Electronic Supervisor: New Technology, New Tensions

DOCUMENT RESUME

ED 299 406 CE 051 015

TITLE The Electronic Supervisor: New Technology, New Tensions. INSTITUTION Congress of the U.S., Washington, D.C. Office of Technology Assessment. REPORT NO OTA-CIT-333 PUB DATE Sep 87 NOTE 152p. AVAILABLE FROMSuperintendent of Documents, U.S. Government Printing Office, Washington, DC 20402 (GPO Stock No. 052 - 003 - 01082 -B: $6.50). PUB TYPE Reports - Research/Technical (143)

EDRS PRICE MF01/PC07 Plus Postage. DESCRIPTORS *Civil Liberties; *Computer Oriented Programs; *Electronic Equipment; Employer Employee Relationship; *Employment Practices; Equipment Utilization; Foreign Countries; *Job Performance; Legal Problems; Legal Responsibility; *Privacy; Public Policy; Supervisory Methods; Telephone Communications Systems IDENTIFIERS *Computer Work Monitoring

ABSTRACT Computer technology has made it possible for employers to collect and analyze management information about employees' work performance and equipment use. There are three main tools for supervising office activities. Computer-based (electronic) monitoring systems automatically record statistics about the work of employees using computer or telecommunications equipment in their jobs. Service observation refers to the practice of listening in on an employee's conversation with a customer. Automatic service observation refers to computer tracking of calls' durations and destinations. Telephone accounting refers to automatic, computer-generated records of the times, durations, and destinations of telephone calls. The use of these technologies is surrounded by controversy. Some praise this new technology as a way of helping employers manage resources, plan workloads, reduce costs, and provide employees with timely feedback. Others fear the potential civil liberties and privacy abuses of these new technologies. Another important concern is that there are, at present, no requirements in United States law that monitoring be fair or that employees be consulted about work standards. (Appendixes include notes on computer work monitoring in other countries and the privacy and civil liberties implications of testing employees in the workplace.) (MN)

***************************** ***** ************************************* * Reproductions supplied by EDRS are the best that can be made * * from the original document. * AA*****************************************************************-*** 2(7\%1 lbec4ronic Supervisor

New Technology

New Tensions

U S. DEPARTMENT OF EDUCATI01. Office of Educational Research andimprovement EDUCATIONA L RESOURCESINFORMATION CENTE (ERIC) o /Ms document has bunreproduced as received from the person ororganization originating it C Minor changes have beenmade to HTIPrO,0 reproduction quality

Points of view or opinions statedth IS document do not necesvinlyrepresent official \r) OERI positron or policy 0

CONGRESS OF THE UNITED STATES OFFICE OF TECHNOLOGY ASSESSMENT Office of Technology Assessment

Congressional Board of the 100th Congress

MORRIS K. UDALL, Arizona, Chairman TED STEVENS, Alaska, Vice Chairman

Senate House ORRIN G. HATCH GEORGE E. BROWN, JR. Utah California CHARLES E. GRASSLEY JOHN D. DINGELL Iowa Michigan EDWARD M. KL-NNEDY CLARENCE E. MILLER Massachusetts Ohio ERNEST F. HOLLINGS DON SUNDQUIST South Carolina Tennessee "LAIBORNE PELL AMO HOUGHTON Rhode Island New York JOHN H. GIBBONS (Non voting)

Advisory Council

WILLIAM J. PERRY, Chairman CLAIRE T. DEDRICK RACHEL McCULLOCH H&Q Technology Partners California Land Commission Brandeis University DAVID S. POTTER, Vice Chairman S. DAVID FREEMAN CHASE N. PETERSON General Motors Corp. (Ret.) Lower Colorado River Authority University of Utah EARL BEISTLINE MICHEL T. HALBOUTY JOSEPH E. ROSS Consultant Michel T. Ila Monty Energy Co. Congressional Research Service CHARLES A. BOWSHER CARL N. HODGES General Accounting Office University of Arizona Director

JOHN H. GIBBONS

The Technology Assessment Boardapproves the release of this report The views expressed in this report are not necessarily those of the Board. OTA Advisory Council,or individual members thereof

1 I 4 Er liC Supervisor

New Technology

New Tensions

1 CONGRESS OF THE UNITED STATES OFFICE Or I EQHNOLOGY ASSESSMENT %....., Washington. 0 C 20510-8025 Recommended Citation: U.S. Congress, Office of Technology Assessment, The Electronic Supervisor. New Technology, New Tensions, OTA-CIT-333 (Washington, DC: U.S. Government Print- ing Office, September 1987).

Library of Congress Catalog Card Number 87-619855

For sale by the Superintendent of Documents U.S. Government Printing Office, Washington, DC 20402-9325 (order form can be found in the back of this report) " Foreword

The Electronic Supervisor. New Technology, New Tensions,deals with the use of computer-based technologies to measure how fast or how accuratelyem- ployees work. New computer-based office systemsare giving employers new ways to supervise job performance and control employees'use of telephones, but such systems are also controversial because they generate such detailedinformation about the employees they monitor. This assessment exploresa broad range of questions related to the use of new technology in the workplaceand its effects on privacy, civil liberties, and quality of working life. This study was requested by the House Committeeon Government Opera- tions and the Subcommittee on Civil and Constitutional Rights ofthe House Com- mittee on the Judiciary. A second report resulting from thesame request, Defending Secrets, Sharing Data: New Locks and Keys forElectronic Informa- tion, will discuss how important developments in computersecurity are converg- ing with technologies and policies for communications security.It also explores the past and future role of government policies in the evolutionof information security technologyparticulary cryptographyfor safeguardingcommunication in government and the private sector. OTA wishes to thank the many people and organizations thatcontributed to this assessment through advisory panels, interviews, reviews, and othermeans of sharing their information and experience withus. The final responsibility for the study, however, rests with OTA.

JOHN H. GIBBONS Director

b iii r Advisory PanelThe Electronic Supervisor: New Technology, New Tensions

Granger Morgan, Ph.D., Chairman Carnegie Mellon University Peter Arment Gary T. Marx, Ph.D. Division of Telecommunications Professor of Sociology State of New York Massachusetts Institute of Technology Robert R. Be lair, Esq. Robert Morris, Ph.D.* Kirkpatrick & Lockhart Chief Scientist, NCSC-C Jerry J. Berman National Security Agency Chief Legislative Counsel Susan L. Quinones, Esq. American Civil Liberties Union Conde llo, Ryan, Piscitelli H. W. William Caming, Esq. Virginia du Rivage Consultant Research Director 9 to 5 National Association of Working Robert H. Courtney, Jr. Women President RCI Forrest Smoker Director Harry B. De Maio Corporate Telecommunications Director, Data Security Program North American Phillips Corp. IBM Corp. Willis H. Ware, Ph.D. John Harris Corporate Research Staff Special Assistant to the President The Rand Corp. American Federation of Government Employees (AFL-CIO) Fred H. Wynbrandt Assistant Director James M. Kasson Criminal Identification and Information Vice President, R&D Branch Rolm Corp. State of California Steven Lipner Digital Equipment Corp.

*Ex officio

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members. The panel does not, howover, necessarily approve, disapprove. or endorse this report. OTA assumes full responsibility for the report and the accuracy of its contents. Project StaffThe Electronic Supervisor: NewTechnology, New Tensions

John Andelin, Assistant Director, OTA Science, Information, and Natural Resources Division

Fred W. Weingarten, Manager Communication and Information Technologies Program

Project Staff Charles K. Wilk, Project Director Karen G. Bandy, Principal Author Jim Dray, Research Analyst Robert Kost, Legal Analyst Mary Ann Madison, Analyst Joan Winston, Analyst**

Administrative Staff Elizabeth A. Emanuel, Administrative Assistant Audrey D. Newman, Administrative Secretary Sandra Serbinoff, Secretary

Contractors Sandra Albrecht, University of Kansas Emanuel Donchin, University of Illinois Steven Deutsch, University of Oregon Paul Faerstein, Hay Group, Inc. Gary T. Marx and Sandford Sherizen Data Security Systems, Inc. Michael J Smith, University of Wisconsin Sharon H. Strom Alan F. Westin, Russell Pipe, and Elaine J. Eisenman Educational Fund for Individual Rights

Until Ape 1986. **After April 1986.

4 Reviewers and Other Contributors

David Alexander Dick Greenwood Diana Wickes Roose George Meany Center for International Association of Oberlin College Labor Studies Machinists and Aerospace Jacquelin Ruff Robert Arndt Workers Service Employees International University of Wisconsin Henry Guzda Union Nicholas A %ford U.S. Department of Labor James Rule Massachusetts Institute of Chris Higgins State University of New York Technology University of Western Ontario Hedva Sarfati Jordon Barob George Kohl International Labor Organization American Federation of State, Communications Workers of David Shamanski County and Municipal America Philip She Maas Employees Jack Landers IBM Corp. Joe Blanding General Services AdministrationDonald Sheridan United Auto Workers Charlotte Le Gates General Services Administration Bobbette Bond Computer and Business Donald E. Soeken United Food and Commercial Equipment Manufacturers Association of Mental Health Association Workers Specialties Andrew Clement David Le Grande Jay Spechler York University Communications Workers of American Express Co. America Bill Cunningham Ken Stavenjord AFL-CIO Jenny Luray U.S. Department of Defense Office Technology Education William Daddio Project Ronnie Straw Communications Workers of Marvin Dainoff Walter Lypka Miami University Graphic Communications America Ben Di Sylvester International Union Paul Trause Robert E. Nolan Co. Charles Milk General Services Administration Richard Dwyer AT&T Gwen Wells George Meany Center for Office and Professional Robert E. Nolan Employees Union Labor Studies Robert E. Nolan Co. David Eisen Leslie Nulty James Woodford Newspaper Guild United Food and Commercial Nayyar Zaidi Donna Euben Workers C&P Telephone Co. 9 to 5 National Association of Karen Nussbaum OTA Reviewers Working Women Service Employees Louis Feldner International Union Lawrence Miike, Senior Federal Communications Dorothy Paleologoa Associate Commission Aetna Life & Casualty Neil Holtzman, Detailee, Benjamin IL Friedman Connie Pond Johns Hopkins General Services AdministrationAmerican Federation of Priscilla Regan, Analyst Lisa Gallatin Government Employees Denise Dougherty, Analyst Office Technology Education Char la Rath Benjamin C Amick, Analyst Project American Society for H. Stephen Gordon Information Science National Federation of Federal Barton Reppert Employees CBR International Judy Gregory Karen Ringen AFL-CIO American Civil Liberties Union

vi Contents

Page Summary 1 Chapter 1. Introduction 5 Chapter 2. Using Computers To MonitorOffice Work 27 Chapter 3. Telephone Call Accounting 61 Chapter 4. Electronic Work MonitoringLaw and Policy Considerations 85

Appendix A. Notes on Computer WorkMonitoring in Other Countries 121 Appendix B. Privacy and Civil LibertiesImplications of Testing Employees in the Workplace 128

vil Summary

Introduction Computer-Based Work Monitoring Computer technology makes possible the Computer work monitoring is affecting a continuous collection and analysis ofmanage- small but growing segment of the office work ment information about work performance and force. It is estimated that around 6 million of- equipment use. This information can beuse- fice workers have part or all of their work evalu- ful to managers in managingresources, plan- ation based on computer-generated statistics; ning workloads, and reducing costs. Itcan be for many others, such statistics may be col- advantageous to employees as well, by provid- lected but are not currently used for evalua- ing timely feedback on performance andan ob- tion. The number of monitored workers can be jective basis for evaluation. Despite thesepos- expected to grow as computers begin to be used sible advantages, however, there is controversy in more office jobs. In addition, computer work about computer-based monitoringon grounds monitoring is also affecting people in non-office that it invades employees' privacy,causes jobs, for example retail sales, as computers are stress, and can be used unfairly by some em- introduced in a greater variety of workplaces. ployers. Privacy.Although many workers have expressed a feeling of privacy invasion when they are "constantly watched" by a machine, com- Tools for Supervising Office Activities puter-based monitoring usually does not raise "Computer-based monitoring" or "electronic issues of privacy infringement in the strict le- monitoring" systems automatically record sta- gal sense. The workplace activities that are tistics about the work of employees using com- monitored by computer are primarily inher- puter or telecommunication equipment in their ently public activities, many of which were sub- jobs. Such statistics might include number of ject to counting or supervision in other ways keystrokes made, types of transactions com- before computers became available. Privacy pleted, or time spent for each transaction, for and access questions may arise, however, re- example. lated to employees' ability to see or challenge records concerning their wori.. "Service observation" refers to the practice of listening in on an employee's conversation Fairness.The central workplace issues with a customer to check on courtesy, correct. raised by monitoring are labor relations ques- ness of information, or other factors. Service tions of fairness, dignity, autonomy, and con- observation is not automatic; it requires a hu- trol, and are greatly influenced by the labor- man supervisor. However, it is often used in management relations climate of a given firm conjunction with computer-based systems that or industry. The effects of computer-based collect information about the duration of the monitoring depend on how it is used. Allega- telephone call or the types of transactions the tions of "unfair" or "abusive" monitoring usu- employee performs during the call. New tech- ally focus on questions like high or increasing nology has made service observation com- quotas, inappropriate work standards or puni- pletely silent, so neither the employee nor the tive use of monitoring information by super- customer may know a supervisor is on the line. visors. Computer-based monitoring appears most likely to be opposed or resented by em- "Telephone call accounting" refers to auto- ployees when they perceive that it is used un- matic, computer-generated reccrds of the time, fairly o when it is imposed without their un- duration, and destination of telephone calls. derstanding or participation. Conversely, in It is generally used to manage telephone costs some workplaces employees accept electronic rather than to supervise the work procf ss. monitoring as a tool that helps them get con- Some employers use telephone call accounting trol of their own work and ensures that their to help reduce their employees' personal use supervisos evaluate them on the basis of fair of office telephones. criteria.

I Stress. An additional issue is the possibil- which are not required to be protected, are used ity that monitoring contributes to employee to track the calling habits of individuals. If call stress by creating a feeling of being watched accounting is going to be used by the Federal o; by creating pressure to work at high speed. Government, development of better guidelines There is some research on effects of computer- on acceptable levels of personal use of tele- based monitoring, but it generally fails to sep- phones, procedures for tracing patterns of ma- arate the effects of monitoring from those of jor abuse, and guidelines for protecting call job design, equipment design, lighting, work- records from unauthorized use are necessary load, machine pacing, and other potentially to minimize intrusions on Federal workers. stressful aspects of work in offices where computerized equipment is used. This area de- Privacy and Fairness in the Workplace serves further research. The uses of technology discussed so far are Service Observation controversial because they point out a basic tension between an employer's right to con- Monitoring the content of messages raises trol or manage the work process and an em- a related set of issues. Some employers say that ployee's right to autonomy, dignity, and service observation (listening to or recording privacy. This same tension is also evident in the content of employees' telephone conver- the use of other technologies for surveillance sations with customers) helps assure quality and testing in the workplace. For example, and correctness of information and protects Controversy over polygraph testing, drug test- all parties in case of dispute. However, serving, genetic screening, and emerging brain ice observation also impacts the privacy of the wave testing illustrates the tension between customer, and workers and labor organizations employers' rights to manage their enterprise, have argued that it contributes to stress of the reduce costs and reduce liability, and ate em- employee, and creates an atmosphere of dis- ployees' rights to preserve individual privacy trust. Service observation is legal when part and dignity. Fairness questions relate to the of a formally established program of evalua- accuracy of the tests themselves as well as to tion. In the Federal Government, employees the criteria for deciding who is to be tested. must be informed that such a program is in Privacy issues include employers' acquisition effect, but do not need to be informed precisely of personal inforr ation not related to work, when a supervisor is listening. and the protection of records generated by testing. Telephone Call Accounting Telephone call accounting (computer-gener- Legal and Policy Implications ated records of the time, duration, destination, and cost of calls) gives employers a powerful There are no legal requirements in U.S. law tool for allocating costs to different projects, that monitoring be "fair," that jobs be well settling billing disputes with telephone com- designed, or that employees be consulted about panies, and discouraging nonbusiness use of work standards, except insofar as these points telephones. Other technologies, including call are addressed in union contracts. Less than blocking, authorization cxies, and levels of 20 percent of the office work force is unionized, service, can be used to limit nonbusiness uses and even where unions are involved, their ef- of telephones, either instead of or in conjunc- fectiveness has been limited because technol- tion with call accounting. ogy choice and productivity measurement are often considered "management rights" under The Federal Government has collected de- the contract. Similarly, workers are not cur- tailed call accounting data on long-distance rently protected by law against stressful work- calls for at least 10 years, but new technology ing conditions, although stress can be a com- would make this information easier for agen- pensable injury under Worker Compensation cies to use on a regular basis. Privacy ques- statutes. tions are raised when accounting records,

4-1 2 Chapter 1 Introduction CONTENTS

Page Background 5 Monitoring and the Legal Context 5 Monitoring and the Labor Relations Context 6 Findings of the Report 7 Historical Background 15 Monitoring in History 15 Privacy in the Workplace 18 Policy Options 20 Sources of Information for This Report 23 Organization of the Report 24 Figure Figure No. Page 1. Categories of Behavior Subject to Monitoring, Measurement, or Testing 13

Ii Chapter 1 Introduction

BACKGROUND This study addresses the capabilities of new call accounting can be a powerful management computer and communication technologies for tool fur allocating telephone costs, checking monitoring employees' activities in the work- the correctness of telephone bills, and reduc- place. New communication technologies such ing personal use of employers' telephones. The as digital private branch exchanges (PBXs), Federal Government, through a recent audit local area networks (LANs), and digital tele- of call accounting records, found that about phony in the switched network provide more 33 percent of off-network long-distance calls capability to monitor calling patterns as well on the Federal Telecommunications System as content of telephone calls. Equipment and were personal calls. software for telephone call accounting (tracking the time, destination, and cost of calls) On the other hand, there are concerns about make up the fastest growing segment of the these practices as well. There are strong argu- telecommunication industry. ments that computer-based monitoring can be abused and that monitoring has potential for The networking of computers, either through invasions of employee privacy, as well as as- LANs or sometimes through the telephone sys- saults on their autonomy, personal dignity, and tem, provides a broad capacity to monitor work health. Computer monitoring of performance that is performed at a computer terminal. Com- provides continuous minute-by-minute records puter-generated statistics provide the basis of of employee performance and could be used to part or all of the work performance evaluation speed up the pace of work or enforce unfair for about 4 to 6 million office workers. For work standards. Service observation, when many millions more, computer statistics of done without notice or warning -an contrib- some sort are collected every time they use ute to a feelng of being spied up and may their terminals, even though these records are have implicat'ons for the privacy of customers not currently used for performance evaluation. as well as employees. Telephone call account- Most of the employees subject to computer- ing could cot ceivably be used to build a "pro- based work measurement are in clerical occu- file" of an employee's personal or professional pations, or in other jobs where work is largely telephone contacts which might be used to har- repetitive. Ultimately, as electronic mail and rass him or her. In general, the concern is that other computer-based technology become more these new information technology tools might pervasive in the office, it is likely that com- give employers powers of surveillance and con- puter-based monitoring will affect a large num- trol in the workplace that might be abused ber of workers at all organizational levels. used simply for the sake of control, beyond what is necessary to organize the work process. Managers say that computer-based monitoring is very useful to employers. Computer monitoring of productivity can help them enhance Monitoring and the Legal Context productivity, maintain production standards, spot bottlenecks, and plan personnel and equip- In general, the law has recognized the em- ment needs. "Service observation," the capa- ployers' interests in organizing work, select- bility to listen in on telephone conversations ing technology, setting production standards, between employees and customers, helps them and managing the use of facilities and re- make sure that customers receive correct in- sources. Although some aspects of working formation and courteous service. Telephone conditions may be subject to collective bar-

15 6 The Electronic Supervisor. New Technology, New Tensions gaining, the vast majority of office workers in be need for a new balance between workers' the United States ere not represented by un- rights to privacy or autonomy in the workplace ions. Thus, employers have had considerable and management's requirements for informa- latitude in making use of new monitoring tech- tion to efficiently control their resources. A ma- nologies; they have generally been considered jor decision for Congress is whether the present merely extensions of traditional management balance between worker rights and manage- prerogatives. ment requirements is reasonable, and, if not, if it can be satisfactorily accommodated On the other hand, the law also provides cer- through stakeholder agreement, e.g., negotia- tain protections to employees, such as the right tion between labor and management in gov- to join unions, to bargain collectively, or to ernment and the private sector. If the use of work in a safe and healthy workplace. One new technology is seen as weakening the voice question that may appear before Congress is of employees in such negotiations, Congress whether employee health, or the quality of may choose to take action to ensure a reason- working life, or employees' rights to privacy able balance. or personal dignity need protection against possible abuses of work monitoring. Monitoring and the Labor It is possible that the present extent of com- Relations Context' puter-based monitoring is only a preview of growing technological capabilities for monitor- Monitoring is an integral part of a larger sys- ing, surveillance, and worker testing in the tem of management, labor relations, industrial workplace. If this is the case, then there may competitiveness, and ethical and legal systems. Much is undergoing rapid change in the United States and the issues of who is working, where we work, what jobs we do, and how we do them today may markedly differ even from the immediate past. Technology is a significant factor in these changes; so are international developments, changing labor-management relationships, and cultural values. Some specific changes follow: The American labor force has changed dramatically in recent decades, primarily due to the najor influx of women, wbo now constitute close to one-half of all working Americans. It is a labor force that is better educated and includes more non-white workers. The shifts away from goods- to service - producing industries has accelerated in the past two decades. The United States is predominantly a white-collar, service v. goods-producing society.

Photo: Cowboy of CAP Telephone 'This section summarizes work In Steven Deutsch, "The Computerization Is transforming jobs like oustcmer Context for Exploring Workplace Monitoring," contract paper service and order processing. prepared for OTA, September 1986.

.16 Ch. 1Introduction 7

Early automation in the1950s&'id1960s fice workers in the private sector are not was largely restricted to manufacturing, unionized. A higher proportion of Federal, but there has been an enormous growth State, and local government employees of office automation in the past decade and are unionized. For union workers, there investment per employee by 1990 may be have been efforts to address the new tech- comparable in office and factory settings. nology, including workplace monitoring, Computers are commonplace in office,and in collective bargaining agreements and in retail sales. through quality of work-life committees. There has been a large growth of clerical A parallel activity has included efforts to employees: from 5 million in1940to 20 pass State legislation protecting workers million in1980,from 1 out of 10 to 1 out on visual-display terminals (VDTs) and ad- of 5 employees in the United States. Only dressing worker privacy issues. Such re- 6.3 percent of males are in clerical jobs forms at the State and local level may well while over one-third of all women work- accelerate in the near future. ers in this country are clericals. Since cler- The challenge for meeting international ical work is increasingly being done on competition has pushed many in govern- computer terminals, women are dispropor- ment, management, and unions to adopt tionately affected by the microelectronic a more cooperative labor-management technology in the office environment. stance and to work towards cooperative The growth of office employment and the approaches for making best use of new risP in office automation makes for a technology. While this trend does not af- greater proportion of the American work fect all firms, where greater labor manage- force in settings where computerized work ment cooperation does exist it has allowed monitoring is possible. better resolution of many issues related Collective bargaining affects only about to technology, job protection, training and 20 percent of U.S. workers, and most of- retraining, and quality of working life. FINDINGS OF THE REPORT Finding #1 what remains to be done, what people and re- Computer technology makes possible the sources are available to work with, the status continuous collection and analysis of man- of partially completed products, and so forth. agement information about work perform- This type of management information helps ance and equipment use. This information them to decide if staffing levels are appropri- is useful to managers in managing re- ate, if more equipment is needed, if bottlenecks sources, planning workloads, and reducing need to be relieved, etc. costs. When it is applied to individual employees, however, the intensity and continu- In an office, the computer is often now the ousness of computer-based monitoring chief tool for carrying out the work process. raises questions about privacy, fairness, and The transformation of the original input data quality of work life. to a final product may require many steps performed by the computer system, a human Information about the progress and status worker, or an interaction between the two. For of work is vital to managers of most organiza- example, between the time a credit card com- tions. Whether their output consists of manu- pany receives a sal 3 record, and the time it factured goods, services, or information-based mails out a payment to the merchant, and a products, managers want reliable knowledge bill to the cardholder, literally dozens of proc- about what has been done, how long it took, essing steps are required. The credit card com-

17 8 The Electronic Supervisor. New Technology, New Tensions pany processes hundreds of thousands of sale visor occasionally checking their work. Privacy records each week, so meticulous recordkeep- can also refer to exercising one's own auton- ing is necessary at each step to keep the proc- omy, even in routine work, there is some per- ess from going awry: most of the recordkeep- sonal variation in work style. Some people ing is done by the computer software itself, work fast for short periods but take lots of because so many transactions go on inside the breaks, others work fast in the morning and computer where they are invisible to the naked slow in the afternoon. These individual work eye. Monitoring software does this by keeping styles may not matter when the basic unit of track of the time, type, and duration of every evaluation is longsay a day or a week. People relevant transaction. Such meticulous record- with widely differing styles might accomplish keeping generates a great deal of information the same amount of work in a day. However, that must be processed to produce reports continuous monitoring offers management usable to human managers. The particular in- more detailed information. If the employer uses formation and amount of detail wanted will de- the information gathered through monitoring pend on the purpose of the report and the level to change the pace or style of workregulating of management. The president of the firm may the number of breaks or requiring people to want to total only transactions and revenues accomplish as much in the afternoon as in the for the day, but line managers will want more morningthen the employee loses a certain detailed information on which to base day-to- amount of control over his or her own job. day decisions. Fairness.Fairness is related to the way Concern about electronic monitoring be- monitoring is implemented in the workplace. comes most intense when it centers on evalu- At some locations, employers and workers ating the work performance of individual em- alike note that electronic monitoring can be ployees. A growing number of firms rely on a fairer basis for performance evaluation than computer-based monitoring to measure the other more subjective moans On the other work of at least some employees. The infor- hand, at other locations monitoring was viewed mation can be quite detailed: How many trans- by employees as an unfair practice. Chapter actions were performed? Of what type? With 2 outlines some of the factors that might be how many errors? When were transactions per- considered in assessing the fairness of a work formed? How long did they take? What were measurement program and also reviews the in- the longest or the shortest? How many breaks terviews done by and for OTA that suggest did the employees take? When and for how there is a range of opinion among workers long? about the fairness of the way monitoring is Although people object to monitoring be- used in their organizations. Among the factors cause it "invades the privacy" of employees, included in fairness are: reasonable standards, the objections to electronic monitoring applied understanding by workers of the extent and to individual employees cannot be phrased in use of the monitoring system, ability of work- terms of privacy alone. This discussion sum- ers to contest or correct records, and partici- marizes them in terms of three headings: pation by workers in the design of the system. privacy, fairness, and quality of work life. The Quality of Work Life.Quality of work life effects in these areas are reviewed in greater is a complex area that is affected by many fac- detail in chapters 2 and 4. tors in the workplace. Two major objections Privacy.Privacy encompasses the right to to electronic monitoring of individual perform- be left alone and to not be intruded upon. Some ance are allegations that it contributes to em- workers complain that electronic monitoring ployee stress and stress-related illnesses and is intrusive because it is making a constant that it contributes to an atmosphere of distrust minute-by-minute record, crating a feeling of in the workplace. While there has been only "being watched" all the time. This, they say, limited direct research on the stress effects of is quite different from having a human super- electronic monitoring, there does seem to be 18 Ch. 1Introduction 9

some evidence that it can contribute to stress, as will be discussed below and in chapter 2. Finding #2 Computer-based systems offer opportunities for organizing work in new ways,as well as means of monitoring it more inten- sively. Electronic monitoring is most likely to raise opposition among workers when it is imposed without worker participation, when standards are perceived as unfair,or when performance records are used puni- tively. Worker involvement in design and implementation of monitoring programscan result in greater acceptance by workers, but despite activities of labor unions in some industries and recent progress in labor -man- Pima Cowboyor canTftPhon agement cooperation in others, most firms Computerization of directory assistance helps operators perform their jobs more efficiently and also provides do not have mechanisms to do this. means to supervise work electronically. OTA's report Automation of America's Offices discussed in detail the ways in which computer systems can change the organization bility of the performance evaluation systems of office work. The introduction of large main- in their firms. frame computers in the 1950s and 1960s prob- Employee Participation.Onlya small pro- ably reinforced the tendency toward central- portion (about 20 percent) of U.S. workersare ized control, routinization of tasks, and unionized. Among office workers this percent- assembly line organization of office work. How- age is even lower. About 12 percent of techni- ever newer trends in office automation, allow- cal, sales, and administrative support working "end-user computing" and communication ers are represented by unions and 17 percent networks that give remote access to central of managerial and professional specialty work- databases, allow more flexibility in workorga- ers.2 In most workplaces, therefore, labor nization. While many firms still use theassem- organizations do not play a role in represent- bly line model, others have discovered thatnew ing employee views about monitoring systems. information technology is allowing them to Even when unions are involved, technological "reintegrate" work. This means that jobsare choice, such as the decision to introducecom- made more interesting, and more effective, by puter equipment with monitoring capability, giving the individual (or sometimes a team of may be considered a management right that individuals) a variety of tasks. is not subject to bargaining, althoughsome No matter how work is organized in the of- union contracts do require employers to bar- fice, electronic monitoring can be applied to gain over changes in work technology orper- the computers and their users. Whether the formance standards. The monitoring issue has work in question is that of a directory assis- sczved as a spur to union organizing in some tance operator, performing a very few tasks previously unorganized firms. in a repetitive cycle, or an insurance company's legal case analyst whose work encompasses dozens of different activities, each transaction 'Statistical Abstract of the United States,Table No. 713 can still be computer monitored. In interviews "Union Membership of All Workers and Median Usual Weekly with supervisors and workers, OTA founda Earnings," 1986, p. 424. The category of managerial and proles- sional specialty includes school teachers, many of whom areun- range of opinion about the fairness and suita- ionised.

19 10 The Electronic Supervisor. New Technoicpy, New Tensions

On the other hand, there is a growing trend tronic monitoring. In general, electronic mon- in the United States, according to some ana- itoring practices were covered by quality of lysts, toward greater labor -managentent coop- work life legislation and by labor-management eration in making dada= about new technol- negotiations. Monitoring in other countries is ogy and haw it is used. This trend is affecting discussed in more detail in appendix A. both unionized and nommionized organizations. While the actual number of firms involved is Finding itll small, observers are encouraged that some of There is reason to believe that electroni- them are very large firms and leaders in their cally monitoring the quantity or speed of particular industries' work centributes to stress and streeerelated Work Monitoring in Other Industrialized illness, although there is still little research Countries. Ina number of other industrial- separating the effects of monitoring from ized countries, where the power of employees job design, equipment design, lighting, ma- and their representatives in making workplace chine pacing, and other potentially stress- decisions 'a greeter than in the United States, ful aspects of computer-based office work. there appears to be greater use of the collec- Some research suggests that there are a tive bargaining process to limit the use of elec- number of possible health problems related to tronic monitoring of individuals. In some coun- the use of computer terminals or VDTs in gen- tries, legislation ensuring employees a good eral, including vision, muscular-skeletal, psy- quality of work life has been interpreted to pre- chosocial, and possible reproductive health clude individual monitoring as an insult to in- problems! Many of these problems can be dividual dignity. In Norway, Sweden, and ameliorated or eliminated through good equip- West Germany for example, electronic moni- ment design, proper job training (e.g., allow- toring is generally used to measure the per- ing frequent breaks or scheduling duties away formance of groups rather than individuals. In from the terminal for part of the day), and Sweden, individual monitoring is sometimes proper training (instructing workers in proper used in cases where the union and management adjustment of screens, lights, and furniture). agree there is an overwhelming need, or occa- In the United States, the way that office auto- sionally for nonunionized temporary workers. mation systems are implemented and used is In addition, in some countries, electronic mon- almost entirely at the discretion of employers, itoring runs counter to other norms for enforc- and there is a wide variation in their adher- ing work discipline. In Japan, for example, elec- ence to good practice in these areas. tronic monitoring of individuals goes against the tradition of teamwork and peer pressure Review of the psychological and physiolog- as a means of encouraging good work and is ical literature suggests a number of reasons therefore not used. One Japanese executive why monitoring could be stressful, and a num- stated that introducing it would offend both ber of studies have shown a higher level of managers and workers. Many Western Euro- stress experienced by monitored workers. pean countries also have strong data privacy These studies are discussed in more detail in laws governing the use of computer-based files chapter 2. The particular stress problems about individuals, but OTA did not find that raised by electronic monitoring are very difficult to separate from other job design or equip these laws were a major factor in limiting elec- ment design factors. For example the job of

See, for example, U.S. Department of Labor, Bureau of Labor-Manepment Relations and Cooperative Programs,U.S. 'For a summary, see, Jeanne Stellman and Mary Sue Labor Law widths Mame I LaborWanagement Cooperation, Medlin,OfficeWar* CanBe Hasercious toYourHealth(New BLMR 104, Waeldngten, DC, 1986; and Nicolas Ashford and York, NY: Pantheon, 1988); Bob DeMatteo,ilormisal Shock Christine Ayers, "Champs and Opportunities in the Environ- The Health Hazards of Video Display Terminal.(Toronto: NC ment for Technology Bargaining," Ilassechusetta Institute of Press, 1986); and U.S. Congress, Office of Technology Assess- Technology, prepared for the Assistant Secretary for Policy, ment,Automation of Amwica's Offices,OTA-C1T-287 (Wash- U.S. Department of Labor, no data. ington, DC: U.S. Government Printing Office, December 1906).

