Would You Work If You Won the Lottery? Tracking Changes in the ...

Journal of Applied Psychology 2010, Vol. 95, No. 2, 349 –357

© 2010 American Psychological Association 0021-9010/10/$12.00 DOI: 10.1037/a0018359

RESEARCH REPORTS

Would You Work If You Won the Lottery? Tracking Changes in the American Work Ethic Scott Highhouse, Michael J. Zickar, and Maya Yankelevich

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Bowling Green State University Although many social scientists and political commentators have speculated that the American work ethic is in decline, the last longitudinal study of this issue was conducted by Vecchio (1980) on data collected over 30 years ago. Vecchio examined whether workers would wish to continue working even if it were not financially necessary (i.e., the so-called lottery question from the National Opinion Research Center survey) and concluded that there had been a significant decline in work ethic since the 1950s. In the current study, the authors used an updated data set that included data from 1980 through 2006 and found evidence for a declining trend since Vecchio’s study, although the decline seems to be leveling out. Demographic characteristics do not account for this apparent decline in shared sentiments about the importance of work for a productive life. The authors caution against drawing definitive conclusions, given the limitations of the lottery item as a measure of work ethic. Keywords: work ethic, work and leisure, nonfinancial employment commitment

80% from Morse and Weiss’s research, Vecchio concluded that the results supported the then current views of social change theorists—that there has been a steady decline in the value attached to work since the middle of the 20th century. In summarizing this idea, he noted that “the present findings suggest that a leisure ethic may be replacing the traditional work ethic in the United States” (Vecchio, 1980, p. 366). The purpose of the present investigation was to follow up on Vecchio’s (1980) study, examining whether, 30 years later, responses to the NORC question show evidence for the predicted trend toward greater emphasis on leisure in the United States. Although investigators from other countries have looked at responses to various versions of what has become known as the lottery question (Harpaz, 1988, 1989; Snir & Harpaz, 2002), there has been no longitudinal investigation of responses since Vecchio’s (1980) study. The nature of this research is descriptive rather than inferential, but we believe that a careful and updated examination of these data can provide insight into the changing nature of work ethic and suggest avenues for future research.

Would you continue to work if you inherited enough money to live comfortably without working? Morse and Weiss (1955) posed this question to a national sample of working men in America and found that 80% of these men indicated that they would indeed continue to work. The authors therefore concluded that people derived much more than economic benefit from work. According to Morse and Weiss, “it is through the producing role that most men tie into society, and for this reason and others, most men find the producing role important for maintaining their sense of wellbeing” (p. 198). Society has changed dramatically since Morse and Weiss conducted their research, and many have asserted that the meaning and value attached to employment has declined among workers in America (e.g., Howard & Wilson, 1982; Lasch, 1978; Yankelovich, 1978). Responding to such assertions, Vecchio (1980) examined responses to a question from interviews conducted by the National Opinion Research Center (NORC) in 1974, 1976, and 1977 to see whether Morse and Weiss’s conclusions held up after 20 years. The NORC asked a random sample of 1,099 male workers in the United States the following question: “If you were to get enough money to live as comfortably as you would like for the rest of your life, would you continue to work or would you stop working?” Vecchio reported that no change occurred over the 3-year period but reported that 72% of respondents across the three data sets indicated that they would continue to work. Because this observed percentage was significantly different from the expected

Declining American Work Ethic? Since the time of the Reformation, Western civilization has been characterized as holding a value system that Max Weber (1904 – 1905/1958) called the Protestant work ethic. This ethic is characterized by hard work, self-denial, and the avoidance of idleness (e.g., Cherrington, 1980). A unique American work ethic has also been proposed (e.g., Applebaum, 1998), even though its defining features seem similar to Weber’s Protestant work ethic. Few would dispute that Americans have traditionally been characterized as unusually industrious. As early as the mid-19th century, immigrants commented on Americans’ lack of tolerance for idleness

Scott Highhouse, Michael J. Zickar, and Maya Yankelevich, Department of Psychology, Bowling Green State University. We are grateful to Robert Gibby for his contributions to this project. Correspondence concerning this article should be addressed to Scott Highhouse, Department of Psychology, Bowling Green State University, Bowling Green, OH 43403. E-mail: [email protected] 349

RESEARCH REPORTS


350

(Stott, 1985). For example, one early immigrant, James Dawson Burn (1865), observed, “There is no such thing as shirking labour either in the workshops or fields. A man must make up his mind to work or play, otherwise he will soon find his level” (p. 280). Although longer work hours and later retirement have kept the image of the American work ethic alive and well in the eyes of the international community (e.g., Benjamin & Horwitz, 1994; Curry, 2003; Hunnicutt, 1996), domestic commentators have continued to allude to a decline in the value Americans place on work versus leisure (e.g., Aley, 1994; Brady, 2002; Howard & Wilson, 1982). Indeed, the shift of emphasis from industry and achievement to personal growth and even self-indulgence was said to begin with postwar societal changes and the growth of human relations in industry (e.g., Highhouse, 2002). Yankelovich (1978) observed that loyalty to the organization had been replaced by loyalty to the self and that the concern for work had been superseded by the concern for leisure. Peter Warr (1982) distinguished between the wanting to work component of work ethic and the wanting to work hard component. Whereas the latter component has been studied extensively in the motivation and personality literature, the former component has been paid considerably less attention. Although measures of job involvement (Brown, 1996; Rabinowitz & Hall, 1977) and work centrality (Hirschfeld & Field, 2000; Mannheim, 1993) have been put forth as indicators of the degree to which work is a central life interest or an all-encompassing part of one’s life, they do not tap the idleness component of work ethic. Warr thus coined the term nonfinancial employment commitment to get at the wanting-to-work aspect of work ethic. The decision of whether to work after a financial windfall is viewed as the quintessential behavioral manifestation of the American work ethic (Kaplan, 1985), and the so-called lottery question has been the primary method for studying nonfinancial employment commitment (e.g., Meaning of Work [MOW] International Research Team, 1987).1

