Listen to a news conference with the author

Despite all of the recent attention given to the needs of working mothers for flexible work schedules, mothers are no more likely than other workers to be able to determine the times they arrive at and leave work, or to decide when to take an occasional day off. Single mothers, who must handle all the responsibility for work and family on their own, have particularly rigid schedules. There is also a pronounced racial difference in work schedule flexibility: black workers are much less likely than white workers to be able to exercise any discretion over their work schedules. In contrast, men, and to some extent women with supervisory or policy-making authority, enjoy much greater flexibility than other workers.

Furthermore, contrary to the expectations of many economists, workers who do enjoy flexible hours earn more, not less, than those with rigid work schedules. Some of this differential is accounted for by organizational power, and some by occupation and industry. The evidence also suggests that unions successfully negotiate greater compensation for workers in rigid jobs. Because women are no more likely to enjoy flexible schedules than men, the freedom to adjust one’s work schedule does little to explain the gender pay gap. Again, this is contrary to the expectations of many economists, who presume that men are paid more, in part, for their willingness to accept rigid schedules.

The inability of mothers to secure flexible jobs, or to earn more for working rigid hours, indicates a need for active public policy to help with work-family conflicts. As is the case in most other affluent countries, the United States can legislate a minimum number of sick and personal days and vacation time that could be used for unexpected family needs. Vigorously enforced affirmative action can help move women and black workers into more flexible jobs. Since unions help workers get higher pay for rigid work schedules, policies supporting workplace organizing can also help families. Firms can also take a number of measures even in the absence of more family-supportive public policy.

The problems with pay and flexible schedules
The double day, once the nearly exclusive problem of low-income women and women of color, has now spread to the great majority of mothers, and some fathers as well. Nearly three-fourths of mothers with children under 18 now work for pay. Seven million of them do not have spouses to share the work of earning a livelihood and caring for the children. There are also more than 1.5 million single fathers (U.S. Bureau of the Census 2000). Included among the working parents are about 1.6 million who have left welfare since 1993 for jobs (Sawhill 2001). These are mostly people without spouses, working in the lowest-wage jobs, while trying to meet their families’ need for their time and attention. Many thousands of other parents cannot even get jobs or cannot work as many hours as they would like because of competing work and family responsibilities.

Faced with constantly conflicting demands from jobs and families, more workers than ever before are seeking greater flexibility on the job. They are looking for extended periods of family leave, high-quality part-time employment, and ways to handle the many unexpected contingencies of family care, including sick children, no-show babysitters, snow days, and doctors’ appointments for elderly parents. Some jobs provide an important element of short-term flexibility that helps workers fulfill their family responsibilities, through flextime programs, personal days, and sick days. In other cases, workers can make these decisions informally if they have an understanding supervisor. But for many workers, the employer’s control over their work schedule is nearly absolute.

Short-term flexibility is unquestionably crucial to working parents. But it is also important for many other workers and is an important dimension of autonomy on the job. Influence over decisions about when to arrive at or leave work, or when to take an occasional day off, is part of the contested boundary of control between workers and supervisors. Short-term flexibility also helps some workers avoid commuting during peak traffic hours. For others, it creates opportunities for workday leisure, such as spending time outdoors during the daylight hours. There are a multitude of reasons for many people to seek short-term flexibility.

However, economists usually assume that it is women who get more flexible schedules because they are primarily responsible for the care of children. Many economists also assume that this explains part of the gender pay gap, since they believe that mothers are more willing than men to trade off higher wages for flexible scheduling. Called the theory of “compensating wage differentials,” this line of reasoning concludes that if there is a shortage of workers to fill jobs with undesirable characteristics – more physically hazardous, more stressful, less flexible – then the relatively few workers who are willing to take those jobs (supposedly men) will be paid more for accepting them. According to this theory, workers are willing to “pay” various amounts (that is, give up higher wages) for various degrees of flexibility in work schedules. The competitive labor market lets them choose an optimal tradeoff between flexibility and wages. Moreover, the theory of compensating differentials suggests that government interventions such as mandatory sick days and personal days can only make workers, their employers, or both, worse off by limiting workers’ choices and/or by increasing costs to firms.

The compensating differential theory can be questioned on a number of fronts. First, are there market imperfections that undermine workers’ ability to choose their optimal combination of wages and flexibility? Second, who actually gets more flexible schedules and better working conditions? Third, do workers with more rigid schedules actually get higher pay?