2 Ch. 1introduction 11

the directory assistance operator is often used as an example of a job where monitoring leads to stress. However, this job is often both monitored and paced by the computer; that is the computer not only measures the amount of time it takes an operator to handle a call, but it also automatically sends the next call as soon as the line is free. In determining why this is a high-stress job, it is difficult to separate the effects of lack of control from the effects of monitoring, and for this reason it is also hard to generalize the experiences of these operators to other types of work. Roar Owes, of AT&T Finding *4 Supervisors regularly check courtesy and accuracy of telephone operators like these through "service Monitorbg the contest of messages raises observation," that Is, listening in on calls Menet set of issues. Some employers with customers. say that service observed°. (listening to or recording the content of employees' telephone conversatis with customers) helps tion of workers rather than to protect the cus- assure quality and correctness of informa- tomer or the firm. The "secrecy" is removed tion and by protecting all parties in case of if the employee and the customer can hear a dispute. However, service observation also "beep" tone or other cue when a supervisor impacts the privacy of the customer, and is on the line. workers and labor organizations have ar- Listening in on or recording employees'per- good that it andramtes to stress of the em- sonal calls or calls outside of a regular service ployee, and creates an atmosphere of dis- observation program have been considered trust. Monitoring the content of electronic eavesdropping by the courts. Service observa- mail messages or personal computer (PC) tion is discussed in more detail in chapters 2 diskettes also raises privacy issues. and 4. Many telephone systems are designed so OTA interviewed several employers about that certain users, usually supervisors or qual- their policies regarding the privacy of PC dis- ity control workers, can listen in on telephone kettes used by employees. All believed they conversations. Service observation is consid- had a right to search employee diskettes for ered an important aspect of quality control in personal material or unauthorized company inmany firms and public agencies that have a formation and would do so if they had cause lot of telephone contact with the public. In to believe inappropriate material was being some cases, employers may be liable for mis- stored on the diskettes. Such audits of PC dis- information given out by their employees; they kettes have taken place in the Federal Gov- therefore want to make sure that all employ- eimnent to ensure that computers were used ees follow guidelines. Service observation, for official business and to check security pro- when part of a formally established program cedures related to confidential information. of evaluation, is legal. There is currently no requirement that employees know precisely Finding .5 when they are being monitored, although, at Telephone call accounting (computer-gen- least in the Federal Government, they must erated records of the time, duration, desti- be informed such a program is in effect. Some nation, and cost of calls) gives employers workers and unions have objected to "secret" a powerful tool for managing the costs of service observation. They argue that the prac- telephone systems. However, it raises pri- tice is sometimes used for control or intimida- vacy questions when accounting records are el 12 The Electronic Supervisor. New Technology, New Tensions

used to track calling habits of individuals. There are a variety of technological and Other cost control technologies can be used administrative techniques that can help busi- to limit nonbusiness uses of telephones, ei- nesses and government agencies cut down on ther instead et or in addition to call account- waste calls. Some of these can be implemented lag. Establishing a policy for use of these without using cail accounting while others are technologies will be especlidly important for more effective if used in conjunction with call the Government as it builds a new Federal accounting. These are discussed further in Telephone System. chapter 3. Call - accounting equipment and software rep- The Federal Government is preparing to m- resent the fastest growing segment of the tele- ate a new long-distance telecommunications communication industry in the past few years. network, and many individual agencies are now Divestiture and deregulation of the telephone planning the purchase of new telephone equip- industry, along with the falling costa of comment, including switching equipment with call- puter equipment, have made it possible for accounting capability. Now, as these new sys- many firms to take closer control of their tele- tems are coming into place, is a good time for phone costs. Call-accounting software can gen- the government to assess the effectiveness of erate not only a listing of all calls, but can its current policies and determine if more work- produce reports that highlight calls made on able guidelines on personal use of telephones particular pb.mes, to particular destinations, might be developed. These options are dis- charged to particular accounts or for a certain cussed further in chapter 3. length of time. All of this information can be useful for telephone systems managers in al- Finding 4 locating costs and planning new facilities, but Electronic monitoring is only one of a as discussed in chapter 3, they raise questions range of technologies used in today's work- of privacy and fairness. Many employees use place to gather information about the work the it employers' telephones for some personal process or to predict work quality based on calls, and some firms have used call account- personal characteristics of the worlms. ing to track and prevent unauthorized tele- Many applications of technology, tacluti:mg phone use, especially for long-distance calls. polygraph testing, drug testing, genetic Call accounting has become an issue particu- screening, and, possibly, brain wave testiag, larly in the Federal Government, where per- illustrate the tension between employers' sonal use of long-distance lines is illegal. A re- rights to manage their enterprise, reduce cent audit performed by the General Services costs, and reduce liability, and the employ- Administration, under the auspices of the Pres- ees' rights to preserve individual privacy ident's Council on Integrity and Efficiency, and autonomy. Recent concerns of employ- found that personal use represents 33 percent ers, labor unions, civil liberties groups, the of the off -network long-distance calls sampled. courts, and individual workers suggest that a range of workplace privacy issues are in Privacy concerns are also raised by telephone need of resolution. call ar"ounting. A great deal of information about a person's personal and professional Interest in the privacy and stress effects of activities can be derived from analysis of a electronic monitoring, while of long standing complete record of his or her telephone calls, in some industries such as the telephone in- even though gathering of such information was dustry, are only now reaching the awareness not the objective of the call-accounting sys- of the general public. At the same time, some tem. Thus, what happens to those records and other hotly contested issues related to work- who has access to than are important consider- place privacy are also receiving public at- ations. Some observers :lave expressed fears tention. that call records could be used to identify or Figure 1 shows a range of types of monitor- harass whistleblowers, union organizers, or ing and testing that raise questions of privacy other dissidents within a firm or agency. and civil liberties in the workplace. Some types

22 Ch. 1Introduction 13

Figure 1.Some Categories of Behavior Subject To Monitoring, Measurement, or Testing

SOURCE- °Mu of TochoologY Assfisoment 11047- of monitoring seem primarily directed toward whether keystrokes are an appropriate meas- measuring work performance or work-related ure for a given job). On the other hand, per- activities in the workplace. OTA has called forming a blood or urine test to determine these "work monitoring" or "work measure- whether an employee has been using cocaine ment." Other types of monitoring and testing seems clearly to be a measurement of personal, seem to focus more on measuring the worker individual characteristicsa case of worker himself or herselfinvestigating activities out- testing. The test reveals the presence or ab- side the workplace or personal characteristics sence of certain chemicals in the bodyit does that might or might not have a bearing on work not show current impairment or measure job performance. OTA has called these "worker performance. This type of testing could be con- monitoring" or "worker testing." sidered predictiveit is used to determine whether a person has potential for poor job per- Counting the number of keystrokes some- formance as a result of drug-induced im- one performs in a day seems on its face to be pairments. an example of work monitoring. It is an objective measure of how much work is being done While the terms "work monitoring" and (leaving aside, for the moment the fquestion of "worker testing" may have some value as

, 2 ,3 14 The Electronic Supervisor. New Technology, New Tensions terms of analysis, it appears that all these settings, was used by the Department of information-gathering techniques are on a con- Defense in the 1970s to identify return- tinuum with no clear boundaries. It is not al- ing Vietnam soldiers with drug problems. ways possible to make a distinction between Now nearly all military personnel, millions work monitoring and worker testing. Even the of private employees, and a growing num- most extreme examples seem to have at least ber of government employees find that some elements of both. For example, even a their jobs depend on passing such tests. keystroke-monitoring system impinges on the Genetic Screening. This is still an emerg- worker as an individual if it reveals when or ing technology for predicting a person's for how long he or she takes breaks, or if the hielihood of developing diseases. It is now way the monitoring information is used causes used only in a few workplaces, usually to stress-related illness. By the same token, if a identify workers who may be hypersuscep- worker's job requires a high degree of coordi- tible to chemicals found in those work- nation, good judgment, or trustworthiness places. However, researchers expect that e.g., law enforcement or air traffic control tests for many common diseases will even. evidence of drug use could be said to be an ob- tually be commercially available; employ- jective measure of unfitness to do that job. re or insurers may want to include them There is a huge gray area between the ex- in preemployment physicals. tremes. Service observation, the practice of Brain Wave Testing.Still in the research listening in on employee's telephone calls with stage are a number of tests based on brain customers, has elements of monitoring both waves. Currently under study is the pos- the work process and the worker. It appears, sible use of brain wave analysis in monitor- however, that the intensity of the privacy de- ing the level of concentration, detecting bate increases as we move from techniques that lies or "guilty knowledge," and predict- focus on the work to those that focus on the ing certain illnesses. A computer-based worker. system to detect drug use by measuring Although this report is about electronic mon- brain waves is already on the market. itoring technologies, most of which appear to In addition, OTA looked at brain wave re- be on the "work performance" end of the con- search, which in the view of some experts prom- tinuum in figure 1, OTA looked briefly at sev- ises improved systems for testing for drug use, eral technologies from the "worker testing" honesty, and susceptibility to disease. (See ch. end to see the sorts of questions raised through 4 and app. B for a more detailed discussion.) their use. These technologies are discussed in Serious questions have been raised about the more detail in chapter 4 and appendix B. They accuracy and reliability of all these tests, as is discussed in more detail in appendix B. Poly- Polygraphs.The polygraph is not a new graph tests have not been shown to have any technology. It has had limited use in law validity in employment screening situations, enforcement for 60 years. Now, however, and research shows that they give a high rate its dominant use is in personnel screening; of false positive results (innocent people iden- of 2 million polygraph tests given annually, tified as deceptive.) Nor has there been research about 98 percent are given by employers indicating that use of polygraphs reduces pil- to job applicants Slid employees.' ferage and other crimes in the workplace. Drug Substance Abuse Testa.Medical screen- tests can be unreliable due to poor handling ing for drug or alcohol use, formerly used of urine specimens, sloppy lab work, or poorly primarily as a diagnostic tool in clinical calibrated test equipment. Regulation of commercial labs is spotty, and there are few mechanisms available to enforce high-quality work Harrison Donnelly, "Privacy in the Workplace,"Editorial Research Reports,Mar. 21, 1986, p. 214, citing figures from Employers who use polygraphs or drug the American Polygraph Aseodation. tests, the types of testing now common, as-

24 Ch. 1Introduction 15

sert that testing is necessary to protect their Fairness. In this context, the concept of businesses and to maintain a safe environment fairness encompasses both the accuracy of the for employees and customers. On the other tests and the concept of "due process" within hand, civil libertarians and others argue that the testing program. Serious doubts have been these gains, to the extent they are actually raised about the accuracy and validity of all achieved by testing, have a heavy cost: undue the tests discussed above. There is also con- intrusion into private lives of employees; cre- troversy about how testing programs are to ation of an atmosphere of fear and intimida- be constructed: should tests be given on a regu- tion in the workplace and false accusation and lar basis to all employees or randomly selected denial of job opportunities for many innocent employees, or should they only be given to People- those who have shown by their behavior that Privacy.Drug testing by urinalysis is there is reason to think they have been using clearly intrusive in that it requires the subject drugs. A number of court decisions have struck to produce a urine specimen under observation. down testing programs that have not relied Genetic tests require removal of a blood speci- on probable cause, or at least a reasonable sus- men. Both the polygraph and brain wave test- picion, that the person to be tested is using ing require the subject to wear electrodes at- -Irugs. However these cases have all involved tached to the skin. Beyond these physical State, local, and Federal Government employ- intrusions, however, is another privacy prob- ees who are protected by the fourth amend- lem. Privacy also encompasses the ability to ment against unreasonable searches. Employ- withhold certain information about oneself, ees do not have this protection from private and some of these tests reveal information that employers. is not only personal but is arguably not rele- Work or performance monitoring tends gen- vant to the employment situation. Drug test- erally to raise debate about stress, fairness, ing by urinalysis cannot determine when the and the quality of work life, including ques- drugs were used or whether drug use actually tions of privacy and autonomy. Worker test- impairs job performance. Polygraph testing ing, which tends to be more intrusive and ex- especially has raised controversy because some tensive, very clearly raises controversy over employers' tests include personal questions individual rights of privacy (i.e., employer and particular questions on religion, sex life, po- employee rights to know and control certain litical beliefs, or union affiliation. A further personal information) and also questions about privacy question relates to the privacy of the the accuracy and reliability of the test results. records generated by the tests, both within the In all cases there seems to be some question firm and outside. Such records, once released of balancing the need of employers to gather to insurance companies, employment clearing- information and the desire of employees to houses, or others, might follow a person keep personal control over some aspects of throughout his or her career. their work and/or private lives.

HISTORICAL BACKGROUND Monitoring in History seems clear that work monitoring has been an integral part of industrial development; in Work monitoring is not new. Employers many ways work monitoring seems to have in- have tried, since the earliest days of organized tensified as industrialization has progressed" human endeavor, to keep track of how well their employees were working or how much °This section draws on Sandra L. Albrecht, "Historical P ack- they produced. The organization and supervi- ground to the Electronic Monitoring of Office Work." contract sion of work have changed over time, but it paper prepared for OTA, August 1986.

25 ImIlm. 16 The Electronic Supervisor. New Technology, New Tensions

Industrialization in the United States has were produced under this system.9 Textile largely been characterized by a separation of production employed the largest number of planning or organizing the work process from home v. orkers, primarily women and children, the actual work itself. Organizational struc- to do spinning, weaving, and production of tures have evolved that rely on division of labor hand cards for combing cotton and we and place primary knowledge about the productioL process in the hands of managers Putting-out is a transition stage between rather than individual workers. This gives rise craft production and factory labor, the precur- to a ne,...d for coordination, control, and stan- sor to mass production. It coexisted for some dardization of work. The search for greater con time with the early mills and factories, but it trol leads to a need for more intense monitor- disappeared by the mid-19th century except ing, whether of processes, or work groups, or in the the garment industry where hom: workers continued to be employed. This industry individual workers, in order to give manage- is currently seeing a resurgence in what has ment feedback to make future decisions. typically been seen as a pre - industrial work form." At the same time, new technologies have been adopted which incorporate certain skills For the most part, deficiencies in the putting - in the equipment, with a corresponding " de- out system gave rise to the factory system. slolling" of work and workers over time.' A One factor in the development of the factory less skilled work force is one which is easier was the issue of work monitoring. With the to manage through intensified monitoring. putting-out system, workers set the pace of This is not to say that deskilling is always the their work day and control of the work process inevitable outcome of technological change. It was in their hands. The factory system can be is important to look at this long-term trend, seen as a social control mechanism, where which may be obscured by unevenness in in- workers were collected together and could be dustrial development. As certain occupations monitored (watched) by supervisors or over- undergo de-skilling others newly created may seers, both to increase work discipline and to require skills heretofore unknown. Case histo- discourage theft. ries of individual occupations, however, show Some view the social control of workers by that over time these new occupations can also employers as the primary reason for the de- undergo a &mauling process .° This trend velopment of the factory. Others focus on the may underlie growth in the use and intensive- inability of the putting-out system to effec- ness of monitoring over time. These trends are tively utilize newly developed machinery that illustrated by the history of work organization required a central power source.i2 This inte- in the United States dating from colonial times. gral interconnection of social control, organizational structure, and technology is a defining The Early Factoey.Although the American characteristic of industrial development. Industrial Revolution dates from the mid- 1800s, a pre-industrial system of home-based production, known as the "putting-out system," already employed home workers for piece 'Alfred D. Chandler, Jr.,The 'risible Hand: The Managerial Revolution in American Business(Cambridge, MA: Harvard rate wages by the late 1700s. Such diverse University Press, 1977), pp. 19 and 53. products as shoes, furniture, lace, and textiles "Edith Abbott,Women in Industry(New York, NY: D. Ap- pleton, 1913), ch. 2; and Victor S. Clark,History of Manufac- turers in the United Stater 1607.1860(New York, NY: McGraw- 'See Andrew Zimbelist led.),Case Studies on the Labor Proc- BM Book Co., 1929), pp. 183 and 539. sm(New York, NY: Monthly Review Pews, 1979), both his dis- "Sandra L. Albrecht, "Industrial Home Work in the United cussion of Bravirmen's thesis in the Introduction and subs States: Historical Dimension', and Contemporary Perspective," *oat chapters on case studies of the de-sldlling process. Economic and Industrial Democracy,voL 8,1982, pp. 413-430. 'Sandra L. Albrecht, "Historic al Background to the Elec- "See discussion in Dan Clawson,Bureaucracy and the La- tronic Monitoring of Office Work," contract paper prepared for bor Process: The Transformation ofU.SIndustry, 1860-1920 OTA, August 1988. (New York, NY: Monthly Review Press, 1980), ch. 2.

20 Ch. 1Introduction 17

The factory system collected works s to- The significance of scientific management gether under one roof and joined them through is its extension of the control of work by man- a cash nexus; labor power was a commodity agement. Prior to scientific management, the sold by workers and bought by employers. overall setting of the workplace and workers Monitoring increased, in that overseers could were monitored: they were concentrated to- count the output and enforce working hours gether and supervised, work hours were deter- of each individual or group, but the early fac- mined, and discipline used to ensure produc- tory maintained a mixture of traditional and tion quotas. But, the actual performance of new worl. fame. New patterns of work hours, work was left in the hands of workers. The cen- work pace, and discipline were instituted; but tral core of Taylor's management philosophy supervirion, though often despotic, was pri- was the idea of the "separation of conception marily indirect. Management was small in size from execution?"16 Decisions about the every- compared to contemporary standards and less day performance of work were removed from knowledgeable about the actual nature of the workers, and centralized in the hands of wan- work process. "Inside contractors," skilled agement, who in turn would determine the workers who understood how bo produce the most rational and efficient method of per- product, were often responsible for hiring em- formance. This brains/hands dichotomy makes ployees and overseeing the process." In cer- managen_dnt the depository of all knowledge tain ways, inside contracting carried on the tra- about the work process, capable of determin- ditions of craft production and brought old ing in minute detail the tasks to be carried out. styles of personal relations into the factory. Workers, divested of this knowldge and con- But, as industry expanded and employers trol over determining work, were responsible looked to greater rationalization, efficiency, only for carrying out the designed tasks. and intensity of labor, this indirect monitoring of work was seen as an obstacle to increased With this new managerial approach, work productivity. Managerial philosophies soon un- monitoring intensified. Every task within a job derwent change. came under scrutiny, and elaborate tally sheets and production forms were developed to rec- Scientific Management and the Ass mbly ord each detail in the operation. M- isures of Line.Frederick W. Taylor, known as the fa- hand and eye movements, spacing between ther of scientific management, began what he worker, machine, and product, time per task, defined as the scientific study of work in the level of efficiency through the day, and the ef- 1880s, but it was not until the 1'.;16s and later fect of rest periods on production were some that his work began to be adopted. Taylor was of the many new calculations performed in an not the first prson to scientifically study work; effort to monitor production more closely. In craft workers had historically valued the addition to the information collected on work knowledge of the labor process as well as its tasks and worker performance, there was in- production. Taylor's work, rather, was the sci- creased emphasis on developing new tools and entific study of themanagementof work, and machinery that would conform to the grow- represented the culmination of managerial ing detailed division of labor tasks. This new ideas developing in Great Britain and the technology bore a design influenced by em- United States throng: rout the 19th century." ployers' interest in increasing control over work and productivity. Jobs became more sub- divided and fragmented This detailed division "Ibid., ch. 3; and Graeme Sideman, Class and the Corpora- of labor which separates various job aspects tion (Great Britain: Fontana Paperbacks, 1981), pp. 37-41. into distinct parts and assigns them to differ "For this discussion, see Harry Braverman, Labor and Mon- ent workers diminishes both the skill and cost opoly Capital: The Degradation of Work in the Twentieth Cen- tury (New York, NY: Monthly Review P113111, 1974), chs. 4 and 5. Frederick W. Tsylor describes nis ideas in The Principles of "Harry Braverman, Labor and Mc-opoly Capital: The Degra- Scientific Management (New York, NY: W.W. Norton & Co., dation of Work in the Twentieth Century (New York, NY: 1911). Monthly Review Press, 1974), p. 114.

27 18 The Electronic Supervisor. New Technology, New Tensions of labor.° Scientific manage: by separat- Most measures of work production depended ing knowledge from performance, in 'eased the on paper and pencil tallies of items completed. ability of employers to monitor the workplace, However, there were also mechanical aids. not through overbearing surveillance, but by Clock-driven "time stamps" were used to re- investing them with the knowledge and deter- cord precisely when clerks received and re- mination of how work was performed. turned measured batches of work. Devices at- The development of the assembly line pro- tached to typewriters for counting keystrokes vides a good example of the changes that were or lines of typing were first used in the early occurring during this period." The Ford Mo- 19008.19 The practice of posting charts or ta- tor Co., established in 1903, began with the bles with each clerk's performance statistics, employment of highly skilled workers, former "to excite the emulation of others," was con- bicycle or carriage mechanics, who built entire sidered a useful tool for increasing produc- attomobiles. As the demand for the Model T tivity.° rose, Ford introduced full assembly lines. While scientific management as a basic ori- Rather than skilled mechanics, unskilled and entation of management philosophy continues semi-skilled workers performed small opera- today, there have been other philosophies of tions in stationary positions along the endless- management with impect on U.S. industry. chain conveyor. This system greatly increased One was the "human relations" approach management's ability to control and monitor fostered by E. Mayo and colleagues at Har- both the pace and intensity of work. Introduc- vard Business School in the 1930s and 1940s. tion of the assembly line, with its skill reduc- This philosophy emphasized the social aspects tion and corresponding wage leveling, met of work and the importance of social support worker resistance even at its earliest stages. from fellow workers in helping determine In 1913, Ford had a labor turnover rate of 380 worker productivity. Variations on this theme percent, and a unionization drive began.18 continue to the present. The human relations Scientific Management in the Office.The ap- approach did not replace scientific manage- plication of scientific management greatly in- ment, and by the 1950s, the iss43 of power and creased the volume of information to be trans- real differences between managerial and em- ferred from the plant level to the office. The ployee interests were accepted in many man- result was a rapid growth in the number of of- agerial theories. The challenge was to integrate fice workers, both managerial and clerical. Sci- work organization goalsharmony, productiv- entific management, originally conceived for ity, profitswith those of the employees. factory employment, was also introduced into the office. Privacy in the Workplace Many offices were restructured according to The idea that a worker should have some ex- a more "industrial" style of organization: jobs pectation of privacy in the workplace is a new were broken down into more detailed tasks, one, one that is only beginning to develop in skilled aspects of the job were separated from American law. Only a little more than a cen- lesser skilled operations, and the tasks were tury ago, the employer-employee relationship distributed among differentially paid employ- was viewed as analogous to the zr.aster-servant ees. Such firms as insurance companies, where relationship; the master had some paternalis- work was repetitive and easily measured, be- tic responsibility for the welfare or moral de- gan to incorporate an assembly-line approach velopment of the servant; the servant owed to the flow of work through the office. obedience and good service. Owners of early 111bid., pp. 72-83. "See, for example, William H. Leffingwell, "This Plan More "Ibid., pp. 146-161; and Melvin Kranzberg and Joseph Gies, Than Doubled Our Typists' Output,"System,voL 30, Novem- Bythe Sweat dale Brow(New York, NY: G.P. Putman's Sons, ber 1961; and William H. Leffingwell, "What 'Scientific Man- 1978), ch. 13. agement' Did for MyOffice," System,vol. 30, December 1961. °Keith Sward,The Legend of Henry Ford(New York and '')William H. Leffingwell, "What 'Scientific Management' Toronto: Atheneum, 1948), p. 32, cited in Braverman, p. 149. Did for MyOffice," System,voL 30, December 1961. Ch. 1Introduction 19

factories believed they had the right, indeed States, once in the 1920s and again in the 1950s the responsibility, to strictly controlmany when employers compiled psychological pro- aspects of their employees' lives, on and off files, employment histories, and other files of the job. In the factories of the 1800s, work rules personal data quite unrestrainedly." governing church attendance, place of resi- dence, and nightly curfews were not uncom- During the changing social climate of the mon. Even as late as the 1910s, the Ford com- 1960s and 1970s, court decisions and worker pany employed a group of 50 social workers protection legislation gave employees some to investigate employees' neighborhoods, protections in how their employers could use home conditions, finances, and habits to de- information about them and placed a greater termine if they were worthy of profit sharing burden on employers to demonstrate scientific bonuses." validity of employment tests. Other legislation, like the Occupational Safety and Health Act, Gradually over the course of the 1800s, U.S. gave employees certain protections as well as courts began to view the employment relation- rights to information about hazards in the ship as analogous to a contract between equals, workplace. Antidiscrimination legislation be- with the employer buying the labor that the gan to limit employers' right to discriminate employee wished to sell. This view gave rise on the basis of race, sex, religion, age, and (in to the notion of "employment at will." Each some States) union activity. A number of party was free to enter or refuse the contract States have passed "mini-privacy acts" to pro- for any reason; if either was later displeased vide some protection of workers records. In for any reason, either was free to bleak it; the addition, a number of court decisions in the employer could fire, or the employee could quit. past two decades have further eroded the doc- The contract analogy does not recognize the trine of employment at will, limiting em- very large differences in bargaining power that ployers' freedom in firing employees. often exist between a single individual anda corporation. The National Labor Relations Act The changes in legislation and in social val- of 1935, which obligated employers to bargain ues in the 1960s and 1970s gave some measure with workers' representatives over hours, of additional power to the individual in an em- wages, and working conditions, marked one ployment relationship, and led people to the early action of the Federal Government to mod- expectation that they had certain rights. in- ify the employment-at-will doctrine. cluding the right to privacy. Workplace privacy is a new right, however, and probably a The concept that ar employee has a right tenuous one. It will be tested on two fronts: to privacyeither to be free from intrusion or by the drive toward higher productivity, which to keep certain information privateis a rela- encourages employers to use electronic moni- tively new one. Throughout the previous cen- toring, and current social concerns, such as tury and up through the 1950s, the right of drug abuse, that encourage employers to employers to inquire into any aspect of an em- gather more and better information about the ployee's life was virtually undisputed. Em- people they hire. Although employees are now ployers could choose their employees in any beginning to feel a right to privacy in the work- way they wished, and were quite free to say place, these pressures to gather information, "We want only this kind of person work- along with availability of the technological ing."22 Worker testing has gone through at means to collect that information, may weaken least two periods of popularity in the United the development of this emerging right. "Robert Ellis Smith, Workrights (New Yc-rk, NY: Dutton, The field of players involved in labor rela- 1988), pp. 18-18. Also, Stephen Meyer III, The Five DollarDey: tions questions is broadening. Traditionally Labor Movement and Social Control in the Ford Motor Com- pany, 10064921 (Albany, NY: State University of New York the parties involved were the employer, em- Press, 1961), pp. 34-85. "Citing Alan Westin incrime Donnelly, "Privacy in the Workplace" Editorial Research Reports, Mar. 21, 1986. ""Can You Pass The Job Test," Newsweek, May 5, 1988.

20 1/4 20 The Electronic Supervisor. New Technology, New Tensions

ployees, and, if it existed, a union. Government bor, women, and environmental organizations ;involvement has been limited to establishment has strengthened around the introduction of of guidelines for union contracts. Government VDTs into the workplace, and on related is- has also become involved through laws that sues such as computer monitoring. Working cover all workers and workplaces regardless together, these groups are largely responsible of the union or the collective bargaining agree- for the generation of legislative efforts in at ment. These include laws on child labor; mini- least 22 States to explore VDT standards. mum wage; nondiscrimination on the basis of While most of these have not passed, some race, religion, sex, and age; and finally protec- have resulted in advisory guidelines, as in New tion of health and safety on tha job. In addi- Mexico. In a few States these same coalitions tion, a number of laws have been passed in the have pushed for laws concerning electronic past decade at the State level governing the monitoring and service observation. privacy of employment and medical records The declining proportion of the labor force and the use of polygraphs in employment. represented by unions is one of the factors in- Such laws were enacted because of heavy lob- fluencing the move toward legislative solutions bying by a range of groups including unions, to worker and workplace problems. There is civil rights advocates, women, environmen- persuasive evidence that efforts to establish talists, community-citizen alliances, health pro- expanded employee rights through State and fessionals, and others. This move to the legis- local legislation will continue, both in the areas lative area to deal with workplace issues has of electronic monitoring and in worker testing. accelerated in recent years, particularly at the Legal and policy questions are summarized be- State and local level. The coalition among la- low and discussed in greater detail in chapter 4.

POLICY OPTIONS Before addressing the problem of how Con- Option 1: gress might act, it is first necessary to con- Take no Federal action concerning work sider whether and when action may be appro- monitoring at this time. priate. Some factors suggest that a "wait and Questions of the fairness of work monitor- SPA" posture may be appropriate; uncertainty PY ing practices would be left, as they are at about whether monitoring causes stress, the present, in the hands of stakeholders, em- lack of judicial precedent, the possibility of pri- ployers and employees. In industries where la- vet aly negotiated restraints on monitoring, bor unions are active, collective bargaining and marketplace checks on monitoring are with regard to technology change, monitoring, among these. Other factors indicate that Con- and methods of evaluation would continue un- gress may want to act now to alleviate grow- der the current practices. ing concern about monitoring in the workplace. These include the lack of union representation Although many unions have adopted posi- in the bulk of the monitored work force, the tions opposing electronic work monitoring, inadequacy of current law to address concerns their bargaining strength with respect to it, over health, privacy, and dignity, the difficul- whether by .nformal negotiations or by formal ties of legislating against powerful economic collective bargaining or arbitration, is prob- interests at the state level, and the increasing ably not great. However, the monitoring that sophistication of the technology itself. Several does take place varies between industries and possible directions of Federal policy are de- companies. An argument can therefore be scribed below. made that, pending the development of a Ch. 1Introduction 21

longer history of negotiations between labor may be successful, giving rise to a variety of and management on this issue, monitoring is legislative or regulatory approaches to deal- best addressed at the company or union level. ing with the issues related to electronic moni- The parties concerned are most familiar with toring. Some may serve as models for Federal the specific problems, and contracts, rather action at some later time, should the need for than national policy, may be the best way of the harmonizing effect of national legislation approaching what appears to be situation- be seen more clearly in the future. specific problems (see ch. 4). Under these circumstances, Ccagress may want to avoid legis- Option 2: lating on the issue of monitoring per se, and Establish whether stress effects of elec- instead make monitoring an item for compul- tronic monitoring are an occupational health sory arbitration or collective bargaining under hazard; if they are, consider creating Fed- Federal labor law. eral legislation or regulations governing the use of electronic monitoring. This, of course, does not necessarily ensure an outcome that is satisfactory for the majority The effect of monitoring on stress and of monitored workers, who are not unionized healthissues which might provide the pol- and are therefore powerless to negotiate fair icymaker with the most direct and least value- monitoring practices, or any other aspects of laden approach to acting on monitoringis in the quality of work life, through the collective a state of scientific uncertainty. There exist bargaining process. Furthermore, an increas- few authoritative studies on the effects of elec- ingly large segment of the work force is made tronic monitoring on health. Some studies and up of temporary workers, who, since they come informal polls of workers have suggested that and go on a weekly or monthly basis, have little monitoring has stressful effects, and there is ability to improve the quality of work life. a certain common sense appeal to the idea that working in fast paced, highly monitored envi- There is the argument that natural "mar- ronments may be highly stressful. However, ket forces" may tend to limit unfair monitor- there is no research separating the effects of ing and preclude the need for congressionalac- monitoring from other office stressors, nor is tion even on behalf of nonunionized workers: much known about the types of monitoring employee backlash, low morale, and high turn- that are stressful, how stress might be reduced, over should dissuade employers from monitor- or how stress due to monitoring manifests it- ing practices that their workers findonerous. self (if at all) in physiological symptoms. Un- If monitoring is indeed stress-producing, then til more is known about the effects of moni- employers who use it will inevitably see the toring on health, policy action under a "stress" effects of stress on diminished quality and ou t- rationale may be premature. The policymaker put of its product or service. The response to may consider it appropriate, therefore, to ini- this is that many monitored jobs are routine tiate studies on stress in the workplace, and work that is subject to and indifferent to a high on the role that monitoring plays in such stress. turnover rate, and in many instances, high attrition works to the employer's benefit (by The National Institute of Occupational lowering the costs of pension, salary increases, Safety and Health would seem to be the logi- etc.). Thus it is not clear that "natural" checks cal agency to supervise or carry out studies will be sufficient to ensure that monitoring is of stress as a workplace hazard. Specific not abused. studies of monitored workers would have to be done with an eye to separating the effects If natural checks are not sufficient, politi- of monitoring from those of other workplace cal action is still available. Unions and other stressors, a major deficiency in existing stud- interest groups have worked to pass State-level ies. In addition, however, it would be useful legislation on monitoring, service observation, to understand more about the phenomenon of or VDT health and safety. These activities will workplace stress in general, given the rising probably continue. Some of these attempts number of worker compensation claims and 3 i Z3 The Electronic Supervisor: New Technology, New Tensions

other evidence of the growing importance of That these issues are not currently addressed stress in occupational health. Research may in law does not mean they could not be. As reveal that other factors in the workplace are is discussed in appendix A, a number of other rixas important as or more important than mon- countries have quality of work life legislation. &ging in contributing to stress-related illness, Such legislation could give guidelines on the and that these should also be covered by pro- rights to health, safety, privacy, constitutional tective legislation or regulation. protections, or information that employees can 0Pdo & expect to enjoy in the workplace. As indicated Consider Federal legislation aimed at earlier in this chapter, the erosion of the doc- gaps in current law. This could be in two trine of "employment at will" through anti- possible direcdona general legislation aimed discrimination, health and safety legislation, at establiddig certain rights for employees and public interest concerns, has already within the workplace, or surgical legislation marked some invohement of the U.S. Govern- aimed at specific monitoring practices. ment in regulating the work environment. The issue of electronic monitoring in offices is prob- There have been few, if any, court cases chal- -bly too narrow to serve as a basis for com- lenging the types of monitoring considered in prehensive work environment legislation. It this report. Two differing conclusions can be should be just one factor of many to be con- drawn from this. The first is that, until the ju- sidered in determining what rights U.2. citi- diciary acts, Congress has no way of knowing zens have in the workplace, both as employers kj- the type of legal inadequacies it should address, and employees. and ought therefore wait to legislate on work monitoring. The second is that current law is inadequate even to form the basis for a law- However, if blanket legislation on work life v. suit, and that Congress must take the lead in quality is neither wise nor desirable, Congress providing rights to monitored employees, might address concerns over specific issues cit- should it decide that certain forms of monitor- through the use of specific amendatory legis- : big are unreasonable. lation. If, for example, telephone call accounting is an area of particular concern, Congress Current worker protection legislation gives might cddress the problem specifically by workers a variety of rights, such as the right amending the Electronic Communications Pri- to a minimum wage, to organize, to bargain vacy Act to comport with what it considers L collectively, and increasingly, the right to know "fair" monitoring practice. The guidelines de- about health and safety hazards that form part veloped for the audit conducted by the Gen- of the working environment. However, U.S. eral Services Administration for the Presi- law has not heretofore involved itself deeply dent's Council on Integrity and Efficiency P- in quality of work life issues nor in issues of might form a template for such legislation, or !. privacy or dignity in the workplace. instead, Congress may mandate alternatives 1=nenolegal right to be treated with dig- to telephone call accounting discussed in chap- sky or as an autonomous person. There is no ter 3 of this report. legal right to a well-designed, interesting job, nor is there law that compels employers to consider employee input in decisims about new Another example of an area of the law not technology or new monitoring procedures. To currently addressed, and on which Congress the extent the law treats privacy in the work- may wish to act, is what might be called trans- place, it looks to a standard of what an em- actional privacy, or the collection of "informa- ployee might reasonably expect to remain pri- tion about information." For example, the vete; as discussed in chapter 4, this standard number of keystrokes, the number of visits to may fail as a guide for action in the face of the restroom, the destination of calls, etc., all ployer's increasing use of monitoring, surveil- provide information about transactions, rather boos, or testing technologies. than about the content of communications or

3 2

11.1111rat-- Ch. 1Introduction 23

activities (see part II of ch. 4).24 Although whether security systems are workingprop- present law, such as the Privacy Act and the erly. The collection of transactional data be- Fair Credit Reporting Act, regulates whatcan comes most subject to controversy when it is be done with transactional informationonce collected about the performance ofan individ- collected, it does not forbid its collectionas ual worker. It may be that Congress would such. As discussed in chapter 4, however, the choose to treat electronic monitoringas a collection of transactional information, particu- "right to know" issue for workers; that is,em- larly if done on an intensive basis,can arouse ployers could have the right to collect what- feelings of having one's privacy, dignity, and ever kind of transactional data they wish about autonomy invaded. Moreover, because of the employee performance, but would be required power of computers to generate profiles and to give employees access to, and if need be, croesmatch many transactions, transactional correct, this information. information can yield informed estimates of the substantive content of communicationsor As this report indicates throughout, how- patterns of behaviorit can be, in other words, ever, the issue of work monitoring cannot be a "back door" for getting at personal informa- adequately understood, nor appropriately ad- tion that existing law regulates. dressed, in isolation from larger labor- Certainly, to forbid or regulatf, the collection management, privacy, and health and safety of all transactional information would be contexts in which it is embedded. Nor will spe- un- cific policy actions taken with respect topar- reasonable. Much transactional data collected ticular forms of monitoring necessarily end the by electronic monitoring software is usedto controversies arising out of the application of monitor equipment utilization, to track totals of transactions made, and to determine new forms of technology to the workplace. The policymaker -,noukl therefore be aware thatan exclusive f )cus on the forms of monitoringcon- "Transactional information, it will be recalled, differs from sidered in tnis report will at best form the ba- substantiveinformation, in that the latter reveals the content sis for a series of patchwork solutions to what or meaning of communications or documents. Transactional in- has been a perennial issue between workers and formation, in contrast, reveals facts about communications or documents. employers.