Updating and Extending Vecchio (1980) Warr (1982) presented one version of the lottery question to British workers and found that overall 69% of the men and 65% of the women in the sample indicated that they would continue working in the absence of financial need—numbers he considered comparable with Vecchio’s (1980) findings from the United States. Itzhak Harpaz and his associates (Harpaz, 1985, 1988, 1989; Snir & Harpaz, 2002) conducted the most recent analyses with a lottery question from the 1981 International MOW project (MOW International Research Team, 1987). If we took their data at face value, we might conclude that the proportion of affirmative responses to the lottery question had risen in the United States an astounding 18% in the decade since Vecchio analyzed the NORC data. However, the differences in the items used in the two studies are so substantial that direct comparison is impossible.2 Needed is an investigation of responses to the same NORC item, tracked since Vecchio’s original study. According to Warr, “A minimum requirement [for establishing a change in work ethic] is the repeated administration over several years of standard questions to large random samples of a population” (p. 298). Unfortunately, no published work since Vecchio’s has used the NORC item to gauge whether there has been a change in responses since the 1970s. This

is especially surprising given that the NORC has continued to survey random samples of Americans with the lottery question since Vecchio’s report. The primary aim of the present study was to update the findings of Vecchio (1980) and to conduct a longitudinal investigation of changes in nonfinancial employment commitment in America since the late 1970s. Because Vecchio focused only on working men, a secondary purpose of this analysis was to investigate whether sex differences exist in nonfinancial employment commitment. Although there is not a theoretical basis for formulating a hypothesis, research on related constructs has given little basis for expected sex differences. Brown (1996) conducted a metaanalysis of studies on job involvement, for example, and found virtually no relation with demographic variables. Mixed results have been found for sex differences in the sparse literature on work centrality (Mannheim, 1993; Mannheim, Baruch, & Tal, 1997). Of interest, a study of actual Ohio lottery winners showed no statistically significant difference between men and women in quitting their jobs (Arvey, Harpaz, & Liao, 2004). However, this study suffered from a very low response rate, combined with a low base rate of actual quitting. In addition to respondent sex, we examined other demographic and attitudinal variables as they relate to responses to the lottery question over time. Like Morse and Weiss (1955), Vecchio (1980) observed that younger respondents were more likely to prefer to continue working than older respondents. Vecchio did not, however, observe differences by age in the relative decline in affirmative responses to the lottery question. Not surprisingly, Vecchio observed less desire to continue working among those who were less satisfied, but other demographic and attitudinal variables did not appear to explain much variance in responses to the lottery question. According to Vecchio, “it must be concluded that individual preference to continue– discontinue working [after a financial windfall] cannot be easily predicted from demographic information” (p. 365). Finally, none of the studies that have investigated the lottery question have tried to link variations in macroeconomic environment to fluctuations in responses to the lottery question. There are precedents in other areas of research for linking individual phenomena to macroeconomic conditions. For example, Hulin (1966) found a link between community prosperity and job satisfaction, and meta-analytic research has linked 1

The lottery question was only modestly related to measures of work centrality, supporting the view that nonfinancial employment commitment and work centrality are different constructs. 2 The item used in the MOW project (MOW International Research Team, 1987) was, “Imagine that you won a lottery or inherited a large sum of money and could live comfortably for the rest of your life without working; what would you do about work?” (p. 225). Response options were to stop working, to continue to work in the same job, or to continue to work but under different conditions. Affirmative responses to second or third options were used to derive the percentages reported for each country. Aside from the different item wording, the absence of the third alternative in the item analyzed by Vecchio (1980) does not allow for direct comparison of results. People explicitly given the option of working in a different job may be more likely to indicate a desire to continue working after a windfall.

RESEARCH REPORTS

turnover to unemployment (e.g., Carsten & Spector, 1987). It is possible that, when the economy is prosperous, people may express greater nonfinancial employment commitment than when economic times are tough.

Method


Sample The sample used in the present study was taken from responses by currently employed, full-time workers to the General Social Surveys conducted by the NORC.3 As discussed in Vecchio (1980), the NORC used a quota-sampling scheme, based on U.S. Census tract information, to collect interview data, allowing for generalizability of results to the American population. More specifically, only respondents answering the lottery question on the surveys conducted during 1980, 1982, 1984, 1985, 1987, 1988, 1989, 1990, 1991, 1993, 1994, 1996, 1998, 2000, 2002, 2004, and 2006 were analyzed in the present study. In total, 7,871 men and 7,549 women provided responses to the lottery question. The mean age of respondents was 39 years; 75.8% of the sample was Caucasian, 14.4% was African American, and 9.9% responded other. Because of small sample size, all further analyses collapsed the African American and other categories. The mean level of education for the sample was 14 years.

