Through a qualitative study of a software development group, I examine how managers control the hours employees work, and therefore the temporal boundary between employees' work and life outside of work. Analysis of field data shows that managers use three types of techniques to exert boundary control over "knowledge workers": (1) imposing demands, by setting meetings, reviews, and internal deadlines, controlling vacations, and requesting extra work; (2) monitoring employees, by standing over them, checking up on them, and observing them; and (3) modeling the behavior they want employees to exhibit. Employees either accept or resist managers' boundary control; those who resist are penalized by the reward system, even when they devise creative ways to schedule and complete their work. Many employees are married, and the demands of their work have consequences for their spouses. Spouses' reactions to the demands that ultimately affect them further influence how employees respond to boundary control. These findings contribute to a theory of boundary control and carry practical implications for resolving work-family conflicts in our society.

The industrial revolution marked a fundamental change in the separation of work and family life. On the farm, families worked together from dawn until dusk, intermingling work and family responsibilities, subject to the particular demands of the day. Similarly, although work and family constituted distinct spheres of life, skilled artisans did not have an externally defined, rigid temporal framework that determined when the responsibilities of either sphere had to be handled (Rock, 1988). For both farmers and artisans, tasks, rather than the clock or cultural pressures, determined the length of their work days. As workers entered factories, they faced external control not only over their activities at work but also over how much time they spent working (Owen, 1979; Landes, 1983). Employees' work schedules became regulated by the technical control exerted by the production process (Bendix, 1956). The later development of bureaucracies led to control embedded in social relations and social structure (Etzioni, 1961). A system of rules emerged governing when employees were to be at work and what types of absences were excused.

More recently, it has become difficult to design jobs as a series of explicit tasks to be performed, with appropriate incentives to ensure adequate output from qualified employees. We are told with increasing frequency that organizational culture, built from underlying values and beliefs about what is important, valued, and rewarded within an organization, assumes and carries crucial control functions (Ouchi; 1980; Wilkins and Ouchi, 1983; Van Maanen and Kunda, 1989). This seems especially true in so-called knowledge organizations, where the work to be performed is of an open-ended, creative, individually styled, and highly demanding sort that cannot be standardized or fully planned out in advance (Bell, 1973). In such work settings, attempts are made to elicit and direct the required efforts of members by controlling the experiences, thoughts, and feelings that guide their actions (Hochschild, 1983; Van Maanen and Kunda, 1989). The intent is for workers to be driven by internal commitment, strong identification with company goals, and intrinsic satisfaction from work (Kunda, 1992). This type of control compels employees not only to do what is expected at work but to conform to norms at work that determine how they lead their lives both in and out of the workplace.

Each of these types of control involves managers governing the temporal boundary between employees' work and their lives outside of work. I refer to the various ways in which managers in organizations cajole, encourage, coerce, or otherwise influence the amount of time employees physically spend in the workplace as "boundary control." As I define this term, boundary control refers to managers' ability to affect how employees divide their time between their work and non-work spheres of life.

Forms of Boundary Control

The nature of control itself has varied with the social class of workers as much as with the dominant structure of work. In a study of department stores, Ouchi and Maguire (1975) found that people at lower hierarchical levels experience more personal surveillance, or "behavior control," whereas people at higher hierarchical levels experience more measurement of outputs, or "output control." Moreover, the overall amount of control ("behavior" and "output" control combined) people experience decreases as they move up the hierarchy. One can also safely assume, however, that increased work hours are expected of those higher in the hierarchy. While people at the bottom of organizational hierarchies are more closely monitored at work, their time is reckoned by the clock rather than by their activities (Thompson, 1967; Clark, 1985); the length of their work day is fixed 'rather than driven by demands of the job. As a result, people at the bottom of the hierarchy surrender control over when they work and what they do at work, exchanging control for predictability in their work lives.

Although many studies have examined employees' loss of control over their time at work as they moved from the farm to the factory and later to the office (Edwards, 1979; Owen, 1979; Barley and Kunda, 1992), an important but often overlooked effect of these changes for people low in occupational hierarchies was the significant increase in predictability that they experienced over the temporal boundary between work and life outside of work. As Zerubaval (1981: 166) has asserted, "The very same institutions that are directly responsible for much of the rigidification of our life--namely the schedule and the calendar--can also be seen as being among the foremost liberators of the modern individual."

People higher in the hierarchy, i.e. managers, professionals, and technical workers, did not experience the same degree of loss of control over what they did at work nor the same increased predictability in their work schedules. In his description of work life in the 1950s, Whyte (1956) labeled senior executives "non-well-rounded men," to characterize his finding that senior executives typically worked nine and a half hours in the office on weekdays, four out of five weeknights, and part of most weekends and therefore had no time for anything else. Kanter (1977: 65), like Whyte, documented the blurring of the boundary between work and life outside of work for people at the top of organizational hierarchies: "Question.' How does the organization know managers are doing their jobs and that they are making the best possible decisions? Answer: Because they are spending every moment at it and thus working to the limits of human possibility. Question: When has a manager finished the job? Answer: Never. Or at least, hardly ever. There is always something more that could be done."

Boundary control can be represented as a spectrum ranging from the imposition of constrained work schedules characteristic of blue-collar work to the all-encompassing schedules characteristic of senior management. In this paper, I focus on the boundary control exerted on the higher echelons, which results in long and often unpredictable hours of work. While this form of boundary control has always been familiar to senior executives, its prevalence is growing rapidly as we approach the twenty-first century.

Boundary Control and Knowledge Work

Since the 1950s, a demand for increasingly complex, analytic, and even abstract work has shifted the division of labor in the United States away from blue-collar work toward technical and professional work (Bell, 1973; Stehr, 1994). The number of professional and technical jobs has increased 300 percent since 1950 (Barley and Orr, 1997). A quarter of all new jobs currently being created in the United States are either professional or technical (Silvestri and Lukasiewicz, 1991). According to management scholars, these new jobs result from a transition from an industrial era to a knowledge era. Savage (1990) claimed that networking enterprises are replacing steep hierarchies, complex, unpredictable work tasks are displacing routines, and functions performed concurrently and more wholistically have eclipsed sequential processing. Some of the names for this new organizational form include the post-industrial firm (Bell, 1973), the information-based organization (Zuboff, 1988), the human networking enterprise (Savage, 1990), and the high-involvement organization (Lawler, 1986). However it is labeled, the new firm is said to process information and knowledge, in contrast to the industrial enterprise, which fabricated goods. Using General Motors to illustrate the resulting shift in expectations, Bennis (1985: vii) wrote: "It used to be that the old fashioned GM philosophy of management could be summed up by this phrase: 'DON'T THINK, DUMMY--DO WHAT YOU'RE TOLD.' Now .... there is a new and very different credo which goes: 'THINK! I'M NOT GOING TO TELL YOU WHAT TO DO.' "Jobs therefore demand workers with capacities for planning, judgment, collaboration, and analyzing complex systems (Dertouzos et al., 1989).

Scholars claim that the people--knowledge workers--who fill these jobs cannot be managed like blue-collar workers. Dertouzos et al. (1989) warned that compulsion will not effectively generate the commitment, responsibility, and knowledge necessary to be effective in highly cerebral organizations. Lawler (1986) and Walton (1985) have advocated, instead, for a new high-involvement management, which gives knowledge workers broader responsibilities, encourages them to contribute, and helps them to derive satisfaction from their work. Drucker (1989) suggested that these employees must be treated as specialists who direct and discipline themselves. Handy (1989) therefore encouraged managers to reinforce, motivate, teach, and counsel knowledge workers, while giving them room to make mistakes. Savage (1990) holds managers responsible for creating an environment of trust and openness.

At the same time, managers are held accountable by senior executives for the performance of the knowledge workers under their command. Because managers cannot easily or directly measure the work output or involvement of knowledge workers, they turn to work hours as an indicator of both productivity and commitment. Moreover, managers recognize that knowledge work is both interdependent and open-ended and that those they manage often need each other to complete their work on time. Managers therefore assume it is best for everyone to be present as much of the time as possible and judge knowledge workers accordingly. As a result, the managerially valued knowledge worker in today's world--a world that demands responsiveness, adaptability, flexibility, and creativity in responding to global markets and to customers--demonstrates total devotion to work. The grueling schedules that used to be typical only of top corporate management and self-employed people are becoming common in one occupation after another. Corporate lawyers, investment bankers, computer programmers, and many other professionals are now expected to work seventy- or eighty-hour weeks routinely, with extra effort during particularly hectic times (Kidder, 1981; Schor, 1991 ).