SOURCES OF INFORMATION FOR THIS REPORT In compiling this report, OTA used a num- monitoring and telephone call accounting. A ber of major sources of information in addi- subcontractor to the Westin project, Dr. Elaine tion to published literature cited in the foot- J. Eisenman, provided a paper summarizing notes throughout the report. her 1984-85 research on employee perceptions One major OTA contract, completed by Alan of monitoring at six private sector sites (three F. Westin of the Education Fund for Individ- unionized and three nonunionized). Her find- ual Rights, includes the results of site visits ings are based on questionnaires and group and interviews of 110 public and private offices workshops conducted with 365 employees and in 1983 and 1985-86 to examine their use of 27 supervisors.° office automation including electronic work OTA also participated in a survey on office monitoring." The Westin report also summa- automation equipment use that was conducted rizes some of the legal implication:- of work "Elaine J. Eiseman, "Employee Perceptions and Supervi- sory Behaviors in Clerical VDT Work Performed on Systems "Alan F. Westin, "Privacy and Quality of Life Issues in Em- that Allow Electronic Monitoring," preparedas part of contrac- ployee Monitoring," contractor report for OTA, 1986. tor report for OTA, April 1986. 3 24 The Electronic Supervisor. New Technology, New Tensions by Hay Management Group in 1986.2' In of unionized and nonunionized workplaces in that survey of 45 large New York area firms the United States." OTA inserted several questions to determine Information about the use of electronic mon- how many of the firms make use of electronic itoring in other countries came from a report monitoring for purposes of planning or individual evaluation. by Russell Pipe and Alan F. Westin of the Edu- cation Fund for Individual Rights, and one by OTA staff conducted semi-structured inter- Dr. Steven Deutsch, of The Center for Work views that encompassed 35 work locations in and Society, University of Oregon. Dr. Deutsch private industry and in the Federal Govern- also provided a paper on the context of labor ment to discuss the use of work monitoring management relations in the United States." and the reactions of managers and workers to it. Information on historical evolution of work monitoring came from a paper by Dr. Sandra In addition to the staff interviews, OTA con- Albrecht, University of Kansas" and contri- tractor Dr. Michael J. Smith, of the Univer- butions of Dr. Sharon Strom. sity of Wisconsin, conducted semi-structured Information on telephone call accounting, in interviows of 41 monitored workers at 5 work addition to published sources, came from OTA sites." These interviews were in support of staff interviews of approximately 12 commu- his report to OTA on behavioral and psycho- nications managers and 3 telecommunications logical implications of monitoring. consultants. Staff also used a mini case study An ad hoc group of representatives from a -4 the State of New York telephone system, number of labor organizations working under based on interviews and documents provided the auspices of the AFL-CIO provided a set by the State telecommunication office. Staff of 34 case examples highlighting employee re- also interviewed officials of the GencPal Serv- action to the use of monitoring at a variety ices Administration and a number of Federal agencies. In addition, staff interviewed two experts on whistleblowing.

"Cited in this report as "AFL-CIO Case Examples," Novem- "Hay Group Inc., "1986 Office Systc me Survey," Septem- ber 1986. ber 1986. "Steven Deutsch, "The Context for Exploring Workplace "Michael J. Smith, Pascale Carayon, and Kathleen Miezio. Monitoring," contract paper prepared far OTA, September 1986. "Motivational, Behavioral, and Psychological Implications of "Sandra L. Albrecht, "Historical Background to the Elec- Electronic Monitoring of Worker Performance," contractor re- tronic Monitoring of Office Work," contract paper prepared for port prepared for OTA. July 1986. OTA, August 1986.

ORGANIZATION OF THE REPORT Chapter 2 of this report discusses the tech- quality of work life issues as well as a discus- nology of computerized work measurement, sion of policy alternatives related to work mon- some of the jobs in which it is used, and the itoring and telephone call accounting. workplace issues raised by its use. Chapter 3 focuses on the use of telephone call account- In addition, this report has tw- appendixes: ing in both government and private firms, appendix A discusses policies toward monitor- along with the use of other technologies to ing in some foreign countries while appendix manage telephone costs. Chapter 4 presents B summarizes the issues raised by worker a discussion of the legal aspects of privacy and testing.

34 Chapter 2 Using ComputersTo Monitor OfficeWork CONTENTS

Page Introduction 27 What Kind of Work Gets Monitored? 28 How Widespread Is Electronic Work Monitoring? 30 Work Monitoring in the Private Sector 30 Work Monitoring in the Federal Government 33 How Technology Assists Selected Business Applications 34 Workplace Issues Related to Work Monitoring 36 What Is Fair 38 Standards of Performance 40 Quality of Work 45 Job Design and Work Process 47 Supervision and Evaluation 48 Personal Computer Monitoring 49 Monitoring and Stress 50 Computer Pacing and Stress 55 Feedback and Motivation 55 The Future of Work Monitoring 57 Figures Figure No. Page 2. Example of a System Status Report for an AutomaticCall Distribution System 35 3. Example of a Supervisor's Display in an AutomaticCall Distribution System 36 4. Example of Agent Statistics From an AutomaticCall Distribution System 36 5. Two Models for Electronic Work Monitoring 37 6. Example of Individual Work Monitoring Report(Performance Summary) 41 7. Example of a Work Monitoring Report for a Group(Volume of Work Accomplished by Group in One Week) 42 8. Example of Individual Work Monitoring Report (HoursSummary) ... 43 Tables Table No. Page 1. Some Office Jobs Currently Subject toElectronic Work Monitoring ... 29 2. Conditiors Most Favorable to ElectronicWork Measurement 29 3. Workplace Issues Related to Work Monitoring 38 4. Key Issues and ProPem Areas in MonitoringWorker Performance ... 39 5. Com?arison of Mowcored and NonmonitoredWorkers 49 6. Rateo of Frequent Health Problems Related toComputerized Monitoring of Work Performance 52 7. Rates of Frequent Health Problems Related toProduction Quotas or Productivity Standards 52 8. Review of Oregon Worker CompensationClaims Involving Mental Stress 53

3 6 Chapter 2 Using Computers To Monitor Office Work

INTRODUCTION Electronic work monitoring is the computer- impossible because of the pace of work; job ized collection, storage, analysis, and report- satisfaction is low, and stress and stress-related ing of information about employees' produc- health problems are the inevitable result of tive activities. Within this broad definition, the having to work under such conditions. primary focus of this report will be on obtaining data about employees directly through Electronic work monitoring is already a daily their use of computer and telecommunication reality for millions of U.S. workers. They seem equipments This type of information gather- to be having varied experiences with it. Some ing has been called "electronic monitoring," view electronic monitoring as a useful tool that "electronic work meastuement," or "telephone helps them get better control of their work, monitoring?* This chapter will also include a ensures that their supervisors give objective discussion of "servic s observation," the prac- evaluations, and helps their company be more tice of listening in on conversations between productive. Others believe they indeed do work employees and customers to evaluate the em- in electronic sweatshops as described above, ployees' courtesy or competence. Service ob- and that monitoring is an unfair surveillance servation is often used in conjunction with used to control them. Still others have mixed electronic work measurement for telephone feelings: they may not mind monitoring per operators and customer service workers. se, but they feel it isn't being used in the fairest or most effective way. Some people have warned that use of electronic monitoring leads to creation of electronic Electronic monitoring is usually used in con- sweatshops. "Electronic sweatshop"the term juction with a work measurement system. Work conjures up images that combine the worst fea- measurement systems usually do four things. tures of both the factory and the office: bor- First, they set standards for the time it should ing, repetitive, fast-paced work that requires take to produce certain units of work. Second, constant alertness and attention to detail, all they monitor the actual time it takes to pro- done under the pressure of constant supervi- duce each unit of work. Third, they analyze the sion and demands for faster work. Worst of variance of the actual time from the standard. all, the supervisor isn't even human. Employ- And finally, they provide data for use in plan- ees must labor at top speed under the view of ning, cost estimates, and productivity improve- unwinking computer taskmasters that record ment.' As more employees use computers in every item of work completed, along with every their jobs, computer software is increasingly mistake, rest break, and deviation from stand- used to monitor actual performauce, compare ard practice. A person's job depends on the performance to standards, and provide plan- computer's comparison of performance to the ning data. standard. Interaction with fellow workers is As more capabilities of the computer and telephone are being explored for office work, 'This definition is very simP or to one found m: Andrew Clem- it is probably natural that some of the same ent, "Electronic Management.: The New Technology of Work Place Stuveillance," Canadian Informatics' Processing Society capabilities are found useful in supervising Session 84 Proceedings. Calgary, Alberta, May 911,1984, pp. work as well. Both work and supervision are 259-288. becoming automated. Work done on a com- 'In this chaptar, "work measurement" is used to include the whole process of developing procedures and standards for job performance, collecting data on actual performance and com- 'See U.S. Department of Defense, Office of The Inspector paring actual performance to standards. "Work monitoring" General, Work Measurement Systems and Engineered Labor mars only to collection of information about actual performance. Standards (Washington, DC: October 198(.

27 3 7 fro Electronic Supervisor. New Technology, New Tensions puter is very abstract. Many information work- duction can be used to estimate future workers no longer handle concrete items like docu- loads, to plan for new personnel, or to justify ments, file folders, paper clips, orders, invoices, new equipment. Performance statistics may rubber stamps, or checks. Those physical ob- also be compared with budgeted costs to de- jects have disappeared into the computer and termine the cost-effectiveness of an operation. have become abstract analogs of their former Such complete and up-to-date information is selves. The various stages of work that trans- necessary in a cybernetic model of manage- form raw materials into final products take ment that requires immediate feedback to place inside the computer, too. A supervisor managers about current activities to be used who could only observe the physical activity as the basis of future decisions. of people at their computer terminals, without knowing what was going on inside the com- Computer work monitoring can also give in- puter, would know little about the work being formation on individual performance. The fo- done. cus of this chapter is computer monitoring of individual performance. Computer-generated The tool that gives a picture of what is hap- statistics can be used as a tool to increase or pening inside the computer is the computer it- maintain levels of employee performance. They self. Computers' capacities for recording and may be used in individual personnel decisions storing information make it possible to keep pay, promotion, retraining, and discharge. detailed records about all aspects of the pro- They can be a feedback tool to help employees duction process. And their ability to sort that gain more control of their own work; conversely, information in different ways means that the they can be used to limit employee decisions information can be put to many uses. about the work process. Like most technologi- Computer records can give a picture of the cal tools, work monitoring per se is neither bad total performance of a work group, or a depart- nor good. Its effects depend on how it is used. ment. Statistics on historical patterns of pro-

WHAT KIND OF WORK GETS MONITORED? Most of the electronic monitoring found by keystrokes, and in these high-production jobs, OTA and other researchers affects office work- counting keystrokes is an obvious way to ers with short-cycle "production" jobs, that measure performance. is, jobs where a limited number of standardized tasks are performed repeatedly to produce However, key entry is not the only job where some information-based end-product. Most production can be monitored electronically such jobs are considered clerical, for example using computer counts. Table 1 lists some of data entry or insurance claims processing. the office jobs that are often subject to work However, monitoring can also be applied to measurement from production data gathered professional jobs with a quantifiable output, through electronic monitoring. The list is by for example computer programmers or insur- no means exhaustive. The table summarizes ance underwriters. a few of the aspects of work that can be electronically monitored for each job. Data-entry jobs are perhaps the epitome of routine, standardized information-processing What do all the jobs in table 1 have in com- work. The operator reads information from a mon? Why is it that they lend themselves to source and enters it on a keyboard or computer monitoring? Jobs that are subject to be recorded on computer tape or to electronic monitoring are generally those disk. Most key-to-disk and key-to-tape data- that are subject to work measurement tech- entry systems are equipped to count operator niques. In work measurement systems, man-

38 Ch. 2Using Computers To Monitor Office Work 29

Table 1.Some Office Jobs Currently Subject to Electronic Work Monitoring

Job What is measured Hcw obtained Word processors speed, errors, time working keystrokes counted by computer Data-entry clerks speed, errors, time working keystrokes counted by computer Telephone operators average time per call each call timed by call distribution system Customer service workers iime per customer each call timed by call distribution system; number and type c I transaction transactions counted by computer Telemarketingiother sales ..time per customer, sales volumeeach call or transaction timed; sales tabulated by computer Insurance claims clerks ....number of cases per unit time time spent on each form tabulated by computer Mail clerks letters or packages per unit timecollected by letter of package sorting machines Bank proof clerks checks processed per unit of collected by proof machine time SOURCE: Office of Technology Assessment, 1967. agement sets standards of production and then perform them (more about work measurement makes records of actual performance in order standard setting below). Measurement is also to compare them to the standard. For many easier and more meaningful when there is a kinds of work, manual recordkeeping systems large volume of work from which to draw, or or physical counts of completed work are used a relatively continuous flow of work. to gather performance information, but the growing availability of computer-based sys- Routine tasks can be performed by inter- tems in offices has led many employers to col- changeable workers. Individual employees lect this information electronically. may come and go (turnover) but the work still gets done. These jobs do not require a worker Table 2 identifies some of the work condi- to have rare personal qualities, extensive tions most favorable to the application of work professional training, or highly specialized measurement The first set of conditions is that skills. The training required for most routine the work be routine and require the repetitive jobs is minimal, and the amount of time needed performance of a small range of tasks. When for a newly trained worker to reach full com- this is the case it is possible to time those tasks petence is usually short. Training for some and establish a standard amount of time in types of data-entry jobs, or for such jobs as which a competent worker can be expected to proof machine operators in banks, can be accomplished in as little as a few days. Other routine jobs, however, require more skills and Table 2Conditions Most Favorable to Electronic longer training. For example, training for tele- Work Measurement phone customer service representatives at Routinlad work firms interviewed by OTA ranged from a few small number of tasks performed by each employee days to 6 week q, depending on the firm and large volume of work relatively continuous supply of work the complexity of the services offered. Interchangeable workers: Although work measurement is most easily relatively low training requirement relatively small difference In the productivity of experi- applied to less skilled jobs, it is increasingly enced and Inexperienced workers, or short time needed being directed to higher level, more skilled tech- to bring inexperienced workers to full capability nical, professional, and managerial positions. tolerance for turnover ample labor supply Even the most complex work has its routine simple dais oollectlow elements, and given sufficient analysis, those employees use information technology as part of their work elements can be identified, grouped together, Information about transactions is already being collected and counted. The jobs of commodities broker, for other purposes data collection is transparent to the users, and making use computer programmer, and bank loan officer, of It Is simple for supervisors for example, could lend themselves to moni- SOURCES: Adapted from Robert E. Nolan, "Work Measurement," In Robed N. toring. They all have a high proportion of rou- Lewer, wore cally PluduciftflY Pow Yolk Nye McGraw mix, ismk also personal ctwnmunlcsllon from James Rule, August 1956. tine elements. But these jobs also require SO The Electronic Supervisor. New Technology, New Tensions

higher levels of training and experience. Work tics can be collected about many routine of- measurement and electronic monitoring can fice jobs that are not computerized, usually by be used in jobs like these, but if workers re- having the employee or supervisor keep paper sent them, the costs of resistance might be un- records. Care must be taken to design a means acceptably high for the employer. A firm's per- of data collection which does not unduly bur- ception of the interchangeability of certain den workers or supervisorsotherwise the types of workers, the ampleness of replacement measurement system may decrease produc- labor, or their own tolerance for turnover are tivity because it takes too much time or re- all relative. They can change over time with duces morale. Further, a work-measurement variations in corporate goals, job markets, system that requires people to take an extra managers' personalities, or internal corporate step to keep performance records may be sub- politics. ject to error or fraud. With electronic moni- As was pointed out in OTA's earlier report toring, manual recordkeeping can often be re- on office automation, the change in work proc- duced or eliminated, even vitae much more ess that takes place when certain types of detailed measurements are being made. professional or technical workers start mak- Jobs involving telephone contact with the ing use of computers, sometimes leads to public are often subject to "service observa- greater standardization or routinization of tion," that is, having a supervisor or quality their work. Some researchers hold that the control specialist listen in on employee tele- increased use of computers to assist profes- phone calls to evaluate courtesy, accuracy, or sional and managerial work will lead inevita- compliance with company guidelines. Service bly to the de-skilling of mental work and the observation is a common practice in telemar- * creation of "intellectual assembly lines. "6 keting firms, direct sales outlets, market re- Many employers are greatly concerned with search firms, companies with large customer getting higher performance from highly paid service departments, and of course telephone professional and managerial workersthe companies. Service observation is not new; tele- "last great frontier" of productivity improve phone companies have been practicing it for ment. Computer monitoring can offer a way over 80 years, as have many other firms. Serv- to make them more accountable and to meas- ice observation is also not automatic. It re ure their performance against performance quires a human listener to make judgments goals. There will be further discussion on elec- on the content of a call. tronic monitoring of professional, managerial, and technical workers later. However, new information technology has been transforming service observation by de- In table 2 the final group of characteristics veloping systems that integrate service obser- typical of monitored jobs is "simple data col- vation with other, more automatic, monitor- lection." As noted below, performance statis- ing techniques and also by improving the quality of new equipment. In older telephone 'U.S. Congress, Office of Technology Assessment,Automa- don sf Amwica's Offices,OTA-CIT-287 (Washington, DC: U.S. systems, for example, a drop in volume or a Government Printing Office, December 1985), p. 106. click might be heard when an observer came `Judith A. Perrone, "Intellectual Assembly Lines: The Ra- on the line. Most modern service-observation *mitigation of Managerial, Professional, and Tecimical Work," paper presented to the American Sodological Association, Wash- equipment is perfectly silent and does not in- bran, DC, August 1985. terfere with the operation of lines.

HOW WIDESPREAD IS ELECTRONIC WORK MONITORING? Work Monitoring in the Private Sector puter use as part of computer security programs, audit trails, or cost allocation programs. Thus, It should be noted that many computers re- in nearly every organization that has a main- cord information about individual workers' com- frame, minicomputer, or integrated word proc-

40 Ch. 2Using Computers To Monitor Office Work 31 easing system, computer-generated records are ceived responses from 40,000 readers of four made when a terminal logs on or off; when a major women's magazines Working Woman, database is accessed; when a file is created, re Ms., Glamour, and Essence. Of the 4,500 revised, or deleted; or when a remote computer sponses randomly selected for analysis, 43 per- is accessed through telecommunication lines. cent reported that they used visual-display ter- Some privacy implications of computer secu- minals (VDTs), cathode-ray tubes (CRTs), or rity systems will be addressed separately in personal computers (PCs). Of these users, 25 a companion report.' However, for the pur- percent were in managerial jobs, 30 percent poses of the present report, such records are were professional and technical workers, and not considered to be work monitoring. 44 percent were in clerical jobs. The clerical There are no reliable figures on how exten- category comNned secretarial, customer serv- sivr:y employers are applying computer-based ice, data entry, and similar job titles. software to monitor individual employee per- One question on the survey asked: formance, or to what extent they are using it to make judgments on individual pay, promo- Is your work measured, monitored, "con- tion, or discipline. No one has such figures, and stantly watched" or controlled by machine or no available basis for calculating them is at computer system. hand. Seventeen percent of all office automation Before trying to estimate how many office users answered "yes" to this question. When workers are subject to work monitoring by broken down by occupation, 20 percent of cler- computer, it is important to clearly define work icals answered "yes," compared with 15 per- monitoring. In this report, work monitoring cent of managers and professionals and and will refer to the computerized collectio: of 13 percent of technical workers. transactions on performance statistics used in Some critics have noted that the sample in individual work evaluations. Based on this def- this survey is self-selected, and that the results inition it appears, as will be discussed in this may not be representative of all women in the .ection, that only a minority of office workers U.S. work force. On the other hand, the posi- are monitored. tive replies to the question on monitoring may According to one work-measurement expert be lower than the actual incidence of electronic in the health insurance industry, some sort of monitoring in the United States. In the course statistics are collected about the computer of doing this study, OTA staff and contrac- transactions of nearly everyone who uses a tors often found it difficult to ask about work computerabout 80 percent of the people in monitoring with a simple yes or no question. the industry. However, he estimated that only Some people simply did not understand the for about 20 percent of the people were these question without further explanation. Work- statistics actually used to measure individual ers in some locations did not know that they performance.' This 20 percent represented were being monitored. In other firms, computer primarily the low-skill end of the clerical work use information was collectel but not used for force. but individual evaluation. Some information on monitoring was col- One small survey of office automation use lected in the 1984 National Survey on Women at 45 large New York firms, conducted for and Stress, conducted by 9 to b National Asso- OTA, suggests that electronic monitoring is ciation of Working Women. This survey re- still not widespread at those firms. Only eight firms s :8 percent) reported using information 'See Office of Technology Assessment, "Federal Policy on collected through electronic monitoring as a Electronic Information Security: Emerging Issues and Tech- basis for individual performance evaluations, nological Trends," forthciming, 1987. "Fred Friedman, Director of Operations Strategy, Blue Cross and six firms (13 percent) used it for team or Association, interview, May 29, 1988. work group appraisal. Fourteen firms (31 per 32 The Electronic Supervisor. New Technology, New Tensions

cent) used automatically collected data for Service observation, which is usually done planning work force requirements .° for jobs where employees have a great deal of telephone contact with customers, is often On the other hand, another survey of 110 combined with electronic measurements cf mganizations in 1982-84 found that the great productivity as well As mentioned above, serv- majority of firms (80 to 90 percent) collected ice observation is a standard practice in the individual performance statistics for at least telephone industry, and most of the Nation's some of their workers .° Most of the jobs af- 226,000 operators and service representatives fected were the clerical jobs, but some were are evaluated in this way. in addition, the great professional or technical. About one-quarter growth of telemarketing and telephone cus- of the firms collecting performance statistics -nraar service in the past decade means that said they did so only for assessing group per- an increasing number of employees are affected formance, to plan for peaks and valleys of work by service observation. A few firms inter- demand, and to cost-justify their use of the of- viewed by OTA reported that they had service- fice systems. The remaining Lhree- quarters observing capabilities in their telephone sys- were using individual operator statistics to tems but did not use them. make some sort of individual evaluation whether for base pay, incentive pay, promo- Westin estimates that the great majority of tion, or trainingfor some of their work force. clerical employees working on computer ter- In some cases, machine statistics were "almost minalsin the 65 to 80 percent rangeare not the entire basis" for such judgements, and in currently being monitored by computer and others it was "one factor in five or six factors evaluated for pay, promotion, or discipline on used to evaluate performance." that basis." Further, most professional, technical, and managerial workers-95 percent or A survey of the sac-9 110 organizations in moreare not currently evaluated based on 1985-86 revealed no increase in the percent- computer statistics. However, if 20 to 35 per- age of employers using computer measurement cent of clerical employees are being monitored, for personal evaluation. However, a majority this means that 4 to 6 million employees are of the firms now reported that they had cre- being evaluated in this manner. The addition ated a "more formal system" for setting work of professional, technical, and managerial standards and letting employees see the results workers could add another million or two to of monitoring if they wished.' the total, and this number could grow steadily larger over the next 5 to 10 years. To this Those organizations not collecting statistics total also should be added retail sales workers at all were usually either: and grocery clerks, whose speed and sales vol- 1. organizations using older word processors ume are sometimes monitored via electronic or microc3mputers that did not have soft- cash registers. ware for measurement; The clerical work force is predominantly fe- 2. organizations with new applications in male, and the low-skill end of the clerical work early implementation; force has a disproportionate number of minor- 3. non-profit organizations or universities ity women." Similarly, women are more who "just don't do that"; or likely to be employed in the lower levels of 4. State and local government agencies which professional work, such as routine computer "saw no need to compile those records. " programming or routine insurance underwrit- ing, rather than in higher levels of those profes- 'Hay Group, Inc., "Analy% of Customized Items for the Of- fice of Technology A mumps 1986 Hay Office Systems Sur- sions. Because monitoring is most likely to be vey, September 1986. 'Alan Westin, "Privacy and Quality of Work Life Issues in "Ibid. Employee Monitoring," contractor report prepared for OTA, "U.S. Congress, Office of Technology Assessment, Automa- 1086. tion of Americas Offices, OTA-CIT-287 (Washington, DC: U.S. "Ibid. Governmert Printing Office, December 1986), pp. 300-304. Ch. 2 Using Computers To Monitor Office Work 33

applied to precisely these lower level jobs, work The Department of Labor's Office of Work- monitoring is a topic that especially affects ers Compensation monitors the work of claims women and minorities. examiners and bill examiners For claims examiners the statistics include time elapsed Work Monitoring in the Federal from case creation to case adjudication, num- Government" ber of adjudications, number of wage loss claims processed, and elapsed time from receipt The work-monitoring practices in the Fed- of claim to decision. Bill examiners' statistics eral Government tend to follow some of the include number of bills paid per day, time same patterns as the private sector. There are elapsed from receipt of bill to payment. Claims no reliable statistics re how many Federal em- examiners are in the GS 5-11 range; bill ex- ployees have their work monitored by computer. aminers are typically GS 4-6.18 In general, Westin's survey, combined with The Internal Revenue Service, which has re- interviews by OTA, found the same sort of dis- cently implemented its Automated Collection tribution in the use of work monitoring in the Service, employs about 2,300 contact repre- government as in the private sector. Some sentatives who speak with delinquent tax- agencies, or specific work groups within agen- payers by telephone, negotiate payment sched- cies, used performance statistics only at the ules, and update taxpayer files. Performance aggregate level and only for planning or bud- data is collected by computer (time per trans- geting purposes. Some used monitoring sta- action, time logged on and available for work); tistics as part of individual evaluation of some in addition supervisors are required to listen workers, usually clerical workers. Some, espe- in on calls to monitor for courtesy and correct- cially small agencies and those with older ness of information. Employees know that equipment, did no monitoring at all. service observation is performed, but do not know specifically when they are being listened The following are a few specific examples of to. According to IRS sources, service obser- applications of work monitoring in the Federal vation is fairly infrequent and used primarily Government. For example, performance sta- for training of new employees. tistics are automatically collected for data tran- scribers at the Department of Agriculture's At the Social Security Administration's Tele- National Finance Center. Statistics include a service Centers, service observation capabil- total time on machine, keystrokes per hour, ity of the new telephone system has become and errors. Supervisors get daily, weekly, and a matter of dispute between SSA and the union monthly reports; operators get feedback representing 1,500 teleservice workers (Amer- monthly and some also maintain manual ican Federation of Government Employees). records of their own performance." Similar The union is attempting to negotiate specific performance criteria are used for data tran- time periods during which service observation scribers at the Bureau of the Census, but be- will be used. At present it can be applied at cause of the design of the computer system, anytime and employees do not know whether keyers must record the time manually. When they are being monitored. a new system is installed in 1988, all information will be collected automatically." Given similar levels of computerization in the Federal Government and the private sector," it seems likely that the number of mon- "This section is based on a survey performed by Westin in itored office workers in the Federal work force 1984 of 44 Federal agencies. See Alan Westin, "Privacy and is similar to the private sector-20 to 35 Quality of Work Life Issues in Employee Monitoring," contrac- percent. tor report prepared for OTA, 1986, pp. 3963. In addition, OTA staff interviewed a number of Federal agency managers and union officials. "Based on OTA interviews. "Based on OTA interviews. "Based on OTA interviews. "See OTA,Automation of America's Offices,op. cit. 4J 34 The Electronic Supervisor: New Technology, New Tensions

HOW TECHNOLOGY ASSISTS SELECTED BUSINESS APPLICATIONS The technology used in computerized work easing department. Itemized reports were sent measurement is not especially esoteric. Most to managers, who found that these reports people whose work is electronically measured were useful for other purposes, such as plan- are working on mainframe or minicomputers ning personnel schedules, justifying requests or are using telephone systems controlled by for new staff or equipment, and measuring the minicomputers. Stand-alone personal com- time it took individual employees to complete puters generally do not have the power to main- transactions." They are now being used as tain sophisticated work-measurement software work-monitoring tools in at least some of the while also carrying out the desired application. participating departments. [See later section on "Personal Computer Mon- itoring'1 Office systems that interlink a num- In most locatioaa, work-measurement tools ber of computers may sometimes have one were developed explicitly for the purpose, but computer dedicated to monitoring and meas- the idea is still the same. They collect and re- uring workflow. analyze information that is already being recordeu about computer utilization or busi- Some work-monitoring software systems are ness transactions. For example, travel agents available commercially, but in many cases, at some large agenciek: work on terminals con- firms with large data processing departments nected to a network that includes a nearby develop their own work-monitoring software minicomputer and a mainframe computer at in-house, or with the help of work-measurement some central location. The applications soft- consulting firms. Commercially available sys- gal e in the mainframe allows them to check tems in common use include those associated schedules and make or cancel reservations. The with key-to-disk data-entry systems, auto- mkicomputer, called in some organizations the matic call distributors, and "back office com- "back office computer," records details about puters" for travel agents, which will be dis- comput _r utilization. It notes who logged on cussed later. to whit terminal at which time, and it also makes records of the time, type, and amount Individual work-measurement statistics can of each transaction that the agents perform be developed using information that is already on the mainframe. Thus the back office com- being collected for some other purpose. Devel- puter provides a local audit trail and sales rec- oping work-measurement statistics from this ord for the whole office. This information can information is simply a matter of being alle be used in a numbs of ways, for example, it to retrieve some or all of this information, store allows the local printing of tickets and itiner- it in a separate file, and perform statistical aries for customers. It also can be used to de- analyses. velop individual performance histories, since Because it is based on analysis of eadev it has a complete record of all the computer available information, work monitoring ..laz activities of each agent." even begin "accidentally," without a specific In other firms interviewed by OTA, work management plan to introduce it. For exam- monitoring software was an integral part of ple, in one Midwestern bank, the data proc- the application software. That is, the same essing department installed a "black box" to computer software package that helps an in- record the time, type, duration, originating ter- surance claims examiner to key in client infor- minal, and user for every inquiry or transaction mation and calculate the amount of a payment, in on-line databases. The purpose was to monitor the speed of computer response time and "OTA interview with Senior Vice President for Operations, to make sure other departments were getting a midwestern bank, May 1986. "OTA interview with three travel agencymanagers, June the level of service promised by the data proc- 1986. Ch. 2Using Computers To Monitor Office Work 35 also automatically tallies the number of claims ysis of daily or monthly work volumes can help of each type that each examiner completes in managers better understand cycles in their a day. In several cases these tallies were then business so they can predict busy periods when transferred to another computer program,per- they must hire temporary workers or offer haps in a personal computer, which does sta- overtime. tistical analyses, compares performance to established standards, and prints reports for The example in figure 2 shows such a status supervisors. report. Service level ("sery level") is the percentage of incoming calls that were answered Collection a data about employee perform- within the specified time (typically 20 seconds). ance can in m^ny eases be made transparent Calls offered is the total number of calls, in- to the user, that is, information can be collected cluding those that were lost, delayed, or without interfering with the work that is be- diverted as shown in the following columns. ing done. From the viewpoint of theuser, an Positions manned ( "poi manned") means the automatic call distributor (ACD) simply routes number of agents jacked in and ready to work; incoming calls to individual telephones. Inac- the following columns indicate whether the tual fact, however, the distributor is also auto- agents are on incoming or outgoing calls. Aver- matically recording the type of call (insideor age delay is the average time in seconds for outside line), the time the call arrived, the iden- a call to be processed, whether answered, de- tity of the employee to whom it was routed, layed, or lost. The next columns show the num- the number of seconds before the employee ber of calls waiting (CW), the maximum num- picked up, the time the call started, the time ber of calls that were waiting at any time in the call ended, the number of times the caller this period, and the number of seconds that was put on hold and for how long, the exten- the current longest call waking has waited. sion to which the call is transferred, thenum- ber of seconds before that person pickedup, Here again, reports about individual per- and so on. In addition, it can show thesuper- formance are fairly simple to develop based on visor at any moment which operators are busy, information that must be collected in any case. which are waiting for work, which areon break. In order to route calls, the computer control- At the end of the day it can provide summaries ling an automatic call distributor must keep of individual and group activities. track of which telephones are busy, which are available, and which are unattended at any Properly organized, this can be very useful given time. However, this information must management information. For example, an be sorted and averaged in order to be of use. ACD can report th number of secondscus- A supervisor could make no sense of all the tomers were "delayed" before someone was detailed information that a computer collects available to help them or the number of cus- about each call. The work-monitoring software tomers that "abandoned" callshung up with- sorts, totals, averages, and summarizes the in- out speaking to anyone. If these figures get formation so that a supervisor can see activi- too high it may indicate the need for more tele- ties of the entire work group, or totals and aver- phone lines and more customer agents. Anal- ages of each individual's activities for a given

Figure 2.Example of a System Status Report for an Automatic Call Distribution System

SeryCallsCallsCallsCalls Pos. CallsCalls Avg. MaxOdly Gate leveloffrd.dlyd. Lost div. manned in out delayCWCW time

1 47 15 8 7 1 6 0 2 1 0 1 0 2 43 7 4 4 0 3 1 2 2 3 3 91 4 67 6 2 1 1 6 0 0 7 0 1 0

52 28 14 12 2 15 1 4 3 SOURCE: Adapted from Solid State Systems Inc ,The Smart Telephone System ACD Supervisor User's Guioe," Marietta Georgia, 1962. 36 The Electronic Supervisor. New Technology, New Tensions timed period, or exception reports, that show to come on the line. The time column shows particular calls or particular employees that how long (in minutes) the individual has been are far out of the average range. in the current status. The last column shows that three agents have calls waiting for them Figure 3 shows the type of information typi- on other lines. cally available on the VDT screen of the supervisor of a group of telephone customer service The example in figure 4 shows the sort of workers using an automatic call distribution information that might be included in produc- system. Status options here include vacant tivity reports summarizing periods of a few ("vcnt") or on break, talking ("talk") or avail- hours, or a day, or a week. This example reports able for calls ( "avlb "). "Work" means the agent the number of calls, the total handle time is doing other work related to a previous call, ("hndl") in minutes (the sum of "talk time" perhaps updating the database or preparing and "work time"), and the total time available a letter to the customer, and is not accepting and waiting for calls, in minutes. In addition calls. The agent named Joe is talking and has the report shows the average handle time, talk signaled for "help," requesting the supervisor time, and work time per call in seconds.