Market failures in ability to choose schedules
Some economists and sociologists have long doubted that workers in jobs with less desirable working conditions are paid more than similar workers in more attractive jobs. There are several reasons to expect that markets will not fully compensate workers for poor working conditions. First, workers may have poor information about job characteristics. For example, a worker’s scheduling options may depend on an individual supervisor’s empathy or favoritism; supervisors may come and go; or company policies may change in unpredictable ways. If employees are not really sure about what to expect from their employers, they may not be able to arrange compensation for accepting rigid work hours.

Also, workers often will not complain about rigid schedules or other poor working conditions for fear of losing their jobs or compromising their career prospects. If there is no union protection, every employee will hope that some other employee brings up the problem of work-family conflicts first. That intrepid person would take the heat from management, endure the doubts about her commitment to the company, risk her job or her promotion opportunities, and – somehow, hopefully – resolve the problem. Since every worker faces identical incentives for someone else to raise the issue, ultimately no one is likely to do it. The result is that firms have poor information about worker preferences and are probably unaware of the optimal tradeoff between wages and flexibility (see Dorman (1996) for a more extended discussi

on of theoretical problems behind the theory of compensating differentials).

Sociologists, who tend to pay more attention than economists to the internal hierarchies and networks within the firm, point to another problem with the theory of compensating differentials. They argue that there may indeed be pay premiums for some workers, but usually those workers will be the ones with the greatest power on the job, who privilege themselves with job amenities such as flexibility and greater compensation. Usually this means employees with supervisory or policy-making authority. But in a unionized environment, workers covered by a collective bargaining agreement may negotiate some of these benefits for themselves as well. Thus, those with organizational power get both higher pay and more flexible schedules, contrary to the compensating differential hypothesis (Jacobs and Steinberg 1990, 1995). But for the great majority of workers who have no workplace authority, and no union protection, there is little chance of extracting compensation from their employers for unsafe, dirty, or stressful work, or for inflexible work schedules. The problem is particularly acute for women workers and people of color.

Economists have in recent years begun to pay more attention to authority structures within a firm, through the lens of “principal-agent theory.” Because shareholders in large companies have difficulty monitoring what the managers do, they will not be able to sanction the managers for practices that do not maximize shareholder wealth. So, if the managers give themselves more flexibility to enjoy workday leisure, do not reduce their own compensation for this privilege, and at the same time do not give greater flexibility to their subordinates who might have been willing to trade off wages to some extent, the theory of compensating differentials is clearly violated. In this case, managers would get both higher pay and more flexible schedules.

A final reason that there might not be greater pay for less flexible jobs comes from “efficiency wage theory.” This theory posits that employers cannot costlessly monitor worker effort on the job. Therefore, employers must either spend resources on supervision, or, alternatively, on wage premiums, to induce workers to produce at the level employers desire. Employers will control the pace of work in some jobs, such as production line or clerical jobs, with computerized monitoring, machine pacing, or large numbers of supervisors. But for jobs that are relatively difficult to supervise directly – for example, those which require a lot of independent decision-making, such as managerial and professional jobs – employers will offer wage premiums to motivate these workers to supply the desired level of effort. Hence, for managerial workers, we may see more flexibility (because their jobs are more independent and difficult to supervise) and greater pay, which would offset the compensating differentials (Fairris and Alston 1994; Gariety and Shaffer 2001). Another aspect of efficiency wages is that, while there may be some compensation for bad working conditions, it need not be complete, or even positive (Dorman 1996).

Who gets flexible jobs?
Do women really have more flexible jobs? And among women, do mothers get more flexible hours? A lot of evidence suggests otherwise.

The 1991 Comparative Project in Class Analysis (CPCA) featured two questions about job flexibility (Hout, Wright, and Sanchez-Jankowski 1996).¹ The first asked, “who [usually] decided when the respondent came to and left work.” The second asked, “who [usually] decided whether the respondent could take a day off from work without losing pay or having to claim vacation time, sick leave, or put in compensatory time.” Respondents were given the option to respond “[I] can sometimes decide on my own,” “someone else always decides,” or “something that [I] and another person decide jointly.”2 Thus, one group unambiguously has more job flexibility than another if two conditions hold: they are more likely than the comparison group to say they sometimes decide on their own, and they are also less likely to say that someone else decides.³

Figures A and B show who has more control over coming to and leaving work and taking a day off, by race and sex. In 1991, white men clearly had the most job flexibility. White men were most likely to decide on their own, and least likely to have someone else decide, when they arrived at or left work. Similarly, white men had the greatest control over the decision to take the day off. Black women had more control over these decisions than black men, but blacks generally had much less scheduling flexibility than whites. The racial differences were actually larger than the gender differences.