Current workplace structures, practices, and expectations surrounding knowledge work are based on the notion that employees are willing and able to make work their priority over and above their family, community, or other concerns in their private lives (Christopherson, 1991; Bailyn, 1993). People are consumed by what Coser (1974) referred to as the "greedy institution." This notion is clearly problematic for anyone who has responsibilities outside of work. And, as the demographics of the work force continue to change, more and more employees have such outside commitments. In 1960, for example, 61 percent of married couples had a relationship in which the husband worked and the wife was a full-time homemaker; in contrast, by 1990 this number had dropped to 25 percent of married couples. During the same period, the number of couples in which both spouses worked increased from 28 percent to 54 percent (Hayghe and Cromartie, 1991). Moreover, the number of working women in managerial and professional occupations increased dramatically. From 1972 to 1990, the percentage of women working in executive, administrative, and managerial occupations grew from 20 percent to 40 percent; the proportion of women in professional specialty occupations rose from 44 percent to 51 percent. In particular, the percentage of physicians increased from 10 percent to 19 percent; lawyers from 4 percent to 21 percent, and computer programmers from 20 percent to 26 percent (Hayghe and Cromartie, 1991). Thus, women are not only entering the labor force, creating dual-earner couples, but large numbers of women are entering the professional and managerial ranks and therefore must confront the long and unpredictable hours seemingly inherent in these positions. Furthermore, most women who work outside the home have primary responsibility for the "second shift," as Hochschild (1989) referred to child care and household chores. The demands on their time are therefore exceedingly high. Some men share the responsibilities at home and also suffer from the burden of trying to meet the demands of both work and family. Even those men in dual-earner couples who do not share equally in the responsibilities at home experience much marital stress and demands to help at home (Hochschild, 1989).

In the remainder of the paper, I explore in-depth one high-technology organization's use of boundary control and engineers and their spouses' responses to it. Much research on engineers has focused on the degree to which engineers have autonomy at work (Zussman, 1985; Whalley, 1986; Meiskins and Watson, 1989). I focus, instead, on engineers' lack of autonomy across the boundary between their work and life outside of work.

Research Site

I studied a product development team at Ditto (a pseudonym), a large, high-tech corporation.(1.) I chose Ditto because of its reputation as a leading-edge firm in terms of awareness of employees' work-family concerns. A corporation concerned with these issues provided the potential opportunity to observe a wide spectrum of accepted work-family arrangements.

Ditto employs 100,000 people worldwide, 3,000 of them at the site I studied, which is Ditto's primary site for design and manufacturing. The team I studied was developing PEARL (a fictitious product name), a color laser printer positioned to sell for $10,000. Prior to PEARL, this team made much larger electronic machines that sold for closer to $100,000. It was hoped that PEARL would not only prove profitable but would also position Ditto in this new market. There were plans to follow PEARL with a whole product family.

The PEARL product development team consisted of 45 people. The product manager reported to one of Ditto's seven division vice presidents. In turn, eight managers, including a software manager, reported to the product manager. I focused my data collection on the seventeen members of the software group--the software manager, three project team leaders, an individual contributor, and twelve software engineers--and their three senior managers: the product manager, the senior software manager, and the division vice president. The status of these twenty people is hierarchically organized as follows: the twelve software engineers report to the three project team leaders. In turn, the three project team leaders and the individual contributor report to the software manager. He, in turn, reports to both the product manager and the senior software manager, both of whom report to the division vice president. Figure 1 shows this reporting structure. In this paper, I refer to everyone above the software manager as "senior managers." I refer to the software manager and the three project team leaders as managers. I refer to the software manager and everyone below him as part of the "software group."

Figure 1. Organization chart of Ditto division.

Table 1 shows the demographics of the software group and its senior managers. The software group consists of four women and thirteen men. All of the women are married with children. Eight of the men are married, and seven of them have at least one child. The average age in the software group is 32 years old. The seventeen members of the software group are highly experienced. They have been at the company for an average of seven years. The shortest period of time that any one of them has been at the company is three and one-half years, and all three engineers who arrived at that time came from other software development jobs. Furthermore, over half the group has Masters in computer science or related fields, and four were pursuing Master's degrees part-time at the time of the study.

The members of the software group have a high degree of control over their time at work. They are free to come and go as they please. They do not have to report to anyone if they leave for a few hours to attend to personal issues such as automobile maintenance, a dentist or doctor's appointment, or a child's Little League game. If they are absent for a whole day, they merely have to inform a more senior member of the group. There is no set limit on the number of personal days that individuals may take. Rather, on an ad hoc basis, absence from the office for accommodation to demands outside of work is tolerated.

Beyond what I refer to as "ad hoc flexibility," the members of the software group also have a high degree of operational autonomy, although not strategic autonomy. Bailyn (1985) differentiated between operational autonomy--freedom to attack a designated problem by means determined by oneself within given organizational constraints--and strategic autonomy--freedom to set one's own research agenda. The twelve engineers each have a list of deliverables for which he or she is personally responsible, and each of the three project team leaders is further responsible for the output and integration of the work of the four engineers he or she manages. Engineers and project team leaders, therefore, are provided with a clear sense of what they have to do (low strategic autonomy) but have freedom in how they accomplish these goals (high operational autonomy).

I did not choose to study this software group because it is typical in terms of the autonomy its members have at work. Rather, managers of the division chose the group for me, based on my desire to study a small group of engineers with a large number of seasoned employees in a mix of family situations. I expected to find variability among successful employees in terms of how they managed work and life outside of work, especially given the apparent ad-hoc flexibility and operational autonomy. Instead, I found a group of engineers all under enormous pressure and with little choice about how they allocated time between their work and their lives outside of work.

Data Sources

My field work at Ditto took place over a complete product development cycle. My aim was to develop an understanding of how a small group of individuals manage demands at work and outside of work. I collected data from participant observation, interviews, "shadowing" employees, yearly performance evaluations, and interviews with employees' spouses.

Participant observations. I observed the software group for nine months, from the date funding was committed to PEARL in September until PEARL launched in June. I spent an average of four days a week on site observing members of the software group at work in their cubicles, in the labs, in meetings, and in hallway conversations. I also engaged in social activities with the group members: I regularly ate lunch with them, attended many company parties, joined in several "happy hours" at a local bar, and traveled with them on a two-day bus trip to New York City to take part in the unveiling of their product. When I was on site, I typed field notes throughout the day (as time permitted) and for several hours each night. I recorded all observations and reactions to the day's events.

Interviews. I engaged each of the seventeen members of the software group in an interview lasting one to two hours. These interviews provided background information about group members and allowed me to gain an understanding of their perceptions about the demands they face at work. I conducted an additional fifteen interviews with other members of the division. These interviews served a dual purpose: as with the interviews of the members of the software group, I asked questions about these individuals' backgrounds and about their perceptions of their own work demands. I also asked questions about these individuals' relationships with members of the software group and their perceptions of whether the expectations of the software group are similar to or different from the rest of the members of the division. The people interviewed included the product manager and the division vice president. I also interviewed three of the eight direct reports to the product manager (in addition to the software manager). I further interviewed two mechanical engineers and three system engineers on the product team. Finally, I interviewed five software engineers who were in the division but did not work on the PEARL product team.

Shadowing. I "shadowed" all seventeen members of the software group for at least half a day. I observed everything the individual did, and I wrote down each activity as it occurred. Shadowing group members provided me with an in-depth understanding of how these people spend their time at work. I shadowed one member of the team for three days, five members for one day, and eleven members for half a day.

Performance evaluations. At the end of each year, Ditto senior managers decide the categories of possible raises for that year and the percentage of employees who should be placed in each category. The year I was on site, Ditto senior managers determined that the top 10 percent of employees would receive a 6-percent raise, the middle 70 percent would receive a 3-percent raise, and the bottom 20 percent would receive no raise. The division managers then decided how to put their employees into these categories. They first ranked their employees and then determined each individual's raise. I had access to this information. Managers are subject to a different system of reward and recognition. The year I was at Ditto, it was decided that due to the company's poor earnings that year, managers would not receive a raise. The top 10 percent of managers, however, would receive a one-time, 10-percent, lump-sum bonus. I had access to information on whether each of the managers I studied received this bonus.

Family interviews. Since I wished to understand spouses' responses to the repercussions of work demands on them as well as the effects of their responses on the employees to whom they were married, I requested to go to the homes of each of the married group members and interview their spouses. Most of the engineers met my request with enthusiasm. Of the twelve software engineers I studied, seven are married, and five of the seven allowed me to interview their spouses. Furthermore, the three project team leaders, the independent contributor, and the software manager are married, and all five of them welcomed me into their homes to interview their spouses. I also visited the homes and interviewed the spouses of the product manager and the division vice president.

I made a total of twelve home visits. Each visit lasted from two to six hours. In the case of the male Ditto employees, I visited all of their homes, except one, in the afternoon and interviewed their wives before the employee came home. In two of those cases, I joined the families for dinner after the interview. In the case of the four female Ditto employees, all of their husbands work full-time, and I therefore went home with the women after work, took part in picking up their children from day care, preparing and eating dinner, and putting their children to bed. I interviewed these spouses while the women were home but conducted the interviews in another room. In the case of the seven homes I visited where both spouses were present at the end of the interview, I also engaged in three-way discussions with the couple about work-family trade-offs they were each making. These discussions gave me (and them) an opportunity to hear and respond to each other's opinions.