Figure 3.Example of a Supervisor's Display in an Automate Call Distribution System

AGENT AGENT STATUS TIME 10:32 5110187 SUPERVISOR GATE NO. NAME SUPVS 1

1 513 SAM VCNT 6 3 CW = 4 1 512 JANE TALK 0 3 1 514 JOE HELP 1.2 1 560 BILL WORK 4.7 2 570 FRED AVLB 6.3 2 510 SUE TALK 12 2 513 SAM VCNT 6.3 3 512 JANE TALK 0.3 CW = 1

SOURCE Adapted from Solid State Systems Inc The Smart Telephone System ACD Supervisor User's Guide,' Marietta Georgia 1982

Figure 4. Example of Agent Statistics From an Automatic Call Distribution System

14:32 10114182 AGENT STATISTICS AGENT NO. HNDLTALKWORKAVLB AVG AVG GROUP 4 GATECALLS TIME TIME TIME TIMEHNDLTALKWORK

GEORGE 454 1 0 00 0.0 0.0 0.0 HARRY 455 1 0 0.0 0.0 0.0 0.0 JERRY 582 1 6 84 8.4 0.0 6.9 84 84 0 SUPV1 580 1 5 3.5 3.5 0.0 5.7 53 53 0 SUSANNAH501 1 5 10.0 10.0 00 5.3 120 120 0 BILL 503 1 4 3.6 3.6 0.0 3.2 54 54 0

SOURCE Adapted from Solid State Systems Inc The Smart Telephone System ACD Supervisor user's Guide." Marietta Georgia, 1982

WORKPLACE ISSUES RELATED TO WORK MONITORING People in some organizations perceive work ment at all levels to put out a good product. measurement and service observation to be At other organizations, use of the same tech- useful tools that help employees and manage- nologies is resented and feared as "Big Brother

46 Figure 5.Two Models for Electronic Work Monitoring

UNION Individual All data Known pro- Employee can Group produc- lQuality/Problem Standard pay Individual short- MODEL performance available to cedure for pace work tion quotas factors fall leads to sampling employee challenging recognized training or group "record" discussion

"TAYLOR" Constant Data not No procedure Machine paces Individual Quantitylspeed Piecework pay Individual short- PRODUCTION machine readily for challenge viork quotas standard or bonus fall leads to MODEL monitoring available discipline

PRIVACY-RELATED ISSUES' LABOR RELATIONS/EMPLOYEE RELATIONS ISSUES "Intrusiveness" and "Subject Access" "Fair Work," "Fair Pay," "Fair Performance Evaluation"

SOURCE Alan F WestinPrivacy and Quality of Life Issues in Employee Monitoringcontractor report for OTA 19%

46 47 98 The Electronic Supervisor. New Technology, New Tensions

Table 3.Workplace Issues Related to Work Monitoring

Privacy and Labor relationsHealth/quality access related or "fairness" of life Is monitoring constant or Intermittent? X Can employees see their own records? X Can the employee challenge, explain, or correct records? X Does the employee or the machine pace the work? Do employees understand performance criteria and use of information? Are quotas set on an Individual or group basis? Are quotas fair, allowing work at a reasonable pace? Is pay standard or based on performance? What happens to employees falling short of quota? SOURCE: Office of Technology Aeseeement. surveillance." The difference seems not to be chose to call the first model the "union" model, so much in the specific measurement technol- since it represents a blend of features included ogy, but in the politics of. how it is introduced, in model contract language suggested by some how it is used, and what is done with the in- U.S. and international unions. The second he formation collected. called the "Taylor production" model; it is Although many workers' complaints about based on an extreme form of an industrial engi- monitoring focus on its intrusiveness, a closer neering approach to work measurement, one look shows that privacy is only one of a com- which places virtually all information and plex of issues raised by electronic work moni- power in the hands of management. In apply- toring. Table 3 outlines some possible charac- ing these models to the organizations inter- teristics of a work-monitoring program and viewed for OTA, no pure examples of the "union indicates the kinds of issues that are raised model" were observed. The "Taylor produc- by them. Privacy and access issues cover such tion model" in its pure form was observed in questions as whether employees know they are action in a few government and private sector being monitored and whether the employees organizations. Most organizations used meth- have access to records about their own per- ods representing a blend of the features of the formance. The second set of issues relates to two models, with about two-thirds of the orga- the perceived fairness of the monitoring sys- nizations interviewed by Westin tending toward tem and the way the employer uses it in evalu- a modified version of the "Taylor" model. ating and rewarding employees; these are ques- In most cases, employers introduce electronic tions of employee relations. The final set of monitoring unilaterally, only informing em- issues, overlapping the other two, relate to ployees of the change after all decisions have stress, health, and the quality of working life. been made. Often, too, monitoring is only one Whether the effect of monitoring is perceived of a number of changes in work process or job as intrusive, unfair, dehumanizing, or un- design that take place when new office equip- healthy often depends on how management ment is purchased. As was discussed in detail structures the work-monitoring program, what in OTA's report, Automation of America's Of- it does with the late it collects, and how those fices, employee participation in design end im- actions are perceived by employees. plementation is often a key to successful implementation of new office systems. What Is Fair Ensuring employee participation can require effort on the part of managers, as few U.S. Westin used some of the same elements dis- workplaces have mechanisms for employee in- cussed in table 3 to construct two models of put in areas of technological change or evalu- work monitoring shown in figure 5.9' Westin ation procedures. Nevertheless, Westin found that the difference between employees pretest- "This section draws heavily on Alan Westin, "Privacy and Quality of Work Lib Issues in Employee Monitoring, " contrac- ing over "Big Brother surveillance" and em- tor report proposed for OTA, pp. 10312 (draft). ployees perceiving work measurement as rea-

tiO Ch. 2Using Computers To Monitor Office Work 39

sonable, often depends on whether they agree Table 4.Key Issues and Problem Areas In on: 1) the fairness of the standards set; 2) the Monitoring Worker Performance fairness of the monitoring process employed; Key issues/problem aspects and 3) the fairness of the way measurements Fairness of work standmds: are used in employee evaluation. This agree- Do standards fairly reflect the average capacities of the ment was usually made through worker-man- particular work force? Will they create unhealthy stress for many employees? agement discussions before monitoring was Do they take into account recurring system difficulties and implemented. Such agreements are possible other workplace problems? where there is genuine involvement of employ- Do they Include quality as well as quantity goals? Do they represent a "fair day's pay" for a "fair day's work?" eeseither through joint labor-management Do employees share in any productivity gains achieved committees in unionized organizations, or through introduction of new technology? through employee involvement techniques in Fairness of the measurement proms: natation settings. Where management started Do employees know and understand how the measurements are being done? with the trusting assumption that almost all Can the measurement system be defeated easily, thereby employees were ready to put in a fair day's Impairing the morale of those willing to "follow the rules?" work for a ink day's pay, and where topics such Do employees receive the statistics on their performance as work standards, work measurement, and directly, and in time to help them manage their work rate? is the relation between quality, service measures, and work productivity recognition were matters for open quantity communicated by supervisors when they discuss discussion, introduction of monitoring was problems of performance- levels with employees? Do supervisors communicate clearly that they are taking usually relatively painless. system and workplace problems Into account? Are group rather than individual rates used when particu- One impression that emerged from OTA's lar tasks make such an approach more equitable? interviews is that the way managers and em- is there a formal complaint process by which an operator can contest the way work data has been used by the su- ployees deal with monitoring often closely ps:visor? parallels the way they deal with other work- Fairness in applyIng measurements to employee evaluation: place issues. Firms whose "corporate culture" Are there meaningful recognition programs for these em- tends toward authoritarianism tend to use ployees? Is work quantity only one of a well-rounded and objective monitoring in an authoritarian way. In orga- set of performance criteria used for employee appraisals? nizations where relations between employees Does the employee get to see and participate in the per- and managers are antagonistic, the monitor- fonnance appraisal? ing system is a source of antagonism, but only is there an appeal process from the supervisor's performance appraisal? one of many. In organizations where coopera- is there a performance - planning system that identifies em- tion is the norm, people worked together to de- ployee weaknesses in performance and identifies ways to velop a fair system. remedy such problems? SOURCE: Alan R. Westin, "Privacy and Quality of Life issues In Employee Monitoring," contractor report for OTA, 19110. Recognizing that employee involvement in the design, testing, implementation, andcon- tinuing adjustment of work monitoring iscru- cial to a successful process, it is alsonecessary ation policies" along the lines of a positive an- to deal with the substantive issues, to becon- swer to most of the questions in table 4. sidered in designing such a program. Table 4 shows some of the issues to be considered. The Of the 34 case examples submitted to OTA main categories, and the specific questions in by unions, 28 dealt with electronic monitor- this chart represent recurring themes ina num- ing of office workers (unionized and nonunion- ber of interviews with monitored workers and ized) like the ones studied by Westin. In nearly their managers. all the case examples, employees had little input concerning the monitoring system, and in Westin's sample found that only aboutone- only a few cases was it clew. that they hadac- third of the firms in his sample using electronic cess to information about their own perform- work monitoring for individual evaluations ance or the ability to contest wrong informa- were following what he called "fair work evalu- tion. In nearly all these cases the workers were 40 The Electronic Supervisor. New Technology, New Tensions described as considering the monitoring sys- Sometimes "standards" are really goals or tem unfair." ideals. In one firm interviewed by OTA the standard of 50 completed transactions per Standards of Performance agent per week had recently been established by the national office. The standard was de- Work measurement systems are usually ap- veloped by dividing the average revenue per plied to jobs for which standard end products, transaction into the total revenue the firm or surrogates for end products, can be clearly hoped to generate at each branch office." The identified. That end product might be "cus- new standard was not related to past perform- tomers served," "claims paid," "programs ance levels or analysis of the best way to do written," "interviews completed." Generally the job, but ratherthe amount that must speaking, electronic work monitoring primar- be sold in order to meet revenue projections. ily measures the quantity of work performed. In this cab% 50 transactions per week was far Other methods, discussed in the next section, above past performance; office managers measure quality. hoped that introducing incentive programs An important element in measurement of would inspire agents to achieve the new goals. almost all kinds of work is time. In almost A more formal method of standard setting every case in table 2, the purpose of measure- is the "engineered standard." The Methods- ment is to measure the time it takes the em- Time Measurement (MTM) system or the Ad- ployee to do something, and then to compare vanced Office Controls (AOC) system have the result to a standard. Robert Nolan, an ex- been used in many office settings. In these pert in developing work-measurement systems, methods, a trained analyst, usually an indus- defines a work-measurement system in this trial engineer, observes a work task, selects way: the most efficient method of performing the In the most simple terms, it is a means of task, and then will time the actions of aver age establishing what a fair day's work should be. people performing the task under average it has two main components, a measure of the working conditions. General MTM and AOC volume of work, avid a measure of the employee standards have been developed for nearly every time used up. These two factors can be ex- imaginable motion in an office workplace. For pressed in their only common denominator: example, the MTM standard for fastening the time required to produce one unit of work, sheets of paper with a table model stapler is or what we call a standard." 41 time measurement units (TMU), or about Thus, measurement alone is often of little use 2.9 seconds. Opening an envelope and remov- as a management tool, unless its purpose is ing the contents takes 198 TMU or 14.2 ser to compare the individual or group perform- onds. A trained analyst can combine a num- ance to a standard. ber of these general standards, develop new ones, and adapt them to the special circum- Standards may be established in a number stances of spacial arrangement, work process, of ways. Many are arrived at rather informally or equipment use in a given office." or arbitrarily, perhaps based on supervisors' or managers' estimates of how long it ought A well-designed standard, according to ex- to take to complete certain tasks. In some cases perts, is not one that makes people work as standards are set based on historical perform- fast as possible, but one that encourages good ance levels; managers may take an average of average work. It should include time for per- some past period, and expect that it be main- sonal breaks and allow for personal variabil- tained as an average in the future. 230TA interviews, assistant manager, Washington office, na- si"AFIrCIO Case Examples," November 1987. tional travel firm, June 1988. "Robert E. Nolan, "Work Measurement," in Robert N. Le- "Examples from Robert E. Nolan, "Work Measurement," barer,White Collar Productivity(New York, NY: McGraw Hill, in Robert N. Leherer,White Collar Productivity(New York, 1988), p. 111. NY: McGraw Hill, 1983), pp. 142-148. 5i Ch. 2Using Computers To Monitor Office Work 41

ityworking a little faster at some times and ard. The determination of an "acceptable"pace slower at others. Fair standards must be real- depends on the firm, but a well-designed stand- istic, taking into account system downtime, ard is one where most trained, experiencedem- slow response time, varying levels of complex- ployees will work in the range of 85 to 100per- ity of different tasks, and so on. When stand- cent of standard most of the time.26 ards are not realistic, or when they are notper- ceived as fair by employees andmanagers, they Figure 6 shows part of a weekly work moni- can easily lead to declines in morale, increased toring report for an insurance employee work- turnover rate, and ultimately a decrease in pro- ing under a "predetermined time" work-meas- ductivity. urement system. The report lists the types of tasks done, the standard time to do the task Since AOC, MTM, and other "predetermined once, the number of times the employee actu- time" systems provide a standard time for the ally completed the tasks, and a calculation of completion of each task, an employee's actual the "earned" hours. Figure 7 summarizes work performance can be compared to that stand- of a group of insurance employees. Figure 8 ard. If the standard time for examininga cer- integrates the work monitoring system with tain type of insurance claim is 10 minutes, then a time and attendance report. It shows the an employee who completes 6 of them will have number of hours each employee was available done 60 standard minutes or 1 standard hour's for work ("avail work"). The time available for worth of work An employee who completed measured work ("avail meal " employeesmay 48 such cases in an 8-hour day would be said have other duties that are not captured by the to be working at 100 percent of the standard, system) and the number of earned hours ("earn that is, his or her paid hours would exactly hrs") worth of work completed. Note that there equal standard hours. Faster employees might work at 110 or 120 percent of standard, while "See Robert E. Nolan, "Work Measurement," in Robert N. Leherer,WhiteCollar Productivity(New York, NY: McGraw slower ones work at 80 or 90 percent of stand- Hill, 1983), P. 121.

Figure 6. Example of Individual Work Monitoring Report (PerformanceSummary)

RUN NUMBER - ARM1B140 WORK MANAGEMENT PROGRAM DAT' OF RUN - 10/09/84 INDIVIDUAL PERFORMANCE REPORT TIM OF RUN - 18:00 WEEK ENDING 10/6/84

E" PLOYEE: EMPLOYEE D DIVISION' G CLAIMS AND SERVICE DEPARTMENT:GBCLMS PROC

CALCULATIONOF EARNED HOURS

KVI DESCRIPTION UKVI AVERAGE TIME ITEMS EARNED DV DP CODE NAME FLAG TO DO ONE ITEMx COMPLETED= HOURS

G GB 0040 EOB STUFF 00 MIN 16 SEC 148 0.6 G GB 0050 SUBSCRIBER PREP 01 MIN 48 SEC 2 0.0 G GB 0070 SUBSCRIBER CODING 03 MIN 54 SEC 2 0.1 G GB 0080 PROVIDER CODING 01 MIN 03 SEC 8 0.1 G GB 0140 MEP ADJUDICATION 02 MIN 47 SEC 482 22.4 G GB 0150 VISION ADJUDICATION 02 MIN 33 SEC 68 2.9 G GB 0175 PRIMARY GSCR ROSTER 00 MIN 10 SEC 621 1 8 G GB 0176 GSCR CHK-ADJUDICATOR 00 MIN 08 SEC 454 1.0 G GB 0190 COB PROCESSING 04 MIN 57 SEC 20 1.6 G GB 0210 COB RETURN LETTERS 00 MIN 49 SEC 50 0.6 G GB 0185 PULLING CLAIMS 03 MIN 00 SEC 8 0.4 TOTAL EARNED 32.0

SOURCE Adapted from James S Hogg, Manager, Professional Productivity, Blue Cross & Slue Snistd of Maryland 5 63-982 0 - 87 - QL: 3 -- 2 42 The Electronic Supervisor. New Technology, Now Tensions

Figure 7.Example of a Work Monitoring Report for a Group (Volume of Work Accomplished by Group in One Week)

RUN NUMBER - ARM1B130 WORK MANAGEMENT PROGRAM DATE OF RUN - 10/09/84 VOLUME SUMMARY FOR WEEK ENDING 10/06/84 TIME OF RUN - 18:13

DIVISION: G CLAIMS AND SERVICE DEPARTMENT: GB CLMS PROC UNIT: XXX SECTION: YYY

KVI KVI VOLUME PERFORMANCE EARNED CODE DESCRIPTION REFERENCE HOURS

0010 MAIL SORT 1938 0.0068 13 2 0020 PROCESS INCOMPLETES 219 0.0085 1 9 0040 EOB STUFF 2105 0.0046 9 7 0050 SUBSCRIBER PREP 216 0.0280 6 0 0051 PROVIDER PREP 425 0.0045 19 0060 MEDICARE CLAIMS PROC 391 0.0385 15.1 0070 SUBSCRIBER CODING 457 0 0657 30.0 0080 PROVIDER CODING 1496 0.0179 26 8 0090 SUBSCBR DATA ENTRY 293 0.0224 6 6 0100 PROVIDER DATA ENTRY 1072 0 0185 19.8 0110 VISION PREP & CODE 164 0 0246 4 0 0120 VISION DATA ENTRY 168 0 0236 4.0 0130 PAF REPORT UPDATE 232 0 0100 2 3 0140 MEP ADJUDICATION 846 0.0466 39 4 0150 VISION ADJUDICATION 70 0.0426 3 0 0160 PHONE INQUIRY 382 0.1015 38 8 0175 PRIMARY GSCR ROSTER 1740 0.0030 5 2 0176 GSCR CHK-ADJUDICATOR 926 0.0023 21 0180 ADJUSTMENT CODING 15 0.0927 1 4 0181 PROC. CODED ADJUSTMTS 14 0 0454 0 6 0182 HND.PROC.ADJ/NO FILE 14 0 0102 0 1 0183 ART COMP-HANDLG REFD 15 0.1284 1 9 0185 PULLING CLAIMS 8 0.0496 0 4 0190 COB PROCESSING 36 0.0823 3 0 0195 WRITTEN CORRESPONDENCE 21 0 3801 80 0200 PROC OCC DELETES 43 0 1054 4 5 0210 COB RETURN LETTERS 55 0 0135 0 7

SOURCE Adapted from James S Hogg Manager Professional Productooty Blue Cross & Blue Shield of Maryland is a wide variation in effectiveness ("eff % "), and is to be fair. Changes in the work might which calculates the earned hours as a percent- arise when a new product is introduced (say, age of measured hours. This particular firm in an insurance company, a new kind of pol- does not use an incentive system so all employ- icy), or when regulations change (e.g., requir- ees are paid for their regular hours despite ing a change in the kind of information banks these differences. The report also shows totals must supply to customers), or when the tech- for the whole group. nology changes. Work-management specialists at firms using predetermined-time systems The work processthe set of procedures that note that "maintenance" is a major need if govern what tasks are done and how tasks work measurement is to be applied conscien- interrelatealso has a major impact on many tiously. Work must be periodically re-analyzed predetermined-time work-measurement sys- and standards must be adjusted. tems. Because the "standard" for each task or set of tasks depends on a close study of the Standard setting is often combined with job work performed, any changes in the work redesign, work simplification, or procedural quire a change in the standard, if the stand- changes, because it is difficult to establish a

53 Figure 8.Example of Individual Work Monitoring Raport (Hours Summary)

WORK MANAGEMENT PROGRAM RUN NUMBER - ARM1B140 HOURS REPORT DATE OF RUN - 10/09/84 TIME OF RUN - 18.08 WEEK ENDING 10/06/84 DIVISION: G CLAIMS AND SERVICE DEPARTMENT:GB CLMS PROC

EMPLOYEE TIME SPENT AT WORK TIME NOT MEASURED

REG OVER TIME NOT AVAIL A 'AIL EARN EFF NUM NAME ST TIME + TIME WORKED =WORK PROD RFW UKVI OTH = IvMEAS HRS 0/0

1518 EMPLOYEE A 03 37 5 0 0 23 2 14 30 0 0 0 0 0 0 0 0 14 30 11 6 81 4529 EMPLOYEE B 03 375 00 02 37 30 00 00 00 49 32 40 26 4 81 4669 EMPLOYEE C 03 37 5 0 0 0 2 37 30 0 0 0 0 0 0 0 4 36 90 18 1 49 5206 EMPLOYEE D 03 375 00 02 37 30 00 00 00 19 35 40 32 0 90 5210 EMPLOYEE E 03 00 00 00 0 00 00 00 00 00 000 00 0 5244 EMPLOYEE F 03 375 00 02 37 30 00 00 00 10 :5630 15 1 41 5245 EMPLOYEE G 03 375 00 02 37 40 00 00 00 24 35 00 23 4 66 524E EMPLOYEE H (i3 37.5 0 0 0 2 37 30 C 0 0 0 0 0 5 7 31 60 22 0 69 524/ EMPLOYEE I 03 375 ;) 0 02 37 30 00 00 00 00 37 30 24 7 66 5255 EMPLOYEE J 03 37 5 0 u 0 2 37 30 0 0 0 0 0 0 17 7 19 60 29 1 148 5256 EMPLOYEE K 03 37 5 0 0 0 9 36 60 0 C, 0 0 0 0 0 8 35 80 35 0 97 5257 EMPLOYEE L 0) 00 00 00 000 00 00 00 00 000 00 0 5258 EMPLOYEE M 03 37 5 00 36 9 0 60 00 00 00 05 0 00 00 0 5435 EMPLOYEE N 03 37 5 0 0 1' 2 22 30 0 0 0 0 0 0 1 8 20 50 12 9 62

SECTION TOTALS 450 0 0 0 77 7 372 30 0 0 0 0 0 0 37 2 335 10 250 3 74

COST CENTER TOTALS 450 0 0 0 77 7 372 30 0 0 0 0 0 0 37 2 335 10 250 3 74

PERSONNEL EMPLOYED 14

EQUIVALENT PERSONNEL AVAILABLE FOR WORK 99

SOURCE Adapted from James S Hogg Manag.r Profess.onal Productivty Blue Cross & Blue Shield of Maryland

54 44 The Electronic Supervisor. New Technology, New Tensions

set time for performing a task if everyone is Within many organizations, introduction of free to do it in a different way. Some waysare work measurement and the process of setting better than others, whether faster, less fatig- standards can become a hotly contested labor- uing, or more reliable. Experts in productivity management issue and a major source of em- and work measurement usually counsel that ployee discontent. Where employeesare not employees should be included in the process involved in standard setting, theymay view of changing procedures and establishing stand- a new standard as an unfair "speed up," an ards. TThhey argue that employee involvement attempt by management to make them work not only short-circuits resentment to stand- harder for the same pay. Similarly, work sim- ards that are imposed from outside, but usu- plification or procedural changes that are im- ally leads to the creation of better procedures posed from outside can be viewed as remov- and fairer standards, since employees under- ing variety and autonomy from the job, and stand the work that needs to be done better making it less interesting andmore me- than anyone, including their managers. chanical. An example of employee involvement in standards development is the case of legal case For example, the changes in work standards, analysts at an insurance firm." These highly evaluation, and pay that accompanied work skilled workers handle correspondence and track monitoring for claims examiners prompted a the progress of legal cases on a computer-based unionization drive at Equitable Life Assu. legal diary system (LDS). Although the LDS ante. With the introduction of the measure- software created some internal statistics on ment system, pay was changed from a straight the transactions done, they had not been pre- salary to an incentive program that was based viously used for individual evaluation because on performance. Examiners complained that no standards had been developed for the case they had to work much faster in order to make analyst's work. The work-measurement spe- the equivalent of their old salaries. A fewac- cialist assigned to develop standards found the cepted transfers to lower paying jobs because work very complex and also saw that the le- they could not keep up the pace. The contract gal diary software was very flexible, allowing between Equitable and District 925 of The analysts to use several different procedures for Service Employees International Union certain tasks. Instead of trying to prescribe (SE IU), addresses some of the issues discussed procedures and standards, the work-measure in the section on "What Is Fair." Under the ment specialist held a series of 24 tv, o-hour contract, employees now have access to their seminars in which the analysts talked about own performance records and a procedure for their work. They discussed different tasks, challenging records. Evaluations are based 80 compared their approaches, and decided among percent on computer-based statistics and 20 themselves the simplest and most effective percent on supervisors' judgments. In addi- procedures for each task. They also helped to tion, the contract changed several other work- set the standard times for the tasks. Interest- ing conditions, such as leave policy, that had ingly, the productivity of this department, in been a subject of dispute. terms of dollars recovered through legal actions, began to increase before the final work When electronic monitoring allowsa com- measurement program was in place, presum- plete record of each worker's performance, it ably because the case analysts voluntarily be- becomes easier to pay workers basedon their gan using the improved procedures as soon as output. Some call this a revival of the "piece- they were developed in the sem;!...rs. rate" system and decry it as a form of worker exploitation. Often, however, performance does not translate into pay directly on a per-piece "James Hogg, "The Results of Technical and Professional Measurement in Insurance," in Proceedings, 1988 AOC Users basis. For example, "incentive" planspay a Conference, May 15-18, 1988, Robert E. Nolan Co.; also,per- base rate for acceptable performance and bo- sonal communication, May 29, 1988. nuses for higher levels of performance.

56 Ch. 2Using Computers To Monitor Office Work 45

Incentive programs appear to be fairly com- year based on their previous year's perform- mon in data entry, where there is a wide range ance. Since the number of available jobs had of performance." Operators who arevery fast declined, only above-average keyerswere in- can increase their income by 50 percent or more vited back. However, performance standards above the base rate, depending on how the in- were also raised yearly, presumably to deal centive plan is structured. In some firms where with the workload. Thus, each keyerwas re- standards are based on a predetermined-time quired to make an increasingly greater effort method (i.e., so many keystrokes equalone to remain "above average."" An annual in- standard hour of work), slower keyers can make crease in standards has also been a cause of bonuses if they are willing to work for a longer complaint among key entry operators in Dade time. This also raises fairness questions and County, Florida." worker protection questions, for example, should employees feel pressure to skip lunch or breaks Complaints about job stress in the U.S. in order to improve their performance? Postal Service, which receiveda great deal of publicity in 1984 and 1985,were directed pri- Incentive programs have also been used by marily at fast pace and high work standards employers to increase the performance of a rather than at automated equipmentor the group of employees. In one bank, for example, presence of monitoring per se." Industrial industrial engineers studied the work of check- engineers have noted since the beginning of proof reading operators and found that the the century that there isa limit to how much engineered standard was far above the current a pace can be increased, even if incentives are level of achievement of the department. Rather offered. Beyond a certain point the employees, than insist that operators begin to work to the either individually or asa group, will not per- new standard, management began to pay regu- form any faster on a regular basis,no matter lar wages to those who met the old average what the inducement. Tolerance for perceived and bonuses to all whose work approached the unfair standards may dependon many factors, new standard. As departmental proficiency in- including the availability of other jobs. One creases, the management expects to raise the case example noted that at three Internal Rev- standard and adjust the bonus structure toen- enue Service Centers (where standards have courage even faster performance." been increased over the years), the turnover The practice of "rate busting" or increasing rates for key entry operatorsare very high; performance standards over time is the basis presumably workers left due to the heavy for many objections to monitoring. Ever-increas- workload and fast pace required. However, the ing standards do not have to be related to in- Wilkes Barre Service Center, located inan area centive pay. Standards can rise due tonew of high unemployment, hasa low turnover rate, technology, revised productivity goals, or for but a high incidence of absenteeism." other reasons that lead management to expect better performance from employees. In the Quality of Work well-publicized case of one data-entry center One problem with computerized work operated by the Internal Revenue Service,em- mon- ployees and their union were complaining about itoring is that it focuses mainlyon quantity the stress resulting from the increasedpace or speed of work. Although a well-designed of work. In this instance, workerswere sea- work standard should allow workers time to sonal, and were invited back to work again each do a good job, some standards require such "Alan Westin, "Privacy and Quality of Work Life Issuesin "Data Entry Management Association, "Sixth Annual Employee Monitoring," contractor report prepared for OTA, Member Statistical-Compensation Survey,"DEMA Newslet- 1986; also John Harris, American Federation of Government ter,April 1985. The average rate for U.S. operators is 11,400 Employees, personal communication, February 1986. keystrokes per hour, but the fastest operators can do around ""AFL-CIO Case Examples," November 1987. 25,000. "For example, Peter Pert, "Monitoring by ComputerSparks "Interview with Work Measurement Manager of a southern Employee Concerns,"Washin,Von Post,Sept. 2, 1984. bank. ""AFL-CIO Case Examples," November 1987. 48 The Electronic Supervisor: New Technology, New Tensions a fast pace that workers feel quality must be ing place on an employee's screen. Thus su- sacrificed, or that the pressure to maintain pervisors can view transactions as they are both speed and quality leads to excessive taking place to check them for correctness. stress. In a number of cases, for example, tele- Computerized letter-sorting equipment used phone operators have objected that the pres- by the U.S. Postal Service has similar capa- sure to complete calls within the standard time bility, so supervisors can periodically check prevents them from giving courteous, high- each worker to be sure he or she is keying in quality service. Some customers agree. On the proper zip codes. other hand, because most operators are also For telephone service workers, quality subject to service observation (i.e., a supervi- checks are made by supervisors who listen in sor sometimes listens in on calls to check for on calls to check that employees are courte- adherence to company procedures), they some- ous, are using proper procedures, and are giv- times feel stressed because of the conflict being correct information. Correct information tween quantity and quality imperatives." is of interest to many firms whose representa- In one mail order firm employee morale tives deal with the public, because employers dropped and turnover rose to 80 percent after may be held liable for information their em- monitoring was introduced for VDT operators. ployees give out over the telephone or for ac- "Everything was numbers," one executive tions taken as a result of telephone conversa- recalled, with "no attention to the downtime tions. In some cities "911" emergency calls or and slow-response-time problems of the new utility company "trouble" calls are recorded system, or the changes in volumes operators so that there will be a record of time, address, faced during peak periods, or of the different or other information for possible future use. length and complexity of customer orders." In some o.-ganizations employees know when In addition, the pressure operators felt to speed their supervisor is listening, either because up their work led to mistakes and improperly there is a drop in volume or because a beep tone filled orders. This productivity system was is heard. In other cases, the monitoring equip- scrapped after several years of operation, and ment is completely silent. One organization replaced with a new approach that still collects was so concerned that employees not know individual operator statistics, but has stand- when the supervisor was listening that super- ards geared to actual system operations and visors were required to wear their headsets all load cycles. In addition, as part of an overall day so that employees would not be able to "Quality First" campaign in this firm, the new guess whether they were listening or attend- performance standards stress "order quality" ing to other duties. over "sheer numbers." Several dozen long-term employees interviewed for OTA said that the In some firms quality assurance is consid- first productivity system was a "very bad ered such an important function that a sepa- time" at the company, but that the new ap- rate department handles it. At American Ex- proach is "fair to both company and employ- press, for example, customer service supervisors ees," and they have no trouble in meeting both listen in on calls on an regular basis and rate the quantity and quality standards." the quality based on established criteria. In addition, a separate quality assurance worker Quality evaluation often requires inspection listens in on calls of employees in any unit. Re- by a human supervisor, but even here computer sults are always discussed with the employee technology can be of assistance. Some office within a short time after the call. systems allow the supervisor to view on his or her screen whatever transactions aretak- Some workers object to service observation precisely because it is not necessarily objec- ""AFL-CIO Cue Examples," November 1987. tive. Some firms, in fact, do not have firmly "Alan Westin, "Privacy and Quality of Work Life Issues in established criteria for how often to listen or Employee Monitoring," contractor report prepared for OTA, 1988, p. 72. how to rate quality. At one government agency