Do parents get more flexible schedules? Figure C shows that men with children were the most likely to be able to decide when to come to and leave work, but that women with children were the least likely. Single mothers were least often able to decide entirely on their own (although they were also the least likely to have someone else completely control the decision). Childless workers of both sexes also had more control than mothers did over arrival and departure times.

Finally, Figure D shows that men with children were again more likely than women with children to be able to decide when to take the day off. Single mothers unambiguously had less flexibility in this respect.

A more recent dataset allows us to examine a similar question. In 1998, the General Social Survey asked, “Which of the following statements best describes how your working hours are decided? (By working hours we mean here the times you start and finish work, and not the total hours you work per week or month)” (emphasis in the original). Respondents were given these choices: “Starting and finishing times are decided by my employer and I cannot change them on my own,” “I can decide the time I start and finish work, within limits,” and “I am entirely free to decide when I start and finish work.” The dataset was restricted to those aged 18-65 and not self-employed.

Figure E shows the distribution of responses by sex and race. Again, despite the difference in wording and the difference in the date of the survey, black workers have noticeably more rigid work schedules than white workers of the same sex. Black men have more rigid schedules than black women, but the difference between white women and white men is statistically insignificant.

Figure F examines this same survey data by sex and presence of children. Having children in the household only slightly affects the probability of having flexible working hours. What little variation there is by gender and parental status does not support the idea that mothers have the most flexible schedules. And, in this dataset, single mothers by far have the most rigid schedules. Two other recent sources help to dispel the notion that women workers have more flexible working hours. Golden (2001) found that, in 1997, men were more likely than women to respond affirmatively to

the question, “Do you have flexible work hours that allow you to vary or make changes in the time you begin and end work?”4 The gender difference in scheduling flexibility increased with age. Non-white workers were much less likely than whites to have flexible schedules. Heymann (2000) found that women were less likely than men in 1995-96 to be able to decide when to take breaks, change starting and quitting times, and take days off for sick children.

Heymann and Earle (1998) looked at job flexibility for women who were leaving welfare.⁵ They found that the most vulnerable women workers, that is, those with the greatest need for flexibility, typically have the most inflexible jobs. Mothers who had left Aid to Families With Dependent Children (AFDC) were less likely to have paid sick leave, paid vacation time, and flexible schedules than mothers who had never been on welfare. About 70% of women who had never been on AFDC had paid sick leave; about 80% of them had paid vacation time. Only 41% of women who had been on AFDC for at least five years had paid sick leave, and only 59% of them had paid vacation time. Fifty-eight percent of women who had never been on AFDC had flexible schedules; 46% of the formerly long-term welfare recipients had flexible schedules.

Thus, there is no support from any of these datasets for the proposition that women get more short-term flexibility than men. Since most mothers do have primary responsibility for children, they certainly need more flexibility, but they are not as likely to get it as other workers.

Workers who get flexible jobs are, however, more likely to have a lot of authority and autonomy on the job, supporting the sociological approach and the principal-agent theory. Tables 1 and 2 illustrate the relationship between workplace authority and autonomy and job flexibility.⁶ The CPCA survey asked respondents whether they supervised others as a part of their official job responsibilities. Although the gender gap in supervisory authority decreased between 1980 and 1991, 46.9% of men said that they did supervise the work of other employees, or tell them what to do, while only 35.3% of women did so.⁷ If the response was affirmative, the survey asked about specific aspects of authority over others, such as the ability to discipline subordinates and to influence pay and promotion decisions. The survey also asked about policy-making authority (e.g., control of budgets), and about autonomy on the job (e.g., how frequently workers were supervised by someone else).

Each entry in Table 1 answers the question: how many people with or without a certain aspect of job authority or autonomy determine their own times of arrival or departure from work vs. how many people are subject to someone else’s decision? For example, 44.2% of the men who can determine the tasks of others decide on their own; 33.5% have their arrival/departure times determined by someone else; and 22.3% (100% – 44.2% – 33.5%) participate in a joint decision (not shown). Similarly, 19.3% of the men who cannot determine the tasks of others decide their own arrival/departure times; 36.2% have their times determined by someone else (and 54.5% participate in a joint decision). (Rows in table sum to 100% when joint decision-makers are included.)

