Telework in its various guises has been a common practice for a long time. Office workers have traveled, worked at home during bad weather, taken work home with them, and in other ways worked outside of an assigned office for many years, and the practices and activities associated with those patterns are hardly new. What is new, however, is that telework can now be understood as not just a useful, if marginal, epiphenomenon of office-based work, but an indicator of, and support for, a much more fundamental restructuring of the firm itself. Moreover, the transition of telework from an informal and marginal practice to a formal, indeed even necessary, infrastructure for many firms in a knowledge economy, has significant implications beyond the firm.

There are a number of partial synonyms for telework. Thus, for example, “e-work” emphasizes the importance of Information and Communications Technology (ICT) in enabling telework, but does not cover those situations where ICT is not used, such as taking papers home over a long holiday. “Telecommuting” refers to the classic case of working at home rather than commuting to a centralized office location, but this is only one of many possible telework situations. “Telework” is broader, however, and is coupled to a fundamental trend: the decreasing importance of time- and place-based work in the knowledge economy.

In practical terms, however, the terms associated with telework have not yet stabilized, in large part because the underlying work patterns are not discrete, but constitute a continuous scale. Some people take work home once a month; some people work from “virtual offices” in their homes and never go to a centralized work location. Both are “teleworking”, but the implications of the two patterns - for corporate structure, for individual performance, for the environment, for cost and benefit calculations - are very different. Nonetheless, it is possible to draw some preliminary conclusions regarding telework in the United States.

While the number of teleworkers can vary significantly depending upon the particular definition used, most researchers report very similar teleworker demographics. Teleworkers are classified proportionally in comparison to non-teleworkers across all marital statuses, racial/ethnic, and age groups. They are more likely to be from the Northeast and West, are generally higher educated and earn higher incomes. Moreover, telework is found across all industries and all occupations, from technical/managerial to services such as call centers, although there are obviously some occupations, such as manufacturing line jobs, where telework is not feasible. While teleworkers typically are more highly paid and educated, telework does not require an advanced degree or an executive position - about 22% have a high school degree or less, and about 13% of the administrative support or clerical universe works from home. Telework is predominantly found in very small and very large organizations.

At this point, several important points about the data regarding telework must be made. First, much of the data we will use throughout this paper is from the United States; in general, it is difficult to get data from other countries, and more difficult to validate them. Second, many of the numbers cited, as above, will be estimates and extrapolations by industry consultants: these must be treated with some caution, as they are as often wishful thinking as anything else. Third, definitional confusion is endemic in this area, so numbers that purport to refer to the same phenomenon may, in fact, not be comparable. Finally, many numbers are based on surveys that measure not existing practices but posit hypotheticals, such as “If you could telework, would it make your marriage stronger?” The results can be informative but, as with any other attitudinal survey, should be used carefully.

In fact, it is generally true that there are not a lot of hard data underlying many of the current discussions of telework, and what hard data do exist are usually not relevant to the institutional and operational challenges faced by the practitioner, or those responsible for implementing telework within their firms or organizations. Indeed, this paper is in large part based on, and an effort to introduce, data that AT&T has developed over 8 years of tracking internal telework. These data are survey-based, but they reflect input from real teleworkers and their managers, are internally consistent over time, and are cross-referenced with other company data, giving us considerably more confidence in them than in many of the existing public sources. Thus, when we refer throughout this paper to AT&T data, it is not because we both work for AT&T, but because we believe these data to be among the more reliable available.

One of the difficulties in discussing telework is that it is a practice embedded in several dimensions simultaneously, and which dimension is important changes with the question or issue one wishes to deal with. Thus, for example, it is trivial if a manager takes her work home to finish up, even though work will be done away from the office (trivial, but not necessarily completely unproblematic: Is any of the work government classified or critical intellectual property of the firm, for example?). On the other hand, to move entire offices from a centralized work pattern to virtual is neither trivial nor impossible, but is clearly strategic: it significantly changes productivity and personnel retention rates, family structures, management considerations, and the like - and must be done while maintaining output and function. At AT&T we have successfully done several such flash-cut pilot projects, but they need to be carefully managed, and planned in light of all the dimensions of telework. Thus, it is useful to lay out an overview of several intersecting sets of dimensions for the reader ab initio. These will not be completely explicated here, but are important frameworks for understanding our subsequent discussion.

Telework as a symptom of the shift to a knowledge economy. From this perspective, telework is perceived not as just an activity, but as one of a number of activities supporting the evolution away from a manufacturing mental model, and towards a knowledge economy. In this sense, one can pose questions such as: What do telework and dress-down days have in common? (The answer, of course, is that they both represent the boundary between “work” and “everything else,” which is both clear and absolute in a manufacturing economy, weakening and becoming porous in a knowledge economy). Here, telework is just another facet of a much more fundamental evolution away from manufacturing as the structuring force in developed societies. This approach leads to . . .

Telework is also a lens through which to view these more fundamental evolutionary economic and social trends. This is similar to the “model” perspective, but more inclusive: for example, individuals who move to a virtual office environment may well find themselves exploring different ways of managing, and taking advantage of, their intellectual capital: telework becomes the means by which they perceive, and explore, the much broader question of how they fit into a knowledge economy.

Telework, especially when it involves virtual offices, is a significant change for the family. While in most cases both teleworker and family report enhanced quality of life as a result of implementation of telework, there are a number of situations where telework may be difficult: small children at home, inadequate space for a home office, marital stress, or the like. To avoid dysfunctional situations, it is always a good idea to make telework agreements conditional on both the firm and the employee benefiting (see Appendix A for a sample policy incorporating such requirements).

Telework is a challenge for the implementing unit. Our experience has shown, for example, that the approach that might seem most cautious and responsible - gradually implementing telework, starting with only a few superior employees - is often the most dysfunctional, implying that telework is an abnormal “employee perk,” rather than a strategy to enhance corporate performance and productivity. Moreover, managers under such conditions tend to continue their old practices of judging productivity by “time at desk” rather than by actual output, and fail to make the adjustments necessary to function well in a knowledge economy.

Telework is a transforming practice for the firm. If implemented appropriately, telework decreases costs and operating expenses, increases productivity, allows firms to access knowledge sources they were unable to previously take advantage of, and enhances the quality of life of their employees. But to take full advantage of such opportunities, firms must also evolve themselves from facility-based creatures to intranet-based networks, which, as we discuss in more detail below, has significant implications for virtually every function in the firm, from human resources to security to information technology.

Telework is a transforming practice for society and its underlying culture. In the short term, telework has a number of effects, such as reducing traffic congestion, increasing economic productivity while enhancing quality of life (if implemented properly), and, by dispersing information and workers, making a society less susceptible to disruption by disaster or attack. In the longer term, the effects are more difficult to predict with certainty, but may be even more profound: creation of an “information worker elite,” for example, or enhancing the ability of poorer rural communities to attract and benefit from a wealthier knowledge worker population, no longer tied to any particular locale by corporate fiat.

Across these dimensions of telework cuts a broader organizing principle, that of the so-called “triple bottom line,” or TBL. This principle, which holds that corporations should seek to optimize not just economic performance, but also social and environmental performance, derives from a number of sources. To begin with, it is an implementation algorithm for the even fuzzier concept of corporate social responsibility that has become popular, especially among human rights and environmental non-governmental organizations (“NGOs”). It facilitates management of complex issues by providing a simple categorization process, but the virtue of simplicity can and has led to superficial analysis and wishful thinking in lieu of intellectual rigor. Most fundamentally, of course, incorporation statutes almost universally require firms to be economically responsible (Vagts, 1973); to the extent environmental or social activities would undercut this legal responsibility, tensions are obviously created. Thus, while it is obvious that most firms of any size are socially and environmentally active to some extent, it is equally apparent that the firm as an institution is severely conflicted if it is perceived as the major vehicle by which social and environmental goals are to be achieved, absent regulation of some sort. But that does not mean that firms cannot create substantial benefits along each dimension of the TBL, especially where they align, as telework makes clear (Allenby and Richards, 1999).