58 Ch. 2Using Computers To Monitor Office Work 47 there were even stories of service observation signed. While many firms use computers to being used punitivelyi.e., a supervisor lis- continue or intensify the assembly line work- tened to certain workers almost constantly, in ing conditions of the industrial style of work order to accumulate enough mistakes to dis- organization, some others have experimented cipline them. This would clearly seem to be an with new forms of organization that reinvest abuse of service observation. the jobs of individual workers or teams with Service observation also invokes feelings of more variety and responsibility." invasion of privacy, even though the conver- Some of these experimental organizations sation involved is not really a private one. One make use of what is called a semi-autonomous operator interviewed for OTA said, "When work groupa team of workers who are re- they are listening to me, I'm very upset be- sponsible for not only doing the work, but man- cause you can't stop it."" The privacy aspect aging some aspects of their own work as well. applies more clearly to the customer's side of In these cases, work monitoring data may still the conversation. Some people may object to be collected, but is used by the work group as third parties overhearing their conversations. a tool for assessing its own progress. Two States, West Virginia and California, attempted to legislate restrictions on service ob- One well-publicized example is the HOBIS servation. The West Virginia law required a (Hotel Billing Information System) office at beep tone when the supervisor is on the line Tempe, Arizona, in an experiment worked out as well as a published notice to customers that jointly by AT&T and the Communications calls might be observed. This law was passed Workers of America (CWA) in 1982." This in 1983 but repealed in 1986. A similar law was office of 100 operators was rsaganized accord- passed in the California legislature but was ve- ;ng to the autonomous work group principle. toed by the Governor. It had no first-line supervisors and only one second-line supervisor in the role of advisor. In the case of West Virginia there is evidence Operators assumed the responsibility of super- that both operator productivity and customer visors, rotating through administrative duties. satisfaction remained high during the period when "secret" service observation was not per- The employees changed the traditional work mitted." However, several employers, par- monitoring practices. They eliminated individ- ticularly AT&T objected to the legislation. ual measurement and remote secret service AT&T's threat to build its new credit manage- observation. Average work time (AWT) was ment center in another State was instrumen- measured only for the whole group. Service ob- tal in the repeal of the West Virginia monitor- servation was performed by small groups of ing law." peers by the old-fashioned "jack-in" method, where the observer sits beside the person being monitored, listens to a few calls and then Job Design and Work Process discusses the results with the employee. As discussed in greater detail in OTA 's re- It was generally agreed by CWA and AT&T port Automation of America's Offices, new in- that the Tempe experiment was a success: to- formation technologies sometimes offer firms tal office AWT was equal to or better than that more flexibility in the way office work is de- of traditionally supervised HOBIS offices; M-4-7chael J. Smith, Pascale Carayon, and Kathleen Miegio, S. Congress, Office of Technology Assessment,Automa- "Motivational, Behavioral, and Psychological Implications of tion of America's Offices,OTA -CI T -287 (Washington, DC: U.S. Electronic Monitoring of Worker Performance," contractre- Government Printing Office, December 1985),ch. 4. port prepared for OTA, July 1986. "This description is based on Ronnie Straw and Gregory ""Resulta Summary, Key Service Indicators," C&P Tele- Nichlas, "Office Automation and Autonomy A Comparison of phone Co. of West Virginia as of Sept. 11, 1986, supplied to Choices," Communications Workers of America, 1985; Alan OTA by Communications Workers' of America. Westin, "Privacy and Quality of Work Life Issues in Employee "Teetimony of John D. Landers, AT&T, before the Judici- Monitoring," contractor report prepared for OTA, 1986; Thomas ary Committee, West Virginia House of Representatives, Feb. Taylor, Mountain Bell Telephone, telephone interview, Apr. 16, 12, 1986. 1986. 53 48 The Electronic Supervisor: New Technology, New Tensions there were fewer customer complaints; the em- One researcher who studied data-entry oper- ployee grievance rate was lower and absentee- ators, claims processors, and data collectors ism was lower. In addition, there were consid- (telephone interviewers and collection agents) erable savings in management salaries, and and their supervisors, found no significant pat- some of the money was spent on training for tern of differences between the monitored and employees. The 'Tempe office was closed nonmonitored sites.' The differences she did shortly after the AT&T divestiture for reasons notice were between unionized and non-union- completely unrelated to the experiment. Other ized locations. The workers in unionized loca- joint labor-management experiments in alter- tions were better informed about VDT health native methods of work organization are be- issues and more willing to ask questions and ing sponsored by CWA, AT&T, and some of state opinions during informal workshops held the local telephone companies. For example, after their survey forms were complete. She further experiments with semi-autonomous found, however, their concerns encompassed work groups are being carried out among a variety of VDT health issues, including vi- AT&T operator groups in Columbus, Ohio, and sion problems, workstation design, and repro- south-central Florida° ductive hazards; monitoring did not emerge In a financial services firm interviewed by as the major focus of concern. OTA, autonomous work groups were also con- Another study found that in both monitored sidered successful. Workers were taken out of and nonmonitored sites, roughly half the work- "functional" areas and organized into teams ers (47.8 and 46.3 percent respectively) ex- servicing the needs of certain large clients or pressed satisfaction with the evaluation proc- client groups. Employees "cross-trained" one ess.° Among the monitored workers, 17.3 another in different jobs so that each could do percent were not satisfied and 34.7 percent a variety of work and understand the whole were neutral. Among the nonmonitored work- process. The group met together to establish ers, 28 percent were not satisfied and 25 per- productivity goals. Although this firm main- cent were neutral (see table 5). In reviewing tained a more traditional management struc- the supplementary comments made by inter- ture within each group and still applied indi- view subjects, the authors found clear differ- vidual measurement to some jobs, officials and ences in the causes of dissatisfaction. At the employees believed that the reorganization, monitored sites, nearly all dissatisfaction was teamwork, and greater diversity of work greatly directed at the electronic monitoring system; improved both productivity and quality of at the nonmonitored sites it was directed at working life." a number of causes, including supervisors, lack of standards, unfair evaluations, and the like. Supervision and Evaluation This study also found that workers at moni- A few researchers have attempted to com- tored sites tended to believe that their evalua- pare how perceptions of closeness of supervi- tions overemphasized quantity and under- sion, emphasis on performance measures, and emphasized quality, tended to see their re- job satisfaction differ in monitored and non- wards as closely tied to their evaluations, and monitored workers. This is a difficult t ask be- thought that level of supervision was too close. cause so many other cultural, i ob design, and The majority of workers in both groups felt environmental factors can overshadow the ef- they had little participation in workplace de- fect of electronic monitoring. "Elaine J. Eisenman, "Employee Perceptions and Supervi- sory Behaviors in Clerical VDT Work Performed on Systems That Allow Electronic Monitoring," prepared for Educational "Communications Workers of America and AT&T Commu- Fund For Individual Rights and submitted as a contract re- nications, The Emergence of Second Generation Quality of Work port to OTA, 1986. Life Models in AT&T Caamunications: A Pilot Study, Febru- "R.H. Irving. C.A. Higgins, and F.R. Safayeri, "Computer ary 1986. ized Performance Monitoring Systems: Use and Abuse," Com- 'IOTA interviews, November 1985. munications of the ACM, August 1986.

60 Ch. 2-Using Computers To Monitor Office Work 49

Table 5.-Comparison of Monitored and Nonmonitored Workers

Dimension Monitored Percent Not monitored Percent Total ChiSq (df) (p) Satisfaction: Not satisfied 8 17 3 23 28.0 31 Neutral 16 34.7 21 25 6 37 Satisfied 22 47.8 38 46.3 60 Total 46 82 128 2.25 2 p= 0.3247 Emphasis on performance measures: Quantity: Underemphasized 0 16 20.2 16 Appropriate 19 38.0 45 56.2 64 Overemphasized 31 62.0 19 23.7 50 Total 50 80 130 23.79 2 p =0.001 Quality: Underemphasized 22 44.0 15 20.2 37 Appropriate 19 38.0 45 56.2 64 Overemphasized 5 8.3 1 1.4 6 Total 50 124 15.3 2 p= 0.0005 Characteristics of feedback: Amount of feedback: Not enough 21 42.8 41 45.0 62 Enough 22 44.8 41 45.0 63 Too much 6 12 4 9 9.8 15 Total 49 91 140 0.20 2 p= 0.9048 Usefulness of feedback: Not useful 8 17 0 13 14.2 21 Adequate 13 27 6 20 21.9 33 Useful 26 55 3 58 63.7 84 Total 47 91 138 0.93 2 p =0.6281 Importance of evaluation for rewards: Not important 2 4.3 18 22.2 20 Marginal importance 3 6 5 8 9.8 11 Important 41 89 1 55 67.9 96 Total 46 81 127 8 08 2 p= 0.0176 Closeness of supervision: Not close 16 32 6 38 41 7 54 Acceptable 10 20 4 36 39 5 46 Too close 23 46 9 17 18 6 40 Total 49 91 140 13.14 2 p= 0.0014 Participation in evaluation process: Low participation 30 65 2 50 62.5 80 Average 8 17 3 11 13.7 19 High participation ...... 8 17 3 19 23.7 27 Total 46 80 126 084 2 p= 0.6570 SOURCE R H Irving, C H HiggItiS, and F R Safayeni, "Computerized Performance Monitorng Systems Use and Abuse," Communications of the ACM, August 1986, 0796

cisions. This study found little evidence that frames or minicomputers. Yet personal com- workers opposed computer monitoring in prin- puter (PC) use is growing rapidly, especially ciple; their chief problems were not with the among professional and managerial workers. technology itself but with the way it was used OTA did not find examples of production mon- by management. itoring of workers using PCs, but there was considerable interest and controversy over Personal ,omputer Monitoring privacy of an employee's PC files and the right of employers to inspect them. Most electronically monitored work is per- Three primary areas of employer interest in formed on workstations attached to main- PC monitoring have already surfaced:

BA. 50 The Electronic Supervisor: New Technology, New Tensions

1. abuse of PCs (using company resources tors from GSA and other agencies' Inspector for personal purposes), Generals have inspected PC disks of govern- 2. confidentiality (security breaches), and ment employees. These audits have been done 3. violation of legal/regulatory duties in use both to determine that computers are being of client or employee data. used for official purposes and to ensure that confidential information is being properly used Of the 10 firms interviewed for OTA regard and properly protected." ing PC monitoring, none were doing any inspections or searches of PC-user disks or files. All of them, however, said they felt they had Monitoring and Stress* the legal right to do so, and would not hesi- One area in which electronic monitoringmay tate to do so if a specific rule violation or com- have far-reaching implications is in thearea promise were suspected. A typical comment of health effects. A number of authors have by one information system director was: noted the likelihood of a link between electronic monitoring and physical and psychological We have issued a policy guide for privacy stress. Many of the published stories of oppres- and security compliance in PC use, and have sive, heavily monitored workplaces cite the stated that the company reserves the right to inspect all PC files and materials bought by overwhelming fear, anxiety, hatred, and loss the company and used here for our business of self-image that workers suffer. Many purposes. But we haven't felt it wise or neces- authors have stated that there must bea link sary to swoop down on people and demand to between monitoring, stress and health prob- see what they have on their disks. lems, absenteeism, and lowered productivity." On the other hand, representatives of all 10 firms said their organizations audited trans- Stress is now recognized as a major occupational health problem. Stress-relatedsymp- actions done by PCs interacting with data- toms have been estimated to cost U.S. indus- bases on other computers. These were part of try $50 to $75 billion per year in absenteeism, the regular, user-password-based security procedures of mainframeidatabase management. company medical expenses, and lost time.47 Statistics indicate that claims for worker com- The monitoring consisted of: 1) followingup pensation, based on disability due to gradual on any unusual use patterns indicated in regu- accumulation of stress, have been growing rap- lar audit-trail records; or 2) ad hoc inspections of audit records to identify use levels and pat- idly during the 1980s." For workers under terns. End users are informed (in all the orga- "Interview with Don Sheridan, Office of Inspector General, nizations) that such auditing is conducted. General Services Administration, Dec. 16, 1986. "This section draws heavily from Michael J. Smith, Pascale As for assuring end-user compliance with le- Carayon, and Kathleen Miezio, "Motivational, Behavioral, and Psychological Implications of Electronic Monitoring of Worker gal and regulatory rules governing an organi- Performance," contractor report prepared for OTA, 1986. zation's handling of client or employee personal 'See, for example, Tim Healy and Peter Marshall, "Big data, 4 of the 10 organizations reported they Business is Watching You," In These Times, Feb. 26-Mar.I I, 1986; Arlene Hershman, "Corporate Big Brother is Watching had issued written policies to PC end-users You," Dun's Business Month, January 1984; Robert Howard, restating such requirements (e.g., Fair Credit Brave New Workplace (New York, NY: Elizabeth Sifton Books- Reporting Act; State employee access to per- 1985); Peter Perl, "Monitoring by Computers Sparks Employee Concerns," The Wiehington Poet, Sept 2,1984; Peter sonnel records laws; confidentiality of medi- Perl, "Watching the Workplace: High Tech Methods Boost cal information laws; etc.). However,none of Productivity, But at a Cost," The Washington Poet, Sept. 3, the 10 firms reported having done surpriseor 1984. 'Robert Arndt and Larry Chapman, "Potential Office Haz- announced inspections of disks or other desk- zards and Control," September 1984, report prepared for OTA held file materials. Interviews with officials of project on Preventing Illness and Injury in the Workplace,p. 30. the Inspector General's office of General Serv- 'National Council on Compensation Insurance, "Emotional Stress in the WorkplaceNew Legal Rights in the Eighties," ices Administration (GSA) indicate that inspec- 1985. Ch. 2Using Computers To Monitor Office Work 51 age 40, claims related to stress exceeded claims style, genetic predisposition to disease, and related to other occupational disease in 1985." emotional support from others. These claims are from workers of all kinds, in- Although there has been some research on cluding managers and supervisors, who are less the health effects of office automation, there likely than other workers to file claims for has been little research attempting to draw a physical injuries. To the extent that electronic direct link between electronic monitoring and monitoring is associated with stress, then it stress. There are theoretical grounds for must be viewed as contributing to an impor- postulating a link between monitoring and tant health hazard. stress, and the few studies that have been done According to the view most frequently cited suggest that monitoring may be stressful. Un- in the literature, the presence of stress can be fortunately, none of these studies have success- inferred in an individual from a very general- fully separated the effects of computer-based ized physiological response pattern.50 Symp- monitoring from the combined effects of other toms include increases in adrenaline secretion. stressors. the dumping of sugar into the bloodstream, Although there is no clear scientific valida- and other related physiological processes. tion of a link between electronic monitoring These symptoms can be provoked by a vari- and stress, several surveys have found higher ety of environmental agents and situations, incidence of stress among people in monitored such as drugs, fear, and job ambiguity. While jobs. One survey that attempted to look there is nothing wrong with physiological directly at stress and health outcomes of work arousal per se, it can, if chronic, produce seri- monitoring was the 1984 National Survey on ous degenerative effects due to wear and tear Women and Stress, conducted by the 9 to 5 on the body. Thus, stress provides a basis for National Association of Working Women. As the development of various illnesses called noted above, this survey includes one question "diseases of adaptation," since they are not directly related to monitoring: "Is your work a direct function of the agent or situation that measured, monitored, 'constantly watched' or elicited the response pattern, but a conse- 'controlled' by machine or computer system?" quence of the body's adaptive reaction. When the health problems experienced by A 1982 journal article suggests that job fac- women who answered "yes" to this question tors can create stress and lead to chronic dis- are compared to those of all respondents, as orders." The author states that individuals shown in table 6, they show a consistently may perceive the demands imposed by the higher experience of stress-related illnesses. environment as either stressful or not stress- Respondents whose work was subject to com- ful, depending on factors such as prior experi- puterized monitoring were also more likely to ence, current emotional status, health status, rate their jobs as "very stressful." Forty nine and genetically predisposing features. If de- percent of them rated their jobs as very stress- mands are perceived to be stressful, then acute ful, compared to 33 percent of all respondents. biological and emotional responses occur, Seventy-four percent of the monitored work- which, if they continue to occur with some con- ers reported strain, stress, or pressure "often sistency over a prolonged period of time, can or always" in the previous month, compared eventually lead to disease. Various interven- with an overall rate of 63.5 percent for all re- ing factors, which determine the potential for spondents." disease to develop, include individual coping A related question in the 9 to 5 survey asked about production quotas. Almost half (47.4 per- "Ibid. cent) of the women working under production "Hans Selye, TheStress of Life(New York, NY: McGraw- Hill, 1956). 51L. Levi, "Methodology Considerations in Psychoendocrine 529to 5 National Association of Working Women, "The 9 to Research,"Acta MedicsScandinavia, 1982, 191, Supplement 5 National Survey on Women and StressOffice Automation: 528, pp. 28-54. Addendum," 1984, pp. 4-5. Table 6.Rates of Frequent Health Problems Related to Computerized Monitoring of WorkPerformance Respondents answering the question' _ . Is your work measured, monitored, constantly watched, or controlled by machineor comouter system?" Health problems experienced two to three times per week or more/daily Nausea, Exhaustion, Digestive Chest Anxiety, Medical Heada ies dizziness fatigue problems pain nerves Anger Depression problems Yes (358 of all respondents) 31 0% 9 83% 49 40% 2590/o 9.75% 4034% 39.20% 32 2% 40.5% No (1,705 responses) 24 7 554 38 33 17 8 5 40 31 15 28 44 20 0 32 0 Rate for all respondents . 24 2 6 '0 39 70 19 0 560 31.90 28 50 20 8 31.6 Chi square probability ratio.. 0 0201 0.0033 0 0001 0 0008 0.0333 0 0001 0 0001 0 0001 0 0019 SOURCE Nine to Five National Association of Working Women The Nine to rive National Survey on Women and Stress Office Automation Addendum 1984

Table 7.Rates of Frequent Health Problems Relatedto Production Quotas or Productivity Standards

Respondents answering the question "How often does the followingstatement describe your Job? I am required to complete a certain amount of work per hour or per day, e g, a certain number of keystrokes, forms, or items to process." Health problem two to three times All Never Sometimes Often per week or more/daily Always If ever respondents (3,135) (761) (334) (443) (1,531) Eyestrain . . 20.2% 178% 23 3% 234% 274% 250% Headaches. . 24 0 22 3 25 3 28 0 Nausea/dizziness 28 7 27 14 62 56 65 76 89 74 Insomnia 176 154 21 5 20 9 24 2 22 1 Muscle pain. 38 6 35 5 42 6 45 2 49 3 43.2 Exhaustion/fatigue 39 7 35 7 44 4 47 6 56 0 48 4 Digestive problems 188 16 7 20 8 22 8 26 8 23 0 Chest pain . 57 44 70 78 109 83 Nerves, tension, anxiety 31 2 28 0 34 3 39 1 41 8 37 7 Anger/irritability 27 4 24 7 30 3 32 5 37 6 32 9 Depression . 20 0 16 7 24 3 29 3 28 7 26 6 Medical problems 31 6 29 9 34 7 34 4 36 3 35 1 Lost work time 20 9 193 23 4 24 1 25 4 24 1 Base is an currently employed respondents SOURCE Nine to Five National Association of Working Women The Nine to Five National Survey on Women and Stress Office AutomationAddendu n 1984

0 1 16 G 2Using Computers To Monitor Office Work 53 standards reported that their work was meas- State of Oregon shows that a little under one- ured or monitored by a computer. For the fifth of the total claims were made by people others, work is presumably counted by super- in occupations where monitoring is common. visors or by the workers themselves. Women Worker compensation records do not release working under production standards were the detailed cause of injury or the detailed job more likely to rate their jobs as very stressful description of the claimant, so it is impossible (48.1 percent if always under standards, 41 per- to determine if electronic monitoring was ac- cent if often, 29.2 percent if never). Their ex- tually a factor. Of the 542 cases listed, about perience of stress-related illness, as shown in 102 (18.8 percent) were in occupations where table 7, was higher than the experience of all electronic monitoring is fairly common. These respondents." occupations include clerks (of various kinds), These results are consistent with other re- insurance adjustors, bank tellers, telephone search suggesting that monitoring induces operators, dispatchers, and retail sales work- pressure to perform. Some managers may feel ers. The rate of acceptance and denial of claims that this is a desirable effect, since it implies is shown in table 8. The acceptance rate for high production. But occupational stress re- potentially monitored office occupations was search indicates that excessive work pressure roughly the same as for all jobs, 34.2 and 35.2 is not conducive to good long-term perform- percent respectively. ance and brings about adverse health conse- Other studies have found a high incidence quences!' In fact, there are a range of stress- of stress-related illness among workers most ful working conditions that may be related to Maly to experience electronic monitoring, even electronic monitoring of employee perform- though monitoring itself was not examined as ance. These include heavy workload, especially a variable. For example, a study by the Na- of repetitive or machine-paced tasks; routinized tional Institute of Occupational Safety and work activities; lack of control over timing, Health found that secretaries had the second- speed, and variety of tasks; and social isola- highest incidence of stress-related illness tion, including lack of peer social support, re- among 22,00C workers. The Framingham heart duced supervisory support, and fear of job loss. study, released in 1985, found that women cler- There are many potential causes of stress ical workers develop coronary heart disease at in the workplace, and it is not clear from worker nearly twice the rate of other women work- compensation claims that work monitoring is ers." Researchers have commented that the a dominant one. However, a review of mental stress-provoking factors in these jobs are rapid stress worker compensation claims from the work pacing, including machine pacing, monotonous or repetitive work, and lack of discre- "Ibid. tionary control. I4C.L. Cooper and J. Marshall, "Occupational Sources of Stress: A Review of the Literature Relating to CoronaryHeart Disease and Mental Ill Health,"Journal of Occupational Pay. 55Working Women Education Fund, "Health Hazards for Of- chology49, 1976, pp. 11-28. fice Workers," April 1981.

Table 8.Review of Oregon Worker Compensation Claims Involving Mental Stress January 1985 Through September 1988

Number of Percent Occupations claims AcceptedacceptedDenied All occupations 542 191 35.2 351 Possibly monftored occupations: Office' (percent of total 13.4) 73 25 34.2 48 Retail sales (percent of total 5.3) 29 11 37.9 18 &Occupations,In order of demising frequency, are. clerk (39), insurance adjustor (10), dispatcher (9), administrative support (5), computer operator (4), data army (2), bank teller (2), telephone operator (2) SOURCE Oregon Worker's Compensation Department, Research and Statistics Section, "Accepted and Denied Claims In. volving Mental Stress, Ore900. 11.54188."

s 6 54 The Electronic Supervisor. New Technology, New Tensions

Time pressure, such as having to meet dead- has a propensity for reducing worker feelings lines, is another significant factor in stress. of job involvement and may in this way in- Studies have shown increases in stress level crease worker distress. The chances to partici- as difficult deadlines draw near." The experi- pate and be involved in the job process may ence of deadline pressure on a constant basis, be diminished in work systems that are driven as might be the case in a fast-paced monitored by employee performance monitoring. job, may be more damaging than deadline pres- Reduced coworker support can also contrib- sure experienced on an occasional basis. ute to stress. Monitored workers in several Two organizational. 'actors have been shown studies, and those interviewed by OTA stated to be of special significance for increased job that, due to their production standards and stress and decreased worker health. These are: the electronic monitoring system, they had no 1) job involvement or participation; and 2) or- opportunity to interact with coworkers 80 ganizationa: support, as reflected by supervi- One study of work monitoring in the tele- sory style, support from managers, and chances for career development. Lack of participation communication industry suggests that the pos- in work activities has been demonstrated to sible connection between monitoring and job- result in increases in negative psychological related stress is through the changed struc- moods." In terms of organizational support, ture of the work. In this study, no direct correlation was found between electronic moni- it has been shown that close supervision and toring and stress-related illness. However, a a supervisory style characterized by constant correlation was found between monitoring and negative performance feedback are related to low job control which hrs been found, in other high levels of stress and poorer worker health." The implication of these findings is that ex- studies, to be associated with stress-related illness. The conclusion reached by the research- cessive, impersonal electronic monitoring of ers is that when jobs are redesigned to facili- employee performance that produces close su- tate computerized monitoring of work perform- pervision and constant negative performance ance, they are also reshaped in ways that feedback could promote worker stress. increase the degree to which management It has also been demonstrated that workers' directs both the pace and the method of work. feelings of lack of involvement are related to stress and that prolonged stress can be related and Health3, 1977, pp. 165-182. R.A. Karasek, Jr., "Job Deci- to health complaints." Electronic monitoring sion Latidute, Job Design, and Coronary Heart Disease," in G. Salvendy and M.J. Smith (eds.),Machine Pacing and Oc- "M. Friedman, R.H. Rosenman, and V. Carroll, "Changes in cupational Stress(London: Taylor & Francis, 1981), pp. 45-56. the Serum Cholesterol and Blood Clotting Time in Men Sub- R.D. Caplan, S. Cobb, J.R.P. French; R.V. Harrison, and S.R. jected to Cyclic Variation of Occupational Stress,"Circulation, Pinneau,Job Demands and Worker Health(Washington, DC: 1958, pp. 852-861. U.S. Government Printing Office, 1975). M.J. Smith, B.G. Co- "B. Margolis, W.M. Kroes, and R. Quinn, "Job Stress: An hen, and L.W. Stammerjohn, "An Investigation of Health Com- Unlisted Occupational Hazard,"Journal of Occupational Medi- plaints and Job Stress in Video Display Operations,"Human dne18,1974, pp. 854-661. R.D. Caplan, S. Cobb, J.R.P. French, Factors23,1981, pp. 3f.17-400. B. Gardell, "Technology Aliena- R.V. Harrison, and S.R. Pinneau,Job Demands and Worker tion and Mental Health,"Acta Sociologica19, 1976, pp. 83-94. Health(Washington, DC: U.S. Government Printing Office, B. Margolis, W.M. Kroes, and R. Quinn, "Job Stress: An Un- 1975). M.J. Smith, B.G. Cohen, and L.W. Stammerjohn, "An listed Occupational Hazard," op. cit. S.G. Haynes and M. Fein- Investigation of Health Complaints and Job Stress in Video leib, "Women, Work and Coronary Heart Disease: Prospective Display Operations,"Human Factors23, 1981, pp. 387-400. Findings From the Framingham Heart Study,"American Jour- "R.D. Caplan, S. Cobb, J.R.P. French, R.V. Harrison, and nal of Public Health70, 1980, pp. 133-141. M.J. Colligan, J.J. S.R. Pinneau,Job Demands and Worker Health(Washington, Smith, and J.J. Hurrell, "Occupational Incidents Rates of Men- DC: U.S. Government Printing Office, 1975). M.J. Smith, B.G. tal Health Disorders,"Journal of Human Stress3, 1977, pp. Cohen, and L.W. Stammerjohn, "An Investigation of Health 34-39. Complaints and Job Stress in Video Display Operations,"Hu- "For example see R.H. Irving, C.A. Higgins, and F.R. man Factors23, 1981, pp. 387-400. Safayeri, "Computerized Performance Monitoring Systeme: Use "World Health Organization,Psychosocial Factors and and Abuse,"Communications of the ACM,August 1986. In- Health: Monitoring the Psychosodal Work Environment and terviews in Michael J. Smith, Pascale Carayon, and Kathleen Workers' Health(Geneva: 1984). J. Rutenfranz, W. Colquhoun, Miegio, "Motivational, Behavioral and Psychological Implica- P. Knauth, and J. Ghats, "Biomedial and Psychoeocial Aspects tions of Electronic Monitoring of Worker Performance" con- of Shiftwork,' 'Scandinanvian Journal of Work Environment tract report prepared for OTA, July 1986. 67 Ch. 2Using ComputersToMonitor Office Work 56

This lack of personal control, in turn, places have an even greater stress effect than tradi- workers at significantly greater risk of ill tional factory pacing." health." Feedback and Motivation" Computer Pacing and Stress Perhaps the best use of information about Machine pacing is different from computer an employee's performance is to give it back monitoring. The work of a directory assistance to the employee. One advantage that electronic operator offers an example of a "low control measurement can offer to workers is accurate job," one that is both paced and monitored by and timely information about their own per- computer. A compt.: m-controlled distribution formance. Studies of feedback, whether related system passes a call to an operator. He or she to simple sensory feedback or to higher levels greets the customer, hears the name to be of feedback related to knowledge of results, all looked up, and keys it into a cuiliputer termi- indicate that ,eople want to know about their nal. Once the proper information appears on performance and will seek out such knowledge the screen, the operator presses a key to re- when it is absent." lease the call. A voice synthesizer actually Immediate sensory feedback helps employ- reads the telephone number to the customer. ees to exert better control over skilled actions Once the cell is released, the distribution sys- and to correct errors." For example, the feel tem presents another call U., the operator. of the keyboard and the display of the charac- The job is monitored, L. that records are kept ters on the video screen help a data entry or on the operator's performance within each word processing operator to know that data callthe time to respond to the call, locate the is being 1, ...Jed properly. This type of feedback proper information, and release the call (sum- is continuous throughout the task. marized as average work time or AWT). In A higher level of feedback, knowledge of re- addition, the job is also machine paced in that sults, occurs when a task is completed and the cycle time betweer. ..alls is controlled by evaluated against some external standard, and the computer, not by the operator. Work pres- the results are fed back to the employee; "You sure increases if that cycle time is very short. have produced 10 percent over the rraduction Its been roted that new technology has goal today," or "Your output had 2 percent turned the job of directory assistance opera- errors.- This kind of feedback provides director into a literally thankless task. Not only is tion to the worker about future output. the pace hectic, but because the operator re- Feedback about one's own activities can be leases the call before the customer receives the neet:ed information, the operator never hears customers say "thank you." Job design fac- "M.J. Smit'u, B.G. Cohen, and L.W. Stammerjohn, "An In vestigation of Health Complaints and Job Stress in Video Dis- tors, along with the fast pace, probably greatly play ^aerations," Human Factors 23. 1981, pp. 387-400. A. contribute to stress in this job. Caki..4. Reuter, L. Von Schmude, and A Armbruster, Inves- qgations f the Accommodations of Human Psychic and Phys- Machine pacing has been implicated as a sig- ical Pune:one to Data Display Screens in the Workplace (Ber- nificant factor in ill-health among factory work- lin:Institute fur ArbeitswissensAtrft der Technician Universitat Berlin, 1978). ers. Computerswhich c-n operate at high 63This sec on draws heavily from Michael J. Smith, Pascale speeds on a continuous basis--have increased Carayon, tu.Kathleen Miegio, "Motivational, Behavioral and the pacing impact coffice workers. Recent Psychological Implications of Electronic Monitoring of Worker Performance" contract report prepared for OTA, ;July 1988. research suggests that pacing produced by S.J. Ashford and L.L. Cummings, "Feedback as an Indi- computer-driven video display systems may vidual Resource: Personal Strategies of Creating Information," Organizational Behavior and Human Performance 32, 1983, pp. "Steven P. Valles and William V. Calabro, "Occupational 370-398. Cditions and Worker Health in the Communications Indus- K.U. Smith and M.R. Smith, Cybernetics Principles of try," New York Institute of Technology, Human i"esources De- .,earning and Education Design (New York, NY: Holt, Rine- velopment Center, no date. hart & Wi iston, 1966).

Cj 56 The Emcfronic Supervisor. New Technology, New Tensions

a very powerful motivator and has been found they thought they did and 2) what they are to have a strong influence on productivity, and expected to do. In some cases it may correct in some cases on job satisfaction as well. Peo- misconceptions or inaccurate perceptions about ple like to do a good job, but without informa- what they are doing. In one example, airline tion they often don't know whether they are reservation clerks were provided with profiles doing one or not. of their verbal behaviors based on sample Because computer technology is adept at recordings of their telephone conversations gathering and correlating information, it can with customers. One clerk commented on see- be very useful in giving timely feedback to ing the feedback: workers in a useful form. For example, office When asked previously whether I used the systems can be designed to give performance cw,tomer's name I would have saidand information to workers as well as to supervi- bedeved "Of course, we were trained to do sors. In some firms, for example, customer that." I was really surprised when I saw ob- service representatives can get private access *live evidence on how little I was actually to their own recent performance by keying the doing it." proper code into their workstations." Any in- As a result of the feedback, use of the custom- formation that is available to their supervisor er's name by the clerk, rose by 87.5 percent, is available to the individual workers, as well while the clerks' interrupting of customers (a as comparisons with the gro average and habit the employer wished to discourage) the standards. nearly disappeared. A recent study reviewed lase histories Feedback is an effective modifier of behavior and found that without excel.: ple per- if it is seen as a valueci commodity by there- formed better when they were ip some ob- cipient and if it is timely. It t Rkes on value to jective, quantitative feedback about their own the individual when it is effective (relevant, un- performance or output. This study did not fo- derstandable, accurate, useful) and when it cus on electronic work measurement, but comes from a trusted or highly regarded rather on both manual and electronic methods source. Although feedback need not be imme- in a variety of work settings, including banks, diaie or continuous, it should be given fre- payroll offices, reservation offices, manufac- quently. The longer the delay, the less effec- turing facilities, and health care facilities. The tive it is in affecting performance." A form in which feedback was given also varied number of the workers interviewed for OTA from one setting to another. The three meth- expressed a desire for more frequent feedback ods most commonly used were private individ- about their work. They also thought that feed- ual feedback, public individual feedback, and back information from the electronic monitor- public group feedback. In some cases, objec- ing system could be better designed to help tive feedback was combined with other inter- them gain more control of their work. ventions such as praise, public recognition, or additional training; but positive results were If employees perceive that rewards and/or also noted where feedback alone was punishments could ensue from an evaluation provi ded.67 of their performance, they are especially interested in feedback. They want to understand Feedback may serve as both a motivator and the basis of rewards and punishments, and an instructional device. When people receive feedback helps to resolve feelings of ambiguity what they perceive as objective feedback, they or uncertainty. can compare what they actually did to: 1) what p. 176, citing Stephen A. Allen, "Aer Lingua Irish "Interviews at Americt.n Express Southern Regional Over- (B)" me #9-477-640 (Boston, MA: Intercollegiate Case Clearing- ations Center. house, 1976), p. 7. "Richard E. Kopelmsn, Managing Productivity in Organi- "N.J. Smith, B.G. Cohen, and L.W. Stammerjohn, "An In- zations: A Practical, People-Oriented Perspective (New York, vestigation of Health Complaints and Job Stress in Video Dis- NY: McGraw Hill, 1986), pp. 163-187. play Operations," Human Factors 23, 1981, pp. 387.400. Ch. 2Using Computers To Monitor Office Work 57

At the same time, feedback of performance pie know how to pace their work. When the can create feelings of anxiety, frustration, and display shows all zeros, agents feel freer to take lowered self-esteem. Since feedback fulfillsan a little extra time with a difficult caller, to catch error-correction function as well asa perfor- up on paper work, or to take a break. When mance-appraisal function, it can indicate toem- many calls are backed up, they can make an ployees that they are not doing their jobas extra effort to finish a call quickly, or perhaps well as they would like to,or as well as the to defer a break for a few minutes rather than employer would like them to. Thiscan create abandon their colleagues in a crunch. stress, even when it resolves the stressasso- ciated with uncertainty and ambiguity. In Team spirit and friendly competition can be short, feedback is a two-edged sword insofar powerful motivators, and both employers and as stress is concerned. employees can benefit if they are not abused. However, workers can also perceive employers' Feedback is necessary to groupsas well as use of feedback and social pressure to be un- individuals, and computer monitoring systems fair and manipulative. A Pacific Western Air- can also give workers immediate information lines (PWA) productivity campaign drew union about the work environment letting them know protests when company posters urged reser- how their work group is doing rightnow, and vation clerks to how they can best contribute. Compare yourself with your friends. Com- To take telephone customer service againas pare yourself with ones who aren't your an example, some offices have clearly visible friends. Are you pulling your weight at the of- displays on the wall that show the number of fice? When the monthly statistics are pub- incoming telephone calls waiting to bean- lished, ensure you're not dragging down your swered and the age, in seconds, of the oldest team and your office. call. Such displays could be usedas weapons The union newsletter charged PWA with set- of callous management to keep constantpres- ting workers against each other, and called the sure on agents of understaffed offices. How- campaign a "new low in...degradation."" ever, in a properly staffed office, where peaks of incoming calls occur fora few minutes at a time, a few times a day, status displays be- "Lawrence Archer, "I Saw What You Did and I Know Who come tools in the hands of the work group. Peo- You Are," Canadian Business, November 1985, p. 81.