Analyses

Following the guidelines suggested by Glaser and Strauss (1967) and Miles and Huberman (1984), I developed empirically grounded sets of categories related to work demands (how these demands are made, the implications on individuals' lives outside of work, the consequences for individuals of not responding, etc.). In doing so, I followed an iterative process, fitting observations to categories. As new observations arose, I continually made choices as to whether to retain, revise, or discard particular categories or category sets. Periodic analysis throughout the data collection process helped sharpen questions, focus interviews and observations, and ground evolving theory. At the conclusion of field work, I analyzed transcripts and field notes to address the following three questions: (1) How is boundary control maintained by managers? (2) How do employees respond to these controls? and (3) How do spouses respond to the resulting demands on them?

In answer to the first question, I found three overarching techniques used by managers to exert boundary control on subordinates: (1) imposing demands, through meetings, requests, reviews and internal deadlines, controlling vacations, and training; (2) monitoring by standing over, checking up on, and observing employees; and (3) modeling the behavior they want employees to exhibit. I discuss these below.

To answer the second question, regarding employees' responses to boundary control, I focused specifically on the evidence for each of the seventeen members of the software group and analyzed whether each individual's behavior reinforced or resisted boundary control. Some individuals accepted boundary control and did whatever they could to conform. I refer to these people as "acceptors." Others had limits beyond which they could not be pushed. I refer to these people as "resisters."

To answer the third question, involving the responses of employees' spouses, I drew on the data from the spouse interviews. Spouses varied in the degree to which they were responsible for their home and family. Further, even among spouses who had similar responsibilities at home, there were differing degrees of tolerance for boundary control; some accepted and others resisted the responsibilities that they were asked to assume as a result of their spouses' work demands. I refer to those who accepted the level of responsibility for which they were responsible, whatever that level was, as "acceptors." I refer to all other spouses as "resisters."

Three techniques are used by managers to pressure individuals to work long hours. Senior managers impose demands on both project team leaders and engineers, monitor the work of project team leaders and engineers, and act in the desired ways, modeling the behavior that they expect from more junior managers and engineers. It is those who are most responsive to the demands put on them, demonstrating that work is their first priority at whatever cost to their lives outside of work, who are highly rewarded.

Imposing Demands

Although there are no set times at which work officially begins or ends, senior managers impose demands on more junior managers and engineers which ensure that these people are at work during certain hours. The ways in which managers impose demands take a variety of forms.

Meetings. Scheduling meetings is a prime example of the imposition of boundary control. If meetings occur, individuals are expected to be present. Yet there is no thought about the implications of scheduling meetings early in the morning, late at night, or over the weekend. Every morning at 8:30 there is a "Sunrise" meeting. If the agenda involves a topic relevant to an individual's work, he or she is expected to be present and often must attend a pre-sunrise meeting to prepare for the subsequent meeting. Similarly, when senior managers feel that a weekend meeting is necessary, they do not ask their subordinates about their availability but, rather, inform them that a meeting will take place. As one engineer explained, "You can only say no so many times. You need to think carefully before you say the word 'no.' And when you do, it had better be for a good reason."

The engineers and project team leaders may be asked for their input as to the time of a meeting, but even those responses are hardly taken into account. One Thursday afternoon, the software manager informed his direct reports that they would meet the following Saturday. He asked if anyone had a preference for the time. One employee said that he would prefer the afternoon. The software manager responded, "Oh, you have baby-sitting responsibility in the morning?" The employee answered, "This is true," and then immediately added, "Don't worry about my preference, I can easily hire a baby-sitter." He seemed embarrassed that he had expressed a preference. In the end, his request was ignored, and the meeting was set for 8 A.M. on Saturday morning, which is exactly when the employee's wife worked.

Another common phenomenon is for meetings to run late. During one meeting I attended, it became clear around 4:30 P.M. that they would never be able to get through all of the agenda items by 5 P.M. At this point, the manager leading the meeting suggested that they extend the meeting until 7 P.M. He asked the obligatory "Will that be OK with everyone?" One woman agreed but then left the room immediately following the decision, to make alternative arrangements for her children. When these situations arise, sometimes employees can make alternative arrangements at home at the last minute. Other times, however, they simply cannot and must leave as planned, risking the consequences at work, both in terms of negative stereotypes and missed information or decisions that may affect them.

Requests to individuals. Sometimes, an employee is requested to work extra hours. When managers make these requests, employees have little choice but to agree if they want to be perceived as productive and committed. Managers become familiar with those who respond to such requests and return to them with future requests.

Managers register the conditions under which their subordinates are most likely to consent but also expect them to respond whenever they are needed. For example, one engineer will work all night any week night but prefers not to work on the weekends. According to his manager: "1 try to monopolize on [this engineer's] willingness to work during the week .... I don't refrain from asking over the weekend .... I just know that it is better not to .... Recently, I called him at home on a Saturday and had to go through the third degree from his wife before she would even put him on the phone ... but he still helped .... In general, it is easier to ask on a weekday, but he will always say yes if I need him." Such willingness to respond characterizes the organizationally successful Ditto employee. Employees tend to get caught in a cycle either spiraling upward or downward: the more often one accepts managers' requests, the more one is asked to do; the less often one accepts these requests, the less one is asked to do. Typically, only those who accept extra work are rewarded.

Reviews and internal deadlines. A less direct way of imposing control on engineers and project team leaders is through the scheduling of reviews and release dates to assess progress. Managers at the division level and above set a series of reviews for the product. The threat to group members is that if they fail a review, the product may be canceled and their employment may be subject to question. Beyond the reviews for the product, the software group also has a set of internal deadlines for which it is supposed to release portions of the software for further development by other members of the product team. Failure to meet one of these deadlines results in a highly visible failure to meet expectations.

Prior to both reviews and internal deadlines, managers expect employees to increase their work hours. Usually, an extra investment of time is required because employees are worried about falling short of the goals they are expected to achieve. Even if they, personally, are on schedule, however, employees are expected to put in extra effort around these times. One project team leader said about an engineer, "He may have been done with his work, but the expectation is that he... help the group during a crunch time. Before a major release we are all supposed to pitch in." Engineers seem to agree that if there is a deadline that involves their work, it is their responsibility to stay late. As one engineer put it, "When it is crunch time, no one refuses. You know you just have to do it."

Managers tend to take advantage of this norm. They perpetuate what can be labeled a crisis mentality to appeal to individuals' sense that they must be present under these circumstances. Referring to the software manager, one engineer said, "If the work is due in two days he will tell us it is due in one, just to ensure we get it done." If the team slips behind schedule, additional reviews and deadlines are often inserted in the schedule to further intensify the pressure on group members. As the division vice president reckoned at one point in the product's development, "1 fear we are slipping behind schedule so I added a review .... This will give me a better sense of what is going on .... It also makes people work harder. People work harder when there are deadlines coming up." Also common is for managers to assign work based on a "real" deadline, but if the deadline changes, managers do not inform their subordinates. Rather, managers leave engineers under the impression that the original deadline remains intact, to ensure that the work is completed. One engineer spoke about a project he worked on:

Monday morning we were given notice that something was due for a senior manager at 3 P.M. on Tuesday. Three of us stayed all night. I grabbed the output from the printer and ran to our 12 P.M. meeting on Tuesday, only to find out that yesterday afternoon our 3 P.M. deadline had been indefinitely delayed .... When I asked why no one had told us, he [referring to his manager[ just said, "We are going to have to do the work some time, so it is better to have gotten it done."

Managers use deadlines as a way of maintaining a high level of work intensity. They set dates on which work is due without consulting the engineers, who often regard these dates as arbitrary and unfair.

Restricting vacations. Another way of imposing boundary control is to limit the times during which it is acceptable for employees to take vacations. Despite the three to six weeks of vacation that group members are entitled to each year, few of them use all of their vacation time. Managers pressure subordinates not to take vacation during periods of crisis. Yet slow periods are rare and hard to predict. As one engineer explained, "1 would like to take a vacation .... I have not had a vacation in a year. But to take a vacation would require I plan ahead, and with a sense of urgency, that is impossible.., vacations are not the norm around here."

Vacations, if planned, tend to be delayed, canceled, or aborted. Employees try to plan vacations strategically to follow major deadlines. The expectation is that, following a deadline, the demands at work should temporarily slow down. The schedule tends to slip, however, and well-planned vacations often end up occurring at critical times, making them highly problematic from managers' perspectives. As a result, employees feel they must cancel their vacations at the last minute. According to one engineer, his manager told him, "If I insisted on going, I should consider shortening my vacation by a week." This engineer further explained, "When you are told to think about something, that means you better do it. If you decide not to listen, you risk being scorned by your manager and may even be called back early." One engineer's experience exemplifies the reality of this assessment. He, his wife, and their two young children scheduled a family trip four months in advance. Although they were supposed to leave Thursday night for ten days, crises at work prevented them from leaving until Saturday afternoon, by which time the engineer's wife had grown impatient. On Sunday, the engineer received a call that he was needed back in the office. Someone else on the project team could have solved the problem, but since everyone was dealing with crises, it was more efficient, from his manager's perspective, to bring this engineer back rather than use anyone else's time. He took the first plane back Monday morning. He spent three days in the office, while his wife vacationed with their two young children. According to the engineer, "Luckily, she was at home with her parents, who could help with the children.., otherwise, I would have had a real dilemma leaving her to return to work .... She certainly would have made it more difficult for me to go."