Table 2 shows how many people with or without job authority can decide to take a day off by themselves vs. how many must get permission from someone else.

On every single dimension reported here except union membership, men with power in the workplace are more likely to have flexible schedules than men who do not have power. For men who have the ability to influence pay and promotion decisions, to discipline subordinates, to determine the size of the workforce, and to participate in policy decisions, the relationship between authority and flexibility is especially strong.

For women, the relationship between authority and flexibility is not as strong. A few of the relationships, such as how often someone checks on their work, are contrary to expectation; for example, women who are never checked on are more likely to have someone else decide their arrival/departure times than women who are checked on frequently. But even for women, there is a clear overall pattern of greater authority associated with greater flexibility.

To sum up so far, the conventional theory does a poor job of predicting who gets flexible schedules. Women, mothers, and single mothers in particular, do not get more flexible jobs. Mothers who have left welfare fare especially poorly. Black workers generally have more rigid schedules than white workers. Those who do get flexible schedules are generally those who have authority and autonomy on the job, especially if they are male. Since men, whites, and people with authority and autonomy also tend to be paid relatively well, the next question that should be posed to the defenders of the compensating differential hypothesis is: do workers with flexible schedules really give up money for this benefit?

Job flexibility and pay
The data on pay and flexibility do not support the view that workers with more rigid schedules are compensated with higher pay. A number of studies have been conducted to investigate this question, on a wide variety of datasets, over a long period of time. Randall Filer, one of the most influential proponents of the compensating differential hypothesis, found that “freedom to take time off,” “satisfactory days,” and “satisfactory hours” were not significantly correlated with pay8 (Filer 1985). Brown (1980) summarized a variety of studies on compensating wage differentials from the 1970s; not one found significant pay differences related to “control of overtime,” “freedom to reduce work hours,” and freedom to “take time off.” Duncan and Holmlund (1983) asked whether “punctuality is important on the job,” whether “the use of a punch clock is required,” and whether “[it is] not possible to run an errand for half an hour without telling supervisor.” Again, the analysis failed to support the compensating differential hypothesis. Duncan and Stafford (1980) found that workers who believed it was “not easy to get a couple of hours off from work” did not have significantly higher pay. Kruse (1992) found that workers who punched time clocks had lower wages, but the difference was not significant. Workers who could leave work for a half hour without informing their supervisors and workers who had flexible hours enjoyed significantly higher pay (Kruse 1992).

The 1991 CPCA provides a unique opportunity to examine the relationship between job flexibility, authority, autonomy, and pay. Because the survey’s designers were interested in authority and autonomy on the job as components of an empirical analysis of class structure, they collected many unusual variables on job flexibility and workplace authority, as well as data on the usual earnings and human capital variables.

According to this survey, hourly wages were indeed higher when workers had more flexible schedules.⁹ Figure G illustrates what happens to the pay premium for flexible work and for gender when taking into consideration various factors.¹⁰ The baseline estimate controls only for education, age, job tenure, job training, region of the country, full-time or part-time work, government/nonprofit employment, marital status, presence of children in the household, sex, and race. Workers who are

identical with respect to all of these characteristics but who differ with respect to job flexibility are represented by the black bars in each group.¹¹ Workers who can usually decide on their own when to arrive at or to leave work earn 16.6% more than workers who have no control over this decision, and workers who make the decision jointly with someone else make 5.1% more than workers who have no control. Being able on one’s own to decide to take the day off is associated with 9.3% higher pay, and making a shared decision about taking the day off is associated with 4.4% higher pay.

Figure G also shows what happens to the pay premium for flexible work when additional controls are added. For example, the figure shows what happens when industry and occupation are controlled for, in addition to the variables in the baseline model. The effect of being able to decide one’s own arrival and departure times (or to decide jointly) falls when industry and occupation are controlled for (the black bars are shorter than the dark grey bars).¹² However, the effect of being able to decide to take the day off is generally unaffected by industry and occupation.

It is not exactly clear why industry and occupation should make such a difference for the pay premium associated with control over arrival and departure times. The answer might be found in the efficiency wage hypothesis. If jobs in certain industries and occupations are more difficult to monitor because of independent and flexible work, they will offer higher wages in order to induce worker effort (e.g., see Krueger and Summers 1988). 13 It could also be that some workers have access to industries and occupations with organizational power that is not measured by the variables in this study, and these jobs also permit flexibility.

