Sign In

Communications of the ACM

Technical opinion

Does Telecommuting Improve Productivity?

Productivity is an important concern for organizations considering the adoption of a telecommuting option for their employees. If an organization has evidence that telecommuters are at least as productive as their traditional counterparts, telecommuting becomes an attractive option. A number of studies have shown that telecommuting increases productivity. However, many of these are based on self-reported data rather than actual measures. As a result, these findings have been called into question. For example, Bailey and Kurland's meta-analysis of 80 previous studies found little evidence that telework increases job satisfaction or productivity, despite assertions to the contrary [1].

Here, we examine the effect of telecommuting on productivity by analyzing longitudinal data (spanning nearly five years) for call center representatives at Kentucky American Water Company (KAWC), a company serving over 280,000 people. Specifically, we sought to determine whether the following claims about telecommuting could be substantiated in the case of KAWC:

  • Placebo or Hawthorne effects generate temporary telecommuting productivity gains [4];
  • Telecommuting reduces absenteeism;
  • Telecommuting productivity gains are an artifact of the process used to select telecommuters [4]; and
  • Telecommuters can negatively affect other employees by requiring extra support from those employees [4].

Call center statistics were provided for a five-year period beginning in April 1998 and ending in March 2003 (at which time KAWC was consolidated into a national call center in Alton, IL). Two distinct rounds of data collection were conducted—one ending in May 2000 (this is the data reported in [2]), the other in March 2003. During these periods, the number of telecommuters varied slightly from three to four, and the number of in-office agents varied from six to nine. Seven months of data were not available (June 2000–December 2000). KAWC also provided information regarding costs associated with the project.

Back to Top

Effects on Temporary and Long-Term Productivity

Measuring changes in productivity resulting from any IT implementation is difficult. Westfall [3] suggests that telecommuter productivity can be measured based on four factors: amount of work; intensity of work; efficiency of work; and adjustments for additional costs associated with telecommuting (such as expenses for equipment, technology support, training, managerial support, and so forth).

Loy et al. determined there was an increase in productivity for the telecommuters in the year immediately following the "go live" date of the project [2]. To examine whether this claim held true over a longer time horizon, we used Westfall's [3] model to compare the productivity of telecommuters and non-telecommuters over the entire five-year period. Our measures are described here.

Amount of Work. Number of hours each employee worked per month.

Intensity of Work. Number of calls each employee handled per month.

Efficiency of Work. Westfall identified two factors regarding the measurement of efficiency [3]. First, telecommuters may be more productive because they receive additional training and/or are selected to become telecommuters because they are already the most productive employees. Conversely, because telecommuters devote "in- office" time to organizing resources for when they are "out of office" they may be less productive than their counterparts when they are working in the office. KAWC telecommuters did not receive additional training and were not the most productive employees prior to project implementation. Our study uses the actual number of calls answered and the actual number of hours worked, thus how telecommuters spend time "in office" versus "out of office" is irrelevant. The important issue at KAWC is the amount of call volume a customer service agent handles, which is captured by the intensity factor. Therefore, efficiency is not a required factor in KAWC productivity calculations.

Adjustments. For KAWC, cost adjustments include additional management supervision, monthly ISDN service fees, and additional equipment; savings include reduced expenses for office space and parking, and a reduction in complaints handled by the managers. The costs and benefits are summarized in the table here. Assuming that: telecommuters stay with the company for an average of three years; telecommuters keep their equipment at the end of that time; and there are three telecommuters at the start and end of the project, the total cost of these adjustments per telecommuter is $4,306 per year, and the total savings from these adjustments per telecommuter is $5,667 per year. Because there is a net benefit of $1,361 per telecommuter per year, there is no need to include adjustments for additional costs in the productivity model.

Given that efficiency does not need to be included in productivity calculations for KAWC and there are no negative adjustments, productivity (calls per hour) can be calculated as the ratio of the intensity of work (number of calls per month) to the amount of work (number of hours worked per month).

Loy et al. [2] found the average productivity of telecommuters in the 13 months immediately after the "go live" period increased by 154%, while the average productivity of in-office agents fell by 13.3%. Because overall customer service improved considerably (the percentage of abandoned calls decreased from 12.2% to 3.6%), Loy et al. concluded the overall effect on total productivity was positive.