Training. Requiring that engineers spend extended periods in training to advance their careers is a further way of imposing control on their time. For example, after receiving a promotion from engineer to project team leader, one is expected to attend a two-week management training course seven hundred miles away from the Ditto site I studied. Several years ago, one of the engineers in the software group declined the "opportunity" to attend training because she, her husband, and their three children (all under five years of age) were moving during that same two-week period. She requested to attend a training session at a later date. She has never been given the opportunity, nor has she ever been forgiven for missing training. Managers still refer to the fact that she declined training. They never mention her reason, only that she did not attend and, therefore, implicitly, that she did not display the expected devotion to Ditto. A second engineer who was recently promoted to project team leader was told that to continue her rapid career progress she would have to get an MBA degree. She was given two choices. She could spend six weeks at an intensive management training program off-site, close to six hundred miles from home. Or she could go to school part-time for the next two years, taking two classes a semester on top of her job at Ditto and her responsibilities at home for three young children. Clearly, it would have been more efficient for her to accept the six-week intensive program and finish sooner, but this option would have involved leaving her family for an extended period of time, and her husband was unwilling to take on this responsibility. Consequently, the woman enrolled at the local business school, and she struggles to manage her increased load.

Monitoring

In addition to imposing demands on subordinates, managers monitor subordinates' presence in a range of ways, including standing over them trying to "help," constantly checking their status and their plans, and standing back and observing the hours that they work.

Standing over. The most visible form of monitoring occurs when managers stay late at work to "assist" the engineers and therefore ensure that the engineers themselves stay late. One day, two engineers were struggling to solve a critical problem. They stayed late to work on it. The software manager also stayed late. According to the software manager, he stayed to help the engineers get the work done, but one of the two engineers explained: "He [the software manager] was slowing down the process. He should have gone home. It doesn't help us get the work done to have someone constantly standing over us .... We work best when we are left alone .... The managers just don't get it. They stand over us constantly asking how they can be of help. All they are doing is distracting us. They can help the most by not trying to help at all." On a different night, an engineer and his manager stayed late to solve a critical problem. According to the engineer, "Around 9 P.M. when I complained I had a sore throat, he [his manager] encouraged me to work another hour. At 10 P.M., again I mentioned not feeling well, but it wasn't until 11:30 P.M. he suggested I go home and get some rest." To ensure that the work gets done, managers stand over their engineers continually asking about their status, offering to help, but pressuring them to keep on working.

Checking up. Managers also monitor engineers by frequently asking for a progress report, both verbally and in writing. Early each week, engineers must submit an oral update on their progress to their project team leaders. In turn, each Wednesday, the project team leaders present a written status report on their engineers' progress to the software manager. The software manager then provides a written status report to the senior software manager on Fridays and to the product manager on Mondays. In addition, the software manager asks people how they are doing every time he sees them. If he is concerned about their progress, he often seeks them out and inquires multiple times in a given day. As one engineer put it, "He [the software manager] increases the stress level by constantly asking how we are doing."

Furthermore, when managers are concerned about the progress of their engineers, they often question the engineers about their recovery action plans. For example, when one group was two days behind schedule, members were asked to write a memo explaining the slippage and stating their plan for recovery. After a 45-minute discussion with his four engineers, the project team leader drafted a memo that identified the cause of the delay and simply stated as a plan of recovery, "We will work two days overtime." The recovery action steps never involve rethinking the work processes but merely state plans to work more hours.

Checking-up provides managers with information on where engineers are and further pressures the engineers to work long hours. Moreover, engineers often end up feeling that they must commit to finishing work by dates and times that they know are unrealistic. In one example of checking-up, the software manager called one of his engineers at home. He wanted to know what had happened that night at work and when the engineer would be in the next morning. The engineer recounted:

I would not have gone in until probably close to 9 A.M., but after he [the software manager] called I made sure to be there by 7 A.M .... What he said to me last night was: "1 want to make sure we have our release ready for Sunrise in the morning," which is at eight-thirty, "because I want to be able to go in and say 'you're wrong, we have our release ready.'" ... He [the software manager] always assumes that everything is going to go OK. And nothing ever goes perfectly smooth, especially when you try to rush something and get it done really quick. Then you always fuck it up and have to do it again. I don't think he realizes that. So he just assumed that if I came in real early and gave the code to John, and he made the proms, and we plugged them into the machine, and then they would be ready to go, and he could go to Sunrise at eight-thirty and say, "Here's the release." But, it turns out that we didn't have it working until when?... Eleven thirty or something like that. I knew it would never be ready by eight thirty--that would have taken a miracle.

This engineer, however, never mentioned his well-founded doubts to his manager. Rather, he simply agreed to arrive early, give it his best effort, and thereby display the expected level of commitment to the project.

Observing. Managers also monitor by standing back and taking note of the hours that engineers work. Engineers are well aware that they are being observed. They note that the division vice president walks around conducting "bed checks" to see who is still around at night. Moreover, managers themselves stress the importance of perception. According to one of the project team leaders, the senior software manager told him: "There is an advantage of being perceived as really killing yourself when you are first on a project .... He told me about the importance of perception .... He said perceptions mean a lot around here. It is important to be perceived as a person who will sacrifice personal life for work .... He specifically mentioned sacrificing family time for work as an indication of commitment." According to another manager, "A star performer is one who doesn't know enough to go home at night." One woman embodies the star performer. One day each week, she arrives at 2 A.M. and works straight through the day until 5:30 P.M. She does this because it impresses her manager. She explained: "[He] was really impressed when he ran into me at 2 A.M. one morning .... I used to just go to the kitchen table and use my PC, but after the reaction I got from my manager I decided it was important to do that early morning work in the office .... It is better to be seen here if you are going to work in the middle of the night."

At the other extreme, not being present usually creates a bad impression. With ten days to go on a major software release, two engineers stayed late one night. Their project team leader's boss, the software manager, stayed late to help as well. Even the senior software manager stopped in as he was leaving work around 6:30 P.M. to check on the team's progress. The project team leader, however, left at 6:00 P.M. His parents were visiting, and a family gathering was planned. He felt that it was unnecessary for him to stay at work: "There was nothing that I could do to help. I lacked the technical expertise that was necessary.., so I left." According to the engineers, their project team leader's departure presented no problem to the team: "There is nothing that he could have done to help out .... His presence would have been more of a hindrance than anything." Nonetheless, the project team leader was accosted the next morning by the senior software manager, who questioned him about why he had not stayed late the night before to support his engineers. His response to the censure was not anger or defensiveness but, rather, guilt. He offered the senior software manager an apology. He remarked to me, "1 am such a schmuck. I should have stayed around and done busy work. It is terrible that I left. It looks really bad .... If I had stayed, he [the senior software manager] would have seen me here." To the engineers, "being present" has little to do with performance or substantive output. All of the engineers felt there was no practical reason for their project team leader to have stayed. In fact, had he stayed, his presence might have slowed the team's progress. Still, from his perspective, leaving constituted a grave mistake. He feared that he would now have to live with an image of lacking commitment to the team--"l left them in a time of crisis," he said.

At the end of the calendar year the managers ranked software engineers to determine individuals' raises. The names of each member of the software group were listed, followed by a comment explaining each ranking. The comment following the engineer ranked first on the list read: "works 80-100 hours/week, top quality work." Similarly, the second on the list was noted for working 80-100 hours a week. Comments about others in the top included: "works days and nights" and "works 80 hours per week." In contrast, the comments about those at the bottom of the list included: "average contribution," "light work load," and "minimal contribution." Senior managers clearly notice the hours that the engineers work and use these observations as a criterion for ranking them. Moreover, the engineers know the range of raises given and therefore get a relative sense of where they stand in their managers' eyes.

Modeling Behavior

Managers expect the same behavior of themselves as they expect of their engineers. They work long hours, in most cases, longer hours than the engineers. Managers come in early, stay late, and work most weekends. The senior software manager, for example, typically works from 6 A.M. to 7 P.M. five days a week, plus at least six hours on both Saturday and Sunday. He makes it clear, "No one works harder than I do." Furthermore, when managers put in the time, they publicize their hours. They often refer to the hour at which they have done a certain piece of work or had a certain thought. The product manager, for example, will often refer to a thought he had at 4 A.M., when he was on the treadmill in his personal gym. The senior software manager often refers to the early morning and weekend hours he was in the office and remarks that no one else was there. One project team leader told a story about his chats with the senior software manager at 6 A.M. He emphasized that he, too, is in by 6 A.M. every morning. Managers point with pride to the long hours that they work. They often say, "1 would never ask for anything I wouldn't do myself." This expectation translates into a tremendous number of hours spent at work.