Measures During the course of the interview for all years listed earlier, each respondent answered the following question: “If you were to get enough money to live as comfortably as you would like for the rest of your life, would you continue to work or would you stop working?” For ease of interpretation, responses to the lottery item were coded 1 (continue to work) and 0 (stop working). Respondents also reported their job satisfaction, life satisfaction, occupational prestige, and general happiness. To assess job satisfaction, respondents were asked, “On the whole, how satisfied are you with the work you do—would you say you are very satisfied [1], moderately satisfied [2], a little dissatisfied [3], or very dissatisfied? [4].” To determine life satisfaction, respondents were asked, “In general, do you find life exciting [1], pretty routine [2], or dull? [3].” To gauge respondents’ general level of happiness, they were asked, “Taken all together, how would you say things are these days? Would you say that you are very happy [1], pretty happy [2], or not too happy? [3].” Respondents were asked to describe their job activities, which were then coded by NORC researchers. Prestige data were available for two thirds of the respondents (10,400 out of 15,420 total respondents). When appropriate, we ran analyses with and without the prestige variable to make sure that results were not biased by the reduction in our sample. In addition to the demographic and attitudinal variables outlined earlier, each respondent provided information on financial status relative to American families in general. Responses to this question were coded 1 (far below average), 2 (below average), 3 (average), 4 (above average), and 5 (far above average).

351 Results

Descriptive Statistics Means, standard deviations, and intercorrelations for all study variables are presented in Table 1. Table 1 shows that nonfinancial employment commitment was significantly correlated with job satisfaction (r ⫽ ⫺.10), life satisfaction (r ⫽ ⫺.06), education (r ⫽ .04), and occupational prestige (r ⫽ .04). Although statistically significant, these are considered small by conventional standards (Cohen, 1988). Thus, it appears that attitudes and education explain variance in nonfinancial employment commitment, but the relations appear trivial in magnitude. Income and happiness did not correlate significantly with nonfinancial employment commitment. Next, we consider demographic variables separately.

Change Over Time The next step in the analyses was to examine changes in nonfinancial employment commitment since the 1980 investigation. Analyses were conducted to establish the presence of trends in the data in response to the lottery question. The linear and quadratic trends were statistically significant (Table 2). Taking a closer look at the trend over the 26-year period, we see that from 1980 through approximately 1993, there was a relatively steady downward trend in affirmative responses to the lottery question (Figure 1). This downward movement appears to level off from 1994 to 2006, consistent with the significant quadratic trend. For the time periods just described, 72.8% (1980 –1993) and 68.1% (1994 –2006) of people (Z ⫽ 6.79, p ⬍ .01) responded that they would continue working.

Demographic Variables and Response to the Lottery Question Next, we examined differences in age, sex, and race in response to the lottery question. For ease of interpretation, age was divided into the same categories presented by Vecchio (1980). Mean responses to the lottery question for each level of sex, race, and age are presented in Table 3. To determine whether there are significant differences in responses for sex and race, we performed independent samples t tests. Results revealed that women were more likely to state that they would discontinue working if it became financially unnecessary. Compared with men (M ⫽ .73, SD ⫽ .45), the lower mean response of women to the lottery question (M ⫽ .68, SD ⫽ .47) was statistically significant, t(15, 420) ⫽ 5.64, p ⬍ .01, d ⫽ .11, but this difference was minor. Caucasians (M ⫽ .70, SD ⫽ .46) and minorities (M ⫽ .71, SD ⫽ .45) did not differ significantly in their responses to the lottery question. A univariate analysis of variance with age as the independent variable and response to the lottery question as the dependent variable demonstrated that there was a significant differ3 We did not include the 65⫹ age group in the analyses, because we could not be sure how this group understood the lottery question (i.e., given that 65 is the traditional retirement age, and their responses were not consistent with the general age trend). This group constituted only approximately 4% of the survey respondents.

RESEARCH REPORTS

352

Table 1 Means, Standard Deviations, and Intercorrelations for NORC and Economic Variables


Variable 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Age Sex Race Education Income Job satisfaction Life satisfaction Happiness Prestige Lottery question Unemploymenta Inflationa Gross domestic producta

M

SD

1

2

3

4

5

6

7

8

9

10

38.83 1.49 1.20 13.63 2.98 1.77 1.60 1.85 44.30 0.70 6.19 3.91

11.52 0.50 0.40 4.31 0.94 0.94 0.89 0.80 13.92 0.46 1.47 2.65

—

.01 —

⫺.06ⴱⴱ .05ⴱⴱ —

.01 .00 ⫺.05ⴱⴱ —

.08ⴱⴱ ⫺.07ⴱ ⫺.10ⴱ .16ⴱ —

⫺.10 ⫺.01 .07ⴱⴱ .01 ⫺.12ⴱⴱ —

.01ⴱⴱ .01 .07ⴱⴱ ⫺.05ⴱⴱ ⫺.07ⴱⴱ .11ⴱⴱ —

⫺.02ⴱⴱ ⫺.01 .10ⴱⴱ ⫺.04ⴱⴱ ⫺.12ⴱⴱ .19ⴱⴱ .20ⴱⴱ —

.11ⴱⴱ .01 ⫺.10ⴱⴱ .31ⴱⴱ .23ⴱⴱ ⫺.15ⴱⴱ ⫺.12ⴱⴱ ⫺.10ⴱⴱ —

⫺.15ⴱⴱ ⫺.05ⴱⴱ .00 .04ⴱⴱ .00 ⫺.09ⴱⴱ ⫺.06ⴱⴱ ⫺.01 .04ⴱⴱ —

6,969.96

2,924.89

11a

12a

13a

.39 —

.64ⴱⴱ .36 —

⫺.59ⴱ ⫺.72ⴱⴱ ⫺.57 —

Note. 1 ⫽ male, 2 ⫽ female; 1 ⫽ Caucasian, 2 ⫽ other. NORC ⫽ National Opinion Research Center. a Non-NORC variables. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.

ence between the age groups, F(3, 14986) ⫽ 103.42, p ⬍ .01, ␩2 ⫽ .02. Tukey’s honestly significant difference post hoc analysis revealed that each age group had a significantly higher mean response to the lottery question than the age group before it (Table 3). This indicates that, compared with older respondents, younger people were more likely to report a desire to continue working when it is not financially necessary.