The social benefits also cut across all levels. The individual gets higher quality of life; the firm increases its economic value while reducing the externalities that working practices impose on the environment; and society manages to increase economic productivity and quality of life at the same time. A significant social benefit at all levels is also the inclusiveness enabled by telework: seniors, the disabled, or others that find place-based work practices difficult are able to contribute in a telework environment without having to travel.

The environmental benefits are also apparent. These may be quite subtle; for example, not only does a teleworker eliminate her emissions if she doesn't commute, but she also reduces traffic congestion marginally for every other car on the road, thus improving the emissions profile even for those who are commuting. Similarly, most automotive infrastructure is sized for peak period traffic: reduce the peak, and in many cases you will reduce the need for new construction, thus avoiding the energy, resources, and impacts on local biota which would otherwise be necessary.

A useful and simple way of visualizing some of these dimensions at play is provided by Figure 1.1, the Telework Compass (Allenby, 2002). The horizontal axis goes from the already familiar short term, tactical approach to the almost completely unknown, longer term, strategic approach. The tactical end of the scale is comfortable to most managers today: most organizations are already teleworking to some degree, and many employees and managers already travel significantly, spending much of their time on the road rather than in their offices. At the strategic end of the scale, however, it is apparent that telework is an important component of a fundamental shift away from the manufacturing structure that still characterizes the U.S. economy, even as it has become much more of a knowledge-based, networked structure (Castells, 2000). This is a critical understanding from a policy as well as an implementation perspective. It explains in part why virtually all institutions, from firms to governments, maintain practices and regulations that are increasingly anachronistic, if not dysfunctional. For example, most firms still have some form of time reporting, usually based on the fiction of a 40 hour week, and even those firms that are familiar with telework, such as AT&T, require that it

be coded as an exception in those systems. In reality, of course, it has been many years since the typical manager worked 40 hours a week, and the fiction is maintained because it was the standard manufacturing week. Similarly, for many managers the default option for measuring productivity is TAD (“time at desk”), a relic from manufacturing plants where time on the manufacturing line does directly equate to productivity. In a knowledge economy, of course, TAD is not only irrelevant, but dysfunctional; the critical function, thinking by the knowledge worker, is not limited to any particular time or place. TAD remains a management favorite because it is easy and, more subtly, because no one knows how to quantitatively and objectively measure the productivity of a typical knowledge worker.

Thus, the telework compass reminds one that even as the details of telework may be completely prosaic and familiar, telework itself reflects a fundamental change in economic structure. Boundaries fixed since the beginning of the Industrial Revolution, with its emphasis on fixed, large-scale manufacturing of identical commodities, are melting. Telework thus aligns with other visible examples, such as casual clothing in the workplace, flextime, increased reliance on contractors and project-specific workers, and increasingly fluid relationships among firms (Castells, 2000), that also reflect these changes. Moreover, telework and those technologies to which it is tightly coupled (e.g., broadband to the home, advanced personal computational platforms of all kinds) form a suite of capabilities that enable the evolution of the netcentric firm, as facilities-based asset driven organizations turn into knowledge networks. Of course, physical assets won't disappear, but new information structures and knowledge will be critical in utilizing them to their fullest capability, and maintaining competitive parity.

Telework thus occupies a spectrum between the mundane practices of today, and the only dimly perceived knowledge economy of tomorrow. But that is not the only relevant axis. It is also important to recognize that different forms of telework require very different approaches, and involve very different degrees of change both for the firm and for the individual. There is a significant difference for the individual and the firm between teleworking one or two days a week, and being completely virtual. In the former case, meetings, information technology access, fixed office space, and other structural elements of the traditional workplace remain available, and occasional teleworking thus requires relatively little adjustment in existing practices. Once an employee is primarily virtual, however, meetings must be carefully planned and can occur anywhere. Moreover, information technology issues such as network security, hardware and software maintenance and upgrade, control of intellectual property, and access to broadband “pipes” so large files can be manipulated, become critical. From a management process perspective, mechanisms to ensure continuing communication between the virtual employee and the organization must be planned and monitored over time for success. With part time telework, managers can still use “time at desk” performance metrics, but with virtual employees, they must actually understand what the value of the employee to the firm is. Moreover, the level of trust that must exist within a virtual organization is much higher than that required for occasional telework. From the individual's perspective, significant adjustments to family patterns may be necessary. The challenges of managing one's time and workload as a VO employee are qualitatively different than working at home one or two days a week.

Thus, the telework compass reinforces the complex nature of the challenges posed by a practice such as telework, and the difficulty of constructing an appropriate conceptual framework. But the importance of doing so should not be minimized, either. While it is not necessary to consider every dimension when considering particular telework issues, it is important to know what dimensions are applicable to the problem at hand. Those who only grasp one element of telework are quite likely to have difficulty implementing it, or fail to encourage all the benefits that telework can provide. Ignorance may be bliss, but it is also suboptimal.

Paradoxically, the transformation of the entire firm depends upon the choices made by individual employees. Forcing employees to work at home - mandatory telework - isn't a valid option for those people who cannot do so for any number of legitimate reasons (such as the lack of a quiet home office), while independent contractor, full-time piece-work-at-home jobs are more scam than reality. Employees decide to telework (and most teleworkers are employees (Nilles, 2000)) based on a variety of cultural, managerial and technological factors extant in both in the traditional office and at home. Managers can encourage telework, and the ICT group can install remote connectivity, but at the end of the day it is very difficult to force an employee out of the company's offices.

Non-mandatory telework programs, on the other hand, must address the potentially endless number of excuses by those who argue that their jobs absolutely must be done in the office - it's very difficult to tell someone that their critical issue doesn't merit an in-person meeting. As an example: We worked with a sales manager who had both teleworking and non-teleworking sales personnel in his organization because of a merger. There was very little middle ground between the two camps. The teleworkers from Company A (including the manager) were almost fully virtual, while the non-teleworkers from Company B worked from home very little, if at all. They all sold the same set of products to the same customer segments, under the same lines of management. Since one of the driving forces for mergers is increased efficiency, the manager began to eliminate expensive office space by making the non-teleworkers move out of dedicated company office space and into home offices. A quiet but effective mutiny resulted. The manager was flooded with persuasive, forceful justifications for office workers to continue working from corporate offices, with copies to his manager, his peers and supportive customers. In the end, the manager had to dramatically decrease the speed of the transition, keeping the office space in place while increasing occasional telework to the point at which a smoother transformation to a virtual environment could be made.

Why do individuals choose to telework, given that they're able (that is, they have access to the opportunity and the required technologies)? The answer is clearly the increase in the teleworker's quality of life. These benefits, which primarily involve a better balance between work and family, are well documented and reported. Doherty, Andrey, and Johnson (2000) acknowledge a “large body of research, including both employee surveys and program evaluations, indicat[ing] that non-work factors, particularly those related to work/life balance, are a primary motivation for employees to opt for telework arrangements.” Lovelace (2000) identifies the three main employee benefits as flexibility in work schedule, decreased lost time commuting, and increase in quality of life, while a better balance of work and family has been the top reported benefit in our AT&T studies of employee teleworkers over the last several years.