All employees face boundary control, but analysis of the data reveals that their responses to it vary. Not all employees accept the boundary control imposed on them. Some resist it. Moreover, many employees are married, and the demands of their work have consequences for their spouses. Spouses' reactions to the demands that ultimately affect them further influence how employees respond to boundary control. For married employees, it is not clear whether their own reaction comes first and in turn affects their spouse's reaction or whether their spouse's reaction influences their own initial reaction. What is clear is that different employees expect their spouses to play very different roles, and some spouses accept and others resist this role. Below I provide a description of employees who are "acceptors" and "resisters" of boundary control, as well as spouses who are "acceptors" and "resisters" of the role that results for them at home. Table 2 summarizes the responses of employees and the spouses I interviewed.

Employee Responses

Acceptors. Employees who accept boundary control make work their first priority, always trying to meet work expectations. One acceptor said: "The work is crucial, and I am willing to put in whatever hours the work demands." He continued, "It is not the company's job to manage crises. It is the individual's responsibility to accommodate." When confronted with a crisis, he commits all his waking hours to finding a solution. He arrives at work early in the morning and stays until 10 or 11 P.M. It is not unusual for him to work an 85- or 90-hour week. The stash of food, candy, and tea found throughout his office testify to his habit of sacrificing meals to enable extra periods of work. Another engineer characterized as an acceptor further exclaimed, "Right now at work, I need to be single-minded .... I must dedicate my life to this battle.., and only when it is over can I celebrate."

Although the quintessential acceptor is one who commits fully to work, some are in dual-career relationships and have responsibilities at home. One woman does everything she can to care for her home and family without failing to meet the demands at work. During one two-week period when her husband was away, she was assigned to a short-term, high-visibility project. The project demanded extraordinary hours. For two straight weeks, she would leave work every evening at 5:30 P.M., pick up the kids from day care, feed them, play with them, and put them to bed. She said she would then take a short nap and get up around 12 A.M. to start the next day's work. She worked at home until 5 A.M. and then would get the kids up, dressed, fed, and off to day care before coming into the office to work until 5:30 P.M. She did this repeatedly, never mentioning to her manager that her husband was away. Such behavior is characteristic of acceptors.

Resisters. In contrast to acceptors, who may never get away from their work, employees who resist boundary control make themselves unavailable to work at certain times. Some who resist boundary control do so because they simply do not have a burning desire to succeed, like one of the male engineers, who is single and has few responsibilities outside of work. He works regularly from 7:30 A.M. to 4:30 P.M. with a half-hour for lunch. He is very focused on his work while he is at work and spends little time conversing with colleagues, but he leaves regularly at 4:30 P.M. He is not on the fast track, nor does he aspire to be. For him, work is a job, not a step on a career ladder.

Others who resist boundary control do so because they have demands on their time outside of work that they consider more important. According to one female engineer, "1 work because I want to be intellectually stimulated during the day .... I want to have colleagues to talk to .... I would go crazy staying home all the time." Her primary commitment, however, is to her family. She will not work unlimited hours. She takes her work seriously and will occasionally stay late if the situation is desperate, but in general she works a 45hour week. As she says, "Family is my first priority."

In contrast to resisters like these engineers, who make sacrifices at work to have more time outside of work but resign themselves to the consequences, some employees resist boundary control because they perceive that they have no choice, yet they express frustration about their situations. These engineers often search for alternative ways of working, to make it possible to meet the demands of both a successful career and their responsibilities outside of work. One male engineer must leave work three days a week by 4:30 P.M. His wife works evenings, and she insists that he be home to care for their son before she leaves. When he is behind schedule, he works Saturdays instead of Mondays to increase his productivity. According to this man, "Saturdays are at least twice as productive as Mondays... there are no interruptions." One female engineer also tried to create an alternative way of working so that she could limit her hours and still meet her work demands. Six months before I arrived, she was a project team leader. She created a rotating position such that one day a week one of her engineers would fill in for her. This enabled her to work from home on this day, saving two and one-half hours of commuting and providing her with uninterrupted time during which to work. Moreover, her group members each had the unusual opportunity to act as project team leader while she was home. As a result, they developed managerial skills, and the project did not suffer. Her six-month performance review noted that her group was on schedule. Managers, however, never acknowledged her innovation. Instead, just before I arrived, she was reassigned to a position as an engineer working with confidential data that could not leave the lab. She could no longer work from home.

At the end of the year, the two engineers who sacrificed work for family without trying to compensate each received a 3-percent raise, with which they were satisfied. The male engineer who tried to make a small change also received a 3-percent raise, although he had hoped for higher. In contrast, the female engineer who opted to use a flexible work schedule dropped from a rating in the top 10 percent of employees the previous year to the bottom 20 percent in the year of the study. She ended up receiving no raise. The major difference between her resistance and that of the male engineer was that he framed his change as temporary and benefiting the work, whereas she framed her change as permanent and benefiting herself and her family. Resistance, especially when motivated by a need to meet family obligations, hinders employees' recognition and promotability. These findings reveal the entrenched nature of boundary control and the costs to employees who try to go against the flow. Satisfying output requirements simply is not enough.

Classification of employees. Of the twenty employees I studied, ten accept boundary control, while ten resist it. Of the ten who accept it, nine are male, seven of them married, and the other is a married woman. Moreover, of the eight individuals in the top four levels of hierarchy (project team leader and above), only one is a resister. These numbers make sense. To succeed, one must accept boundary control; therefore, those who accept boundary control would be expected to be the most successful.

In contrast to the ten acceptors, ten employees resist boundary control. Of these ten, seven have resigned themselves to the fact that they must selectively resist certain of their managers' expectations if they are to restrict the time they spend at work. Five of them are men, three of them single and two married, and the other two are married women. The priority for three of these employees is family, for one it is other work responsibilities outside of Ditto, and for three it is simply lack of commitment to spending time at work. In contrast, the three other resisters (two men and one woman, all married) are more frustrated by the tradeoffs they must make. They want to succeed at work, but they cannot put in the necessary time. Two of these employees have family responsibilities, and the third is investing much of his time pursuing a Master's degree.

All acceptors share an ambition to succeed but not all of those with this ambition are acceptors. The frustrated resisters share this ambition, but they are unable to live up to the associated work demands. Neither ambition, nor gender, nor marital status, nor the existence of children can therefore explain whether an engineer accepts or resists boundary control. All of the acceptors and three of the resisters share an ambition to succeed. Nine of the acceptors and seven of the resisters are male. Eight of the acceptors and seven of the resisters are married. And four of the acceptors and eight of the resisters have children under eighteen.

To understand individuals' responses, it is important to consider the demands on their time outside of work. The five single employees I studied were male. Three lived alone, one had roommates, and one lived with his parents. Only one had a serious girlfriend. Although none of these single men felt major obligations to anyone else, it certainly could be the case that a single person would have such responsibilities (i.e. child care or elder care). For married employees, the division of labor at home has a profound impact on their ability to meet work demands. Some are married to a spouse who expects minimal amounts of help at home. Other spouses expect the responsibilities at home to be shared. I refer to spouses who accept the demands put on them at home, at whatever level those demands may be, as "acceptor" spouses. I refer to spouses who resist the demands placed on them as "resister" spouses. A full understanding of married employees' responses to boundary control requires considering the reactions of their spouses to maintaining the home and caring for children.

Spouse Responses

Acceptor spouses. One female spouse embodies the definition of an acceptor spouse. She is the mother of two and has a part-time job (three days a week) as a social worker at the university hospital. When asked what role her husband plays at home, she can tell many stories about events her husband has missed, responsibilities he has shirked, and tasks she would like him to do. She says: "He used to be responsible to empty the dishwasher in the morning before he went to work. Now, even that is more than he can handle .... I used to try to force him to help around the house, but it would always result in an argument. Now, I don't even try .... I have decided it is just easier to do things myself rather than getting angry." Despite her resentment, she accepts these demands. This behavior is common for acceptor spouses. They accommodate the chaotic schedules of their spouses but harbor resentment with the distribution of tasks that results at home.

Although more acceptor spouses are female than male, not all are female. One man, for example, picks up his daughter from the baby-sitter, brings her home, makes dinner, feeds her, bathes her, plays with her, and puts her to bed, while his wife works. He loves his daughter and appears to be a committed, caring father, but he notes, "It gets hard having to be a single parent every night of the week." He says: "1 love her [his daughter] and spending time with her, but sometimes it just gets to be too much. Sometimes I want some time to myself, time in the yard, time to finish all the projects that I want to do .... Sometimes I just feel cheated that I don't have a wife who has dinner on the table when I get home .... I don't really understand what she [his wife] is doing, but I always give her the benefit of the doubt."