THE FUTURE OF WORK MONITORING The OTA report on Automation of America's talk directly to each other, thus reducing the Offices pointed out some trends in the growth need to keyboard data. of computer-based office automation equipment that have implications for the futtre of Another trend was the growth in the intro- work monitoring. duction and use of office automation equip- One trend was the inevitable movement ment and its rapid adoption by all sectors of the economy. It is estimated that by theyear toward direct machine-to-machine communi- 1990 t 'aere will be one computer terminal for cation. Increasingly, data will be captured in machine-readable form at the point of origin, every three workers in the United States; by the year 2000, terminals may be ascommon customers will enter their own data (as with in offices as telephones. automatic teller machines) information will be recorded using optical scanning and voicerec- Both of these trends suggest possible cha:ges ognition, and different computer systems will in the population of workers that will be af-

7 u 58 The Electronic Supervisor. New Technology, New Tensions fected by computer-based work-monitoring over time, however. If at some future time man- technology. For example, the prime example agement determines that the benefits to be of the monitored job today is that of the data- generated from monitoring a particular job cat- entry operator, but over the next 20 yeus the egory will outweigh possible costs in higher growth rate of data-entry workers is expected turnover, monitoring systems are likely be in- to slow or perhaps decline. Those that remain troduced. And while professionals may be able will probably still be monitored, but they will to defeat their current monitoring system, a be a smaller proportion of the office work force. system that automatically collects correct in- The other trendtowards wider use of com- formation could be designed if their employer puter-based office equipmentsuggests that ever decides it is worth the cost. more jobs will be at least partly automated or The growing use of computer-based manage- dependent on the use of a computer. As a re- ment information systems also means that sult, more types of jobs will be possible candi- more managers will be subject to closer moni- dates for electronic monitoring. Although the toring, simply because more of their day-to- characteristics of monitored jobs listed at the day decisions will be revealed to superiors beginni1ig of the chapter (repetitive tasks, high through the computer system, rather than volume of work, low training requirement, high waiting for monthly or quarterly reports. tolerance for turnover, ample labor supply) de- If there is a growth in computer monitoring, scribe ideal conditions for monitoring, they are or a spread to other types of work, it does not not absolutes. It is already possible to apply necessarily mean a devaluation of office work. electronic monitoring to some highly skilled Computer-based monitoring can offer advan- professional and management positions. Some- times monitoring has not worked well in high- tages to employees, for example, improved feedback and better control of their own work. level positions. Employee resistance may have Professional and managerial workers may be caused management to back down on imple- able to use their bargaining power with em- mentation plans, or, as in the case of bank loan ployers to participate in decisions about the officers interviewed by Westin, employees may redesign of their jobs or the implementation have found ways to "game" the system by of work measurement and monitoring, as to feeding it false information." The costs and assure fair use of monitoring. As with other importance of employee resistance can change examples of technology in the workplace, many nontechnological factors, including management and employee attitudes, corporate cul- "Alan Westin, "Privacy and Quality of Life Issues," Alan Westin, "Privacy and Quality of Work Life Issues in Employee ture and relative power relationships, will gov- Monitoring," contractor report prepared for OTA, 1986. ern how the technology is used.

7 Chapter 3 Telephone Call Accounting CONTENTS

Page Introduction 61 Managing Telephone Costs .. . . 62 Telephone Call Accounting .... 63 How Call Accounting Works ...... 63 Call Accounting and Telephone System Management ... 64 Call Accounting and Employees' Personal Use of Telephones 65 Active Cost Control Methods 68 Least-Cost Routing 68 Calling Restrictions 69 Authorization Codes and Levels of Service 69 Timed Signals 69 Studies of Unofficial Use of Government Telephones 70 The PCIE Review of FTS Utilization 71 Results of PCIE Audit 73 Future Directions for Government Telephone Management 75 Need for New Policies 75 Establishing a Policy on Personal Use 75 Exception Reports and Personal Use 76 Future Use of Call Back Audits 77 Local and Suburban Calls 78 Calls to Press, Union, or Public Interest Groups ..... 78 Eavesdropping and Service Observation 79 Other Methods of Cost Control 80 Figures Figure No Page 9. Sample Call Detail Report ...... 64 10. Sample Summary of Calls ...... 65 11. Sample Exception Report ...... 67 Tables Table No Page 9. Results of FTS Intercity Off-Network Call Sample by Agency . . 73 10. Results of Commercial Long-Distance Call Sample by Agency .. 74

7) Chapter 3 Telephone Call Accounting

INTRODUCTION The personal use of government telephones, personal use in the privatesector that range ai the controversy about whatto do about from 10 or 15 percent" toas high as 30 or it, was in the news throughout 1985and 1986. even 50 percent.' News stories carried headlines like:"U.S. Agencies Use High Tech To Curb Workers' Phone Use: Savings Estimated at $300,000 a Employees' personal Year,'" '.U.S. Phones Raise Issue of use of telephones has Privacy: been going on foryearsa common practice New Equipment Would ProvideDetailed that many firms and agencies haveignored as Records of Calls?* "U.S. To Use ProgramTo being o: little importance. However, Audit Federal Employees' Calls, "" "Planned in the past Phone Audit Brings Blast From few years new technological toolsto measure Several and control telephone costs have Groups,'" "Toll Calls Abused by U.S.Em- come on the market and are being enthusiasticallypro- ployees,"`and "Lieten Up Government moted by a growing segment Workers: You May Be Allowed One Call."' of the telecommunication industry. Deregulation of thetele- Personal phoneusein government wasex- phone industry has forcedmany firms to pay amined in an audit conducted by theGeneral greater attention to the costs andmanagement ServicesAdministration (GSA) under the of their telephone systems. Althoughlong-dis- President's Council on Integrity and Efficiency tance rates have declined steadily for thepast (PCIE). That audit reported in the springof decade, telephone costs remaina major busi- 1987 that an average of about 33percent of ness expense. Many firms have adoptednew off -network calls sampledon the FedertlTele- technologies, such as telephone call-ac cotmtliv communications System (FTS)were "unoffi- software, in an effort to further controlthese cial," i.e., made for personal reasons.' costs.

Of course personaluse of employers'telephones is not a phenomenon limitedto gov- Controversy arises becauseuse of call- ernment. Telephone use has been calleda accounting software may impingeon the "phantom job benefit" becausemany employ- privacy of people using the telephone ees, in both public and private sectors, believe system. Although its best and mostcommon use is as they have a right to makesome calls from the a statistical tool to analyze patterns of tele- telephone on their desk. Reliable dataare not phone use for a firm or office, call-accounting available for the private sector, but telecom- records can also provide detailed information munications experts have given estimates of about each ;JO% -lewd call, whetherofficial or personal. The software automaticallycan re- 'Interview with Edward Horrell, Mitchell & Hornell,Inc., cord the information law enforcement officials June 24, 1988. sometimes gather using "pen registers"the 'New York Ilssee. Mar. 17, 1986. Vomputorweeld, Mar. 25, 1985. exact time, date, originating extension, and 'Federal Times, Mar. 25, 1985. destination number of every calllocalor long Washington Poet, June 21, 1986. 'Washington Poet, Sept. 11, 1988. 'President's Carrel! on Integrity and Effie cy, "Consoli- 'Interview with Edward Horrell, Mitchell & Herren,Inc., dated Report on Mind Talsoonimunications S.ern (FTS) Uti- June 24, 1986. lisation," prepared by the General Service. Adtmnistration, Of- 'Judith Ilevemann. "Toil Calls Abused by U.S.Employees," fice of the Inspector General, Mar. 16, 1987. Washington Post, June 21, 1986.

7 81 62 The Electronic Supervisor: New Technology, NewTensions distance. Such information, properly exam- subscribers. Good telecomm,Inications man- ined, can sometimes provide considerable in- agement begins with underst anding the needs formation about the caller." of the firm and making basic decisions about the size of the system needed, what long-dis- This chapter reviews the growing trend in tance carriers to use, whether to lease private use of call-accounting software byemployers lines or to own private switches, and so on. Be- in and out of government. It also attempts to cause of the wide variety of servicesoffered place call accounting in the context of other and the different rate structures of the long- telecommunication management tools. Call distance telephone companies (carriers), wrong accounting is only one of a number of technol- design decisions can be costly. Telephone call ogies gaining use in firms and government accounting, as discussed later in this chapter, agencies to control telephone costs. This chap- can be useful in giving a clear pictureof tele- ter also discusses other techniques that can phone use on which to base management de- be used instead of or in addition to call cisions. accounting. Telephone fraud perpetrated by outsiders Managing Telephone Costs can be costly to an individual firm,but most of its costs fall on the telephone industry and Controlling personal calling is not the only users as a whole. Fraudulent activities range way to reduce the telephonebill. Industry ex- from individual hackers using private systems perts estimate that 35 to 40 percent or more for their own calls to multilevel marketing of all long-distance calls are "waste calls." schemes that sell illegally obtained authoriza- These are sometimes characterized as a com- tion codes to consumers. If the legitimate user bination of four factors: (individual or corporate) notices charges for these calls on the phone bill and denies making poor system design, them, the cost is usually absorbed by the long- fraud (unauthorized use by outsiders), distance carrier. The cost of telephone fraud abuse (personal use by authorized users), was estimated at $500 million in1985." The and telephone industry is attempting to combat misuse (overuse or the use of a high-cost fraud by improving system software's ability service when a low-cost alternative is to detect and investigate illegal users. As will available)." be discussed later in this chapter, telephone The proportional importance of each category call accounting can help firms reduce costs due varies from one organization to another, but to fraud by giving them an accurate listingof abuse and misuse are often the largest, accord- telephone calls independent of their telephone ing to some experts. Several technological ap- bill. proaches for combatting them are discussed There are two basic approaches to reducing in this chapter. unwanted calls The first is to keep careful Poor system design means that telephone track of all calls so that problems of misuse equipment and service are not suitable to the or personal use can be tracked andpeople who particular calling patterns of the firm. In the make the calls can be identified. This is called past 10 years, a bewildering variety of alter- a passive approach to telephone management, natives has developed for business telechcne and the principal tool is telephone call accounting. The second approach is to design the telephone system so that unwanted calls are diffi- K'For a discussion of the use of pen registers in law enforce cult or impossible t^ make, and so that calls went, see U.S. Congress, Office of Technology Assessment,Fed- eralGovernmem, Information Technology: Electronic Surveil- that are mace are of reasonable length. This lance and Civil Liberties, OTA-CIT-293 (Washington. DC: U.S. Government Printing Office, October 1985). "Interview with Edward Morrell, Mitchell & Morrell, Inc., '2Walter G. Frier, "Combating Long Dintance Service June 24, 1986. Abuse," Telephony, Aug. 11, 1986, pp. 69-70. Ch. 3Telephone Call Accounting 63

active approach to telephone systemmanage- tions about the privacy of individuals who ment is aided by such technological toolsas make the calls. With the advent of improved least-cost routing, call blocking, andtimed records on all long-distance calls, thegovern- signals. ment finds itself faced with the need to review and revise some outdated policies relatedto As the Federal Government begins theproc- the use of its telephone system. Before discuss- ess of revamping its telephone system, it will ing these policies in detail, however, it would undoubtedly make use of several of these tech- be good to look at how active and passive tele- niques. The one that has aroused the mostcon- phone management tools work and how they troversy is call accounting, because of ques- are used in government and private industry.

TELEPHONE CALL ACCOUNTING In the past 3 years, telephone callaccount- It is estimated that about 19,600 stand-alone ing has become one of the fastest growingseg- call-accounting systemswere sold in 1985, ments of the telecommunications market, but amounting to revenues of about $206 million call accounting is nothingnew. Every con- for their manufacturers. This market isgrow- sumer engages in telephone call accounting at ing at about 50 percentper year and there are home by reviewing the monthly telephone bill currently about 130 firms that eithermanu- to make sure that all the long-distance calls facture a call-accounting device, write call- listed were actually made bysomeone in the accounting software, or provide call-account- household. ing services." In addition,many private branch exchanges (PBXs), thecomputerized Telephone bills for businesses, while theyare switching systems that route telephone calls sometimes more complex than residential bills, in many offices, have built in call accoundng provide essentially thesame information: the capability. date, time, duration, destination, and cost for calls. Usually all this information is provided How Call Accounting Works to business customers as a matter ofcourse for direct-dial or operator-assisted long-dis- Raw data about callsthe time, duration, tance calls. However, long-distance calls made called number, originating extension, and on Wide-Area Telecommunications Service estimated costcan all be collected bya de- (WATS) lines (which are billedon an average- vice called a stationmessage detail recorder cost-per-call basis) and local calls usuallyare (SMDR) that can be attached to the telephone not reported in detail to business customers, system. SMDRs can producean enormous vol- unless specially requested and paid for. ume of information that is of little use until it is processed and analyzed. Businesses and government agencies are be- Probably the oldest typc of call accounting coming more aware of the value ofan accurate is offered by computer service record of calls in managing and reducing their bureaus, which came into existence around 1970. The service telephone costs. Even if call detail is provided bureau uses mainframe computers to by the carrier (long-distance telephonecom- process pany), the firm may want an independent the magnetic tapes produced by SMDRs and rec- provides the customer with monthly ord of telephone calls in order to verify thecar- or quarterly reports. The cost for sucha service varies riers' bills or allocate telephone coststo widely. Depending different departmt is within the organization. on the number of lines, the Businesses can get such accurate, up-to-date fee can range from $1 to $4per telephone." call accounts either from a "service bureau," "Daniel I. Strusser, "Good News in the Call Accounting or through a call-accounting system on their Market,"Telecoanect,March 1986, p. 62. "Daniel I. Strusser, "The Six Kinds of Call Accounting," own premises. Teleconnect,March 1986, pp. 66-71. ...- 7 0 64 The Electronic Supervisor: New Technology, New Tensions

Figure 9.Sample Call Detail Report

Date: 03107184 Page. 1 Time: 11:56:20 Report Period: 2/24 - 2/29 Name: Dan Jones Division Telecommunications Department: Engineering Ext: 1551

Date Time Duration Charge Number called Facil (1) City ST (2) Acct. Code MI 2/24 08:01 00:12:15 0.06 616-429-2998 Local St. Joseph VA 2/24 11:35 00:25:00 5.86 703-620-0880 WATS Roanoke 2/25 08:46 00:00:30 0.06 616-429-4151 Local St Joseph Mi 2/25 08:52 01:12:30 25.90 212-829-4272 DDD New York NY 2/25 10:57 00:07:30 0.10 Incmg 2/25 12:57 00:10:56 4.10 714-525-5252 MCI Anaheim CA 2/25 14:00 00:16:01 6.27 312-577-7901 FX Chicago IL 2/25 14:07 00:01:30 0.10 incmg 2/27 09:43 01:05:03 35.12 714-525-5252 DDD Anaheim CA 2/27 12:55 00:01:00 .42 703-620-0880 WATS Roanoke VA 2/27 13:14 00:10:00 0.06 616-429-6241 Local St Joseph MI Totals: 03:42:15 77.15 Fixed: 5.00 Calls: 11 Cost/Min: .33 82.15

There are about 30 firms in the United States Call Accounting and Telephone System that provide this service. Service bureaus have Management generally been used only by large firms with high volumes of calls and multiple sites." The report-generating capability of the system is important. While the call-accounting de- Advances in computer technology are now vice keeps track of all calls in the order they making call accounting more economical for are made, a simple printout of all call records smaller firms. Call-accounting software is now may be of little use, especially in a largefirm available for direct use by the customer, and with thousands of telephones and dozens of it can be run on personal computers, minicom- locations. puters, and mainframes. About half of the service bureaus, along with dozens of other com- Figure 9 is a sample printout from a call- panies, lease or sell call-accounting software accounting system. This particular system is for customers to use. In addition, some PBXs designed for small companieswith perhaps come with built-in capability to record rawcall 100 to 500 telephones ar d runs on a personal data with a SMDR and to process call- computer. Raw call data is transferred to the accounting reports. Prices for call-accounting personal computer from the SMDR through software vary widely, from as low as $800 up a RS232 connection (like the modularphone to $40,000, depending on the size of the tele- jack on most telephones). Once the call records phone system and on the special features that areiloaded on the computer, the call-accounting might be desired." Many software packages software can produce a number of standard and produce not only a detailed listing of all calls, customized reports. Figure 9 shows all the calls but also allow the development of a number of a particular extension (1551), including the of standard and customized reports. date, time, duration, cost, number called, and city and State of destination number. This information is similar to that found on a tele- "MCI Education Center, Gaining the Competitive Edge: phone bill, except that there is somewhat Network Design, p. L-6. "Daniel I. Strusser, "The Six Kinds of Call Accounting," greater detail. To some extent the level of de- Teleconnect, March 1986, pp. 66-71. tail to be used in reports can be chosen by the Ch. 3-Telephone Call Accounting 65

Figure 10.-Sample Summary of Calls

Date: 03/07/84 Time: 11:43:37 Page: 1

Report Period:2/24 - 2/29

Total Fixed Local L.D. OutDur Incom InDur Ext. Name cost cost cost cost HR:MN Cost HR:MN 101 Jackson, John 16.54 0.35 0.00 16.19 0:25 0.00 0:00 102Cheever, Chuck 4.48 0.27 0.00 0.75 0:05 0.40 10:32 103Berg, Wendy 9.36 0.42 5.61 3.33 0:11 0.00 0:00 104West, Ellen 4.00 4.00 000 0.00 0:00 0.00 0:00 111 Cassidy, Mike 16.88 0.56 0.00 13.62 0.18 2.70 5:12 125 Ryan, Pete 3.25 3.25 0.00 0.00 0:00 0.00 0:00 150 Potts, Karl 1.20 1.20 0.00 0.00 0:00 0.00 0:00 155 Jones, Dan 83.05 5.00 0.18 77.67 3:33 0.20 0:09 Totals: 138.76 15.05 5.79 111.56 4:32 3.30 15.53 system manager. For example, this particu- calling patterns and system utilization. Forex- lar firm has chosen to list and assigna cost ample, a summary of all calls by trunk (or type to local calls as well as long-distance ones, and of service) would enable a telecommunications to include a listing and charge for incoming manager to compare the number of calls and (Incmg) calls. relative expenditures for direct dial and WATS Figure 10 shows a summary report byex- lines to determine if the firm has the right fa- tension, summarizing the costs and activities cilities to meet current needs. Or the system of all telephones in a particular department. could produce a report of the 50 most fre- Similar detailed and summary reports could quently called numbers, in order of frequency. be generated by extension, by caller,or by ac- The telecommunications manager mightuse count code, for each department or division in this information to determine whethera pri- the firm. vate line connection would be a more economical way to carry calls between the main office More sophisticated cost-accounting reports and a frequently called branch office. A report are also useful for equitably allocating telecom- on trunk utilization, by day and hour, can also munication costs. Based on reports generated be useful in analyzing the level ofuse of the by the call-accounting system, costs can be al- telephone, and might also be useful evidence located to the proper department, project,or in case of disputes with airriers about the custlmer account. Law offices, for example, amount of the telephone bill. which must keep accurate records of each attorney's expenditure of time and resources for Call Accounting and Employees' each client, can generate accurate reports of telephone calls related to each case. This might Personal Use of Telephones be done either by having staff members dial At some firms and governmen t agencies, a cost code before dialing each number, or by analysis of the most frequently called numbers having the call-accounting system store tele- turned up a large number of calls to off -track phone numbers known to be frequently used betting, "Dial-a-Porn," the weather report, and for each client. Call accounting software for many long-distance calls to locations that did hotels and hospitals produces phone changes not do business with the organization. for inclusion in client bills. Employees' personal use of employers' tele- Another advantage of modern call-account- phones has become a concern in the past few ing software is the ability to process mountains years and reference to the money being spent of raw call data into useful information about on personal calls is a major sales tool for ven-

76 66 The Electronic Supervisor. New Technology, New Tensions

dors of call-accounting equipment. Telecom- Despite the long-term decline in long- munications trade magazines (and advertise- distance telephone rates, and the sharp decline ments of cost-accounting system vendors) are since divestiture,19 telephone costs remain a full of anecdotes about abuses uncovered when major expenditure for many firms. Technologi- firms first start keeping track of their tele- cal tools that promise to further control these phone calls. Stories include, for example, the costs are attractive to managers, and vendors' secretary who placed a one-hour long-distance assurances of reduced telephone costs have call during lunch every day in order to listen fueled the sales of call-accounting equipment to a soap opera on her mother's television. Or and software. the man who ran his personal business from Personal use of an employer's telephone has the office telephonea business that required been called a "phantom job benefit." Many hundreds of long.-distance calls weekly. Or the woman who used the call-forwarding feature people consider personal use of the telephone on their desk to be a reasonable perquisite, and of her office telephone to receive many hours of long-distance calls at home in the evening. the question of when this personal use becomes "abuse" is sometimes difficult to decide. Many On the other hand, many other people are people would agree that employees who place using employers' telephones in much less ex- several hours of personal long-distance calls pensive, but still pervasive ways. Employees per day are outrageously misusing their em- are human beings with concerns beyonu the ployer's facilities. Many of the same people workplace, and they sometimes have personal would think that an employer that doesn't al- business that must be somehow completed dur- low parents to call home each day is insensi- ing work hours. People with toothaches have tive to employees' needs. Reasonable behavior to call the dentist. People with car trouble have on both sides is somewhere between these ex- to call the mechanic to see if the work is done, tremes, but where should the line of "reason- and then call a neighbo: to ask for a ride home. able use" be drawn? Two local calls per day? Working parents need to know if their children Ten? One local call and one short long- distance have arrived home from school; indeed, par- call? ents in windowless offices may need to con- Often firms recognize a need to balance good sult the weather report first to know what in- management of the firm's resources with the structions to give their children. biblical injunction against "binding the One survey of Fortune 1000 firms estimated mouths of the kine that tread the grain." They that employees spent an average of 14 9 min- are also aware of their own interestthere are utes per day on personal calls (about 3 percent other productivity factors to consider in addi- of an 8-hour day), or the equivalent of 11/2 work tion to the cost of the telephone call. People's weeks of personal telephone calling per minds are clearer to focus on work if their per- year." The numbers may be suspect, since sonal problems are settled. Some calls simply they are based on estimates by personnel man- have to be made, one way or al Cher. It may agers, but they show that perception of a prob- be better for the firm's total productivity to lem is widespread. As mentioned earlier, some let people take care of personal buenecs dur- telecommunications experts have estimated ing a short break at their Ue1211.s than to require personal calling in the private sector to range them to wait in line at the pay phone. from 10 to 50 percent of calls.18 Organizations differ in their official policies of employee use of telephones. OTA interviewed telecommunications managers of sev- 'Employees Spend Over One and A Half Weeks of Job eral large firms. Some say flatly that office tel- Time on Personal Phone Calls Each Year, Nationwide Survey Reports," Sandford Teller Communications for Accountemps, New York, Aug. 30, 1984. "Tor example see "FCC Orders AT&T, Local Phone Farms "Judith Havemann, "Toll Callr Abused by U.S. Employ- to Lower Long-Distance Rate of Return," Wall Street Journal, ees," Washington Post, June 21, 1986. Aug. 8, 1986, p. 3. Ch 3Telephone Call Accounting 67

ephones are for business use only. However, late and process that information, becomespar- the view of one telecommunications manager ticularly important. For example, a number of was common, "We say company phones are long-distance calls from a department that has for company business, but actually we don't no out-of-state business might indicate that care about local calls as long as they keep it personal long-distance calls are being made. reasonable. "2D A few organizations had no Many calls after business hours might indicate problem with personal long-distance calls as that the security or cleaning crews are mak- long as they were of reasonable length and the ing use of telephones. The call-accounting sys- employee reported the call and reimbursed the tem may be programmed to produce a report organization. of calls to certain prefixes, for example, in And of course, regardless of official policy, many cities, all 976 number are assigned to there is a wide variation in the enforcement "audio text" services like "dial-a-prayer," of the policy. Offices that do not keep track "dial-a-joke," figure 11 shows a sample excep- of their telephone usage through call account- tion report of all calls over $5.00 in cost or 30 ing have little idea whether the policy is being minutes in duration. observed or not. The use of exception reports to find major Even firms using call-accounting systems offenders is effective because most peopleac- seldom find it cost-effective to evaluate every tually make few personal calls. Despite esti- call to see if it is official. The common practice mates of the "average" amount of time spent is to use the call-accounting system to gener- on the telephone, common sense and evidence ate "exception reports," reports that indicate from a few studies suggest that there isa wide unusual call patterns that might result from variation in personal behavior. For example, misuse. Here is where the ability of the call- an examination was made of 1,400 unofficial accounting system to track all calls by time long-distance calls (all to audio-text services) and originating telephone, and then to corre- made from the U.S. Department of Education in Washington. The Department has about 5,000 telephones, but two-thirds of these audio- 2°Interview with telecommunications manager of a financial text calls came from just 41 telephones; 45per- services organization, December 1985. cent (650 calls) came from just 11 tele-

Figure 11.Sample Exception Report

Date: 03107184 Time: 11.48:33 Page: 1 Report Period: 2/24 - 2/29 Calls $5.00 or 30 min

Date Time Duration Charge Number called Facility City State

Extension:226 User:Tom Best 2/24 09:44 00:31:30 S 9.45 714-964-6732 WATS Anaheim CA 2/27 13:57 00:36:30 .10 Incmg Extension: 1466 User Joseph Carr 2/24 10:46 00:31:45 $17.00 616-983-5555 WATS St Joseph MI 2/24 15:19 00:35:30 0.10 Incmg 2/27 12:37 00:17:32 5.99 212.888-1357 C/O New Yorl: NY

Extension:1533 User:Ellen 2/25 09:54 00:22.26 S 8.12 312-685-7863 WATS Evanston IL 2/27 15:33 01:34:12 0.00 616-429-8589 LOCAL St Joseph MI 2129 13:42 00:45:01 4.89 702-734-4444 MCI Las Vegas NV

80 68 The Electronic Su, -visor: New Technology, New Tensions phones." Similarly, a 1984 study by the De- use of the reports. When employees are made partment of Energy found wide variation in awav that records are being kept, their per- the level of unofficial calling in different sonal use of telephones tends to go down. The offices.22 Telecommunications managers have reverse can also be true when they know there found that by relying on "exception reports" are no records. One firm interviewed by OTA to seek out patterns of misuse, they are more removed its call-accounting capability for sev- likely to find habitual major offenders. This eral months while making the transition from is also considered fairer than closely scrutiniz- one telephone system to another. The total ing the telephoning habits of every employee. number of calls increased dramatically during that period, although it was impossible to tell Just the existence of a call-accounting sys- from which phones the calls were made. Once tem can have a deterrent effect on personal tele- the new system was running, each employee phone use, even if management makes little was sent a memo containing a reminder of company policy and a list of the previous weeks' "U.S. Department of Education, Office of the Inspector Gen- eral, letter report ACN 11-40100, July 25,1984. The 1,400 calls calls from his or her telephone. Nothing fur- were made over a 12-month period from November 1982 to Oc- ther was necessary to cause the volume of call- tober 1983. ing to drop to its former levels." 27U.S. Department of Energy, Office of the Inspector Gen- eral, "Review of Abuse of Long Distance Telephone Service (FTS) in the Department of Energy," DOEIIG -0217, Mar. 22, "Interview with telecommunications manager, financial 1985. services organization, December 1985.

ACTIVE COST CONTROL METHODS Active methods of telephone cost control can a different price." Decirting which is the make use of technology to reduce the cost per cheapest method of calling a given destination telephone call and also to prevent unnecessary at a particular time of day can be a complicated calling. Techniques include least-cost routing, problem, one which would be inconvenient for calling restrictions or blocking, authorization an employee to solve every time he or she codes and levels of service, and timed signals. needed to make a long-distance call. The least- cost routing feature makes choosing the right Least-Cost Routing route "transparent" to the user. The employee merely dials; a compnter prcgrani searches Least-cost routing automatically connects through a table of available lines, times, and a call with the least expensive line available. rates to pick the least costly route for each call. Many modern PBXs are equipped with this feature, which requires a computer program At busy times of day, when the cheapest fa- to "hunt" through the available lines to find cilities are all busy, several options are avail- one appropriate for the call being placed. For able. The system may automatically queue the least-cost routing to be most effective, the firm call, and signal the user when a line is free, a, should first study its telecomunication needs the system may signal the user to try again to make sure that it has access to the proper later. Yet another option is to give the caller assortment of different carriers (AT&T, MCI, a warning tone, indicating that the low-cost and Sprint, for example) and different types lines are all busy. If the call is urgent, the user of facilities (WATS lines, leased lines, and di- can hang on, and the call will go through at rect dial) to match its calling pattern. a higher cost.

There are at least 40 different ways to call "Interview with Edward Horrell, Mitchell & Horrell, Inc., from New York to Richmond, VAeach with June 24, 1986.

8 Ch. 3Telephone Cad Accounting 69

Calling Restrictions to make lc-ig-distance calls may do rio from any telephone in tile system. The code is still valid The call-blocking feature allowsMC* .2 telephone systems to be programmed to restrict if the user moves to another office. On the other the type of calls made by certain telephone-, hand, I .1 unauthorized person cannot make or certain callers. For example, the switch may calls on any phone, unless he or she discovers be programmed to block any calls to exchange a code. In addition, authorization codes can "976" in order to restrict theuse of "audio be easily en& ed. For example, if the user's text" (weather, time, dail-a-juke) calls. Tele- job changes '.o require a different level ofserv- phones in departments that do not deal with ice, or if it is discovered that in unauthoriz-ed the public can be programmed to make only person is using a code, the old authorization code can be canceled and a new one issued in in-house calls. Telephones of workers withno out-of-town business can be programmed to short order. provide only local service. Certain telephones A disadv-mtage of authorization codes is can be authorized to make long-distance calls that they require the user to dial five toseven only via the ..,vest cost service, where other additional digits at tre beginning of each long- may be able to override the least-cost routing distance _all. This is annoying to most users feature nd make a long-distance callm'en and a real hardship for those who need to make when low-cost lines are busy. many calls in a day. The State of New York has overcome this problem by makinguse of Authorization Codes and the speed dialing feature of modem telephone Lovels of Service systems. Speed dialing allows the user to store P "st of frequently called long-distance num- Telephone systems can also be programmed oers (in some systems up to 60 numbers per so that no telephone will put through long-dis- user) in the telephone's memory. Eachnum- tance calls unless preceded by an authoriza- ber is then referred to by a two-digit code- tion code that should be known only to people When coffin::: any of these numbers the authorized to make calls. The code also allows user the system to chrzge the call to a particular must still dial the authorization code in full, but only dials the U.).figit code to reach the person or account, which is usef,i1 for cost al- frecp...mtly called r location purposes. .oer. Thus the total number of digits dialed per call is reduced toa man- Authorization codes can form the basis for ageable number. different levels of service. Workers with a need Becaus a authorization codes are generally to make international calls can b assignedan used to allocate costs, they authorization code that permits such calls. are usually used in conjunction with a call-accountingprogram. Those who only need to make calls withinone State can be given a code that all ivis thismore Timed rignals restricted level of calling. Similarly, the telephone system can be programmed tc allow Several firms and government organizations some classes of users to make calls by the use timed signals to remind callers of the time lowest cost service only, while other usersmay they are spending cm telephone calls. Telephone have an "executive overriie" staus that al- systems can be programmed, for example, to lows calls on higher cost lines if low-cost lines give users a tone after some predetermined are busy. The State of New York, for exam- reriodsay 4 or 5 minutes. Whileno penalty ple, has 26 different levels of service toaccom- accrues to the user who continues to talk be- modate needs of different classes of users.m yond this point, the feedback is often useful in reducing the average length of calls." Peo- The advantage of authorization codes is that they are independent of the individual ple sometimes have no idea how long they have been talking, and a 5- minute warning reminds phone instrument. A person who is authorized them that long-distance calling does coat *Interview with Peter Arment. State of New York, Division money elTulecommunications, September 1988. The 26 levels ofse v- dude both voice 1114 data transmissions. 82 70 The Electronic Supervisor. New Technology, New Tensions