As Senge (1990) notes, traditional organization structures either consciously or unconsciously create conflict between work and family. Increased success at work often leads to increased work hours and time away from the family in a self-reinforcing cycle. By shifting the location of the work, telework helps break this cycle, but the more important element is time. For one thing, improved productivity (i.e. getting more done per unit time) has historically immediately followed work/family balance in the pantheon of advantages cited by AT&T employee teleworkers, and having more work to do than hours available for that work is a condition that affects families as well as businesses. But it may be time shifting that contributes the greatest work/family benefit: The teleworker gains the ability to schedule and control the time required to manage all the different responsibilities he or she faces. Knowledge work isn't bound by the clock and many times can take place asynchronously, while family responsibilities cannot all be attended to outside of the traditional 9-to-5 workday. Pratt (1999) found that the three most common reasons for absenteeism are family issues, personal illness, and personal issues, and that the flexibility in scheduling due to telework saves companies about $2000 per teleworker per year in decreased absenteeism.

Because the work is now as close as the nearest personal computer, teleworkers also may work longer hours. ITAC reported in 2001 that about 60% of teleworkers agreed or strongly agreed with the statement that “because you work at home, you work longer hours”, but also found about the same percentage disagreed or strongly disagreed that telework “interferes with other activities in your personal life,” reinforcing the work/family balance advantage. The study goes on to note (Davis and Polonko, 2001, at 4):

“Some believe that work and family roles are more likely to come into conflict for those who work at home. They neglect to consider, however, that work-family conflict may emerge for anyone that takes work home, for example, teachers that take home tests to grade, not just teleworkers. . . . Both non-teleworkers and teleworkers that work at home experience work-family conflict, but their type of conflict is different. While teleworkers may experience some personal costs, such as working longer hours because they work from home, they also experience significant benefits compared to non-teleworkers in terms of less interference between work and family roles. These findings suggest that critics of telework have been simplistic, and perhaps as a result of training and experience, teleworkers may become better able to manage at least several important aspects of work-family conflict.” (Davis and Polonko, 2001)

In terms of work hours, overwork due to telework is more fear than fact; only about 13% of AT&T teleworkers cited it as a significant disadvantage of telework in 2001, while 71% say that telework demonstrates that the company cares about the employee. Our data suggests that the number of hours in a particular individual's workday is roughly the same in the office and at home. In 2001, for example, our employee teleworkers reported working only 18 minutes more per day at home than they did in the office, while non-teleworkers only worked 6 minutes less per day than teleworkers as a whole. On the other hand, a very significant difference between home and office does exist in the number of productive hours per workday, or the amount of time a teleworker feels that he or she is accomplishing personal objectives or contributing to the accomplishment of team or organizational goals.

An often overlooked benefit of telework is the feeling of trust it engenders in the employee. To be free of line-of-sight supervision is a powerful statement of confidence. Telework is trust in action - tangible evidence of confidence and reliance that can't be duplicated in an office. This trust operates at multiple levels as well. It is not just increased trust between manager and teleworker, but the teleworker and their peers. As such, this trust is not just good for morale, and developing the initiative and responsibility of employees - it also seems to encourage the ability to team with others across organizational and geographic boundaries, and thus supports movement away from unnecessary hierarchy.

In this regard, the image of teleworkers as independent contributors or loners in the corporate world is not correct. Despite working at a location away from their peers, almost 4 out of 5 (78%) AT&T teleworkers indicated a “very high” or “high” level of task interaction with co-workers in order to carry out their work activities (Golden, 2001). 17% indicated a moderate level of task interdependence, while only 4% said their level of task interdependence was low. And these high levels of task interdependence seem to be independent of the relative penetration of telework into the work group. About half of the teleworkers said that 30% or less of their organization teleworks, while about one-quarter said that 70% or more of their work unit works from home.

As we'd expect with this potpourri of benefits, a teleworking employee is a more satisfied employee. This must hold true in our voluntary, self-selected model, and it does. The percentage of AT&T teleworkers reporting much greater job satisfaction (i.e. rating their satisfaction 8, 9 or 10 on a 10 point scale) than before teleworking has never dropped below 50% in the last five years, despite a significant amount of organizational and environmental change. The numbers reporting much greater satisfaction with their personal and family lives are equally as high. Many sources have also reported similar findings. Davis and Polonko (2001) found that almost 70% of teleworkers reported that they are quite satisfied or very satisfied with their jobs, with almost 80% saying they are quite or very committed to their organization. Pratt (1999) found that 55% of teleworkers were more satisfied, 7% were less satisfied and 33% were neutral in terms of pre- and post-telework job satisfaction.

These quality-of-life benefits to the individual from telework result in recruitment and retention benefits to the enterprise. Even when flattening hierarchies result in fewer promotion opportunities, firms are better able to retain valued employees. As one example, more than half (56%) of AT&T teleworkers in 2001 who had received competing job offers factored the ability to work at home into their decision not to leave the company. And, if these teleworkers were told they could no longer work from home, one out of three (33%) say they would look for another job within the company - or quit. In the competitive market for high tech employees, firms are finding that it is the companies with more non-traditional work environments that are the most successful in recruiting knowledge workers. The desirability of a telework job is, perhaps unfortunately, underlined by the sheer number of spam-driven work-at-home employment schemes. For the last several years, most messages received through our public website's (www.att.com/telework) feedback mechanism have asked: How do I get a telework job with AT&T?

Finally, one important societal benefit that has not been adequately appreciated is the potential for telework to emerge as a powerful means to engage those who are marginalized by current work-related technologies, practices and cultures. As telework practices and technologies loosen the coupling between work and location, traditional perceptual and physical barriers arising from age, disability, and/or physical appearance also are weakened. Telework offers a powerful way for seniors, as one example, to generate income and remain active in the knowledge economy (and for the economy to benefit from them), as well as to feel useful to society and to continue to be intellectually engaged and challenged. Having reached traditional retirement age, seniors face a quandary. They are too young to truly retire, but are focused on quality of life more than pure income (given the amount of wealth they control). Work on farms and in factories left bodies worn out by age 50 or 60. With knowledge work, that's no longer the case. Seniors are able to contribute effectively for many more years - if they can come to arrangements with the employer. Many older managers and technical professionals - knowledge workers - have been urged into early retirement as a cost-cutting measure. Younger workers are believed to cost the company less, while possessing the very latest skills. However, the number of managers and professionals tends to increase with production or sales, so there doesn't seem to be a net increase in productivity associated with younger employees, while wages for younger managers tend to catch up to that of older employees within two years. (Drucker, 2001)

In the last two years, higher-paid employees have become less likely to be downsized than lower-paid employees (Tischelle, 2002). As we'd expect in a transition to a knowledge economy, companies seem to be placing a premium on the higher levels of business knowledge and insight held by their top-paid managers, while those who are less knowledgeable find themselves at greater risk of losing their jobs. Given that seniors are as productive as younger employees while presumably possessing much more wisdom and experience, does this mean companies will begin to embrace this labor pool? They may not have a choice. As the population ages, so does the workforce. Those firms who understand how to take advantage of this trend will develop a competitive advantage, while those firms who do not will find themselves competing for an employee pool that is steadily drying up.

To begin with, neither the AT&T experience nor the few studies that even consider the issue indicate that teleworkers are motivated to work from home because of the benefit to the environment (e.g., Bailey and Kurland, 2001). Rather, the process of introducing telework within AT&T suggests that the environmental benefits of telework are essentially a by-product of something done for other cultural and technological reasons. In many ways, the obliviousness of environmental regulators and interest groups to telework is surprising, given that the environmental enterprise in the United States is so very large and powerful. Regardless of why a teleworker teleworks, however, the environmental benefits are significant.