Resister spouses. In contrast to acceptor spouses, who enable employees to accommodate boundary control, some spouses exacerbate the difficulty faced by employees who attempt to manage both work and family responsibilities. For example, one female spouse has two children, and she works part time. She expects her husband to be home at night and to do his share of the household chores and child care. "If he isn't paid for overtime," she insists, "he should not work it." Although both husband and wife agreed that her job was secondary, this assessment did not translate into his work ruling their lives. On the contrary, she established strict constraints on what she was willing to accept in terms of his work demands. In the odd crisis, he accommodates work, but he does not regularly make the accommodations made by other employees who are eager to succeed at work. Several male spouses set similar limits for their wives. One male spouse says:

After 7:30 I am too tired to deal with the kids. I find being with them very stressful. I pick them up from day care and bring them home. I spend time with them until she [his wife] gets home but bed time is her responsibility .... I can do the things I do for the kids, but I am not their mother. These things should be a mother's responsibility. When she is not around I do them out of necessity. But when she's around she does them .... She makes sure to be home by 6 P.M. every night.., otherwise we have a fight.

The female engineer in this case gets her children up, dressed, and ready for school. She also does all of the evening chores. Still, her husband resents that he is left to drive the children to school, pick them up, and spend an hour with them each night before she arrives home and takes over.

Employee Responses Revisited

Employees may be categorized as acceptors or resisters. Their spouses also may be categorized as acceptors or resisters. It is important to note, however, that the categorizations of employees and their spouses are not independent. For example, a married engineer may be a resister precisely because his or her spouse resists anything less than a major commitment of time at home. Similarly, a spouse may be an acceptor because his or her spouse, the engineer, already resists boundary control and helps at home to a level of satisfaction. Of the twelve spouses interviewed, eight were acceptors and four were resisters. Of the eight acceptors, six were female. Of the four resisters, two were female.

Employees and their spouses. Puttig together the employees with their spouses, four types of married employees result, as shown in figure 2. I term them careerists, compromisers, jugglers, and rejecters. When both employees and their spouses are acceptors, I refer to the employees as careerists. The employee accepts the demands at work, and his or her spouse perpetuates this acceptance of boundary control by assuming much responsibility for household chores and child care. These spouses' responses enable the employees to devote themselves fully to their work. For example, one male engineer's wife supports his work and tolerates the burdens that result from his erratic work schedule. She goes out of her way to care for the home and asks nothing of him in return. Her response alleviates much pressure for him, allowing him to accommodate work demands. In fact, he is so caught up in responding to the demands of his work and the effects it has on him that he fails to notice the burden he puts on his wife. All careerists in this study are male and occupy the level of project team leader or above. Women employees who had an accepting spouse and themselves accepted work demands would fall into this category, as well, but there were no such women in this study, nor is this scenario common in our society.

In contrast to the five careerists, who are in relationships in which both members accept boundary control and its repercussions, in three relationships, the spouse accepts boundary control, but the employee resists it. In two of these three couples, the employee is female; in the third, the employee is male. I refer to these employees as compromisers. It is possible that these spouses would accept the demands on them regardless of the response of their partners, the employees, to boundary control. Listening to these acceptor spouses, however, it sounds as if they accept boundary control precisely because their partners resist it, and if their partners, the employees, did not resist it, the spouses would not accept their current set of responsibilities. It therefore seems not only to be employees' willingness to participate at home, but their spouses' expectation that they participate at home that results in employees becoming compromisers. For example, one of the male engineers accepts that he will have to make trade-offs at work if he wants to be with his family. His wife accepts the demands at home that result for her. She willingly takes care of much of the child care and household chores, but, she says, "1 know he [her husband] will always come home when I need him." She told a story about one day when she felt depressed because of a recent death in her family. She called her husband at the office, and he came home immediately to comfort her. She says, "I treasure the security of knowing he will always be there when I need him."

Figure 2. Employee's experiences of boundary control. The twelve employees represented in this figure include the ten married members of the software group whose spouses I interviewed, plus the product manager and the division vice president.

The compromisers all seem to experience pressure from their spouses not to work all of the time. Their responsiveness to these expectations helps alleviate ill feelings at home. Their responsiveness, however, may also negatively affect their career potential. A project team leader is the highest-ranked employee in this category, and he acknowledges his expectation that he will stop climbing the organizational hierarchy as a result. He says, "I would never be able to work these reduced hours if I still wanted to be a success at work.., but I am happy at the level I have achieved." As would be expected, he did not receive a bonus given to top managers. The other two compromisers were engineers who both ranked in the middle 70 percent of employees.

In contrast to the eight employees married to acceptor spouses, four employees are married to spouses who resist the demands placed on them. One of the four employees married to a resister spouse still tries to accept boundary control. I refer to this type of individual as a "juggler," because the employee must balance a spouse's resistance and the responsibilities at work. This is a demanding position to put oneself into and leads to a high degree of anxiety about the trade-offs one must continually make. This employee's husband expects her to maximize her time at home and to assume primary responsibility for their children's care. She tries to appease him while also meeting and exceeding work demands. One weekend she had to finish some work, but she did not want to leave her husband at home with the children because, she says, "He would have been resentful." Therefore, she played outdoors with her children Saturday morning and afternoon, fed them dinner, bathed them, and put them to bed. Once the children were asleep (and her husband was sitting happily in front of the TV watching a hockey game), she felt she could go to work. As she recalled, it was about 8 P.M. She worked throughout the night and said she felt good about the work, since she had accomplished it without "sacrificing... her family." But she is tormented by the trade-offs that she must constantly make. She describes her priorities as:

... wishy washy... I want both a successful career, and quantity time with my children .... I worry that I am not spending enough time with my kids. I am with them after work, but the time is so pressured. I feel like I need more time, more time at home and more time at work .... If I truly thought we had enough money on one salary, I would quit working, at least for a few years. But I am afraid that I would never get the same opportunities again .... Family is my first priority. My heart is at home and, as a result, my work is suffering .... I wake up often in the middle of the night and think to myself, it does not matter if you're a superstar; if you are not a premium employee .... I want to be less stressed and more happy I just want to be average.

But then she admits, "I am too internally competitive for that. I cannot stand not being at the top." At the same time that she struggles to manage responsibilities at home, she achieved a lump-sum bonus at work, placing her among the top 10 percent of managers at Ditto.

Women who strive for success at work are expected to be a man's equal in a man's world (Gordon, 1991: 7). These women must be acceptors. At the same time, they are expected to do the majority of the second shift (Hochschild, 1989), which means that they likely have a resister spouse. Women striving for success therefore tend to be jugglers. The fact that only one woman in my sample was a juggler is not indicative of how common this scenario is among professional women in our society who strive for success (Swiss and Walker, 1993); however, it is indicative of how difficult it is to manage both a career and family and of the reason many employees search for alternatives.

I refer to the last category of employees as "rejecters." In these couples, both the engineer and the spouse resist boundary control. I found three rejecters in my sample, two male and one female. All three of these employees did more than just resist boundary control; they each tried to create an alternative way to accomplish work within constrained amounts of time. Two of their change efforts were described above. One of the male engineers worked Saturdays instead of Mondays in an attempt to accommodate his own ambition to succeed and his wife's expectation that he be home at night. Another engineer worked from home one day a week to balance work, a long commute to work, and her husband's resistance to family responsibilities. The frustration the three rejecters experienced led each of them to create new ways of working that enabled them to do the same work, or sometimes even more, in constrained amounts of time, but none of them benefited from these efforts in terms of career progress. I found no rejecters in managerial ranks or headed toward such a position. Moreover, each of the three rejecters received little or no increase in pay.

Gender explanations. A further examination of the four types of married employees by gender reveals that all five careerists are male and that the only juggler is female. These findings corroborate the theory of gender ideology proposed by Hochschild (1989): men strive for success, and women reinforce their husband's effort by accepting responsibility at home. From Hochschild's theory, however, one would expect all men to be careerists and women who are ambitious enough to endure the responsibility and the stress of managing both a home and a career to be jugglers. All other female employees would be expected to resist boundary control. Despite some support for gender identity theory, not all of the data are consistent. Three of the six married male employees are resisters. Furthermore, of the four male spouses married to female Ditto employees, three would be classified as resisters to their own organizational demands, if one were to classify the men instead of their spouses (the employees). (c)f the twelve men included in this sample of married couples, half are resisters of work demands and therefore not careerists.

Moreover, as more married women enter the work force, the number of employees who are careerists will likely continue to decline, both because married women themselves are not likely to have an accepting husband and because as more married women are working, men are less likely to have a wife who assumes most of the responsibility for the household chores and child care. This inference is substantiated by the fact that the average age of the careerist in my study is 42 years old, compared with the average age of the non-careerist, which is 31 years old. While it could be the case that the careerists were themselves non-careerists when they were 31, evidence from them and their spouses suggests they have always been careerists. This implies that younger employees are not following in the footsteps of their seniors in terms of the roles they and their spouses play. Currently, men are more likely than women to be acceptors, not just because they are men but also because they are more likely to have an accepting spouse who makes it easier for them to be an acceptor. Marriage to an acceptor spouse which currently is highly characteristic of men--is a critical factor in determining how one responds to boundary control. With the changing nature of relationships at home, there will likely be a shift in response to boundary control, with more individuals, both men and women, resisting it.