Discriminating Variables A discriminant analysis was performed for the dependent variable of desire to continue working. This was intended to parallel Vecchio’s (1980) analyses and to determine whether there has been a change in which variables are the most consistent discriminators. He included respondents’ age, education in years, race, income, job satisfaction, and prestige. Table 4 presents the results of the discriminant analysis, with associated discriminant function structure coefficients for the predictor variables. Because the demographic composition of the current sample has changed since Vecchio’s (1980) article was published (women were not part of the early sample), we wanted to determine whether sex would be a significant discriminator as well. Both discriminant functions were significant: without sex, ␹2(6, N ⫽ 5380) ⫽ 360.02, p ⬍ .001, and with sex, ␹2(7, N ⫽ 5380) ⫽ 380.21, p ⬍ .001. In the first analysis, the variables with the largest structure coefficients were respondent age and job Table 2 Source Table for Analysis of Change in Responses to Lottery Item Over Time Trend term

df

MS

F

Linear term Quadratic term Cubic term Error

1 1 1 15,403

8.51 3.20 0.15 0.207

41.08ⴱ 15.47ⴱ 0.73

ⴱ

p ⬍ .01.

satisfaction. This mirrors the findings of Vecchio. In the second analysis (i.e., with sex included as a discriminator), age, job satisfaction, and sex were identified as discriminators. Classification results conducted with prior weights of group size show that the classification accuracy was 68.9% of all respondents, only slightly better than the base rate of 68.4%. Although the discriminant function is significant, these classification results suggest that the demographic variables explain small amounts of variance in the lottery question. The classification results are consistent with the small canonical correlation of these functions (.19).

Moderators of the Relationship Between Time and the Lottery Question On the basis of the findings of the discriminant analysis, we decided to perform logistic regression to determine whether the previously identified discriminators (age, sex, and job satisfaction) moderated the relation between time and people’s response to the lottery question (see Table 5). It may be possible, for example, that responses to the lottery question declined less quickly for women than for men. In testing the moderating effects of sex, age, and job satisfaction, we also controlled for education, income, prestige, race, sex (except in the sex regression), and age (except in the age regression). All three models were significant; however, none of the moderator items had significant standardized beta weights. Furthermore, the interaction terms were not statistically significant. The estimated R2 statistics were small; the highest estimated Nagelkerke R2 was .06 with job satisfaction as a moderator. The incremental R2 statistics were negligible (i.e., less than .01), indicating that trends over time were not different across race, age, or sex.

Supplementary Analyses In the absence of significantly meaningful demographic moderators in the present data set, we turned to economic data to better under-

RESEARCH REPORTS

353

0.85

Would Continue Working


0.80

0.75

0.70

0.65

1980 1982 1984 1985 1987 1988 1989 1990 1991 1993 1994 1996 1998 2000 2002 2004 2006

Year Figure 1. Proportion of people who reported a desire to continue working after a financial windfall (1980 – 2006). Error bars represent 95% confidence intervals.

stand fluctuations over time.4 The yearly unemployment rate, the inflation rate, and the gross domestic product (GDP) were used as proxies of national economic fluctuations. The unemployment rate is presented as percentage of unemployment nationally (U.S. Department of Labor, Bureau of Labor Statistics, n.d.). The inflation rate is presented as the average yearly percentage of the change in the Consumer Price Index from a year earlier

Table 3 Mean Responses to the Lottery Item by Demographic Variable and Comparison Tests for Between-Subject Variability Variable Sex Male Female Race Caucasian Other Age 21–34 35–44 45–54 55–64

n

M

SD

7,871 7,549

.73 .68

.45 .47

12,326 3,094

.70 .71

.46 .45

5,748 4,313 3,128 1,801

.77 .71 .66 .56

.42 .45 .47 .50

t

Cohen’s d

Discussion 5.64ⴱ ⫺0.54

.11 ⫺.00

Source

df

MS

F

␩2

Age Error

3 14,986

21.17 0.21

103.42ⴱ

.02

ⴱ

p ⬍ .01.

(see Financial Trend Forecaster, n.d.). The GDP is presented in billions of dollars and is obtained from the U.S. Department of Commerce, Bureau of Economic Analysis (n.d.). We correlated the mean response to the lottery question for each year with each of the economic data indicators. As shown in Table 1, response to the lottery question is significantly related to inflation rate (r ⫽ .64) and the GDP (r ⫽ ⫺.59). Although caution should be used in interpreting the magnitude of correlations based on only 17 data points, the covariation between the lottery question and these economic indicators suggests that people are more likely to report a desire to continue working after a (hypothetical) windfall when economic times are least favorable.