The most cited environmental benefits are the gasoline saved and associated reductions in carbon dioxide, hydrocarbons, NOx and other emissions. For example, AT&T teleworkers avoided driving 100 million miles in 2001, the equivalent of about 5 million gallons of gasoline. Since one gallon of gasoline produces 19 pounds of carbon dioxide (CO2), the savings equates to almost 45,000 tons of CO2,, or about 1.8 tons per individual teleworker. There are similar reductions in NOx and hydrocarbons. Since companies tend to locate close to large workforces (i.e. major urban areas), these environmental benefits also tend to occur where air quality and traffic congestion during rush hours are issues. Thus, the teleworker and his or her family directly benefit from the improved air quality, while individual telework also increases the commuting efficiency of others in the community. By keeping the teleworker off the road, traffic congestion is reduced, and so the roads are more open for others, who then spend less time in traffic, consume less gasoline, and emit fewer air pollutants. Telework environmental benefits, in other words, are network phenomenon.

The energy usage differential between the home and traditional office is an interesting environmental factor. Many teleworker energy flows take place irregardless of location. For example, the personal computer, printer, task light, telephone, cable modem, and so on, all must be powered for work to take place. When the teleworker is at home, however, additional energy must be used to connect that worker to the applications and information resident on the intranet network in other places. In an office, connectivity is generally through a LAN (Local Area Network), which does not require classic telecommunications technologies and infrastructure such as switches and multiplexers located in central offices. Teleworkers, on the other hand, must primarily access the corporate network through the traditional telecommunications network, and so additional energy is required. A rough analysis indicates that this additional energy is in the range of 5 kWh/day, which is about three times less than the estimated energy consumption of 17 kWh/day required to commute to work with an automobile (Allenby and Richards, 1999). This analysis is predicated upon the traditional telephone network being used for intranet access (that is, dial up). Further study is needed to understand the impacts to the environment of remote access technology as it evolves, e.g. broadband access via cable modem, or voice over IP (Internet Protocol). On the whole, however, the 3:1 ratio indicates that we're probably very safe in assuming it takes less energy to move bits of information along a wire or fiber than it does to move a 3000 pound automobile 25 miles.

But what about the endpoints of that wire? As the frequency of telework increases from the occasional Friday to full-time virtual officing, workers cross a threshold; they suddenly have the flexibility to work from anywhere they desire. In industrial structures, employment is inseparable from location. People must often move to find work, in many cases leaving behind friends, family and other social support systems. In knowledge-based structures, on the other hand, workers may move wherever they choose to live. The environmental impacts of the resulting demographic patterns over time are yet to be fully understood.

The first cut to be made in understanding the question is to separate part-time teleworkers and full time virtual office workers. The former must still arrange to have a reasonable commute, while the latter, in theory, can live anywhere. Arguably, part-time teleworkers might choose to move further away from the office while keeping their commuting hours and miles per week or month the same; in this case, the environmental effects of the commute - emissions into the air, wear-and-tear on the infrastructure, and so on - may not change significantly. In reality, there are usually strong economic drivers incenting the part-time teleworker to stay where he or she is, and so a strong driving external force in addition to telework (such as a family matter) must be present to incent any significant move.

Virtual office employees are a more interesting case. In those households where two or more people are employed, location still matters unless everyone works from home. The stress of changing jobs and job locations is reduced for dual-income households, because the virtual office worker is able to remain employed as a “trailing spouse”, but employment and location - and thus environmental impact - remain coupled for this household. But what about the virtual employee who does have the flexibility to move his or her household solely to improve their quality of life? There's no firm evidence that telework plays a large part in increasing urban sprawl (Nilles, 2000), which, in any case, has far more powerful drivers than telework, including an automobile-oriented culture, concerns about personal safety and the poor quality of education in urban areas, movement of major corporations to suburban locations (driven in many cases by tax differentials), and the creation of office complexes and parks in greenfield areas. While further study is needed, we have never met an AT&T employee who moved a virtual office further out into the suburbs surrounding an urban area, but we know dozens who have moved to different states and regions of the country. We see telework as playing a much larger role in the revitalization of small rural communities than in increasing urban sprawl.

There are direct financial costs and benefits to the individual teleworker. Over 60% of AT&T employees in our last research cycle said that the monetary savings for the individual and for the company were two significant benefits of telework. The most obvious place the average employee teleworker saves money is in reduced commuting cost. The 100 million miles those teleworkers avoided driving in 2001 works out to roughly $250 per year each in gasoline alone (at an average mileage of 20 miles per gallon and cost of $1.25 per gallon). If wear-and-tear is added, approximated by the IRS business expense reimbursable amount of $0.31 per mile, the per teleworker average benefit jumps to over $1200, or $100 a month. Add parking, tolls, a larger inventory of clothing, dry cleaning, and the like to this amount, and the monetary savings to the individual become significant very quickly.

On the other side of the equation, of course, are any increases in cost borne by the individual. Most of the infrastructure costs, such as increases in electricity and other utilities, potential increases in the size of the home, and incidental expenses such as light bulbs and cleaning, are extremely hard to quantify and, in our experience, generally not reimbursed. About 60% of teleworkers have separate or dedicated home office space (with unused bedrooms being the most often utilized space), and own homes that are about 500 square feet larger on average (Nilles, 2000). The question of whether teleworkers actually purchase larger homes because of telework is still unresolved. For one thing, the causality is unclear; the presence of adequate space might be part of the conditions that trigger the desire to begin working from home. Teleworkers can also afford to purchase larger homes, since they have higher incomes, and, as most people who have bought homes can testify, affordability is paramount in the purchasing decision. On the whole, however, the presence of good quality office space in the home creates a favorable environment for telework and is a leading indicator of the transformation of the firm.

The most interesting cost data is in those expenses that the employee will voluntarily bear in order to gain the benefits of telework. About 36% of Americans would choose a job with a lower salary in order to telework (Winston, 2002), and AT&T's internal experience suggests that employees will gladly pick up the cost of the required equipment and connectivity. For example, about 25% of frequent AT&T teleworkers use their own personal computer to work from home, while only about 40% have a company-paid voice line and about 20% have a company-paid data line. In part, this is due to the lack of a required home office “bundle”: While the company makes recommendations on home office equipment and connectivity, the actual decisions are left to the teleworker and his or her manager. The embedded base of personal equipment, coupled with the individual's preference and needs, tends to drive many of the purchasing decisions. For example, during a large scale virtual officing initiative, in which a standard technology bundle had been defined, teleworkers resisted the provisioning of company-supplied printers. Many employees already had personal printers, and had no room for another, while other employees preferred to purchase their own new printer with more features than AT&T would buy.

Over the last several years within AT&T, we've observed a bifurcation: Frequent teleworkers appear to be working from home even more frequently, while occasional teleworkers spend more time in the office. We believe that this represents a shift in telework from an employee perk (as in working a Friday at home in order to attend a child's ball game in the afternoon) to a productivity tool. Large amounts of organizational change may cause some employees to feel more comfortable in the office, where they can see and be seen by the powers-that-be, while other employees may turn to increasing their results against quantitative and qualitative business objectives, something that is arguably done better at home than in the office.

In fact, the largest benefit of telework to AT&T - and, by extrapolation, to society - is in fact productivity, both because the teleworker can now use previously non-productive commute time and because teleworkers report being more productive per unit of time. For example, teleworkers report 7.4 productive hours (the amount of time he or she is accomplishing personal objectives or contributing to team or organizational goals) in a regular workday at home, almost one hour more than the 6.5 hours of productive time achieved when working in the traditional office. Across AT&T, about 62% of all teleworkers reported higher productivity at home while only 6% reported higher productivity in the office. These data are self-reported, but well supported by other research.