STUDIES OF UNOFFICIAL USE OF GOVERNMENT TELEPHONES Parsonal use of Federal Government tele- Some agencies' studies have also tried to phones is not only ..ontrary to "company estimate the loss to the government in terms policy " it is illet, ae Federal Information of wages paid for time spent in personal call- Resources Management Regulation (41 CFR ing. The Department of Energy, in its study 201-38.007) specifically forbids the use of FTS of phone use, added to the $3 million per year or other government-provided long-distance cost of personal calls, an additional $6 million service for personal reasons, and provides for per year for lost wages. This was calculated fines, suspension, or dismissal of offending em- by multiplying the total minutes of (tans dur- ployees. Furthermore, & CFR 735.205 prohibits ing work hours (8 a.m. to 12 r..mai and 1 p.m. the use of government property generally for to 5:30 p.m.) by the average wage rate for the personal recsons. Some employees and contrac- department." The figu"e is probably inflated tors have -'been indicted under Title 18, Sec- as the calculation does not consider that em- tion 641 -ublic Money, Property, or Records), ployees could have made calls during their which provides criminal penalties for the theft breaks or other slack periods when they had of a thing of value" from the government. For no other work. However, in the ewe of major example, 4 employees and 25 contract employ- offenders, for example a person running a ees of the Department of Energy were indicted private business from a government phone, for personal telephone use in 1981. Under a pre, wages lost to the government could be signifi- t'ial diversion, the defendants repaid the gov- cant. In the Richland, Washington case men- ernment $38,487.2" tioned above, the 29 defendants were required to repay lost wages along with other fines and Despite the illegality, government employ- the cost of tocalls themselves. ees use their employer's telephone for personal reasons just as much as private sector employ- The Federal Government is a major user of ees do; some would say more so. A number of telephone wrvices. Its Federal Telecommuni- atudies conducted by indivIdual departments cations System (FTS), establish( ' in 1K-3, pro- in t' .e past few years ha. found that an esti- vides voice and low-speed data telecommuni- mated 30 to 60 percent of long-distance calls cation services throughout the United States, are of an unofficial nature. A more recent the U.S. Virgin Islands, and Puerto Rico. The study, part of a coordinated multi-agency au- system contains about 1.3 million telephones, dit, - -carted personal calls made up an average 1,60U Meal switchboards, 52 major switching of X:, 4aercent of off-network calls sampled in centers, and 15,000 long-distance trunks. the Federal Telecommunications System (see About 88 percent of the long-distance FTS below for a description of FTS off -network service is throurh let cod AT&T facilities, with calls). About 20 percent of calls sampled on the rest provided by GTE/Sprint, MCI, and the government's commercial lines w'.ve other carriers. personal." The General Service! htdminis' ration (GSA mane. oyes FTS and supplies telepl+one service to most Federal agencies as required by the "U.S. Department of Energy, Office of the Inspector Gen Federal Property and Administrative Services eral, "Review of Abuse of Long Distance Telephone Service Act of 1949 (40 U.S.C. 481). Some of these are (FTS) in the Department of Energy," DOE/10-0217, Mar. 22, 1985. "full service agencies," in that they resew "Judith Havemann, "Listen Up Government Workers: You than telephone service through GSA. Otheis, May Be Allowed One Peons Call," Washington Post, Sept. 11, 1986. See allo President's Council on Integrity and Eft dency, "U.S. Jegartme- of Energy, Office of the Inspector Gen- "Consolidated Report on Federal Telecommunication- 'stem eral, "Review of A...v of Long Distance T. '-rvhone Service (FTS) Utilization," prepared by the General Services Magni& (FTS) in the Department of Energy,' DOE/i. )217, Mar. 22, tration. Office of the Inspector General, Mar. 18, 1987. 1985. Ch. 3Telephone Call Accounting 71

including Departments of Energy, Justice, on only a 20-percent sample of FTS calls one Commerce, and the Veterans Administration, call in five. GSA uses the information in this procure their own local telephone service in the sample to calculate quarterly telephone bills Washington area, and rely on GSA only for for each department, and also sends copies of lung-distance services. These are knownas "ex- the call records to agency telecommunications clusive use agencies. " managers. Due to the large backlog of work There are three types of "long-distance" in calculating telephone bills for all theagen- calls: cies, AMA reports have been sent to theagen- cies 3 to 6 months after they are collected.w A commercial call is made by dialing 9 from a government phone and accessing Local calls are billed to each agency monthly the local telephone system; intercity calls through GSA's Telephone inventory Account- made after dialing 9 are billed at theregu- ing System (TIAS). Commercial toll callsare lar commercial toll rate. billed to agencies directly by the long-distance FTS on-net calls are intercity calls be- telephone companies. tween two government telephones. In FTS off -net calls, the originating tele- The PCIE Review of FTS U,ation phone is part of FM but the receiving tele- The previously quoted figures on persona phone is not. In such a case, the call would use of FTS came from an audit recently con- travel ..s far as possible on FTS, and then ducted by GSA at the direction of the Presi- would go off-network and use commercial dent's Council on Integrity and Efficiency lines to reach its destination. (PCIE). It is the most recent attempt by the Off-net calls are generally more expensive than Federal Government to study telephone useon FTS calls; commercial calls are mostexpen- a governmentwide scale. The study involved a sive. On many government telephones, a fra- statistical study of telephone use at 16 agencies ture called Automatic Route Selection (ARS) (basicallyan analysis of exception reports). In automatically transfers commerical calls to the addition, GS A conducted a call-back audit ofa lower cost FTS network whenever possible. sample of telephone calls from 14 agencies.' This is similar to the "least cost routing" feature discussed earlier. For purposes of thi, auditing project, GSA provided a sample of call detail listings of off- FTS handled about 300 million calls in 1986 network and commercial long-distance calls to ar 4 cost the government about $500 million. the Inspector General of each participating Off -network calls represent about 65 percent agency. These call records came from the regu- of the calls and 69 percent of the cost. In addi- lar 20 percent sample of FTS calls and from tion, commercial toll calls cost about $15 mil- telephone company billings for commerical lion in 1986. calls. Personnel from the Inspector General's For FTS calls, GSA pays the long-distance office called back the numbers listed on those carriers, bills the participating agencies records to determine whether anyone at the quar- destination telephone engages or has engaged terly, in advance, and then adjusts for actual in business with the department or agency usage. Usage figures for long-distance calls are making the call. Calls were then classified currently collected by the telephone industry, as using the Autom .cad Message Accounting "official," "unofficial," or "unresolved." (AMA) systems of local telephone companies and long-distance carriers. The detailed callrec- ce' d includes the telephone number of th3 origi- "U.S. Department of Education, Office of the Inspector Gen- nating user, a billing account code, the date eral, letter Report No. 11-40100, July 25, 1984. and time of the call, the telephone number of "President's Council on Integrit,y and Efficiency, "Consoli- the called party, and the duration in minutes. dated Report on Federal Telecommunications Systems (FTS) Utilization," prepared by the General Services Administration, Currently, car detail information is collected Office of the Inspector General, Mar. 16, 1987. 72 The Electronic Sup ryisor. New Technology, New Tension

If the potential use of call accounting in gov- tion. For example, Representatives Don Ed- ernment raises questions of privacy, fairness, wards (Chair, House Subcommittee on Civil and enforcement, a call-back audit, like die one and Constitutional Rights) and Patricia conducted by PCIE, was even more controver- Schroeder (Chair, Subcommittee on Service) sial. Announcement of plans for the audit (in wrote to the Office of Management and Bud- late February 1985) was greeted with immedi- get (OMB) requesting further information on ate statements of concern by civil libertarians, the audit: -anion leaders, and others. 1. What is the source of authority for the pro- One concern was the potential for "selective posed monitoring scheme? disciplinary action against workers considered 2. How would the monitoring program be con- undesirable by agency m anagers."" When ducted, who would conduct it, and what the PCIE study was announced, fear was ex- types of telephone calls would be examined? pressed that, if the personal use of telephones 3. How long would the program last? 4. What types of data or analyses would the was as widespread as GSA believed, then program yield? nearly every employee was to some degree 5. How would the resulting data and analysis guilty. This being the case, telephone audits be used and who would have access to could be used as a potential weapon against them? whistleblowers or other dissidents. One arti- 6. What measure.° would be taken to limit dis- cle noted that in 1982 "investigators in the semination of the data and analyses? Environmental Protection Agency secretly ex- 7. What guarantee is there that the program amined the agency's long distance phone will not be used to discourage whistle- records to determine whether Hugh Kaufman, blowers, to stifle dissent, to limit news me- a government employee who had disclosed in- dia access to information, or for other po- formation the ' lqd to the removal of most of litical purposes? 8. What would happen to the data and analy- the agency's top officials, had talked with news ses after the initial analysis is completed? organizations. "'e Our Subcommittees would like to be assured Union leaders also expressed the fear that on these points, and any others that may be information from the audit could be used to raised as additional information comes to harass union members." Nor were critical light, before any monitoring begins." comments limited to labor groups. Bun Bray, In his reply, Jose-h R. Wright, Jr., Deputy president of the Federal Managers A880Citi Director of OMB, said the purposes of the tion, called the audit "another little deal to pun- lob Feder di employees," and asserted that the PCIE review were "to reveal patterns of misuse of the Federal long distance telephone sys- savings from the program would prove "insig- tems" and to develop "recommendations for nificant and minimal." "If they really want to systemic improvmients in the management of save money, they ought to take those resources and check out General Dynamics and some of these systems." those other defense contractors that are rip- Wright also gave specific answers to ques- ping off the taxpayer," Bray said.'" tions raisedinthe congressional letter. Por These topics aiong with other civil liberties example. with regard to limiting dissemination questions, also attracted congressional eaten- of the audit data he -aid: ...lit) will be limited to the staff of participating Inspectors General. The bulk of the "Bill Montague, "Planned Phone Audit Brings Blast From Several Groups,"Federal Times,Mar. 25, 1985. data will be placed in audit workpapers and "David Burnham, "U.S. Phones Raise Issue of Privacy: New used to suppIrt audit findings which are sum- Equipment Would nevi& Detailed Records of Calls,"New Yana Times,Mar. 17, 1985. "Bill Montague, "Planned Phone Audit Brings Blast From "Mar. 13, 1988 letter from The Renewable Don Edwards and Several Groups,"Federal77nm Mar. 25, 1985. The Honorable Patricia Schroeder to Joseph IL Wright, Jr., Dep- "Ibid. uty Director, Office of Management and Budget.

85 Ch. 3-Telephone Call Accounting 73

mary in nature. Any data which require ini- into guidelines for the audit, the two Members tiation of an investigation will be treated as of Congress suggested a number of principles evidence and will be accordingly protected. to be included in those guidelines. Data which may bring together names and numbers will be filed as part of the IGs' Pri- A detailed memorandum of "Guidance on vacy Act systems of records. the Privacy Act Implication of the PCIE Re- On the question of whistleblowers or the sti- view of Federal Telecommunications Systems fling of dissent, Wright reassured the commit- (FTS) Utilization" was completed in August tees that: 1985. This document included many of the principles outlined in the Edwards-Schroeder . . . long distance calls to news media, congres- letter, as well as safeguards discussed by sional offices, public interest groups, etc., will Wright. These guidelines were adopted as the be considered business calls for the purpose PCIE progressed through 1985 and 1986. of this review.. . . The review is being performed by statutory Results of PCIE Audit Inspectors General who have, among other duties, responsibility under their own ambl- The PCIE audit, conducted by the Inspec- ing legislation to protect whistleblowers. In tor General's office of ea :h participating addition, the Civil Service Reform Act clearly agency, made use of a sample of call detail prohibits the kinds of activities described in your question . . . while there are no absolute records supplied by GSA. They included both guarantees that ail persons will act properly, "off-network" ITS calls and commercial tele- there are ample procedures to deal with those phone calls. Researchers called each 4estina- who are found to have committed such pro- don number in `ate sample to detei mine hibited personnel practices. whether anyone1, that location ergageu business with the department or agency mak- Representatives Edwards and Schroeder re- ing the call. At the conclusion of a conversa- plied to Wright. indicating their appreciation tion with the person or persons at the destina- that the PCIE "is sensitive to preventing un- tion, calls were then classified "official," warranted disclosures of information collected "unofficial," or "unresolved." in any audit and avoiding invasions of privacy." Drawing on concepts in Wright's letter, Table 9 shows results of the STS off-network and their own sense of proper elements to go sample. The weighted average (based on the

Table 9.-Results of FTS Intercity Off-Network Call Sample by Agency

Estimated Unofficial Traffic Agency calls (%) minutes (%)cost (%) 1. Department of Agriculture 30.5 26.5 23.8 2. Department of Commerce 25.5 40.0 37.9 3. Department of the Interior 29.5 40.4 35.8 4. Federal Bureau of Investigations 26.5 30.0 25.1 5. Department of Labor 45.5 45.1 44.8 6. Department of Treasury 42.0 45.3 41.1 7. Office of Personnel Management 36.5 41.1 38.9 8. General Services Administration 39.0 49.7 47.5 9. Environmental Protection Agency 29.0 23.3 22.2 10. Small Business Administration 27.5 39.8 34.9 11. Department of Health and Human Services 35.0 28.7 25.7 12. National Aeronautics and Space Administration 41.0 48.3 43.2 13. Department of Housing and 'lrban Development 27.0 36.5 33.2 14. Department of Education 28.5 41.0 36.8 Simple average 33.6 36.4 33.3 Weig;:ted average 33.6 36.4 33.3 SOURCE: President's tAtuncil on blowfly and Efficiency, "Consolideled Report on Federal Telecommunications System (FTS) Utilization," prepared by the General SSMC'S Administration, Mar 16, 1987.

63-982 0 - 87 - QL: 3 -- 3 'se z.

74 The Electronic Supervisor: New Technology, New Tensions

Table 10.-Results of Commercial Long-Distance Call Sample by Agency

Estimated Unofficial Traffic Agency' calls (%)minutes (%)cost (%) 1. Department of Agriculture 10 3.3 2.3 2. Department of Commerce 26 31.1 25.5 3. Department of the Interior 6 1.0 .6 4. Department of Labor 40 49.3 40.2 5. Department of Treasury 14 11.3 8.3 8. Office of Personnel Management 34 17.2 15.3 7. General Services Administration 26 14.4 3.6 8. Environmental Protection Agency 18 6.2 2.6 9. Small Business Administration 46 40.6 42.5 10. Department of Health and Human Services 22 15.8 15.1 11. National Aeronautics and Space Administration 14 18.5 12.4 12. Department of Housing and Urban Development 36 36.8 35.8 13. Department of Education 44 28.6 25.3 Simple average 25.8 21.8 14.6 Weighted average 19.9 15.8 11.0 Federal Bureau of Investigation cid not participate In the audit of com-ercial telephone calls. SOURCE: President's Council on integrity and Efficiency, "Consolidated Report on Federal 'Telecommunications System (FTS) Utilization," prepared by the General Services Administration, Mr 16, 1957.

size of agencies) of the number of unofficial simple extrapolation. from the audit would sug- calls found was 33.6 percent. These calls made gest that personal calling cost the government up 36.4 percent of the total telephone tir.e sam- around $118 million in 1985. pled, and 33.3 percent of the total cost of the calls sampled. While this extrapolation a rough estimate, there are a number of reasons why it may Table 10 shows that use of commercial lines offer a distorted picture of personal telephone was quite different. The number of calls v.-. as use in the government. For example, calling smaller, with a weighted average of 19.9 per- patterns in Washington may be quite differ- cent of the sample being classified as unoffi- ent than those in other areas of the country, cial. In addition, these calls made up only 15.8 so that personal use stab: Aire found in the au- percent of the total time of commercial calls dit should not be applied to the entire Federal sampled and only 11.0 percent of the cost. This telephone bill. Locations outside the Washing- would suggest that many of the personal com- ton area were not included in the PCIE audit. mercial calls were short duration or low-cost However, two agencies, the Departments of calls. Defense and Energy, sampled some areas out- An estimate of the cost of personal calling side Washington at the same time as the PCIE to the government could be developed by com- audit. These agencies found unofficial use rang- paring the cost of the personal calls in the sam- ing from 40 to 50 percent, suggesting that per- ple to the relevant portion of the government sonal use in Washington may be the same or telephone. For example, 33.6 percent of the a little lower than elsewhere in the country. $345 million spent for FTS off-network calls However, the results are Lot strictly compara- is $116 million, 11.0 pereent of the $15 million ble because the sampling technique and study spent on commercial calls is $1.6 million. Thus, methodology were different.

87 Ch. 3Telephone Call Accrunting 75

FUTURE DIRECTIONS FOR GOVERNMENT TELEPHONE MANAGEMENT Need for New Policies equipment with call accounting, call blocking, and other modern features. Developing and Most of the questions raised by the audit implementing a governmentwide policy may still remain open. The exchange of letters be- become increasingly difficult as the system between Congress and the Office of Management comes more fragmented. GSA, as the govern- and Budget resulted in modified guidelines for mentwide telecommunications manager, is cur- the PCIE audit conducted in 1985-86. In addi- rently trying to revise policies related to tion, OMB issued guidelines related to han- long-distance telephone use. dling of call detail recur `Ji under the Privacy Act. However, the question of the basis for fu- A number of major questions need to be decided on a long-term basis. How will personal ture permanent guidelines on the management use be defined, and what level of personal use of the government's telephone must still be re- will be tolerated? Will there be a continuing solved. role for audits, such as the one recently com- The government has been collecting call de- pleted? Will local calls be included in future tail information for long-distance calls for a audits? What will be the policy toward long- long time. The recent audit relied on no new distance calls to suburban areas? What can be technology. A similar audit could have been done to ensure employees prop& access to the conducted at anytime in the past 10 years. As press, to union representatives? Can a suffi- mentioned above, GSA provides its 20 percent cient level of protection be given to call- sample record tt, all government agencies. al- accounting records to prevent their misuse? though the agencies do not always use this What role, if any, will service observation or detailed call information on a regular basis. The other types of "listening in" play in civern- government has used this information when ment telephone management? What alterna- there is sufficient need or motivation. For ex- tive kinds of management technic des are ap- ample, examination of long-distance telephone propriate for use on government telephones? records was an important element in the in- These questions are addressed in the follow- vestigation of Department of Commerce em- ing sections. ployees suspected of using "inside inform* tion" stock trading deals. A number of employees were required to reimburse the government for long-distance calls as a result of Establishing a Policy on Personal Use that investigation." GSA has been struggling to develop a suc- The opportunity to create a usable govern- cessor to the current policy of "no personal ment policy on telephone use is especially im- calls." A number of alternatives might be con- portant in the light of the development of a sidered. This section will discuss only long-dis- new long-distance system to replace FTS, and tance cads, which are the main focus of con- in the light of plans of a number of agencies cern. Local calls will be considered later. to take control of their own telephone manage- One approach would be to allow personal ment. Several, Mu) the Departments of Trans- long-distance calls at the government's ex- portation and State, have sirezdy begun pro- pense under certain circumstances, for exam- curing their own local service for their ple in case of accident or illness, to check on headquarters offices. In most cases, these babysitters or transportation arrangements. agencies have also purchased or leased new The prthlem here is to create a list that elimi- telephone equipment, including switching nates frivolous calling, but still covers all rea- "Join II. Bony, "Than Cowman Employon Find for sonable drcmostances, including some as yet Fronting From Data," Mahington Post, Jun 13, 1966, p. 1. unforeseen.

8 S 76 The Electronic Supervisor. New Technology, New Tensions

Another approach would be to allow work- calls from the telephones on their desksas long ers to make private long-distance calls on FTS as the calls are bill to a home number or per- on a cost - reimbursement basis, as is the pol- sonal calling card. This approach is quite fea- icy in some private firms, universities, and sible now, given the near universality of call- State agencies. The problem with this ap- ing cards. However, it may be considered proach is the possible bookkeeping burden that "legal, since current regulations prohibit the could be created for an organization as large personal use of the government's telephone as as the Federal Government. It would be nec- well as FTS. Billing calls to a home number essary to establish a method for identifying is technically possible from many government personal calls, billing employees, and collect- phones and could be made possible from the ing money. Organizations that allow reim- others. The costs to government of such a pol- bursement tend to be small offices with a col- icy would be primarily in terms of employee legial atmosphere; identification of personal time (for calls not made during lunch period calls is basically an "honor system," reinforced or break). by supervisory review of monthly call records. Usually, a printout of calls is circulated among Because many personal calls may be calls the staff each month so that each person can home made by Federal workers traveling it initial his or her pc sonal calls. While this ap- of town, GSA has advanced a proposal tc al- proach makes it possible for employees to use low each Federal worker one call home per day the telephone freely, and provides reimburse- of travel. This practice is held to be typical of ment to the employer, it does not protect pri- personnel policies in private businesses. GSA vacy. Not only supervisors, but everyone in has estimated that the cost of such a policy the office typically sees the printout and could, could be as high as $100 million per year, based if they were interested, make note of who called on the total travel days of Federal workers. whom. It is not possible to tell how many personal calls are currently made by traveling workers, Using a call - accounting system, it would be thus it is not known how much moving this possible to develop a private printout or "tele- particular type of call from the "prohibited" phone bill" for each worker's telephone. How- to the "permitted" category would affect the ever, identifying personal calls might still be level of personal use of the system. a problem. Workers would be "on their honor" to claim personal calls; those with a poor memory or an underdeveloped sense of honor would Exception Reports and Personal Use still make calls at the government's expense, Telecommunications managers in many pri- unless some sort of regular audit were made. vate and State government telephone systems Collecting payment for calls would require have found that the use of "exception reports" the creation of an administrative structure and is the most effective way to discover patterns would generate costs that would have to be of personal use. The computer software is used passed on to the users, thus re' sing the cost to select and report on calls that have a high of the calls. Presumably this would be little likelihood of being unofficial, for example, calls different from other instances (bookstores, to audio text numbers, calls at unusual times cafeterias) where government employees pay or to unusual destination areas, long or reoc- cash for a service; however, in most cases these curring calls. The ability to use comput Jr soft- i re usually provided by contractors rather than ware to identify only larger instances of pos- directly by the government. It is not likely, sible personal use would Cleo seem the most in any case, that agencies will want to go into cost-effective approach. Call-back audits, if it the "telephone business" on a regular basis is decided they are appropziate, could than be for their employees. conducted in those specific cases where a pattern of prohibited personal use was suspected, Another approach would be to allow govern- rather than as a general approach to telephone ment workers to make personal long-distance management. Ch. 3Telephone Call Accounting 77

Use of exception reports requires decisions Of greater importance for the government about the thresholds below which possible per- will be establishment of clear guidelines for the sonal use can be tolerated or at least dis- handling of exception reports and other com- regarded. This is generally not a problem in puter-based call records to protect the privacy the private sector, where it is a management of employees and to prevent any possible mis- decision, tempered perhaps by the "corporate use of these records by supervisors or others culture." But in the Federal Government, there within the government. is at least the philosophical difficulty caused by the fact that any personal use is not only contrary to policy, but illegal. Even if some Future Use of Call-Back Audits tacit threshold of "reasonable personal use" were implemented, questions of equity might One major question in the future of the gov- shill be raised. People who were caught exceed- ernment's continuing telephone management ing the threshold could claim they are being is how or whether call-back audits will be used treated unfairly because "everyone else" is also on an ongoing basis. One major purpose of the guilty to some lesser degree. recent PCIE pilot study was to provide a benchmark of current telephone system usage. Many agencies are not currently equipped The primary reason, however, was to develop ) manage their telephone systems using ex- methods for detecting personal use in the fu- eption reports. While some agencies may be ture. There will likely be strong arguments to making regular use of the call detail records make call-back auditing a regular feature in provided by GSA, others are not. One Depart- Federal telephone management. ment of Education study noted that the printout exceeds 1,000 pages per month; there is The PCIE pilot study was coordinated by insufficient staff to study the printout on a GSA, but the actual audit was done by the In- regular basis; in some cases the reports have spector General of each participating agency. remained in unopened boxes for months." Each of these agencies is issued its own report Similar complaints have been voiced by De- and is developing its own action steps to im- partment of Energy and the General Account- prove management of its telephone system. ing Office. A envie listing of detailed call While GSA is the governmentwide manager records is generally of little help as a telephone of FTS, agencies have considerable discretion management tool. The information requires in correcting their own problems. further computer sortingby extension, fru- quently called numbers, long calls, etc., in or- It is possible that within a few years, PCIE der to be of much use. Several agencies are will want to conduct another multi-agency making use of call detail information from study in order to compare the results to the GSA, from their own telephone systems, or recently completed study. Even if this is not directly from telephone companies to develop done, a number of individual agencies may de- their own exception reports. cide to undertnke studies on their own, using the techniques developed in the PCIE pilot C SA is currently planning to make call de- study. tail information from its 20-percent sample more useful to agencies by providing excep- Even more likely is that agencies will want tion reports to agencies using FTS service. to use a modified version of the call-back au- GSA is also investigating ways to make call dit on a regular basis to complement their ex- detail data available to some agencies in ma- ception reports. Agencies that create exception chine-readable form so they can more easily reports to identify patterns of "suspicious" use their own computers to sort the informa- calls (e.g., unusually high number of calls, un- tion and develop management reports. usually long calls, calls to unusual area codes) °U.& Deportment of Education, Office of the Inspector Gen- could use the call-back method to determine eral, little report ACN 11.40100, July 25, 1986. whether or not they are legitimate.

9t) 78 The Electronic Supervisor. New Technology, New Tensions

This use of the method still raises some ques- The private sector telecommunications man- tions concerning the privacy of telephone agers interviewed by OTA, however, expressed users. However, it has the advantage of affect- no interest in accounting for local calls. They ing a smaller populationonly those whose saw excessive local phone use as a management calls are "suspicious," rather than the entire problem, but not a telephone management work force. This approach has some appeal, problem. One said, "That's up to the managers since it targets investigation at specific cases and supervisors. If their people are on the of suspected misuse rather than placing the phone all day, incoming calls can't come in, entire work force under suspicion. As men- so they tell them to keep it short. "39 tioned in an earlier section, it would be neces- Calls to nearby suburbs will require separate sary for the agencies or for GSA to establish consideration. These calls may be functionally guidelines for determining when a calling pat- the same as local ones, but are long-distance tern warrants investigation. calls due to the arbitrary boundaries of local service areas. In the Washington area, for example, employees who telephone from down- Local and Suburban Calls town Washington to their homes in Bethesda, An additional question arises about the on- MD, are making a local call, but those who call going policy with regard to local calls. The re- home to Herndon, VA, are making a commer- cent PCIE audit did not cover local calls, and cial toll call. A large number (the personal GSA does not collect detailed information on commercial calls discovered by the PCIE au- local calls on a regular basis and has no plans dit probably fit into this category. This would to do so (such information could be obtained account for the relatively short duration and from local telephone companies if especially low cost of calls shown in table 10many of requested and paid for. In their second letter them were brief personal business calls of the to OMB, Reps. Edwards and Schroeder ex- "checking on the babysitter" variety. The gov- pressed the opinion that the government ernment might decide to allow these employ- "should not procure or install any aervice, fea- ees one call or a reasonable number of calls per ture, equipment or system that would permit day, as discussed above. On the other hand, it to obtain call-accounting data on individual the goverr .gent might take the position that local calls." New equipment being purchased the cost of calling the suburbs, like the cost by some departments will have the capability of daily transportation to the suburbs, is the to record detailed information on local as well employee's responsibility, and might provide as long-distance calls. In some agencies, it is resonable means (pay phones or the use of call- programmed only to report time and du:lition ing cards) for employees to make these calls of local calls, not the destination telephone at their own expense. number. At the present time there is no govern- mentwicie policy on this topic, and individual Calls to Press, Union, or Public departments have to choose how to make use Interest Groups of this feature of the equipment. A governmentwide policy on this topic would be useful The policy of the PCIE pilot study was to treat any call to a press organization, public Local calls generally make up a much smaller interest group, congressional office, or labor proportion of telephone costs than long-dis- union as "official business" without further tance calls, but they are not completely cost free. Besides the "message unit" charge by the local telephone companies 17.5 cents per call in Washington, there is the cost for additional telephone lines if the volume of personal calls gets too high. In addition, :here is the ques- *Interview, May 1986, with the telecommunications man- tion of employee time. ager of an insurance firm.

91 Ch. 3Telephone Call Accounting 79

examination. However, as Reps. Schroeder and In the absence of clear policies andproce- Edwards point out, dures regarding who hasaccess to this infor- . ..deeming such calls "official" does not ad- mation and what they do with it, the possibil- dress.. .the fact that the Government will ity for misuse of the information is great. The be able to determine who has called whom. The challenge is to develop procedures that willpre availability of such information and the ready tect the privacy and first amendment rights means to analyze it remain in our minds the of Federal employees, without unduly hamper- most troublingand as yet unresolved ing investigation into cases of wrongdoing. aspects of the PCIE proposal. Eavesdropping and Service The question of how these calls will be treated on an ongoing basis remains open, and gives Observation rise to the further question of who will have The PCIE study did not inany way involve access to call-accounting records. the content of telephone calls. "Service There seems to be little doubt that the in- observation"listening in to employees' deal- formation in telephone call detail records could ings with the publicis practiced in Federal be used to identify and possibly harass whistle Government offices such as the Veterans blowers, people who speak with thepress, Administration, Internal Revenue Service, So- union organizers, and dissidents. While harass- cial Security, and other agencies withcustomer ment of such people is a violation of the Civil service responsibilities. Federal Information Service Reform Act, use of time and atten- Resources Management regulations require dance records, for example, to harass whistle Federal agencies to notify both employees and blowers does take place.° New technology the public that service observationmay take call-accounting software makes it much easier place, although there is no requirement to sig- than in the past to isolate the calls made from nal that an observer ison the line. particular telephones. Eavesdropping on other types of callsas a If call accounting anduse of audits must be strategy to reduce personal use of telephones used to protect the government investment in is of questionable value and legality. The in- its telephone system, itmay also be necessary stances where Federal managers have been to take positive action to protect the civil lib- found recording or eavesdroppingon conver- erties of telephone users. This protection might sations for any purpose have caused public in- take the form of clear and enforceable regula- dignation. One recent incident prompted the tions to protect the privacy of call-accounting introduction of legislation to specificallypro- records. This kind of transactional data about hibit listening in on or recording conversations telephone calls, while probably notas personal on the Federal telephone system, except in as the content of the calls themselves, does specified instances such as service observation warrant protection." Regulationsmay be programs (H.R. 502, 99th Cong., Federal Tele- needed to ensure that information remainsun- communications Privacy Act of 1985). der control of telephone systemmanagers, and Further, as is discussed in chapter 4, inat pert' -9 inspector generals, who need it toman- least one case in the private sector,a court has age the telephone system, and that data is not ruled that an employer's listening inon an em- available to supervisors or othermanagers who ployee's private telephone conversation is deal (Erectly with employees. eavesdropping and a violation of Title III of 1°Myrcm Peretz Glazer andd Penina Migdal Glazer. "Whistle- the Omnibus Crime Control and Safe Streets blowing," Psychology Today, August 1986; Donald E. Soeken, Act (le U.S.C. Sec. 2510),even though the em- "J'accuse," Psychology Today, August 1986. Also, interviews with Donald E. Seeks% February 1986. ployee was using a telephone normally included "U.S. Congress, Office of Technology Assessment,Federal in a service observationprogram. Government Information Technology: Electronic Surveillance and aril Liberties, OTA-CIT-293, (Wermngton, DC: U.S.(iov- Aside from the privacy and legal questions, ernmant Printing Office, December If A). eavesdropping would also bea costly and im- 92 80 The Electronic Supervisor: New Technology, New Tensions practical means of managing a telephone sys- programs aimed at increasing use of FTS and tem. It would probably be unfair, as well, since reducing use of commercial lines." only a few workers' calls would be affected. The other strategy is to make use of active cost control measures discussed earlier in this Other Methods of Cost Control chapter, for example "blocking" and "level of service access" features. Designing a telephone Finally, given that government agencies will system that is best suited for doing the gov- be procuring new telephone equipment and ernment's business might be preferable to services over the next few years, it would be scrutinizing call records to catch government prudent to consider the use of other methods workers who misuse a poorly designed system. of telephone cost control that may supplement, As mentioned earlier, modern telephone sys- or even replace, call accounting as a means of tems do allow for the programming of differ- controlling the costs of personal calls. It should ent levels of service, depending on the tele- be possible, given the right technological and administrative tools, to greatly reduce the phone needs of the end user. number of unwanted or unauthorized calls The government currently has some capability in this regard, but it is not used to a great One approach is education. Many employees extent. There are five levels of FTS telephone actually believe that calls on FTS (or an em- service available; these affect only the end ployer's WATS lines in the private sector) are free. The belief has developed over years in user's ability to make outgoing calls, not the which no accounting was made, where wide- ability to receive incoming ones. The levels are: spread personal use was tacitly tolerated, and 1. Standard Service: Can call only government where employers and agencies themselves have telephone in local area. treated telephone service as a free good. The 2. Jommercial Service: Can call government PCIE audit has been a first step in demonstrat- local telephones and commercial lines (dial ing that the government's policy is changing. 9). This includes the ability to access commercial long-distance lines after dialing 9. A first step in changing the calling habits 3. Government Service: Can call local and in- of government workers might be a nonpuni- tercity government telephones, but no com- tive educational campaign to inform employees of the new policy on personal calls once that mercial lines. policy is developed. This campaign should 4. National Service: Full access to both gov- include information on the methods of call ernment and commercial telephone net- accounting that are being used As mentioned works, but no international dialing. earlier, the mere knowledge that an effective 5. International Service: Full access to Gov- recordkeeping system is in place has a damp- ernment and commercial networks and inter- ening effect on personal calling. Educational national dialing. efforts should also fully inform government workers of the kinds of calls that are permitted, and inform them of alternative means of making necessary calls not permitted on the government telephone system. Education programs with respect to telephone use can be effective. Agancies have found that they reduced their employee's lev 42Sse for example, U.S. Department of Education, Office of of misuse (use of an expensive facility whei.. the Inspector General, "Report on Review of Telephone a cheaper one is available) through educational Changes," Sept. 11, 1981. Ch. 3Telephone Call Accounting 81

According to GSA, the distribution of these their advantages and disadvantages. Author- levels of service among government full-service ization codes are more flexible, since they de- agencies was as follows: pend on user needs rather than the location Level of service Main lines of a particular telephone he or she is using at 1. Standard...... 881 ( 0.6%) the moment. On the other hand, authorization 2. C o m m e r c i a l service. .. 532 ( 0.4%) codes require users to dial additional digits, 3. Government service. . 242 ( 0.2%) which can be a burden on those who make 4. National service.. . . 36,267( 25.6%) many calls unless speed dialing is available. 5. International service. .. .. 82,955( 58 7%) Assigning restrictions to specific telephones No level...... 20,443( 14.5%) Total would require that levels of service be changed 141,320(100.0%) every time offices are changed; this is possi- The "no level" category refers to lines that ble with new digital telephone systems. could not be classified or were installed before the classification went into effect." It may be that the government will want to pursue both options on an agency-by-agency As shown above, 84 percent of the lines are basis. In some agencies, the telephones may in the two highest levels of service. While it restrict long-distance calling except for holders is possible that such a large number of Fed- of authorization codes. In other areas where eral workers need access to the full range of authorization codes are unworkable, such as national or international telephone service, it teleservice centers where employees make is likely ': hat the actual number with a busi- many calls all day, telephones can be pro- ness need to make extensive long-distance call- grammed so that codes would not be needed. ing is lower than 84 percent. In the past, the Other agencies may decide to assign levels of relative difficulty of changing the level of serv- service to telephones rather than to an indi- ice on a given line once it was installed made vidual via authorization codes. Some possible it difficult to assign specific levels of service levels of service might in ude: to specific types of users. To ensure that telephone service would be at least adequate to Local Government Service: Can call only the needs of any group of workers stationed government phones in local area. in a particular office, lines have usually been Local Service: Can call local government assigned a fairly high level of service. phones and commercial (dial 9) lines. PBX or local telephone company Centres blocks However, new digital telephone systems will long-distance access. have greater flexibility and could allow more Government Long Distance Allowed: All extensive use of programmed levels of service above plus access to FTS. that can be changed as needed. Meaningful Off-Net Long Distance Allowed: All above levels of service that reflect the calling patterns plus ability to call nongovernment phones of government workers would have to be de- via FTS. veloped, and !ndividual agencies would need Commercial Long Distance Allowed: All to determine which of their employees need ac- above plus access to commercial long- cess to commercial lines, which to FTS lines, distance network. which to off-network destinations via FTS Range Authorization: Users in the previ- lines, and so on. ous three categories could be limited to Restrictions might also be assigned to the calls in certain area codes or zones within telephones themselves, or to their users, based the country. on authorization codes. Both approaches have International Calling: All above plus international direct dialing." "Information taken from President's Council on Integrity and Efficiency, "Consolidated Report on Federal Telecommu- nications System (FPS) Utilisation," prepared by the General "List developed by OTA based on categories in "Detailed Services Administration. Office of the Inspector General, Mar. Description" manual of State of New York, Office of General 16, 1987. Services, Division of Telecommunication. 82 The Electronic Supervisor. New Technology, New Tensions

Clearly, establishing better levels of service needs are made by the individual agencies, not will not eliminate personal calling. It is still by the Division of Telecommunication." possible for the holders of authorization codes The Federal Government's telephone system to make personal calls, to the extent allowed must be nationwide and must serve many by their assigned level of service. Similarly, different agency and program needs. However, should levels of service be assigned to specific telephones, it is possible that the people sit- it appears that the principle of reducing the ting near telephones capable of making long- government's exposure to risk of unauthorized calls by reducing the number of employees with distance calls will use them for personal call- access to full-service telephones is a valid one. ing. However, a call restricting approach could reduce the number of unofficial calls made on GSA, as the governmentwide manager of the the government's telephones. The State of New telecommunication system could greatly aid York found that many employees do not need agencies in making use of new information to make long-distance calls in the course of technologies in two ways. First, it could con- their work and therefore do not need authori- tinue, through its own research, to develop zation codes. Only 60,000 codes are in effect model methods for telephone system manage- (there are 200,000 employees at locations ment and communicate these to the agencies. served by the State telephone system). While Second, it could serve as a clearinghouse for codes are sometimes shared among members sharing innovative and useful approaches de- of a small workgroup, it is still clear that not veloped by the agencies. every employee needs access to long-distance service at work. Only about 20 percent of codes assigned have a level of service that permits "Interviews with Peter Arment, State of New York, Divi- sion of Telecommunication, September 1986. It should be noted full national or international dialing privileges. that the State of New York does use call accounting. Authori- The other 80 percent are restricted ti, govern- zation codes are the backbone not only of the service levels, ment calls, off -network calls within the State, but also of the billing system. Long-distance calls are charged to the authorization code, so the call is billed to the proper de- or commercial calls within selected area codes. pertinent, even if the call is made from another office, or even Determinations about employees telephone from a State office building in another city.