Another recent study has also found that scheduling flexibility is associated with higher pay, even after controlling for experience, education, occupation, industry, union membership, and a variety of other characteristics. Gariety and Shaffer (2001) found that women workers with flextime earned 6-7% more than women without flextime. This was true for both 1989 and 1997. Male workers who had flextime earned about 6% more than other men, but this result appeared in the 1997 data only.

Do the higher wages associated with flexible work have anything to do with organizational power? Figure G also shows what happens when controls are added for authority and autonomy (the variables that are featured in Tables 1 and 2), but not industry and occupation. A sizeable part of the wage premium for flexible work is due to flexibility being associated with organizational power. The premium for deciding on one’s own arrival and departure times falls from 16.6% (dark grey bar) to 10.5% (white bar). This effect of authority and autonomy appears to some extent for all the flexibility variables.

Controlling for all of the baseline variables, in addition to industry, occupation, authority, and autonomy yields a final set of estimates (grey bars). The pay premium for flexibility in arrival and departure times has diminished greatly from the baseline estimates, showing that much of the pay premium is associated with industry, occupation, or organizational power. Nevertheless, a substantial premium remains: 8.3% higher pay for those who can usually decide arrival and departure times on their own. The pay premium for being able to decide to take the day off is not as strongly affected by the addition of industry, occupation, authority, and autonomy; it remains at 7.7% for those who decide on their own.¹⁴

Finally, we can ask to what extent is the gender wage differential related to flexible schedules? In the baseline estimate, single men earn 19.8% more than women with identical measured characteristics; married men earn 9.2% more than that. While the pay of single men was not dramatically affected by the addition of the various controls in Figure G, the pay of married men was. Much of the married male wage advantage is associated with the kinds of jobs they have: industry, occupation, and organizational power.

At this point, it is also useful to ask whether there are distinct flexible and inflexible sectors of the labor market, with different pay structures and different kinds of workers in them.¹⁵ The last part of the analysis addresses this question. A worker was identified with the flexible sector if she was usually able to decide – on her own – when to come to or leave work, or when to take a day off (or both). All other workers were assigned to the inflexible sector.

Table 3 shows the results of an analysis of pay differentials in these two sectors.A’>’ sign indicates that a worker with the characteristic in question would earn more in the flexible sector than she could in the inflexible sector; a ‘

First, unions are important. Union workers are somewhat more likely to have flexible jobs, and they have much larger wage gains in the inflexible sector.¹⁶ Union workers in the flexible sector earned 6.3% more than nonunion workers, but union workers in the inflexible sector earned 21.3% more. There are a couple of possible explanations for the wage difference. First, the conventional explanation would be that workers in rigid jobs get compensating differentials, and for some reason unions are more successful at organizing workers with rigid schedules, perhaps because these workers are more dissatisfied with their jobs. However, two problems confront this explanation. The first problem is that, if people in rigid jobs are fully compensated for the disutility of such jobs, then why are they more likely to want unions? The second problem is that this analysis found that union members were more likely to be in flexible jobs (although we don’t know if this is because the jobs were more flexible to begin with, or if it is because unions negotiate to get flexibility for their members). A second possible reason for higher union wages in the inflexible sector might be that unions bargain more aggressively for workers with little control over their schedules, ensuring that they get more flexible schedules (e.g., through such contractual provisions as personal days), or at least receive compensating differentials for the disutility of rigid work.¹⁷ This is exactly as one would expect if organizational power determines who gets compensating differentials.

A second interesting result from the analysis of sectors is that black workers earn a lot less (21% less) in the inflexible sector, but that racial wage differences are insignificant in the flexible sector. There are also evidently dramatic differences in employment opportunities available to black workers between the two sectors. Employers who pay black workers less may also believe blacks need to be intensively monitored and tightly controlled. There is other support from several datasets for this idea (McCrate 2001b). According to the CPCA data, not only do black workers have more rigid schedules than whites (Table 1), but they are also supervised more intensively than whites. According to Table 4, blacks do work that is easier to check on, and, in fact, they are checked on more often. This is easily understood when employers’ beliefs about black workers are taken into account. Studies in the early 1990s that utilized interviews of Chicago employers seeking to fill unskilled, entry-level positions found that the “employ

ers view[ed] inner-city workers, especially black men, as unstable, uncooperative, dishonest, and uneducated.” They characterized black workers as having a “bad work ethic,” creating tensions with co-workers, being “lazy and unreliable,” and having “a bad attitude” (Kirschenman and Neckerman 1991, 204, 213). Similarly, in their interviews of employers of entry-level workers in four major U.S. cities, Moss and Tilly (2001) found that by far the greatest complaint about black workers was that they have “lagging motivation”; employers agreed with that statement more often than “blacks have lagging hard skills” or “blacks have lagging interaction skills.” Employers who think like this are likely to be unyielding about when black workers can come to or leave work, or take a day off.