Thus, many other authors and researchers have found similar productivity increases. Nilles (2000) found a mean productivity increase of almost 15%; multiplying the telework-related productivity change by the respondents salary resulted in an estimated impact of $9,172 per teleworker. Pratt (1999) found that 47% of teleworkers reported higher productivity when working from home, 42% said their productivity remained the same, and 10% found they were less productive when working from home. While other researchers have questioned the reported productivity increases from telework, mainly because of the lack of a control group in the office (see, for example, Doherty, Andrey and Johnson, 2000), these researchers err in fundamentally assuming that a hard line may be drawn between office workers and teleworkers as two separate and distinct groups. The matter is far more complex and blurry; for example, given that teleworkers are generally either self-selected or allowed to work from home because of higher performance, creating an equivalent control group in the office may not be possible.

One approach to a control group, however, is to define the universe at a corporate, rather than organizational, level. Across AT&T, virtual office managers are more likely to be rated in the very highest performance management category - as measured by the formal, managerial appraisal - than their office-bound peers. While the causality might go the other way (i.e. more highly rated employees are more likely to be allowed to telework), this possibility lessens as we move the firm more and more into a fully virtual configuration. For example, only those individuals with poor performance (the bottom 20% of the management universe) are ineligible for participation in many organization-wide virtual officing projects. In an independent study that also looked at the corporation as one entity, AT&T teleworkers received larger annual raises than non-teleworkers (Golden, 2001), and telework was related to increased performance levels as indicated by those higher salary increases.

At least in the current configuration, individuals, not organizations, are the driving force behind telework and virtual officing. They choose to work from home primarily because of an enhanced work/family balance brought about by the disassociation of location and time from work. Freeing the worker from these industrial-age constraints increases personal productivity, which in turn increases business efficiency. The advantages to the employee outweigh any incurred costs; in fact, many employees end up paying for business infrastructure in order to be able to work from home. Taken together with the benefits of telework at an organization and company level, we see the emergence of a true triple bottom line initiative.

At first glance, implementing telework within company subunits might seem to be simply a smaller, less complex case of implementation compared to a firm-wide initiative. Implementing organizations can simply build off the company-wide infrastructure (policies, tools and technology) that is presumably already in place, capturing best practices from those that have gone before and helping to pull along those who aren't quite ready. In practice, however, implementing organizations exhibit a surprisingly degree of diversity when it comes to telework and virtual officing, even when bound by the same functional, cultural and technological constraints existing inside a single firm. In this chapter, therefore, we address telework in organizational subunits of the firm, from small, single-location organizations to larger functional organizations (such as Environment and Safety groups). While we draw on the broader literature, our discussion centers on a number of AT&T telework pilots at various scales, which give us pragmatic experience to supplement what is too often speculative literature packaged as analysis.

While a firm's total employment may be too large to be comfortably located inside one single building, this is far less likely to be true at an organizational level, especially as what we define as an organization gets closer and closer to a single manager and his or her team of direct reports. Thus we see that while a company may already be geographically dispersed, and therefore have a management system that takes collaboration and communication across distance into account either formally or informally, some organizations within that company may be so tied to a single location that they equate telework with the end of the managerial world as they know it. Indeed, it is common in our experience for different organizations within an enterprise to exist at different points of the location-based spectrum, from inseparable to fully virtual. Field support and sales organizations, for example, are normally more decentralized than the overall firm, while groups that are located closely together in physical space sometimes cannot conceive of working any other way, even if the work is almost entirely computer- and telephone-based.

This diversity helps and hinders the adoption of telework within a firm. Highly virtual organizations can be used as public and private role models, are an important source of innovation for the enterprise, and tend to be very attractive to job seekers. Conversely, recalcitrant managers can bring corporate-level programs to a dead stop in their organizations, holding firm against the wishes of their employees - and company policy - by using their managerial discretion to preserve increasingly dysfunctional, but still comfortable, manufacturing era work practices. In actuality, it is difficult in practice to identify the true issues in such cases: is it a real need for face-to-face communication, a fear of disruption and the unknown, the ego of the manager who (naturally enough) enjoys walking into the office in the morning and viewing her or his empire of people, or simply the aversion to change that characterizes individuals and all human organizations? Motivations are usually complex and entangled in the real world, and usually implicit and even unconscious, but trying to understand the significant drivers of individuals and identifying leverage points is important to achieving change, and thus a worthwhile exercise.

This diversity within organizations can be illustrated by Figure 3.1, the Cultural Readiness / Technology Penetration matrix. This matrix suggests that, for an organization or enterprise, virtual officing and telework are gated and driven both by technology (specifically, the degree of transparency between the home and the traditional office, essentially a function of the firm's netcentricity and the worker's remote access to it) and culture (the ability for the organization to embrace individual and team work that are independent of location). For example, one of the reasons that AT&T has been successful with telework is that the firm grew up with employees working remotely from one another: after all, the genesis of the firm's core technology was Bell speaking to Watson from another room. With thousands of staffed locations, the odds are high that any given employee will be working with another employee somewhere else (or a customer, or supplier, or partner, for that matter). Managers have had to manage beyond their line-of-sight, regardless of telework. In other words, AT&T's culture - like those of many multinationals, which are similarly dispersed in time and place - was already primed for the transformational abilities of information and communications technology.

Moving to the technology axis, one of the striking aspects of the AT&T intranet is the sheer size of it. As of December, 2002, the AT&T internal search engine indexed about 250,000 .html web pages, a count which doesn't include the pages created on-the-fly by interactive applications. In all practicality, a project really isn't a project, or a work group really a work group, until it or they have a website. Driven by the need for business efficiency, both managerial processes (such as human resources, purchasing or payroll) and job functions (such as network operations; or marketing; or environment, health and safety) have moved almost entirely to the AT&T intranet. The combination of a receptive culture, the evolution of netcentric information systems, and remotely accessible technology are leading to, and indicative of, AT&T's evolution towards the virtual enterprise it is becoming.

This matrix underlines the significant differences between work structures that support frequent telework - even as often as one or two days a week - and those that support virtual officing. Almost anyone can work at home part of the time, even within organizations that are consciously or unconsciously modeled on manufacturing-era structures from a cultural and technological perspective. The structural elements of the traditional workplace mostly remain in place - face-to-face meetings, office-based voice and data access, and fixed office space are good examples, along with other, less visible structures such as security and political systems. Employees aren't absent from the office enough to cause any cultural tension, and they manage to configure the equipment and connectivity available to them enough to work remotely to some level of adequacy. Thus, even frequent telework requires relatively little adjustment to the formal and informal systems used to produce work output.

In a virtual office environment, however, the boundaries of location and appearance essentially disappear. It is not a simple matter of increased telework participation, but rather a fundamentally different view of an organization and how it works. Meetings and information exchange can occur anywhere at any time. Management processes must be based on performance, not time at desk or in-person charisma; managers must actually understand the value of each knowledge worker's contribution to the enterprise. The level of trust in a virtual organization must be much higher than in a frequent telework environment; managers must be comfortable knowing that the employees they support are working away despite being out of sight, and employees must feel comfortable that their manager is looking out for them, even when they're not immediately outside her or his office.