To investigate whether these results were placebo or Hawthorne effects we examined productivity for telecommuters over the long term, comparing productivity in the year immediately following implementation to productivity in the next 27 months. Average productivity for telecommuters increased from 9.4 calls per hour the year after implementation to 10.5 calls per hour in the second year, dropped to 10.2 calls per hour in the third year, then increased again to 11.1 calls per hour in the final three months of operation. Because telecommuters sustained increased productivity in the long run, there is no evidence of placebo or Hawthorne effects for the telecommuters in the case of KAWC.

Back to Top

Effects on Absenteeism

A common claim is that telecommuting reduces absenteeism. We found that, on average, telecommuters worked 3.98 more hours per month than non-telecommuters. Because this difference is not statistically significant, we conclude that, in the case of KWAC, telecommuting did not reduce absenteeism.

Back to Top

Effect of Process Used to Select Telecommuters

To assess the possibility that productivity gains for telecommuters were an artifact of the process used to select them, we compared the productivity of the agents selected for telecommuting to those not selected prior to project implementation. Agents who were later selected as telecommuters handled an average of 3.7 calls per hour compared with 6.0 calls per hour for those who were not selected. Accordingly, agents selected to become telecommuters were not the most productive in the office prior to project implementation.

Although KAWC considered length of service and knowledge/skill levels in selecting employees to become telecommuters, the selection process was primarily driven by the limitations of the technologies being employed. The deciding factors were the quality of two-way radio communication at the employee's home (enabling them to dispatch repairmen) and the proximity of the employee's home to the KAWC office (for dropping off the materials and equipment). We conclude there is no evidence that productivity gains are a result of the selection process.

Back to Top

Effect of Telecommuters on the Productivity of Other Employees

During the time period in which average productivity of telecommuters increased by 154%, the average productivity of in-office agents fell by 13.3%. This decrease may provide some evidence that telecommuting adversely affects the performance of other employees. However, further study is needed to understand the reasons for the decrease in non-telecommuter productivity.

Back to Top


We found positive support that telecommuting increases productivity and, more importantly, that this increase is sustainable over time. We found no evidence that productivity gains are an artifact of the process used to select telecommuters and little evidence to suggest that telecommuters negatively impact the performance of other employees. We also found no evidence that telecommuting reduces absenteeism. An important strength of our study is that these findings are based upon measures of actual worker productivity rather than self-reported data.

Because our research was conducted within a single organization, generalization of the results to other organizations is limited. We encourage other organizations to conduct similar research in order to determine if our results can be generalized.

It is also possible that several months of missing data in the middle of the project could have affected our results. However, because there is a consistent pattern of improved productivity in the two years before the missing data and in the 27 months after the missing data, the likelihood that inclusion of this data would have led to significantly different results is minimal.

Back to Top


1. Bailey, D.E. and Kurland, N.B. A review of telework research: Findings, new directions, and lessons for the study of modern work. Journal of Organizational Behavior 23 (2002), 383–400.

2. Loy, S.L., Brown, S., and Butler, E.S. Telecommuting at Kentucky American Water Company. Journal of the International Academy for Case Studies 9, 5 (2003), 53–60.

3. Westfall, R.D. Does telecommuting really increase productivity? Commun. ACM 47, 8 (Aug. 2004), 93–96.

4. Westfall, R.D. Does telecommuting really increase productivity? Fifteen rival hypotheses. In Proceedings of the AIS Americas Conference (Indianapolis, IN, 1997), 405–407.

Back to Top


E. Sonny Butler ( is an associate professor in the Department of Information Technology at Georgia Southern University in Statesboro, GA.

Cheryl Aasheim ( is an assistant professor in the Department of Information Technology at Georgia Southern University in Statesboro, GA.

Susan Williams ( is a professor and the chair of the Department of Information Systems at Georgia Southern University in Statesboro, GA.

Back to Top


UT1Table. Adjustment analysis for telecommuters at Kentucky American Water Company.

Back to top

©2007 ACM  0001-0782/07/0400  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2007 ACM, Inc.


No entries found