Third, the analysis of sectors shows which occupations and industries offer higher wages by sector. Looking at occupations first, managers, professional and technical workers, and sales workers earn more in the flexible sector. Workers in farming, forestry, fishing, laborer, and operative jobs earn less than other workers in both sectors, but they do not do quite as badly in the flexible sector. Only service workers and craft workers do better in the inflexible sector. Looking at industries, all workers earn more in the flexible sector, but the difference is not always statistically significant.

Finally, the gender pay gap is about the same in the flexible and inflexible sectors. Single men make a little more in the flexible sector, but married men make a little less. The gender pay gap remains about 20% after controlling for all the baseline variables (occupation, industry, authority, and autonomy) and after taking into consideration differences in pay between the flexible and inflexible sectors. The entire analysis of flexible schedules and pay did almost nothing to explain the gender wage gap, contrary to the expectations of the compensating differentials theory.

Policy directions
The theory of compensating wage differentials predicts that workers will be fully compensated for the inconvenience of jobs with rigid schedules in perfectly competitive labor markets. Workers who prefer flexible jobs (believed to be women) will get them, but they will have to sacrifice some income. Other workers who prefer more income (believed to be men) will sacrifice flexibility. According to this theory, there is a clear tradeoff between income and flexibility, but workers are able to strike a balance that works for them. In this world, policy interventions such as statutory sick days or union-negotiated personal days would make workers worse off (by restricting their choices), would make employers worse off (by raising costs), or both.

But in a world where shareholders cannot easily keep an eye on what kind of opportunistic behavior their managers might be indulging in, and in a world of efficiency wages, the people with flexible jobs could also easily earn higher wages, which is exactly what this analysis finds. Conversely, people in rigid jobs tend to earn less. Part of the wage advantage for people with flexible jobs is due to occupation and industry, but part of it also is due to the authority and autonomy they exercise on the job.

One of the most important findings in this study is that women typically do not have more flexible jobs, despite the rhetoric surrounding flexibility and the needs of working mothers. Single mothers have significantly more rigid work schedules. Not surprisingly, the gender pay gap is also not much affected by who has flexible jobs. Another important finding is that there are clear racial differences in flexibility, and black workers in the inflexible sector experience a larger racial wage deficit than black workers in the flexible sector.

Since the world actually looks quite different from the one suggested by competitive market and compensating wage differential theories, policy intervention by government and by unions is in order. Policy can seek both to increase the number of flexible jobs and to really compensate workers in those jobs where flexibility is more difficult to implement.

A good place to start would be a statutory minimum number of sick days, personal days, and vacation time. These are all ways that family members try to deal with unexpected problems associated with child care or elder care. The United States is alone among affluent countries in not requiring a minimum number of sick days and vacation days. Some countries additionally require days off specifically for the care of sick family members.

Firms, however, do not necessarily have to wait for more enlightened public policy. Drago et al. (2001) found in a small sample of teachers that there was widespread willingness to pay for one week per year of family care leave, through modest payroll reductions, both by employees with families and those without.

There is also some evidence about the relationship between firms’ competitive strategies and work-family programs. Osterman (1995) found that about 40% of American establishments now offer flexible hours for their core workers, and another 15% either have plans to do so or are considering it. The implementation of such flexible schedules is on the rise partly to attract and retain qualified employees; but partly to reduce stress, absenteeism, and other problems that impair productivity. Osterman also found that firms seeking to implement high-commitment work systems, incorporating employee involvement and quality programs, are more likely to offer some type of work-family program (including flexibility).

Cafeteria benefit plans could also help. People who value flexibility more highly can forego something else (e.g., health insurance, if they are covered by their partner’s plan). However, cafeteria plans must be structured so as not to reduce the overall level of benefits (which would address much of the union skepticism about these plans).

Finally, important labor market institutions, such as unions and equal opportunity law, have a role to play. Since unions are an important mechanism for workers to receive compensation for rigid work schedules, support for union organizing is important, especially in occupations and industries that tend to have inflexible jobs. Since blacks and women are less likely to have flexible jobs, equal opportunity and affirmative action enforcement are also needed to ensure that these workers have equal access to the kinds of jobs that already feature flexibility, or which offer compensation for inflexibility.