This study describes a form of boundary control that has long been in existence but is becoming increasingly prominent and problematic. This form of control is best adapted to by employees who are either single with no responsibilities outside of work or are married to a spouse who takes care of the responsibilities outside of work. For these employees, all of whom were men at Ditto, the choice at least exists as to how they want to respond to boundary control. Most employees, however, do not even have this choice. Rather, they are married to a spouse who also works and will not relieve them of all of the home and child care responsibilities. These employees feel pressured from both sides: pressured to put more time into their work and pressured to share more responsibility at home.

There is some room for resistance at work. Those who resist, within reason, do not lose their jobs, but they do sacrifice reward and recognition at work. Nippert-Eng (1995) documented two strategies available to employees struggling to balance work and life outside of work: one strategy is to physically separate work and home life ("segmenting"); the other strategy is to intertwine work and home ("integrating"). In my study, resisters quite clearly adopt a segmenting strategy, conducting work at the office but not from home. Resisters often make statements such as: "1 leave work behind me when I pass the guard booth on the way out the door. I have twenty minutes on the drive home to transition into family life. I don't think about work again until I pass the guard booth in the morning." In contrast, the married acceptors (although not the single ones) follow an integrating strategy: they describe working from home late at night after their children have gone to bed and on weekends. The evidence suggests, however, that regardless of how individuals organize their lives, only a relatively small number are in life situations that enable them to comply fully with work demands.

To date, most research on time and organizations has limited its focus to the use of time within organizations (Zerubaval, 1981; Barley, 1988; Bluedorn and Denhardt, 1988; Fine, 1990). The research at Ditto explores, instead, the nature of the temporal boundary between work and life outside of work. I focused on the imposition of temporal demands at work but considered the effects on the structuring of both work time and nonwork time. For example, when managers schedule a meeting that runs late into the evening, they simultaneously increase the amount of time that engineers spend at work and decrease the amount of time that engineers spend outside of work. By focusing exclusively on time within organizations, researchers previously have not explored the effects of time usage at work on life outside of work.

In the introduction, I described the progression of types of control over the temporal boundary between work and life outside of work. While the types of control are linked to particular organizational structures (Ouchi, 1977), however, they are not necessarily isomorphic with them. Multiple types of control may co-exist in the same firm (Edwards, 1979; Wilkins and Ouchi, 1983; Van Maanen and Kunda, 1989), with each type aimed at achieving roughly the same end (Czarniawska-Joerges, 1989). At Ditto, cultural norms define how individuals are expected to respond in different circumstances. For example, no formal rule states that engineers must stay late whenever an internal deadline is approaching. Nonetheless, engineers are aware that others in the organization (managers and peers alike) expect them to do whatever it takes to meet work demands, and engineers care a good deal about what managers and peers think of them. As a result, when a deadline approaches, they stay. Achieving this form of boundary control depends on affecting values, loyalties, sentiments, and desires of employees that determine how they respond to demands on their time. Despite clear use of culture control, however, managers at Ditto also use traditional bureaucratic control to maintain engineers' temporal boundary between their work and their lives outside of work. Engineers are exposed to the hierarchically based social relations of the organization and the concomitant sets of company rules and policies that reward compliance and punish noncompliance (Edwards, 1979). Career advancement, higher pay, and interesting assignments-rewards that go to those who work the longest hours--are all powerful motivators for Ditto employees.

It is not altogether clear whether it is bureaucratic or cultural control that influences engineers to accept long hours and unpredictable demands on their work time. The control system at Ditto is structured so that whether one is motivated by genuine commitment to the organization or by one's own self-interest, one's behavior will convey commitment to the organization. It may be that employees display such commitment because they are social beings motivated by a deep concern for the organization. More likely, though, they are rational actors with instrumental orientations who recognize that perceived "selfless" commitment to the organization leads to personal success. Even if engineers' actions are intentionally self-promotional, the bureaucratic control in play relies heavily on underlying cultural controls. By and large, people want to be considered successful or star employees, and raises, promotions, and bonuses are the means by which such success is measured, both individually and organizationally. Bureaucratic control is thus premised on cultural understandings of success. In the end, the form of boundary control at Ditto seems to be both cultural and bureaucratic. Culture provides a largely unquestioned image of the star employee as one who does whatever it takes to get the job done, who does not question managerial perceptions of the appropriate means to this end, and who does not allow familial or other nonwork interests or responsibilities to intrude on his or her performance and presence in the workplace (Bailyn et al., 1996). Bureaucracy aligns engineers' self-interests (in being a star employee) with company goals that, when met, provide reward and recognition.

It is worth noting, too, that bureaucratic forms of boundary control may be crucial at lower or junior levels of the organization yet largely unnecessary at higher or senior levels, where culture control takes over. Engineers attribute their long hours to the demands of lower-level managers, who, in turn, attribute their long hours to the demands of senior managers. But senior managers are rewarded on the basis of firm performance and are, officially at least, not rewarded for the hours they clock at the office. Yet senior managers also put in long hours. It seems likely that their actions are determined by a set of deep-seated, cultural assumptions. They put in long hours because they assume--and, in fact, say so continually in both word and deed to underlings in the organization--a perfect correlation between hours worked and output. To ensure their own organizational success, which is of top priority to them, their company must succeed, which they assume requires that they themselves, their peers, their direct reports, and so on down the long line of command, all put in long hours. Senior managers therefore continue to work long hours both because they associate their own success with this work style and because they want to model the work style they desire from everyone else involved.

To alter a control system so deeply ingrained in the work culture requires challenging widely shared assumptions. Senior managers who work long hours because they assume that long hours lead to better performance might alter their behavior if provided with enough compelling evidence to convince them that their current actions do not lead to the best performance, by themselves or those they manage. They would likely be difficult to convince, however, given that they have climbed to the top by acting in the very way that is suddenly being subject to question. Moreover, at lower levels, where long hours are not only considered important in and of themselves but are perceived as a requirement of interdependent work and as an indicator of productivity, change would be even more difficult. Elsewhere, I have described exploratory studies with software engineers that indicate that despite the interdependent nature of their work, they don't all need to be present all the time to facilitate each other's work. I found that structuring work so that large blocks of time are set aside for individuals to work independently improved the organization's productivity (Perlow, 1999). For these blocks of time, it should not matter where individuals work or when they put in this time. To address the requirement that everyone doing interactive and collaborative work be present at all times, therefore, should not involve changing the nature of the work but, rather, only the perception that to best complete this type of work everyone needs to be present at all times. Changing the way people are evaluated, however, would be more complicated. Alternative ways of assessing people's output would need to be established. If further evidence is found to support the claim that long hours hinder productivity (Perlow, 1997), then at least managers should have an incentive to help identify these alternatives.

Finally, while altering boundary control would require deepseated cultural change, such change would have potential implications for resolving conflicts between work and family (Rapoport and Bailyn, 1996; Perlow, 1997). Because work-family conflict is usually conceptualized as an individual-level problem of managing work and family responsibilities, solutions provide ways to help individuals cope. Increasing numbers of corporations are implementing work-family policies (e.g., flex-time, flex-place, job sharing, and part-time work) to ease the burden for those who are trying to balance work and family (Levitan and Conway, 1990; Ferber and O'Farrell, 1991; Schwartz, 1992). These policies help employees manage the demands resulting from boundary control. Ditto, as a company, has a set of progressive human resource policies that provides employees with opportunities to alter their work schedules in multiple ways. Yet, despite the best of intentions, knowledge workers who take advantage of these policies often suffer in terms of career progress (Shamir and Salomon, 1985; Perin, 1991; Bailyn, 1993). The problem is that these options enable knowledge workers to resist boundary control, but they do nothing to change managers' perceptions about employees who do resist.

Work on this paper has been supported by grant #910-1036 from the Ford Foundation. Lotte Bailyn, Andrea Campbell, Gideon Kunda, John Van Maanen, Stephen Barley, and three anonymous ASQ reviewers offered invaluable advice and suggestions.

1. This research was conducted in conjunction with a larger project sponsored by the Ford Foundation. Team members included Lotte Bailyn, Deborah Kolb, Susan Eaton, Joyce Fletcher, Maureen Harvey, Robin Johnson, and Rhona Rapoport, as a consultant.