The first point to be made when examining the benefits of telework at an organizational level is that our research indicates that those benefits increase on a per-person basis as participation rises, providing an exponential increase in the total benefit to the organization. Telework is an excellent example of network economics, in that increases in the number of teleworkers cause each new teleworker to receive more benefit than the one before. Thus, it is interesting but not surprising that larger employers now lead in the adoption of telework, despite a slower start, than do small- and medium-sized businesses (ITAC, 2000). Higher telework participation within a work group increases individual teleworker productivity and job satisfaction, which drives even higher participation, which increases the benefit, in a self-reinforcing cycle. Moreover, higher participation validates, and encourages, greater investment in ICT and migration towards a netcentric firm structure, for not just telework, but the evolution of a netcentric firm, is an autocatalytic and synergistic process. On a small-scale basis, the first teleworker in a work group is an oddity, and often has a difficult time succeeding due to the lack of widely deployed, robust information and communication technologies needed to perform the job, along with the forces associated with breaking away from an in-person, facilities-based culture. When somewhere around half the people in that work group are working from home, however, the communication and management patterns (and, by implication, the culture) change dramatically and the required technologies have obviously been put in place. Telework is no longer an unusual event; it becomes embedded in the business and, as such, creates the foundation for greater increases in organizational efficiencies and effectiveness, or, alternatively, greater decreases in the issues or problems associated with telework. The AT&T-wide data shown in Table 3.1 are from the 2000 AT&T employee teleworker survey:

Note how barriers such as reduced visibility, isolation and loneliness have all decreased as participation has increased, as expected - for example, visibility now begins to refer to presence on the network instead of attendance within a building. As the team moves to a more virtual configuration, then camaraderie also becomes virtual, in the form of telephone conference calls rather than hallway conversations, or instant messaging rather than over-the-cube-wall chat. Loneliness has historically been one of the smallest problems associated with telework in our research; we suspect this is because work is only one part of the social environment surrounding an employee, which would also would include family, friends and community groups. No teleworker is an island.

In many cases, the triple bottom line benefits of telework at an organizational level are more or less a proportional subset of the benefits at the level of the firm, modulated by network economics and other factors relating to the relationship between the organization and the enterprise(s) it belongs to, such as the greater potential for an organization to be located in a single building or work area. For example, job satisfaction and work/family balance are two important social benefits of telework that increase with participation (both in theory and according to our data), and so members of highly virtual organizations are likely to have a higher quality of life and more work freedom than the worker in the firm who must migrate into an office each day. In this case, the organization is probably the more controlling variable, since it encompasses the employees and manager that are closest to the teleworker within the reporting structure, but the participation levels of the firm as a whole also play a not insignificant role. For one thing, employees move around from job to job within a company, and thus both highly virtual organizations and highly place-based organizations are subject to an incoming flow of workers from the more moderate larger firm. The higher levels of satisfaction with job and work/family balance associated with telework offer a recruiting advantage to the organization that embraces it, but retention is a double-edged sword: Employees are less likely to leave a teleworking organization and open up opportunities for others to move in. Virtual organizations also may be more able to identify key talent and access that talent without regard to location or employment status (e.g. hire retired employees as contractors) due to their increased experience with remote work, providing another advantage over other, less-telework-friendly organizations within the firm. And the greater potential for organizations to be based at a single location means that those groups can combine a high degree of virtual officing with the ability to have in-person meetings as needed, producing societal benefit based on having the best of both worlds.

Enhanced security is another societal level benefit that differs at the level of the organization compared to the overarching firm. Broadly speaking, a geographically dispersed information and employee structure is more resilient to a point-based threat or natural disaster than a structure concentrated in one location. A building can be shut down because of a snowstorm or bomb threat, as examples, while a network of teleworkers cannot. Thus, increased business and operational continuity is an important societal-level benefit of telework, as is the quality of life improvement to those employees (and their families) who may be uncomfortable working in a large office building and prefer the familiar surroundings of their home. Geography, then, is the first key difference in telework security for the subtending organization as compared to the firm. The short-term functioning of the company is dependent upon the operations of a certain minimum set of organizations within the firm. A single organization may be disbursed - and therefore more secure in itself - but if the rest of the critical functions for the enterprise are still concentrated at a small set of physical locations then the risk profile for the firm as a whole may not change significantly. Moreover, since an individual work group arguably has a tendency to be co-located in a single location, a location-based organization represents a point of risk for the entire firm, especially since that firm is likely to be dispersed across a number of locations anyway. It follows that any strategy to reduce the firm's portfolio of risk through a telework and virtualization strategy must start by addressing the diverse cultural and technological needs of individual organizations.

Another important difference between organizations and overarching firms is the management system for security. Common information security policies, procedures and layers of technology that are independent of location must be in place for the entire company. A tangled web of organizational level policies and procedures could lead to the unattractive scenario of higher risk coupled with higher cost; for example, allowing each organization to put in place its own means of access to the corporate intranet, instead of relying on a company-wide Virtual Private Network (VPN). Another important difference is documentation management. Sensitive and confidential documents such as personnel files need not, and should not, be stored in hundreds of garages and bedroom closets. Organizational processes to store these documents must be created; accordingly, many work groups create a “telework coordinator” position to serve as a centralized, secure point of repository for records which must be highly controlled, a practice we highly recommend.

Economically, the benefit trinity of productivity, real estate and personnel recruitment/retention is very important at an organizational level. As they do with society overall, network economics and organizational co-location significantly affect organizational productivity, but because corporate subunits have very differing scales, structures, functions, and responsibilities, the specific productivity gains from telework and the general migration to a netcentric environment vary with organization. Even granting the difficulty of measuring the specifics of knowledge worker productivity, however, our experience overall strongly supports the hypothesis that moving an organization virtual can result in significant productivity gains. For example, one of the earliest telework pilots involved the AT&T Environment, Health and Safety organization, which over the period 1996 to 2000 went from a traditional, office-based organization to completely virtual. Over the same period, a strategic imperative to move the organization's functionality to the corporate intranet was enforced, as it rapidly became clear that telework could be truly successful only if envisioned as part of a broader netcentric strategy. Without, for example, establishing net-based environment and safety tools, data, functionality, and communications systems - AT&T's e-EH&S (sm) web-based system is among the most advanced products in its field - the efficiency gains that telework could generate would not have been fully realized. Similarly, the organization shifted from a hierarchical model characteristic of many technical corporate functions, to a flattened, more team- and project-oriented approach, again enabled by access to appropriate networking technologies, and leading to more efficient operating processes. As part of this component of the strategy, all of the organization's managers assumed both substantive and managerial responsibilities; authority within a particular team was predicated on substantive expertise rather than corporate rank. And, of course, the top management of the organization led by example: three of the top four EH&S managers are virtual, and the fourth teleworks at least three days a week. As a result, over this time period the organization was able to shrink by about two-thirds in expenses while maintaining the same level of output by using a dual strategy of moving information and processes to the web and employees to their homes.

In operation, these organizational pockets of higher productivity may begin to serve as role models for the enterprise as a whole, acting as operating expense benchmarks, for example, or becoming known for being able to quickly respond to urgent issues (the EH&S experience operated as just such a catalyst within AT&T)). This tends to increase the amount of virtuality or telework across the business as managers compete for better results. Because it is easier to identify who is doing what with whom at the smaller scale of an organization, work groups may find it easier to realize the productivity gains of telework in an environment where work volume is shrinking and, conversely, to adapt more quickly to rising work volumes. Efficiency and effectiveness increases at the level of the organization, however, may be gated by the performance of the firm, or by other organizations in the firm. For example, a sales group may be composed of the world's best sales personnel but be unable to sell anything because of higher costs imposed by inefficient service delivery organizations.