Most religious and secular definitions of work ethic place central importance on productive activity for providing personal and social meaning in people’s lives. Perhaps no citizenry has been more associated with this ethos historically than Americans (de Tocqueville, 1840/2000). The American work ethic, however, has been under attack by commentators who suggest that Americans have lost their belief in working for work’s sake—a belief that propelled the country forward during its early industrialization and through the postwar boom years. In our study—a follow up on 4

We thank an anonymous reviewer for suggesting these supplementary analyses.

RESEARCH REPORTS

354


Table 4 Results of Discriminant Analysis Variable

Analysis 1

Analysis 2

Canonical correlation Wilks’s ␭ Chi-square df p Structure coefficients Age Race Education Income Job satisfaction Prestige Sex

.19 .97 368.91 6 .00

.19 .96 386.23 7 .00

.76 ⫺.05 ⫺.19 .05 .52 ⫺.20 —

.74 ⫺.05 ⫺.19 .05 .50 ⫺.20 .23

deviation lower in nonfinancial employment commitment may be surprising to people who assume that traditional work roles would make women substantially more likely to leave the workforce if they were financially independent. The apparent absence of large sex differences is consistent with research suggesting that women do not differ appreciably from men in job involvement (Brown, 1996), as well as with a large body of literature in the social sciences showing negligible gender differences (Hyde & Plant, 1995). Similar to Vecchio’s (1980) findings, we found that younger respondents were more likely to report a desire to continue working compared with older respondents. Also consistent with Vecchio’s study, however, was the finding that age differences were not related to the downward trend in affirmative responses to the lottery item. Like Vecchio, we also did not observe race differences in responses to the lottery item. This finding is not consistent with some employers’ attitudes about perceived race differences in orientation toward work (see Thomas, 2003). In general, our findings are in concurrence with Vecchio’s conclusion that demographic variables are not driving the observed decline in desire to continue working after a financial windfall. It is notable that fluctuations in responses to the lottery question seem to be linked to economic conditions. When economic times are difficult (e.g., high inflation, low GDP), the mean level of responses to the lottery question increases. Although one might expect less focus on leisure when the economy is relatively weak, we found the opposite. One possible explanation for this is that people may feel more comfortable giving up work when the economy is prosperous. When times are tough, the decision to give up work may be viewed as irrevocable. These significant findings should be confirmed in further longitudinal studies. In addition, cross-sectional studies that link regional economic variations with regional variations in work ethic might be another avenue for pursuing the role of

Vecchio’s (1980) study of NORC data from the 1970s—we examined whether people would continue to work if they could live comfortably without working. The lottery item served as a proxy for nonfinancial employment commitment, colloquially referred to as work ethic. Compared with Vecchio’s findings, we found that Americans are less likely to say that they would continue to work if it were not financially necessary. Moreover, the general trend of decline in nonfinancial employment commitment is remarkably similar across different aspects of the U.S. working population. Thus, existing longitudinal data appear to tentatively support the assertion that the American work ethic has declined, though this decline seems to have recently stabilized. Because Vecchio (1980) did not include women in his initial analysis of the NORC data, one purpose of our study was to examine whether men and women differed in their desire to work. Our finding that women were, on average, less than one tenth of a standard

Table 5 Logistic Regression Analysis of Response to the Lottery Item Over Time With Sex, Age, and Job Satisfaction as Moderators ␤

SE ␤

Wald

df

Constant Year Sex Year ⫻ Sex Total model

44.39 ⫺0.02 ⫺25.57 0.01

31.25 0.02 19.074 0.01

2.02 1.87 1.68 1.66

1 1 1 1

Constant Year Age Year ⫻ Age Total model

⫺4.86 0.00 0.25 0.00

Constant Year Job satisfaction Year ⫻ Job Satisfaction Total model

7.27 0.00 ⫺0.67 0.00

Predictor

Note.

36.11 0.02 0.86 0.00

22.09 0.01 11.08 0.01

0.02 0.03 0.08 0.102

0.11 0.05 0.00 0.11

1 1 1 1

p

e␤ (odds ratio)

Sex as moderator .155 1.9E .01 .20 .20 Age as moderator .75 .89 .77 .75

Cox & Snell R2

Nagelkerke R2

Model ␹2

df

p

.024

.034

228.34

8

.00

.024

.034

226.80

8

.00

.034

.048

319.92

9

.00

⫹ 19 1.02 0.00 1.01

1.00 1.00 1.28 1.00

Job satisfaction as moderator 1 .742 1,441.83 1 .82 1.00 1 .95 0.51 1 .97 1.00

Education, income, prestige, race, sex (except in the sex regression), and age (except in the age regression) were controlled for in each equation.

RESEARCH REPORTS

economic conditions in shaping individual work-related perceptions. Researchers should also examine the possible impact of immigration waves on work ethic. The possible link between macrolevel phenomena and individual-level perceptions is a potential bridge between psychology and economics.


Research Agenda Before speculating about the causes and consequences of a declining American work ethic, it is necessary to put forth a research agenda for psychologists interested in better understanding a person’s commitment to working for work’s sake. This agenda begins by (a) figuring out what this commonly used lottery question actually measures and (b) developing theoretical connections between nonfinancial employment commitment and other constructs. Table 6 summarizes these recommendations. Measurement issues. Whereas the NORC put considerable effort into population sampling to attain statistical generalizability, relatively little work was done on construct measurement to attain theoretical generalizability (see Chow, 1987). Although some have argued for the efficacy of single-item measures of global constructs (e.g., Wanous & Hudy, 2001; Wanous, Reichers, & Hudy, 1997), the risk of construct deficiency is high for a construct as controversial as work ethic. The NORC version of the lottery question may be subject to different interpretations. One respondent, for example, might assume the question refers to working in the same job, whereas another might interpret it to mean working in any job. Also, the item does not distinguish between paid and unpaid work. The item is posed in the course of an interview that inquires about the respondents’ feelings about their current job. Thus, it is possible that respondents interpret the question to simply mean working in their current job. Can we equate leaving work with having a leisure focus? One may leave paid work to pursue volunteer activities or

Table 6 Summary of Directions for Future Research Issue Measurement issues Theoretical issues

Understanding the decline in responses to the lottery question

Note.

