note that
there is evidence of longer workweeks in manufacturing and managerial positions. They also note that when workers are
asked if they would prefer a shorter workweek, most respond by saying "yes." When workers are asked a slightly different
variant in which they are told that a shorter workweek would provide lower income, a majority of workers
prefer their current mix of income and hours to alternatives. Bluestone and Rose raise concern over the growth of involuntary
part-time labor for both men and women. Using data from the Panel Study of Income Dynamics, they find that for workers with
low levels of education, the increase in weeks worked partly compensates for reduced average real wages. Households headed
by high school dropouts experienced declining family incomes despite increased hours worked between 1973 and 1988.
- The Simplicity Forum, "Take Back Your Time"
http://www.timeday.org/
This website is designed to provide information on "Take Back Your Time" Day (October 24, 2003). This group is involved in a campaign to reduce
the length of the average workweek. This date is selected because it is approximately the date at which a typical U.S. worker
has worked as many hours as a typical European worker works in a year.
- John De Graaf, "Why Americans Don't Have Time to Be Healthy"
http://www.organicconsumers.org/Organic/degraaf081202.cfm
John De Graaf, in this August 2002 online article, examines trends in work hours. He notes that in 1999, the U.S. passed Japan and
to become the industrial nation with the longest average workweek. De Graaf argues that overwork has adverse effects on health
and family relationships. He also suggests that a shorter and more flexible workweek would result in lower unemployment rates and a
cleaner environment (due reduced production and greater time availability for recycling).
- Robert Drago, T. Lynn Riggs, Russell Kashian, Darnell Chad, Robert Caplan, David Costanza, Tanya Brubaker, and Naomi Harris,
"New estimates of working time for elementary school teachers"
http://stats.bls.gov/opub/mlr/1999/04/art4full.pdf
In this April 1999 Monthly Labor Review article, Robert Drago et al use time diary survey data to analyze the time
worked by elementary school teachers. They find that, on average, elementary school teachers work approximately two hours
more per day at their school than required by their contracts. It is also observed that elementary school teachers spend
approximately two additional hours working on job-related tasks while at home. (The Adobe Acrobat viewer plugin is required
to view this document. You may download this viewer by clicking
here.)
- HRM Guide: USA, "Overworked Americans can't use up their vacations"
http://www.hrmguide.net/usa/worklife/unused_vacation.htm
This May 31, 2001 online article summarizes the results from a survey that indicates that one-sixth of American workers who
receive vacation time do not use their entire allotment. It is observed that American workers have the lowest average number
of annual vacation days among industrialized countries. Workers in most other industrialized economies receive 2-3 times as
many days of vacation time. The survey also suggests that workers tend to experience high levels of stress on the job.
- Alliance for Work-Life Progress, "National Work & Family Month"
http://www.awlp.org/awlp/nwfm/nwfm-home.jsp
This Alliance for Work-Life Progress website describes the passage of United States Senate
Resolution 210, declaring October to be National Work and Family Month. The purpose of this
resolution was to focus public awareness on the perceived increase in the time pressures facing
households.
- Fast Company, "Neo-Leisure"
http://www.fastcompany.com/online/00/spy0.html
This somewhat tongue-in-cheek article suggests that the roles of labor and work have become blurred in recent years. Workers
spend more hours at work, but have integrated many leisure activities into their worklife. While the article is intended to be
humorous, it makes several useful points about the changing nature of work and leisure.
- Susan P. Choy, "Debt Burden Four Years After College"
http://nces.ed.gov/pubs2000/2000188.pdf
Susan P. Choy examines the debt burden of college graduates four years after the completion of college
in this article appearing in the Fall 2000 issue of Education Statistics Quarterly. She finds that
approximately 50% of all 1992-93 college graduates had relied on borrowing to finance part of their education.
The average borrowing of these individuals was $10,100. Choy finds that students who had borrowed
larger amounts were less likely to immediately enroll in post-baccalaureate educational programs. Those who
did not enroll in further education faced monthly loan payments that were approximately equal to
5% of their monthly income. (The Adobe Acrobat viewer plugin is required
to view this document. You may download this viewer by clicking
here.)
|