Real estate cost reduction, and recruitment and retention of key personnel, are also significant economic benefits of telework for organizations. For one thing, it's easier to manage real estate utilization at the level where presence in the office can be monitored at an individual scale, so an organization can optimize the space actually needed in the traditional office (perhaps assigning most space in a shared configuration, either on a reservation-based or first-come, first-served arrangement). While various methods of determining the amount of required space can be used, including surveys and analysis of employee entry data from automated badge or identification card readers, nothing can replace the classic approach of a manager walking around the office, visually inspecting the rows of (many times, empty) offices. On the other hand, normally only at the level of the firm can entire buildings can be vacated and then sold, leased or otherwise removed from ongoing costs. Incentives should be applied at the organizational level, however, so that managers can directly impact their budget (and, by association, their goal-driven compensation) by turning back even one cubicle to the property management organization. Unused cubicles can be combined together and eventually entire buildings may be vacated. Recruitment and retention are effectively managed at the organizational level, as well; since individual managers generally control the teleworking arrangements of their subordinates, a great deal of the benefit to be had at a corporate level lies in the hands of local management.

When discussing environmental benefit it's important to first note the uncertainties associated with measuring something so complex. Even at the small scale of the organization, telework is a very complex technological, social and economic phenomenon, and so even understanding - let alone measuring precisely - the environmental changes that accrue upstream and downstream of the worker, organization, firm and so on is impossible with our current state of knowledge. This doesn't mean that it's impossible to determine the relative direction of those changes - most studies, and our internal data, conclude that telework significantly reduces automobile travel over and above errands done from home (see, for example, Atkyns et al, 2002, or Shirazi, 2000) - but only that exact measurements require a degree of systems thinking we haven't developed yet. For example, many changes introduced by telework (such as less demand for laundry and dry cleaning services since business attire no longer needs to be maintained), while obvious in hindsight, weren't predicted or well understood beforehand (Allenby, 2001). We can, however, identify several unique aspects of environmental benefit at the organizational level as compared to the firm. Since organizations are more likely to be concentrated in one geographical area and go to work at the same time, moving to a virtual office configuration can have a more concentrated effect on such variables as commuting times and air quality in a particular area. Since congestion occurs across a network of roads, removing even one car from the highway during peak traffic times increases the efficiency of all the other cars on the road, and it follows that organizations have even more leverage in shaving and leveling peak infrastructure demands. Indeed, organizations are more incented to implement teleworking initiatives if they are located in cities with high commuting times and distances, such as Atlanta. Employees can work from home most days, commuting to the office perhaps once a week or once a month to attend in-person meetings.

While individual employees and their managers make the decision to telework, it is at the level of the organization that the primary linkage to business objectives takes place - and where the tactical issues and problems surface. The critical mass needed to create a functioning virtual network is more easily assembled at the work group level, but first the inertia of building-based institutions and infrastructure, and manufacturing era mindsets, must be overcome. Cultural and technological factors are generally common across the organization, potentially addressable with standard approaches but also capable of collapsing the program if unresolved. Communication up, down and across the organization is more easily accomplished than at larger scales, yet still may not be effective enough. Organizational leadership is more visible, and so plays a larger role in the transition to a virtual configuration for better or worse.

Telework policy at the organizational level must support and be consistent with the overarching corporate telework policy (see Appendix A) as well as other firm-wide policies that govern functions such as procurement, security and human resources. Despite this web of existing documentation, there is a place in the firm's management system for the organizational-specific telework policy. For one thing, the services and equipment being reimbursed may vary from organization to organization, based on the needs of the work that organization does. Common expectations and baseline conditions can be codified at the very “local” and specific level that is needed to address and resolve cultural and managerial issues. The common understanding that results from the sheer act of putting a policy together is an important output, even if the document is referred to infrequently afterwards.

A key issue for telework policy is employee eligibility. Generally, all employees should be eligible to telework until proven otherwise. Starting with a defined list of jobs that are “teleworkable” begins to create a separate structure for telework, rather than moves the current structure to the network; assume all jobs are teleworkable until the worker and the manager say they aren't. Essentially, a "good" employee or manager in the office is a good employee at home - a person's work ethic or results focus (as examples) do not change when she or he changes her or his desk. However, many organizational policies do not allow employees with low performance ratings on their appraisals to move to a virtual office, or sometimes telework at all. In any case, the manager's view of the arrangement is critical; part of creating a successful structure for telework is reassuring managers that they still have authority over their workgroups. An important factor in managerial performance is trust. If the manager trusts the employee to be getting work done even when she or he can't be seen, and if the employee trusts the manager to take her or his needs into account even when they aren't right outside the office door, then obviously there's a greater pool of “eligible” teleworkers. If there is a lack of trust in either of these dimensions, then the eligible pool shrinks accordingly. Of course, such a lack of trust tends to be dysfunctional regardless of whether an organization is virtual or office-bound; telework simply forces it into the open where it can be dealt with.

Similarly, the corporate telework policy must be constructed on a platform of faith in local management. The actual telework participation decision is in the hands of the local manager, along with many other tactical decisions such as the variances in the equipment provided each worker. The manager is in the best position to understand each unique job, worker, and business environment, and to manage the culture change involved with successful implementation of telework. While recalcitrant managers can hinder or halt telework adoption, moving teleworker approvals and program decisions to corporate telework organizations or senior management does nothing to alleviate the problem, and arguably makes it worse; programs which require teleworker approval by vice presidents in our experience have less participation than programs which require teleworker approval by direct managers.

The organizational policy and administration should strive to maximize the benefits of telework. Real estate utilization, for example, can really only be effectively managed at a very granular level - of all the people in the firm, the manager and the employee are the best at knowing when attendance in the office is required, or how often the currently assigned office space is vacant. Organizational policies may spell out guidelines for shared, common and dedicated office space for teleworkers. They may also address performance management issues, such as a formal declaration that office workers and teleworkers will be treated equally in terms of managerial appraisals, or laying out quantitative productivity or performance expectations. Issues of recruitment and retention may also be spelled out in organizational policies. Different workgroups may have different mechanisms and procedures for acclimating new members to the team's culture, or accommodating employees who move to another state after going virtual, for example.

Security practices as a rule do not belong in organizational level policies. Like personnel issues such as fair hiring practices, security must be managed at a firm-wide level so that differing interpretations do not result in gaps in coverage. Organizational telework coordinators, whose job it is to maintain files that shouldn't be allowed to become dispersed across hundreds of virtual offices, can help in complying with the firm's information security and intellectual property policies but are not intended to replace those policies. Similarly, data and network security requires a seamless and layered company-wide approach, not a patchwork of organizational policies that leaves openings for theft or misuse. Security at the organizational level is founded upon employee awareness of corporate requirements.

One of the most important things leadership can do to support telework is to continually and effectively communicate that support to their organization. Without such encouragement from above, middle managers in our experience tend to be concerned that it may not be safe to telework. Fears about missing out on promotions or attractive assignments do not surface verbally in discussions about telework; managers simply stay in the office, and most likely keep their teams there with them. Moreover, most organizational managers are sensitive to the “flavor of the day” phenomenon, where whatever management practice has been most recently flogged by a leading business school is temporarily adopted as corporate practice. Implementing a long term, fundamental change impacting management style, real estate demand, personnel morale and productivity, and ability to function, when it might be just another passing fad, is something no good manager would do. Accordingly, there is a barrier - an activation energy - that must be overcome before managers will accept any new practice, especially a fundamental one, as valid and thus worth implementing. In this regard, the entire “Quality” movement, which by some was regarded as the quintessential management fad, was nowhere near as radical in terms of organizational structure as telework. No wonder managers balk.