Future direction Develop a better understanding of respondents’ interpretations of the NORC lottery item. Study the relation between more complex measures of work ethic and the lottery item. Determine the relation between different traits (e.g., conscientiousness) and motives (e.g., desire for meaning) and nonfinancial employment commitment. Focus on the role of age and temporal factors in the decision to continue working after a financial windfall. Understand the implications of early retirement and alternative retirement scenarios on nonfinancial employment commitment. Does the decline reflect a declining work ethic or a change in response tendencies? Why do economic conditions impact intended desire to continue working after a financial windfall? How has the leisure ethic changed over the same time period?

NORC ⫽ National Opinion Research Center.

355

to care for relatives. Such behavior may not be indicative of a poor work ethic. Our observed age differences on the question may also point to differences in interpretation between age groups. Younger workers simply face more years of idleness if they retire. In addition, younger people do not have the retirement assets of older workers and, as such, may not view a financial windfall in the same way. Assuming construct validity, however, younger people may simply assign less value to idle time than people closer to retirement— younger people may be more idealistic about the value of work, regardless of generation. It is worth noting that our age effect is consistent with the one observed among Israelis in both 1981 and 1993 (Harpaz, 2002). To better understand all of these possible different interpretations, it may be necessary to interview respondents after completing the lottery item and inquire about their own thought processes. Many have argued for a multidimensional view of work ethic (reviewed by Furnham, 1990; Miller, Woehr, & Hudspeth, 2002), proposing as many as four to seven subdimensions. Understanding the relation between these more complex measures of work ethic and the single-item lottery question is important. In contrast to these approaches, ter Bogt, Raaijmakers, and van Wel (2005) adopted a unidimensional view based on Weber’s (1904 –1905/ 1958) notion that a central component of work ethic is the belief that a purposeful life requires engagement in work (i.e., one should work). This is similar to defining work ethic as work for work’s sake (Lessnoff, 1994) or as the value placed on work per se (Blau & Ryan, 1997). The NORC item seems appropriate and meaningful from these perspectives, but measurement and interpretation issues must be addressed before consequential research can progress on this topic. Theoretical issues. We have taken Warr’s (1982) position that work ethic is reflected in the desire to maintain employment. This is distinguished from traits that focus on one’s beliefs about dependability and hard work (e.g., conscientiousness) and from attitudes that combine multiple values and needs (e.g., intrinsic motivation). Understanding the value placed on employment for nonfinancial reasons is important for economists and management theorists who find that perspectives such as agency theory cannot account for nonfinancial work incentives (Frey, 1997; Gill, 1999). The desire to continue work for work’s sake may be predicted by motives for achievement, affiliation, dignity, or a desire for existential meaning. Understanding work ethic requires understanding the relative contribution of such motives to nonfinancial employment commitment. Our data suggest contextual motives as well. On average, people seem to express a desire to continue working when the economy is more unfavorable. It is also interesting to note the steady decrease in nonfinancial employment commitment across the career span. Regardless of time period, older people are less likely to report a desire to continue working than younger people. Although this age effect did not explain the decline across time (i.e., the different age groups generally exhibited less commitment to working across the years), it does seem to contradict the stereotype of younger workers as having less work ethic. We noted earlier that one possible explanation for this is the longer period of idleness faced by younger workers who retire early. Another possibility is temporal discounting, wherein people further away from retirement tend to discount its positive aspects. It is, however, possible that younger


356

RESEARCH REPORTS

people are simply more likely to find meaning in working, as compared with their elders (Harpaz, 2002). Much more attention has been paid over the past 30 years to retirement and the issues faced by those who no longer work. Much of this is likely due to changes in state and federal laws that eliminated mandatory retirement age, thus giving Americans more say in their decision about when to quit working (see Adams & Beehr, 2003). Although retiring after a financial windfall is not the same as retiring after a lifetime of work, the nonfinancial factors influencing the retirement decision may be similar. Wang, Zhan, and Liu, (2008) noted that the retirement literature provides a number of nonfinancial explanations for obtaining bridge employment. These explanations include the retiree’s desire to maintain a work-role identity, the desire to avoid stressful disruptions, and the absence of a strong social network outside of work. Research might be done to examine the relative influence of these factors on nonfinancial employment commitment across the life span. It would also be helpful for theory building in this area if researchers had a sense of which of these factors are most likely to change along with societal changes. It may be possible, for example, that some of the decline in nonfinancial employment commitment is because people are less afraid of the loss of a work-role identity. Increased marketing attention to the aging baby boomers may have also increased awareness of the possibility of a good life without work. These kinds of macro-level attitude changes require research approaches informed by both sociological and psychological theory. Nonfinancial employment commitment may influence important work outcomes as well. For example, employee motivation is often seen as being composed of arousal, direction, and maintenance (see Mitchell, 1982). Arousal is addressed by need theories and fairness theories. Direction is captured by theories of goal setting and decision making. The maintenance component of motivation, however, has been more difficult to predict and influence. Nonfinancial employment commitment is one possible predictor of maintenance or persistence at work. Also, recent conceptualizations of work meaning emphasize the distinction between viewing one’s job as a career or as a calling (Wrzesniewski & Dutton, 2001). It is likely that nonfinancial employment commitment is an important outcome of viewing work as a calling.