-May 2002

The author acknowledges ICPSR for providing the data for this analysis.

Endnotes
1. In 1991 a team of sociologists fielded the CPCA survey in the U.S., Russia, and the Komi Autonomous Soviet Socialist Republic. In the U.S., it sampled people who spoke English and resided in the 48 contiguous states. For the purposes of this study, the universe was further restricted to currently employed persons, aged 18-65, who were not self-employed. The data file provided sample weights.

2. It is apparently important to be specific in the question about who is making the decisions about job scheduling. Both the CPCA and the Current Population Survey (CPS) are very clear about this. Questions on flexibility such as those in the NLSY79 (National Longitudinal Survey of Youth) are less clear – simply asking does your job offer flexible hours – and seem to distinguish between race and gender groups less clearly.

3. In the interest of brevity and clarity, I suppress the percentages for “joint decision.” The percentages for “myself,” “someone else,” and “joint decision” always sum to 100%.

4. CPS (1997).

5. NLSY (1993).

6. A

uthority variables were recoded as follows. The questionnaire first asked a screener, “As an official part of your main job, do you supervise the work of other employees or tell other employees what to do?” People who responded “yes” were asked the series of questions whether they were “directly responsible for” deciding tasks, tools, and pace of work; whether they influenced pay or promotion of their subordinates; and whether they could discipline a subordinate. All people who responded “no” to the screener were coded as zero (“no”) on the subsequent questions.

Also, all respondents were asked about the frequency of someone checking on their work, with six possible responses. For brevity, only the two extreme categories are reported (“never” and “more than once a day”). All respondents were also asked whether they were “personally involved” in policy decisions concerning size of workforce, products/programs/services, methods of production, and budgets. There were two possible ways of participating in these decisions: people who “participated directly” were coded as one (“yes”) and people who just “gave advice” were coded as zero (“no”).

7. McCrate (2000) found much higher proportions of the workforce in supervisory positions than the conventional census occupational measures. Gordon (1996) interpreted this as a limitation of the census data, not the CPCA data, meaning that many people who do not have supervisory occupational codes do in fact have supervisory responsibilities.

8. However, this could be a result of collinearity with the many other job characteristics he included in the regression.

9. The dependent variable in this regression analysis, the log hourly wage, was constructed from usual monthly earnings, divided by usual weekly hours times four.

10. The effects of flexibility were significant when workers were able to decide on their own, but insignificant when they made the decision jointly with someone else.

11. Workers on flexible schedules work many hours of overtime (Golden 2000). I tried several methods to see if there were compensating differentials for overtime and whether these might be related to the flexibility coefficient. The results did not suggest this, but the sample may have been too small to pick up this effect.

12. This means that part of the apparent effect of flexibility is just because of its association with industry and occupation. So, for example, if managers in banking services have more flexible jobs, and earn more money than other workers, then part of the apparent pay premium for flexibility is actually due to the kind of job they hold.

13. In general, though, it is difficult to disentangle the independent effects of various industry characteristics (Dickens and Katz 1987).

14. For people who could decide on their own when to come to and leave work, or to take a day off (or both), the pay premium was 4.6%. For people who could decide both of these questions on their own, the pay premium was 17.7%.

15. This involves a statistical technique that also controls for unobservable characteristics associated with non-random selection into the two different sectors (McCrate 2001).

16. The finding that union members are more likely to have flexible jobs was taken from a structural probit analysis. It partly reverses the association between unions and flexible jobs in Tables 1 and 2. Other analyses of flexible work have found either modest support for the idea that union workers have more flexible schedules (Golden 2001) or no support (Osterman 1995).

17. This would be similar to an analysis that found that unionized workers are more likely to get compensating differentials for hazardous work (Dorman and Hagstrom 1998). It also is similar to an analysis that found that union differentials are larger for shift workers (Kostiuk 1990).

References
Brown, Charles. 1980. “Equalizing Differences in the Labor Market.” Quarterly Journal of Economics 94(1): 115-34.

Dickens, William T., and Lawrence F. Katz. 1987. “Inter-Industry Wage Differences and Industry Characteristics.” In Kevin Lang and Jonathan S. Leonard, eds. Unemployment and the Structure of Labor Markets. Oxford, England: Basil Blackwell.