Legend:
MarSt = Marital status
Chdrn = Children under 18

Title                             Employee(sex)  MarSt    Chdrn

Division vice president              Jed m)        M       -
Product manager                      Dan(m)        M       -
Sr. software manager                 Carl(m)       M       -
Software manager                     Zeth(m)       M       3
Independent contributor              Stu(m)        M       1
Project team leader                  Mike(m)       M       1
Project team leader                  Laura(f)      M       3
Project team leader                  Sam(m)        M       4
Engineer                             Mathew(m)     M       1
Engineer                             Max(m)        S       -
Engineer                             Andy(m)       S       -
Engineer                             Jane(f)       M       1
Engineer                             Sarah(f)      M       1
Engineer                             Larry(m)      M       -
Engineer                             Scott(m)      M       2
Engineer                             Kate(f)       M       3
Engineer                             Jim (m)       M       4
Engineer                             David(m)      S       -
Engineer                             John(m)       S       -
Engineer                             Jeff(m)       S       -

Legend:
Emp = Employee(sex)
A = Acceptor
R = Resistor
Sp = Spouce(sex)

Title                       Emp       A/R        Sp           A/R

Division vice president    Jed(m)      A        Pat(f)         A
Product manager            Dan(m)      A        Kaye(f)        A
Sr. software manager       Carl(m)*    A          -            -
Software manager           Zeth(m)     A        Pam(f)         A
Independent contributor    Stu(m)      A        Natalie(f)     A
Project team leader        Mike(m)     A        Kim(f)         A
Project team leader        Laura(f)    A        Rick(m)        R
Project team leader        Sam(m)      R        Betsy(f)       A
Engineer                   Mathew(m)*  A          -            -
Engineer                   Max(m)      A          NA           NA
Engineer                   Andy(m)     A          NA           NA
Engineer                   Jane(f)     R        Dan(m)         A
Engineer                   Sarah(f)    R        Tom(m)         A
Engineer                   Larry(m)    R        Lisa(f)        R
Engineer                   Scott(m)    R        Heather(f)     R
Engineer                   Kate(f)     R        Paul(m)        R
Engineer                   Jim(m)*     R          -            -
Engineer                   David(m)    R          NA           NA
Engineer                   John(m)     R          NA           NA
Engineer                   Jeff(m)     R          NA           NA

* indicates married person whose spouse I did not interview.

Bailyn, Lotte

1985 "Autonomy in the industrial R&D lab." Human Resource Management, 24: 129-146.

1993 Breaking the Mold. New York: Free Press.

Bailyn, Lotte, Rhona Rapoport, Deborah Kolb, and Joyce Fletcher

1996 "Relinking work and family: A catalyst for change." Sloan School of Management Working Paper #3892-96.

Barley, Stephen R.

1988 "On technology, time and social order: Technically induced change in the temporal organization of radiological work." In Frank Dubinskas (ed.), Making Time: 123-169. Philadelphia: Temple University Press.

Barley, Stephen R., and Gideon Kunda

1992 "Design and devotion: Surges of rational and normative ideologies of control in managerial discourse." Administrative Science Quarterly, 37: 363399.

Barley, Stephen R., and Julian Orr

1997 Between Craft and Science: Technical Work in the United States. Ithaca, NY: ILR Press.

Bell, Daniel

1973 The Coming of Post Industrial Society. New York: Basic Books.

Bendix, Reinhard

1956 Work and Authority in Industry. Berkeley: University of California Press.

Bennis, Warren

1985 "Foreward." In Douglas McGregor (ed.), The Human Side of Enterprise: iv-viii. New York: McGraw-Hill.

Bluedorn, Allen C., and Robert B. Denhardt

1988 "Time and organizations." Journal of Management, 14: 299-320.

Christopherson, Susan

1991 "Trading time for consumption: The failure of working hours reduction in the United States." In Karl Hinrichs, William Roche, and Carmen Sirianni (eds.), Working Time in Transition, 7: 171-187. Philadelphia: Temple University Press.

Clark, Peter

1985 "A review of the theories of time and structure for organizational sociology." In S. B. Bacharach and S. M. Mitchell (eds.), Research in the Sociology of Organizations, 4: 3579. Greenwich, CT: JAI Press.

Coser, Lewis A,

1974 Greedy Institutions. New York: Free Press.

Czarniawska-Joerges, Barbara

1989 Economic Decline and Organizational Control. New York: Praeger.

Dertouzos, Michael L., Richard K. Lester, Robert M. Solow, and the MIT Commission on Industrial Productivity

1989 Made in America. Boston: MIT Press.

Drucker, Peter F.

1989 The New Realities. New York: Harper & Row.

Edwards, Richard C.

1979 Contested Terrain. New York: Basic Books.

Etzioni, Amitai

1961 A Comparative Analysis of Complex Organizations. New York: Free Press.

Ferber, Marianne A., and Brigid O'Farrell

1991 Work and Family: Policies for a Changing Work Force. Washington, DC: National Academy Press.

Fine, Gary A.

1990 "Organizational time: Temporal demands and the experience of work in restaurant kitchens." Social Forces, 69: 95-114.

Glaser, Barney G., and Anselm L. Strauss

1967 The Discovery of Grounded Theory: Strategies for Qualitative Research. New York: Aldine.

Gordon, Suzanne

1991 Prisoners of Men's Dreams. Boston: Little, Brown.

Handy, Charles

1989 The Age of Unreason. Boston: Harvard Business School Press.

Hayghe, Howard V., and Stella W. Cromartie

1991 "Working woman: A chart book." U.S. Department of Labor, Bureau of Labor Statistics, August, Bulletin 2385.

Hochschild, Arlie R.

1983 The Managed Heart. Berkeley: University of California Press.

1989 The Second Shift. New York: Avon Books.

Kanter, Rosabeth M.

1977 Men and Women of the Corporation. New York: Basic Books.

Kidder, Tracy

1981 The Soul of a New Machine. New York: Avon Books.

Kunda, Gideon

1992 Engineering Culture. Philadelphia: Temple University Press.

Landes, David S.

1983 Revolution in Time. Cambridge, MA: Harvard University Press.

Lawler, Edward E., III

1986 High Involvement Management. San Francisco: JosseyBass.

Levitan, Sar A., and Elizabeth A. Conway

1990 Families in Flux: New Approaches to Meeting Workforce Challenges for Child, Elder, and Health Care in the 1990s. Washington, DC: Bureau of National Affairs.

Meiskins, Peter F., and James M. Watson

1989 "Professional autonomy and organizational constraint: The case of engineers." Sociological Quarterly, 30: 561-585.

Miles, Matthew B., and A. Michael Huberman

1984 Qualitative Data Analysis: A Source Book of New Methods. Beverly Hills, CA: Sage.

Nippert-Eng, Christina E.

1995 Home and Work. Chicago: University of Chicago Press.

Ouchi, William G.

1977 "The relationship between organizational structure and organizational control." Administrative Science Quarterly, 22: 95-113.

1980 "Markets, bureaucracies, and clans." Administrative Science Quarterly, 25: 129-141.

Ouchi, William G., and Mary Ann Maguire

1975 "Organizational control: Two functions." Administrative Science Quarterly, 20: 559-569.

Owen, John D.

1979 Working Hours. Lexington, MA: Lexington Books.

Perin, Constance

1991 "The moral fabric of the office: Panopticon discourse and schedule flexibilities." P.S. Tolbert and S. R. Barley (eds.), Research in the Sociology of Organizations, 8: 241-268. Greenwich, CT: JAI Press.

Perlow, Leslie A.

1997 Finding Time: How Corporations, Individuals and Families Can Benefit from New Work Practices. Ithaca, NY: Cornell University Press.

1999 "The time famine: Towards a sociology of work time." Administrative Science Quarterly, vol. 44 (forthcoming).

Rapoport, Rhona, and Lotte Bailyn

1996 Relinking Life and Work: Toward a Better Future. New York: Ford Foundation.

Rock, Howard

1988 "Independent hours: Time and the artisan in the new republic." In Gary Cross (ed.), Worktime and Industrialization: An International History, 2: 21-39. Philadelphia: Temple University Press.

Savage, Charles M.

1990 Fifth Generation Management. Digital Press.

Schor, Juliet B.

1991 The Overworked American. Basic Books.

Schwartz, Felice N.

1992 Breaking with Tradition. New York: Warner Books.

Shamir, Boas, and Irving Salomon

1985 "Work at home and the quality of working life." Academy of Management Review, 10: 455-464.

Silvestri, George, and John Lukasiewicz

1991 "Occupational employment projections." Monthly Labor Review, November: 64-94.

Stehr, Nico

1994 Knowledge Societies. London: Sage.

Swiss, Deborah J., and Judith P. Walker

1993 Women and the Work/Family Dilemma: How Today's Professional Women Are Finding Solutions. New York: Wiley.

Thompson, E. P.

1967 "Time, work-discipline and industrial capitalism." Past and Present, 38: 56-97.

Van Maanen, John, and Gideon Kunda

1989 "Real feelings: Emotional expression and organizational culture." In L. L. Cummings and B. M. Staw (eds.), Research in Organizational Behavior, 11: 43-103. Greenwich, CT: JAI Press.

Walton, Richard E.

1985 "From control to commitment in the workplace." Harvard Business Review, March-April: 77-84.

Whalley, Peter

1986 The Social Production of Technical Work. Albany: State University of New York Press.

Whyte, William H., Jr.

1956 The Organization Man. New York: Simon and Schuster.

Wilkins, Allan L., and William G. Ouchi

1983 "Efficient cultures: Exploring the relationship between culture and organizational performance." Administrative Science Quarterly, 28: 468-481.

Zerubaval, Eviater

1981 Hidden Rhythms. Chicago: University of Chicago Press.

Zuboff, Shoshana

1988 In the Age of the Smart Machine. New York: Basic Books.

Zussman, Robert

1985 Mechanics of the Middle Class. Berkeley: University of California Press.

~~~~~~~~

By Leslie A. Perlow, University of Michigan Business School