Why the Decline? It is worth re-emphasizing that our findings reveal a drop in affirmative responses to the NORC lottery item. Whether this reflects an actual decline in work ethic or something altogether different requires future research. One reviewer suggested that people may simply be more candid in responding to surveys than they were 30 years ago. Alternatively, a change in beliefs about the need to work for a productive life may be a consequence of a more literate and progressive U.S. population. A longitudinal study of the work ethic of Dutch adolescents conducted by ter Bogt et al. 2005) suggested that work ethic is stable across the life span and that it is associated with lower education level and higher cultural conservatism. The National Center for Education Statistics reported that the high-school dropout rates in the United States have trended downward from 6.1% in 1972 to 3.8% in 2006 (Laird, Cataldi, KewalRamani, & Chapman, 2008). Also, parenting changes over time in emphasis on work versus play may have an

influence on later beliefs about the purpose of work (Edwards, 2000). For example, researchers have begun to focus on the leisure ethic as an alternative to the Protestant work ethic (see Buchholz, 1978). One of the reasons for the decrease in affirmative responses to the lottery question is not that the work ethic itself has declined but that the attractiveness of life after work has increased in the United States.

Conclusions Consistent with the findings of Vecchio (1980), the proportion of affirmative responses to the NORC lottery question has continued to decline over the past quarter century. There is some indication that there is a leveling off of the decline as seen in Figure 1. The small correlations with the available NORC variables limit our ability to find clues for why people are less likely to prefer to remain at work after a hypothetical financial windfall. On the other hand, it shows that the issue is more complex than an aggregate decline in perceived well-being, or a function of shifting demographics, or an increase in the number of women in the workforce. In addition, the relatively stronger correlations with economic data suggest that general fluctuations in nonfinancial employment commitment may be influenced by macroeconomic conditions. McCortney and Engels (2003) suggested that the 2001 terrorist attacks on the Pentagon and World Trade Center were colossal enough to cause a positive change in the general work ethic of American workers. Having additional years of data can help pursue provocative questions such as these. It is also important to keep in mind that our study focused on just one (25-year) snapshot of the American workforce. Historians have speculated about shortterm rises and falls in the work ethic throughout the century (e.g., Lutz, 1996), and our data may reflect one such variation. Whether Americans are experiencing a small period of turbulence or an unprecedented decline in work values remains an empirical question.

References Adams, G. A., & Beehr, T. A. (2003). Retirement: Reasons, processes, and results. New York: Springer. Aley, J. (1994, February). Slacker myths. Fortune, 129, 24. Applebaum, H. A. (1998). The American work ethic and the changing work force: An historical perspective. Westport, CN: Greenwood Press. Arvey, R. D., Harpaz, I., & Liao, H. (2004). Work centrality and postaward work behavior of lottery winners. Journal of Psychology, 138, 404 – 420. Benjamin, D., & Horwitz, T. (1994, July). German view: “You Americans work too hard—And for what?” Wall Street Journal, 224, p. B1. Blau, G., & Ryan, J. (1997). On measuring work ethic: A neglected work commitment facet. Journal of Vocational Behavior, 51, 435– 448. Brady, D. (2002, August 26). Rethinking the rat race. BusinessWeek. Retrieved from http://www.businessweek.com/magazine/content/02_34/ b3796646.htm Brown, S. P. (1996). A meta-analysis and review of organizational research on job involvement. Psychological Bulletin, 120, 235–255. Buchholz, R. A. (1978). The work ethic reconsidered. Industrial and Labor Relations Review, 31, 450 – 459. Burn, J. D. (1865). Three years among the working-classes in the United States during the war. London, England: Smith, Elder. Carsten, J. M., & Spector, P. E. (1987). Unemployment, job satisfaction,