Effective communication from work group leaders is thus critical in creating an environment where virtual officing becomes the norm. Whether they truly support the concept or not, middle managers will eventually acquiesce in letting their people telework or go virtual in the face of unflagging upper management support. And, in the longer term, if it becomes “corporate cool” to have a virtual organization, managers will feel pressure (from not only their manager but also their peers) to be “corporate cool” as well - especially if increased productivity and employee quality of life result. But the barriers, personal and organizational, should not be trivialized, and may be the more powerful for being generally unrecognized. Thus, for example, one senior manager in AT&T asked, when discussing telework penetration at headquarters, “What do you tell the person who has worked their entire life to get to this office? That you're going to take it all away?” The office is a visible sign of status and power that disappears in the virtual environment, and many managers, especially those for whom such trappings convey status and prestige, will resent its passing just as they ascend to the appropriate level to enjoy it.

Communication within the work group is fostered by a changing variety of information technologies such as instant messaging, data conferencing and web-based tools and databases. In our experience, however, basic email has proven to be perhaps the most important foundational communication technology for the virtual organization. Email messages have evolved over time from formal memorandums that were simply copies of paper documents to conversational threads that span days or weeks and dozens of people. More than any other ICT-enabled communication method, email has removed time and location from the fabric of business. There are obvious productivity implications: the Pew Internet and American Life's Email at Work study (2002) found that almost 9 out of 10 (86%) of employees who use email at work reported that it saved them time, but the most interesting email phenomenon is the way it replaces and supplements face-to-face communication. The same study found that 59% of work emailers said that it improves teamwork, and almost half (43%) likened it to a virtual water cooler, providing a break during the workday for personal conversations or gossip.

This reinforces the point that the image of teleworkers as independent contributors or loners in the corporate world is not correct. Despite working at a location away from their peers, almost 4 out of 5 (78%) AT&T teleworkers indicated a “very high” or “high” level of task interaction with co-workers in order to carry out their work activities (Golden, 2001). 17% indicated a moderate level of task interdependence, while only 4% said their level of task interdependence was low. And these high levels of interaction seem to be independent of the organization's degree of location-basis; that is, communication and collaboration occur regardless if the parties doing it are at home or in the traditional office . In the study, about half (51%) of respondents reported that 1/3 or less of their work unit teleworks while about one-fourth (23%) reported that 40-60% of their work unit teleworks and a similar percentage (27%) reported that 70% or more of their work unit teleworks. Obviously, information and communications technologies have evolved to replace the face-to-face mechanisms seen by skeptics as critical for teamwork and collaboration.

Organizational culture plays a defining role in telework at the organizational level. In our experience, perhaps the most important attribute of that culture in terms of telework is trust. It is incumbent upon the manager to create a trusting environment if he or she wishes to create a virtual organization. The manager must trust the employees to be productive outside of the directly supervised office environment, but the employees must also trust each other and believe that they are being treated fairly. One of the issues raised by organized labor against telework, for example, is that an isolated worker at home cannot know if his or her peers in the office are “making quota”, or if the worker is being singled out by management for higher performance standards (Kistner, 2002). Indeed, performance management based on results, not appearances or presence in the office, is one of the hallmarks of a telework-friendly organizational culture. Good managers know that appearances may be deceiving; in many ways it is easier to manage a virtual employee because in-person charm and charisma are minimized, while results against business objectives and customer/supplier/peer feedback surface as primary indicators of employee success. Another cultural factor that strongly affects telework is a lack of hierarchical “command and control” management. The employee in a netcentric firm has the capability of using the net (e.g. email) to communicate with anyone in the firm, regardless of level or lines drawn on an organization chart; he or she is no longer bounded by office walls, or by the restrictions imposed by having to pass a sheet of paper up the chain of command. This ability to communicate and work with anyone in the company's network, or to create new networks outside the firm, is absolutely essential for a teleworker. In our experience, the three most important things a teleworker can do to ensure success are to communicate, communicate and communicate.

Technology is also central to telework at the organizational level. While some early teleworkers worked from home (sales people, for example) prior to the rise of the Internet, it is the ability of information technology to create knowledge-based networks that has made telework appropriate for more than just a few jobs. Yet each organization is likely to have unique needs for technology. The technology required is dependent not only upon the functions of the job but also the organization's culture, e.g. the amount of communication and the preferred channel for that communication. While it sometimes seems at AT&T that an organization isn't really perceived to exist until it has a website on the intranet, some organizations implement extremely comprehensive web-based sets of communication and information tools for their members, for example, while others may rely more on email, voice mail and instant messaging. One organization may have volumes of incoming paper documents from a wide variety of sources, and so their home offices require fax machines, while another organization may be able to simply use an e-faxing services, potentially coupled with a scanner, for the small number of paper documents they handle. Corporate standards for technology are extremely important to create economies of scale with suppliers and to minimize the costs of support, but flexibility must be provided for at the organizational level. Within AT&T, for example, a variety of telephones are available on the extranet-based office supplies catalog mutually put in place by the AT&T purchasing department and the supplier, but no organization yet has been able to agree on one single telephone that would satisfy all teleworkers. In implementation, it's proven far easier to simply give the teleworker a spending limit and allow them to pick any telephone off the list they desire than to force-fit all teleworkers into one single choice. In cases where we have managed to settle on a single model of a piece of office equipment (say, a low-cost printer) it soon becomes obsolete due to the high rate of product model change and innovation in consumer electronics.

Any telework implementation must address the dual factors of culture and technology. A good example is the pilot program approach to telework deployment. To manage the perceived risk associated with starting up a program, organizations may begin slowly, allowing only a few employees to work from home for a small number of days. The results of the pilot are almost always favorable, since the small number of select employees who took part in the trial are highly motivated to make it successful. Despite the great results of the pilot, however, the organization soon finds itself back at the beginning, with little if any telework in an almost entirely office-based organization. What went wrong is generally easily discernable. First, beginning with a small group of superior employees signals to the rest of the organization that telework is either an employee perk or a performance reward, neither of which really indicates that the organization is implementing a new structure designed to improve overall efficiency and performance. Secondly, significant cultural barriers and technological issues are unlikely to emerge during a pilot program. The pilot does nothing to surface managerial issues such as judging worker performance by “time at desk”, for example, or to identify gaps in provisioning capacity. These problems lurk until deployment, then tend to jump up and knock the fledging program backwards. Finally, when the critical mass of the organization still resides in the office, workers at home will continually feel the pull of the gravity that mass creates. The office is where the work gets done, and the people at home are simply an aberration.

The solution to these issues, of course, is to remember that bigger is better. Pilot or trial programs that involve a majority of the organization send a strong signal to the group that things will be changing. The real cultural issues emerge; managers learn to manage by results by doing it, and workers learn how to move their face-to-face communication over to the ICT-enabled equivalents. Where possible, standard home office equipment package(s) can be identified to reduce costs and streamline deployment. And, most importantly, the critical mass of employees needed for communication and work channels to move from the building onto the network is in place. During the ideal pilot program, the number of people in the office should be smaller than the number at home, and of less relative importance in the power structure.

On that note, the project manager of the implementation effort should come from within the organization. Outsiders can advise and support, but only a person who “lives” in the organization can provide the insights and influence needed to persuade workers to move out of their offices, to reassure managers that their objectives will still be met, or to settle policy matters such as expenses reimbursed or eligibility. The organizations in the case study below both used internal administrative managers to manage the issues associated with deploying a highly virtual business structure, with support from corporate organizations such as the Information Technology group, Real Estate and Procurement.