Dorman, Peter, and Paul Hagstrom. 1998. “Wage Compensation for Dangerous Work Revisited.” Industrial and Labor Relations Review 52(1): 116-35.

Dorman, Peter. 1996. Markets and Mortality: Economics, Dangerous Work, and the Value of Human Life. Cambridge, England: Cambridge University Press.

Drago, Robert, David Costanza, Robert Caplan, Tanya Brubaker, Darnell Cloud, Naomi Harris, Russell Kashian, and T. Lynn Riggs. 2001. “The Willingness-to-Pay for Work/Family Policies: A Study of Teachers.” Industrial and Labor Relations Review 55(1): 22-41.

Duncan, Greg J., and Bertil Holmlund. 1983. “Was Adam Smith Right After All? Another Test of the Theory of Compensating Wage Differentials.” Journal of Labor Economics 1(4): 366-79.

Duncan, Greg J., and Frank P. Stafford. 1980. “Do Union Members Receive Compensating Wage Differentials?” American Economic Review 70(3): 355-71.

Fairris, David, and Lee J. Alston. 1994. ” Wages and the Intensity of Labor Effort: Efficiency Wages Versus Compensating Payments.” Southern Economic Journal 61(1): 149-60.

Filer, Randall K. 1985. “Male-Female Wage Differences: The Importance of Compensating Differentials.” Industrial and Labor Relations Review 38(3): 426-37.

Gariety, Bonnie Sue, and Sherrill Shaffer. 2001. “Wage Differentials Associated With Flextime.” Monthly Labor Review 124(3): 68-75.

Golden, Lonnie. 2000. The Time Bandit: What U.S. Workers Surrender to Get Greater Flexibility in Work Schedules. Washington, D.C.: Economic Policy Institute.

Golden, Lonnie. 2001. “Flexible Work Schedules.” Monthly Labor Review 124(3): 50-67.

Heymann, Jody. 2000. The Widening Gap: Why America’s Working Families Are in Jeopardy and What Can Be Done About It. New York: Basic Books.

Heymann, S. Jody, and Alison Earle. 1998. “The Work-Family Balance: What Hurdles Are Parents Leaving Welfare Likely to Confront?” Journal of Policy Analysis and Management 17(2): 313-21.

Hout, Michael, Erik Olin Wright, and Martin Sanchez-Jankowski. 1996. Comparative Project in Class Analysis: United States and Russia, 1990-1992 [computer file]. ICPSR version. Berkeley, Calif.: Michael Hout and Michael Burawoy, University of California; Syktyvkar, Komi A.S.S.R.: Pavel Krotov; Moscow, Russia: Valery Mansurov, Russian Academy of Sciences [producers], 1995. Ann Arbor, Mich.: Inter-University Consortium for Political and Social Research.

Krueger, Alan B., and Lawrence H. Summers. 1988. “Efficiency Wages and the Inter-Industry Wage Structure.” Econometrica 56(2): 259-93.

Kruse, Douglas. 1992. “Supervision, Working Conditions, and the Employer Size-Wage Effect.” Industrial Relations 31(2): 229-49.

McCrate, Elaine. 2001a. “Fl

exible Hours, Workplace Authority, and Compensating Wage Differentials.” Working paper. Burlington, Vt.: University of Vermont.

McCrate, Elaine. 2001b. “The Racial Gap in Autonomy at Work.” Working paper. Burlington, Vt.: University of Vermont.

McCrate, 2000. “The Growing Class Divide Among American Women.” In Robert Pollin, ed., Capitalism, Socialism, and Radical Political Economy: Essays in Honor of Howard J. Sherman. Cheltenham, U.K. and Northampton, Mass.: Edward Elgar.

Moss, Philip, and Chris Tilly. 2001. Stories Employers Tell: Race, Skill, and Hiring in America. New York: Russell Sage Foundation.

Neckerman, Kathryn M., and Joleen Kirschenman. 1991. “Hiring Strategies, Racial Bias, and Inner-City Workers.” Social Problems 38(4): 433-47.

Osterman, Paul. 1995. “Work/Family Programs and the Employment Relationship.” Administrative Science Quarterly 40: 681-700.

Sawhill, Isabel. 2001. “From Welfare to Work.” Brookings Review 19(3): 4-7.
U.S. Bureau of the Census. 2000. Statistical Abstract of the United States. Washington, D.C.: U.S. Government Printing Office.

---

For a printer-friendly version of this report, click here.