RESEARCH REPORTS and employee turnover: A meta-analytic test of the Muchinsky model. Journal of Applied Psychology, 72, 374 –381. Cherrington, D. J. (1980). The desire for an enriched job as a moderator of the enrichment–satisfaction relationship. Organizational Behavior and Human Performance, 25, 139 –159. Chow, S. L. (1987). Science, ecological validity and experimentation. Journal of the Theory of Social Behaviour, 17, 181–194. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Curry, A. (2003, February 24). Why we work. U.S. News & World Report, 134, p. 48. de Tocqueville, A. (2000). Democracy in America (H. C. Mansfield & D. Winthrop, Trans.). Chicago, IL: University of Chicago Press. (Original work published 1840) Edwards, C. P. (2000). Children’s play in cross-cultural perspective: A new look at the Six Cultures Study. Cross-Cultural Research, 34, 318 –338. Financial Trend Forecaster. (n.d.). Historical US inflation rates. Retrieved from http://www.inflationdata.com/inflation/inflation_rate/historicalinflation.aspx Frey, B. S. (1997). On the relationship between intrinsic and extrinsic work motivation. International Journal of Industrial Organization, 15, 427– 439. Furnham, A. (1990). A content, correlational, and factor analytic study of seven questionnaire measures of the Protestant work ethic. Human Relations, 43, 383–399. Gill, F. (1999). The meaning of work: Lessons from sociology, psychology, and political theory. Journal of Socio-Economics, 28, 725–743. Harpaz, I. (1985). Meaning of working profiles of various occupational groups. Journal of Vocational Behavior, 26, 25– 40. Harpaz, I. (1988). Variables affecting non-financial employment commitment. Applied Psychology: An International Review, 37, 235–248. Harpaz, I. (1989). Non-financial employment commitment: A crossnational comparison. Journal of Occupational Psychology, 62, 147–150. Harpaz, I. (2002). Expressing a wish to continue or stop working as related to the meaning of work. European Journal of Work and Organizational Psychology, 11, 177–198. Highhouse, S. (2002). A history of the T-group and its early applications in management development. Group Dynamics: Theory, Research, and Practice, 6, 277–290. Hirschfeld, R. R., & Field, H. S. (2000). Work centrality and work alienation: Distinct aspects of a general commitment to work. Journal of Organizational Behavior, 21, 789 – 800. Howard, A., & Wilson, J. A. (1982, Summer). Leadership in a declining work ethic. California Management Review, 24, 33– 46. Hulin, C. L. (1966). Effects of community characteristics on measures of job satisfaction. Journal of Applied Psychology, 50, 185–192. Hunnicutt, B. K. (1996). Kellogg’s six-hour day. Philadelphia, PA: Temple University Press. Hyde, J. S., & Plant, E. A. (1995). Magnitude of psychological gender differences. American Psychologist, 50, 159 –161. Kaplan, H. R. (1985). Lottery winners and work commitment: A behavioral test of the American work ethic. Journal of the Institute for Socioeconomic Studies, 10, 82–94. Laird, J. R., Cataldi, E., KewalRamani, A., & Chapman, C. (2008). Dropout and completion rates in the United States: 2006. Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid⫽2008053 Lasch, C. (1978). The culture of narcissism: American life in an age of diminishing expectations. New York, NY: Norton. Lessnoff, M. H. (1994). The spirit of capitalism and the Protestant ethic: An enquiry into the Weber thesis. Brookfield, VT: Elgar. Lutz, T. (1996). “Sweat or die”: The hedonization of the work ethic in the 1920s. American Literary History, 8, 259 –283.

357

Mannheim, B. (1993). Gender and the effects of demographics, status and work values on WRC. Work and Occupations, 20, 3–22. Mannheim, B., Baruch, Y., & Tal, J. (1997). Alternative models for antecedents and outcomes of work centrality and job satisfaction of high-tech personnel. Human Relations, 50, 1537–1562. McCortney, A. L., & Engels, D. W. (2003). Revisiting the work ethic in America. Career Development Quarterly, 52, 132–140. Miller, M. J., Woehr, D. J., & Hudspeth, N. (2002). The meaning and measurement of work ethic: Construction and initial validation of a multidimensional inventory. Journal of Vocational Behavior, 60, 451– 489. Mitchell, T. R. (1982). Motivation: New directions for theory, research, and practice. Academy of Management Review, 7, 80 – 88. Morse, N. C., & Weiss, R. S. (1955). The function and meaning of work and the job. American Sociological Review, 20, 191–198. Meaning of Work International Research Team. (1987). The meaning of working: An international view. London, England: Academic Press. Rabinowitz, S., & Hall, D. T. (1977). Organizational research on job involvement. Psychological Bulletin, 84, 265–288. Snir, R., & Harpaz, I. (2002). To work or not to work: Nonfinancial employment commitment and the social desirability bias. Journal of Social Psychology, 142, 635– 644. Stott, R. (1985). British immigrants and the American “work ethic” in the mid-nineteenth century. Labor History, 26, 86 –102. ter Bogt, T., Raaijmakers, Q., & van Wel, F. (2005). Socialization and development of the work ethic among adolescents and young adults. Journal of Vocational Behavior, 66, 420 – 437. Thomas, W. F. (2003). The meaning of race to employers: A dynamic qualitative perspective. Sociological Quarterly, 44, 227–242. U.S. Department of Commerce, Bureau of Economic Analysis. (n.d.). Gross domestic product. Retrieved October 2, 2008, from http:// www.bea.gov/national/#gdp U.S. Department of Labor, Bureau of Labor Statistics. (n.d.). Current Population Survey (CPS): Labor force statistics. Retrieved October 2, 2008, from http://www.bls.gov/CPS Vecchio, R. P. (1980). The function and meaning of work and the job: Morse and Weiss (1955) revisited. Academy of Management Journal, 23, 361–367. Wang, M., Zhan, Y., & Liu, S. (2008). Antecedents of bridge employment: A longitudinal investigation. Journal of Applied Psychology, 93, 818 – 830. Wanous, J. P., & Hudy, M. J. (2001). Single-item reliability: A replication and extension. Organizational Research Methods, 4, 361–375. Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology, 82, 247–252. Warr, P. (1982). A national study of non-financial employment commitment. Journal of Occupational Psychology, 55, 297–312. Weber, M. (1958). The Protestant ethic and the spirit of capitalism (T. Parsons, Trans.). New York, NY: Scribners. (Original work published 1904 –1905) Wrzesniewski, A., & Dutton, J. E. (2001). Crafting a job: Revisioning employees as active crafters of their work. Academy of Management Review, 26, 179 –201. Yankelovich, D. (1978, May). The new psychological contracts at work. Psychology Today, 11, 46 –50.

Received February 18, 2008 Revision received November 10, 2009 Accepted November 10, 2009 䡲