9j Chapter 4 Electronic Work Monitoring Law and Policy Considerations

9,3 CONTENTS

Page Part I: Why Is Monitoring an Issue? . 85 Part II: Framing the Debate 89 Purpose 91 Manner and Method ...... 92 Effects 96 Where Is the Future of Monitoring Headed? 97 Part III: An Overview of Applicable Law 101 The Framework for the Legal Analysis 101 The Concept of Employment at Will 101 The Legal Status of Public v. Private Sector Employees .102 Purpose 102 Manner and Method 107 Effects 110 Part IV: Concerns Not Addressed by Law 112 Purpose 113 Manner and Method 113 Effects .113 What Does the Future Hold? 114 Part V: Policy Options 115 Tables Table Ala Page 11. Unions With Positions Against Computer-Based Work Monitoring ... 86 12. Key Issues and Problem Areas in Monitoring Clerical and Customer Service Workers 87 13. A Framework for Addressing Electronic Work Monitoring 90 14. Factors Affecting Electronic Monitoring 97 15. A Framework for Addressing Electronic Work Monitoring 101 16. A Framework for Addressing Electronic Work Monitoring 112

9 ,' Chapter 4 Electronic Work Monitoring Law and Po likly Considerations

With perhaps the exception of marriage, no Between 4 and 6 million office workers have other institution in society is so pervasive as their work measured by computers, and many the employment relationship.' Roughly half millions more are affected by telephone call of an adult's waking hours are spent at work. accounting.' Proponenth of electronic moni- Societal attitudes regarding personal dignity, toring, principally the vendors of monitoring autonomy, and privacy, and the individual's equipment and software and some companies sense of self are tied to the workplace. Elec- that have installed monitoring systems, say tronic monitoringthe computerized collec- that monitoring provides employers with new tion, storage, analysis, and reporting of infor- tools for managing resources, allocating costs, mation about employees' productive activity' improving productivity, controlling quality, may be an increasingly important element and reducing waste, fraud, and abuse of em- in the relationship between employer and employer property. At the same time, however, ployee, and in the overall context in which work critics of work monitoring, principally some gets done. Decisions about monitoring may af- labor unions and civil liberties organizations, fect the setting in which citizens work, and may suggest that electronic work monitoring is de- therefore have a broader social significance. structive to the quality of work life and has damaging effects on workers' privacy, civil liberties, sense of dignity, and health. '"We have become a nation of employees. . . .For our gene ktion, the substance of life is in another man's hands." F. Tan- This chapter ie an attempt to provide the nenbaum,A Philosophy of Labor9 (1951) (as quoted in Blades, policymaker with a conceptual framework for "Employment at Will vs. Individual Freedom: On Limiting the Abusive Exercise of Employer Power," 67Columbia Law Re- addressing concerns over electronic monitor- view1404 (1967). Although the employment relationship is freely ing, now and in the future. It begins by ask- entered into, the need to earn a living, together with limitations ing why monitoring has become a policy issue, the individual's qualifications and the ease with which he or she can be substituted, often place the worker at a disadvan- and what is different aboutelectronicmoni- tage in determining the conditions under which he o: she shall toring that causes these concerns. It then ex- work amines which of these concerns are currently '"Electronic work monitoring," "electronic monitoring," or just "work monitoring" are terms used interchangeably in this addressed by existing law and which are not. chapter and elsewhere in this report to include computer -based Finally, it provides the policymaker with a work monitoring and telephone service observation. Telephone range of options from which to choose. service observation, although not typically computerized, is included because it is often used in conjunction with electronic work measurement techniques. 'See ch. 2.

PART I: WHY IS MONITORING AN ISSUE? Neither work monitoring, nor the applica- "scientific management." While early work tion of technology to measure work, is new. monitoring techniques were confined to pen- As chapter 1 of this report illustrates, the cil and paper tallies of performance, technol- detailed observation and recording of employee ogies such as the time dock, time stamps, and performance has been an integral part of Amer- cyclometers were applied pervasively to the ican ind,41:rialization from the mid-1800s measurement of work.' onward. During the early 1900s, work moni- 'For a more complete discussion of early forms of work mon- toring culminated in Frederick W. Taylor's itoring, see ch. 1 of this report.

85 8 N The Electronic Supervisor. New Technology, New Tensions

Given this long history of work monitoring, Table 11.-20 Unions With Positions Against two questions arise: Computer -Based Work Monitoring 1. why is concern about work monitoring Estimated among management, labor, and public in- Union membership Automobile, aerospace, and agricultural terest groups only now surfacing? and workers (UAW) 1,350,000 2. why is the application of electronics to an Communications workers (CWA) 550,000 old technique unique or different enough Electrical workers 1,000,000 Electronic, electrical, technical, salaried, and to cause this concern? machine workers (WE) 160,000 Federal employees (NFFE) 150,000 Simply put, the questions are "why now?" and Government employees (AFGE) 265,000 "what's new?" The first question is addressed Government employees (NAGE) 300,000 presently, while the second forms the discus- Machinists (1AM) 940,000 Newspaper guild (TNG) 40,000 sion in the remainder of this chapter. Office and professional employees (OPEU) 125,000 Postal workers (AMU) 320,000 Although the question of "why now?" has Railway, airlines, and steamship clerics no simple answer, it is posealle to point to three (13RC) 200,000 broad trends that have contributed to the Service employees (SEW) 650,000 State, county, and municipal employees emergence of monitoring as a policy issue: the (AFSCME) 1,100,000 computerization of office work, the computeri- Steelworkers (USWA) 1,230,000 zation of communications, and the rise of work- Teamsters (IBT) 2,000.000 Telecommunications workers (TIU) 50,000 ers' expectations and rights in the workplace. Treasury employees (NTEU) 120,000 The three types of monitoring considered in Typographers (ni) 80,000 this report computer -based monitoring, tele- 4.4ility workers (UWUA) 56,000 phone call accounting, and telephone service NOTE: Totals are rounded. SOURCE: Westin, Privacy and Otolityof Wcwkilfit issua.: in Employ**Monitoring. observation--illustrate the way in which these contractor report prepared for OTA, May 1966. Bureau of Labor Statistics, IWO; AFL.CIO estimates, supplementary inquiries M trends have propelled work monitoring into the individual unions, 19W. national policy arena. The computerization of office work is less tional TV news programs, a stream of news- than 15 years oldless than half of the work- paper and magazine stories, and several recent ing lifetime of a white-collar worker.6 Com- books. Although it has captured media atten- puterization has lead to rapid, fundamental tion, unions were the first principal critics of changes in the quantity, quality, and organiza- computer-based monitoring.' (Unions having tion of office work. As a result, conflicts are positions against computer-based monitoring emerging between management and labor over are listed in table 11). While the concerns of how work will be designed; who will have a say each union are not necessarily identical in de- in that design; and what the expectations of tails, a broad consensus seems to be that work employee performance, flexibility, and privacy speedups, enforced by close work monitoring, should be. The literature on the impact of auto- are bad because they create harmful stress mation on work is voluminous, and the reader among employees and also compromise the seeking more information is referred to another OTA report., Automation of America's Offices. 'During the comes of this study, OTA spoke with a variety of union representatives having an interest in work monitor- Computer-based monitoring is one example ing, including the American Fedwation of Labor & Congress of how the computerization of office work has of Industrial Organisations (AFL-CIO), the Communications lead to the recent emergence of work monitor- Workers of America (CWA), the Service Employees Interna- tional Union (SEIU),Nine to Five:The National Association ing as a salient issue. Since 1982, computer- of Working Women (9 to 6), American Federation of State, based monitoring has been the subject of na- County, and Munidpal Employees (AF9CME), the American Federation of Government Employees (AFGE), and the Graphic U.S. Canvass, Office of Tedmology Assessment, Automa- Artists International Union (GAIU). In addition, 9 to 6 and tise of America's Offices, OTACIT-287 (Washington, DC: U.S. AFGE were represented on this study's Advisory Panel. These Government Printing Office, December 1985), p. 8. unions, and many others, have taken a public stance against nba computer-based monitoring.

q1/4J- 411._ 4 Ch. 4Electronic Work Monitoring Law and Policy Considerations 97

quality of work provided to the public. As an Table 12.Key Issues and Problem Areas in issue of job stress, unions see work monitor- Monitoring Clerical and Customer Service Workers ingand particularly computer-based monitor- Key issues/problem aspects ingas linked to quality of worklife issues, 1. Fakness of work standards worker solidarity, job design, and health con- Do standards fairly reflect Me average capacities of cerns. Worker privacy is also emphasized as the particular work force? Will they create unhealthy stress for many employees? a concern of the unions, particularly the rights Do they account for recurring system difficulties and of workers to see what records are being col- other workplace problems? lected about their work performance. Do they include quality as well as quantity goals? Do they represent "fair day's pay for fair day's work? Unions represent only a fraction of the esti- Do employees share in productivity gains achieved mated 4 to 6 million monitored employees, but through new technology? 2. Fairness of measurement promise from a field study of 110 business, government, Do employees know and understand how the and nonprofit organizations conducted for measurements are being done? OTA, it is also possible to offer some observa- Can measurement system be defeated, impairing morale of those willing to follow the rules? tions on the attitudes of employees in general Do employees receive statistics on performance with respect to computer-based monitoring. directly and in time to manage work rate? The field study revealed that fairness in mon- Is relationship between quality and quantity communicated by supervisors when discussing itoring is a critical factor in clerical and cus- problems with performance levels? tomer service employee acceptance" The em- Do supervisors communicate clearly that they are phasis on fairness highlights the fact that both taking system/workplace problems into account? Are group rather than individual rates used when such process and substance are involved in how em- an approach is more equitable? ployees respond to electronic work monitor- Is there a formal complaint process for contesting the ing. In addressing process, employees seek gen- way work data is used? uine involvement in the design, testing, 3. Fairness In applying measurements to evaluation Are there meaningful recognition programs for superior application, and subsequent adjustment of new performance? office systems technology. Substantively, the Is work quantity only one of a well-rounded and perceived reasonableness of monitoring de- objective set of appraisal criteria? Does employee get to see and participate in pends on: 1) the fairness of the standards set, performance appraisal? 2) the fairness of the meastnement process em- Is there an appeal process from supervisor's ployed, and 3) the fairness of the way meas- performance appraisal? urements are used in employee evaluation. A Is ther a performance-planning system to identify and help performance problems? breakdown of these key issues is given in ta- SOURCE: Office of Technology Assessment 1987 ble 12. The computerization of communications (PBXs), for example, permits station message refers to the fact that the information moving detail recording (SMDR), a form of telephone within modern communication systems, such call accounting that records from which tele- as the telephone, is increasingly transmitted, phone a call was made, what access code was routed, stored, and processed in digital form used, where the call went, and how long it by electronic computers. Because of the com- lasted. The call-accounting system can then puterization of communications, the use of em- generate detailed, comprehensive reports of all ployers' telephone systems can be tracked with of the telecommunication activities of every greater precision and comprehensiveness than employee in a firm.' was possible previously. Computer software Telephone call accounting is an example of in some modern Private Branch Exchanges how the computerization of communications has placed work monitoring on the public pol- 'Alan Westin, and The Educational Fund for Individual icy agenda. A pilot study of unofficial use of Rights, Privacy and Quality of Work Life Imam in Employee Monitering, contractor paper praptied for OTA, May 1988; field study conducted during 1982-84, and updated at all 110 sites 'See ch. 3 for a more detailed discussion of telephone call- during 1985-88. accounting systems.

100 88 The Electronic Supervisor. New Technology, New Tensions

Federal Government long-distance telephone Finally, heightened public and work m expec- usage has been completed," and the implica- tations regarding privacy, health, and work- tions of this study for employee privacy caused life quality help to explain why monitoring has concern in both Congress and the press." Al- emerged only recently as a public policy issue. though most of these concerns were over the The period from 1965 to 1986 saw the growth manner and method by which this study would of concern over privacy in the public conscious- be carried out (and are therefore addressed be- ness," in the courts and legislatures," and in low), at least some of the concerns are over the the scholarly literature." The concern over purposes for which it might be used. One con- privacy daring the past 20 years largely tracks cern, in particular, is that call accounting might the introduction and proliferation of the com- be used to discourage whistleblowers, to sti- puter as a basic tool for the emerging infor- fle dissent or union activity, or to limit news mation economy. It is Da surprise, therefore, media access to information." Plans are be- that privacy issues have made their way to the ing made to audit Federal employees' long-die office environment, where computers have had tance calls on a regular basis, raising additional their most pervasive influence. As we will see, concerns about the vigilance with which pri- however, the concept of privacy may be inade- vacy concerns are addressed. The Office of quate to address most of the issues involved Management and Budget (OMB) has issued in work monitoring. proposed guidelines on compliance with the Privacy Act, in contemplation of a permanent At the same time as privacy became an im- call-accounting capability in executive nee- portant theme in public policy, there were ris- cies." In the private sector, where about ing medical, media, and public concerns about 30,000 call-accounting systems have been the health effects of stress at the office work- sold," there are similar concerns over privacy place. Studies showed that stress among of- and the potential for misuse. The issues sur- fice workers was a contributing cause of ad- rounding the purposes of call accounting, in both the public and private sector, are consid- "See e.g., The Dimensions of Privacy: A National Opinion Research Survey of Attitudes Toward Privacy, conducted for ered in more detail below. Sentry Insurance by Louis Harris & Associates and Dr. Alan F. Westin, 1979. The survey revealed, among other things, that strong majorities of full-time employees believed that it was "The unofficial use of the Federal Telecommunications Sys- no longer proper for employers to ash job applicimts abcut many tem (F S) or other government-provided long-distance services topics that had once been traditional to collect (e.g., informa- is Mega 41 CFR 20148.007 and 5 CFR 735.206. For a detailed tion on an applicant's spouse, neighborhood, membership in discussion of the "Telephone Call Reduction Initiative," con- organizations, residential status, arrest, and similar matters). ducted by the President's Council on Integrity and Efficiency A 1983 Survey, also by Harris & Associate., reaffirmed the im- (PCIE), General Services Administration, and the Office of Man- portance of privacy in the public mind. agement and Budget., see ch. 3. "Over 20 Acts of Congress have been passed since 1970 to "In March of 1985. Rep. Don Edwards (Chair, Subcommit- address problems of individual privacy. See U.S Congress, Of- tee on Civil and Constitutional Rights) and Patricia Schroeder fice of Technology Assessment, Federal Government Informa- (Chair, tkibcommittee on Service) sent letters questioning OMB tion Technology: Electronic Record Systems and Individual Deputy Director, Joseph R. Wright, on the privacy problems Privacy, OTA-CTT-296 (Washington, DC: U.S. Government involved with the study. The ACLU, The New York Times ("U.S. Printing Office, June, 1986), p. 15. Most recent legislative ef- Phones Rakes Issue of Privacy: New Equipment Would Pro- forts to protect privacy include the Counterfeit Access Fraud vide Detailed Record of Calls," Burnham, Mar. 17, 1985), The and Abuse Act of 1985 and the Electronic Conummwations Federal Times ("Planned Phone Audit Brings Blast From Sev- Privacy Act of 1988. None of these statutes are addressed spe- eral Groups," Montague, Mar. 25,1965) and the Washington cifically at employment, however. At the State level, 21 States Post ("U.S. Agencies Use High Tech To Curb Workers," May have passed legislation dealing specifically with the confiden- 9,1985) and considerable televisions coverage all brought pub- tiality of employee records, and 34 States have statutes con- lic attention to the subject. awning the use of the polygraph in employment Compilation "Letter at Rep. Edwards to Joseph Wright, question No. 6. of State and Federal Privacy Laws (1984-85 ed.) (Washington. Mr. Wright's response was that the program would not be look- DC: Privacy Journal), and January 1906 supp. ing at local calls, and that long-distance calls to news media, "See, e.g., Robert Ellis Smith, Workrights, Westin, Com- congressional offices, public interest groups, etc., would be con- puters. Personal Administration, and Citizen Rests. U.S. Na- Wend business cells for purposes of the study. tional Bureau of Standards Special Palladian 500-50 (1979), l'Notice by the Office of Management and Budget, 51 Fed- and the report of the U.S. Privacy Protection Study Commis- end Register 18982 (Friday, May 23, 1986). sion, Personal Privacy in an Information Society (1977). See 14See ch. 3. also OTA, Electronic Record Systems, op. tit.

10.k Ch. 4Electronic Work Monitoring Law and Policy conorderations 89

verse health impacts, such as heart disease, environment." Accompanying shifts in the and that clerical workersbecause of their structure of American industry, from heavy "high demand/low control" working situations industry to service and information sectors, were among the occupations in office work was a growing recognition of workers' legal most "at risk.'" Although the studies did rightsquite apart from those obtained by col- not find harmful stress dependent on computer lective bargaining. These new rights are beuse (high levels of stress did show up in high- ing introduced primarily at the State level, and production, closely monitored clerical work include right-to-lmow, privacy, safety, and dis- that was not computer-based), the growing crimination laws. What impact the increasing number of "machine paced" and "machine alarm over drug abuse and subsequent drug monitored" computer-based clerical workers testing will have on this trend toward greater generated similar concerns over stress and legal protection for workers is uncertain. health. Of course, understanding the factors that The workplace context in which privacy and have made monitoring into a public policy is- health concerns fermented was also changing. sue is of little help in understanding what the Women comprise roughly half of the work force specific issues are, and why electronic moni- in America today and are especially vulnerable toring raises problems that differ from conven- to the impact of microelectronics on the work tional forms of work monitoring. Part II addresses these questions. "Of 130 common occupations, the following 12, ranked from most to least stressful. were said to be most stressful: 1. Laborer 7. Managerladminitrator "See: Hartmann, Kraut, Tilly Oda), Computer Chips and Pa- 2. Secretary 8. Waitress/waiter per Clips: Technology and Woman's Employment (Washing- 3. Inspector 9. Machine operator ton, DC: National Academy Press, 1986). 4. Clinical lab technician 10. Farm owner "See, e.g., "Beyond Unions: A Revolution in Employee 5. Office manager 11. Miner Rights is in the Making," Business Weak, July 8,1986 (cover 8. Foreman 12. Painter story); ma '1'he New Industrial Relations," Business Week, (as reported in Meeks, "Worker's Compensation and Stress," May 11,1981; and more remedy, U.S. Labor Law and the Fu- 37 CPCU J. 171, 174-75 (1984) [based m a 1977 NIOSH study. ture of Labor-Management Cooperation, BLUR 104 (Washing- See ch. 2 of this report, which discusses the NIOSH study in ton, DC: U.S. Department of Labor, Bureau of Labor-Manage- greater detail ment Relations and Cooperative Programs, 1986).

PART II: FRAMING THE DEBATE As seen in the first section of this chapter ing engenders concerns over privacy. Clearly, and elsewhere in this report, electronic moni- electronic monitoring is a multifaceted issue, toring raises a variety of distinct concerns with no simple term of analysis. ranging from worker participation in job design, to worker solidarity, to privacy, to stress Furthermore, upon close scrutiny, objections and health, to worker dignity, to quality of to electronic monitoring resist categorization worklife, and more. Yet not all of these con- in terms of traditional legal and normative cerns are of the same type; some relate to the principles. As the legal analysis in part III of way that electronic mg_mitoring is implemented this chapter shows: in a given work environment, some to the use Except when monitoring is used for ille- of mcnitoring to drive the worker, some to the gal ends, even some of its more onerous use of information gained in monitoring, and forms (e.g., machine pacing) are entirely some to the very fact that monitoring is conducted at all. Moreover, these concerns differ, The concept of privacy, whether based on depending on the type of monitoring being dis- law or on ethical considerations, seems too cussed; computer-based monitoring raises the narrow to address many concerns over the issue of stress, while telephone call account- types of employee monitoring considered 90 The Electronic Supervisor. New Technology, New Tensions

in this report. The performance of tasks Table 13.A Framework for Addressing Electronic at work is, for the most part, an inherently Work Monitoring public activity, which is done on behalf Concern Criteria of the employer at the place of employ- Purpose of monitoring Fairness Relevance ment. An employee would hiely find it dif- Completeness ficult to assert a right of privacy in his Targeting Manner and method of or her performance at tasks such as com- monitoring AutonomyIntensiveness puter claims processing. Dignity Intrusiveness Although some legal doctrines may be im- Privacy Visibility Type plicitly aimed at 4-xlicating a person's Leakiness claim to bodily or mental integrity, auton- Permanence omy, or dignity, the law recognizes no Effect of monitoring Health Frequency Stress Continuousness "right of dignity" or "right of autonomy" Regularity as such.n Control Not only does monitoring escape the "net" SOURCE: Office of TechnologyAseeeement 1987. of what is normally considered private infor- a computer to do much the same thing. As mation, its infusion into the workplace seems chapter 1 explained, some form of work moni- so gradual an extension of past practices that, toring has always been a part of employment, if there is no real basis in doctrines of privacy and the fact that technology introducesnew for objecting to the proverbial supervisor with or more efficient ways to monitor work may a clipboard, there seems to be none to using not be in itself an obvious incursion on privacy. i'Coneider, for wimple. the common law torts of Some reason must therefore be found why (1) Battery, which is an intentional and wassented-to con- monitoring work by means of microelectronics tact, and in which "Mb, dement of personal indignity has always been given considerable weight": is significantly different from past forms of (2) Asserit, which stems from an interest in freedom from monitoring, and what this difference means in a p p n b s o r i o n of a harmfuloroffsnehm contact (as distinguished terms useful for formulating policy. from costae* This individual is protadad against a purely mental distnebsoce of his personal integrity. Demagog are recordr- To that end, OTA has de -eloped an analyti- able for mental disturberice (fright, humiliation, etc.) as well as any physical illness that flows from it, but an seam* must cal framework that draws on familiar concepts crate -n apprehension of immediate physical harm; and and applies them to the new capabilities and U Indletiss of Mental Distress, an action in which the inflic- characteristics of electronic monitoring. Table tion of mental injury itself became vindictable. It is most often farad in cases of intentional. flagrant acts, where "Warne out- 13 summarizes this framework. In general, rage" of a defendant's act allows recovery ("your husband has most claims abc'it the deleterious effects of been in an accident" or situations in which there is repeated hounding or threatening of the plaintiff). Mental distress must electronic monitraing can be understoodas exist and be severe, and no recovery can be obtained for more statements about its purpose, its manner and profanity, obscenity or abuse. method of implementation, or its effects. It is W.L. Prosser, Prosser on Torts, 4th ed. (St. Paul, MN: West Publishing Co., 1971), pp. 802 at seq. this framework that guides the discussion in Some court opinions suggest a close correlation between com- the rest of this chapter. In this part of the chap mon law rights of privacy and individual dignity, autonomy, ter, the purpose/method/effect breakdown is and personal freedom. For example, in (Realty, "Redefming examined, in light of the characteristics of mon- Privacy," 12 HarvardCivil Rights and avil MortiseLaw Jour- nal288-296, the author reviews several minicar dealing with itoring technologies, to show why electronic the common law right of privacy, and concludes that monitoring may present unique problems for fir meson of deniagespearl or specineeas be arrived et mid we eelmewhdp that it wee sot the mezawed touch - the relationship between employee andem- * ar the emmereeted march as nth that ceased the Wary. Whet ployer. Then, in part III, the variety of legal was injured, rather, wee that maelisr aspect at MOO end tree- dem lemeted i _cress's espeetidiem et pelmey. mechanisms for addressing problems in the Itat 268 lemphesis added). Although it may be the case that purpose, method, and effect of monitoring are privacy rights entail interests hi dignity and freedom, the con- examined. Finally, the chapter looks at the verse does not neoseserily follow: in dignity, freedom. and autonomy are not necessarily privacy interests. To vindi- types of claims the law does not address, and cate inane*. in dignity and autonomy, in other words, requires explores the options Congress may wish topur- a separate and independent basis in law, such as privacy. sue in light of these unresolved issues.

1 I) 3 Ch. 4Electronic Work Monitoring Law and PolicyCona'derations 91

PurPose Relevance.A work monitoring technique that is relevant is one that measures perform- Concerns about the purpose of monitoring ance related to the goal that the monitoring refer to the ends that the employer seeksto seeks to further. Thus, if billing customers in further through a given monitoring technique. a timely fashion is the goal, a relevant meas- Sy and large, electonic work monitoringmay ure would be whether customers were billed be used to measure and documenta variety on time. A less relevant measure would be the of employee transactions, forpurposes of: number of "fields" in a customer account database that are filled in per hour. planning and scheduling personnel and Completeness. A monitoring system is com- equiPment plete if it takes into account all, rather than evaluating individual performance and some, of the performance parameters relevant personnel decisions (promotion, retrain- to a given goal or behavior. If, therefore,a job ing, discharge, etc.); entails both talking with as many customers increasing productivity by increasing in- as possible during a given time and handling dividual performance (feedbackon speed, customer needs in a satisfactory way, a moni- etc., and work pacing); toring system that measures only the number providing security for employer property of customers handled is incomplete. (including intellectual property) andper- Targeting.A monitoring system that gen- sonnel records; erates information on particular individuals investigating incidents of misconductor within an organization, rather than thegroup crime, or human error; and or wo-k process of which that individual is a increasing management control, discour- part, is a targeted system. A monitoringsys- aging union organizing activities, identify- tem that reveals only aggregate organizational ing dissidents, etc. performance is "untargeted" or categorical. Attacks on the purpose of a monitoringsys- Using these definitions as measures of the tem can be understood as complaints about fairness of monitoring technologies, one can its fairness. While illegal monitoringpurposes begin to understand why new technology present little difficulty for finding the practice raises issues of fairness of purpose. For exam- unfair (see part III), itsuse for currently legal ple, in computer-based monitoring, where a purposes is more problematic. In general, computer is used to tabulate total keystrokes employees and unions oppose theuse of moni- during a given period of time, the question of toring for purposes which, while legal, theyre- the relevance and completeness of keystroke gard as unfair. For example, electronicevalu monitoring to the overall task can become a ations of employee performance that reflect point of contention. In contrast to nontechno- inadequately or arbitrarily the task theem- logical methods of measuring keystrokes, such ployee is performing, or that place demande. as a typing test (where typing speed may be on workers' time and energy that are unrealis- relevant only to qualifying for a job), computer- tic or unduly burdensome,are likely to raise based keystroke monitoring may make typing objections by employees and unions. speed an end in itself, without regard to the purpose for which speed is valuedmeeting Because electronic monitoring representsan a deadline as part of an overall project goal, unprecedented ability to measure job perform- for example. An overreliance on typing speed ance exhaustively and in great detail (see chs. might also become an isolated, incomplete 2 and 3), several monitoring system charac- measure of job performance. teristics become key items in ensuring fairness; particularly the relevance, completeness, and In telephone call accounting, issues ofcom- targeting of the monitoring.n pleteness and targeting become important to "Huse factors are illustrative of the types ofconcerns raised ensuring fairness. If, for example, thepurpose because of new technologies; they are hardly exhaustive, and of the audit is to reveal excessive numbers of the reader may have his or her own in mind. long-distance calls, failure of the call- account- It, 92 The Electronic Supervisor: New Technology, New Tensions ing system to also reveal extenuating circum- complaints about a loss of autonomy in job stances may be deemed unfair. If the call-ac- decisionmaking, about the indignity of being counting audit targets specific individuals, "watched" by a machine, or the invasive feel- rather than "cost centers," as abusers of long- ing of having one's every move at work distance service, failure to implement special recorded, reflect deeply held societal values. procedures for giving notice to that individ- In a work environment, we expect, and indeed ual, hearing explanations, and allowing chal- hope, that our performance will be evaluated lenges may give rise to charges of unfairness. by our superiors, yet we may balk at the thought that someone will be constantly In service observation monitoring, fairness watching "over our shoulder." may require providing safeguards to subjective supervisory judgments about the opera- The reason why concerns about autonomy, tor's quality of customer service. In other dignity, and privacy are raised in electronic words, if the purpose of the monitoring is to monitoring has to do with the fact that com- evaluate overall employee performance, it may puters are ever-vigilant; unlike human super- be claimed unfair unless a more complete visors, they do not tire of observing and record- method of evaluation is used. ing the minutiae of employee performance. In some cases, computers are also being used to Many of the complaints reported to OTA pace workers to speed their work rate.23 In about electronic monitoring suggested that the process of using computers as surrogates monitoring may itself change what counts as for immediate human supervision, employees a relevant or complete measure of job perform- may complain of "dehumanization" and iso- ance. If, for example, a job previously entailed lation. They may perceive themselves as a com- finishing a given batch of insurance claims by ponent of a system. rather than as human ac- the end of a week, a monitoring system that tors involved in and concerned with a larger only measures the number of claims finished enterprise. It is not difficult, under such cir- per hour may change that job by changing cumstances, to understanc: complaints about what counts as a relevant measure of perform- a loss of autonomy and dignity: ance, and by foreshortening the time in which a goal is to be achieved. The means for assess- The electronic monitoring is one of the most ing performance may often become an end in offensive and pernicious aspects of our jobs. itself. 1984 is nowhere more apparent than in the electronically monitored Equitable office. We Similarly, a monitoring system that is in- "dock in" at 7 a.m. and from then until the complete, or measures only one of several job end of the day, the VDT is counting eviary parameters, may unintentionally change the keystroke. At the end of the day, managers nature of the job itself. If, for example, only have a computer read-out from which produc- quantity is measured, quality may be sacrificed. tivity is determined and then averaged with subjective factors such as attitude to determine our rate of pay. Being watched, counted, Manner and Method and paced by a machine makes it very diffi- Method refers to what information is cult to take pride in your work.' gathered by monitoring, how it is gathered, It should be emphasized, however, that the and what is done with it once gathered. As potential for creating an onerous work envi- such, issues about the manner and method of ronment through electronic work monitoring electronic monitoring reflect concerns about is not always realized. Indeed, whether com- worker autonomy, dignity, and privacy. Care puter-based work monitoring becomes "offen- should be taken in reading these words in a too narrow or legalistic wayparticularly the "See ch . 2. word privacy. As we shall see, few of the con- "From Alan Westin, and The Educational Fund for Individ- ual Rights, Privacy and Quality of Work Life Issues in Em- earns electronic monitoring and privacy can ployeeMonitoring, contract paper prepared for OTA, May 1986, be vindicated in a court of law. Nevertheless, p. 76; taken from testimony before House Subcommittee (711984. 1 ki Ch. 4Electronic Work Monitoring Law and Policy Considerations 93

sive and pernicious" depends crucially on the intrusive even if the information they yield manner and method by which the syr,tem is is information ordinarily observable. administered, and what the overall work envi- Intensiveness.Intensiveness is the ronment is like. For example, one of many in- amount of detail about a worker's per- terviews conducted for OTA, by Alan Westin, formance that monitoring reveals. An em- told of a suburban newspaper's circulation and ployee log of personal calls made on an em- classified ad department that monitored records ployer's phone reveals only the number via visual display terminal (VDT) in a way that of personal calls. A telephone call-account- minimized complaints: ing system reveals much more detail; i.e., The management has a daily job chart that length of call, destination of call, the num- records each operator's time on and off the ma- ber called, which phone was used, time of chine, errors made, and accounts handled. "I day, etc. don't mind that at all," Alice said. "They don't Visibility.Visibility refers to the degree judge us an the numbers here; they take into to which monitoring is apparent to the per- account changes in the business service we are son being monitored. A computer-based making, and the way customerssome of monitoring system that reports back to themneed more service than others. Its not the employee information on the number a Big Brother thing." She also noted that of keystrokes entered is more highly visi- management's attitude led employees to co- ble than one that reports this information operate informally to take heavy loads off one only to supervisors. In general, the more another when the calls piled up at one or two stations. Alice also said that the pay was visible the monitoring system, the more "OK" by not "great" at this newspaper, but control the employee has in matching his she liked the job very much because "the ben- or her performance to expectations. Visi- efits are excellent, you can take courses at bility is important in part because of its night and have the company pay for it, and influence on the psychology of power rela- people you work with are fun to be with." tionships at work. Whereas unaided monitoring by a supervisor may require a face- Several characteristics of electronic monitor- to-face confrontation with the employee ing systems seem to be key to preserving which both Won= the employee that he worker autonomy, dignity, and privacy: or she is being monitored and "humanizes" Intrusiveness.Intrusiveness is concerned the monitoring by allowing explanations with the degree to which monitoring in- and personal interactionelectronic mon- volves probing the individual's body or itoring allows the supervisor to remove mind. A monitoring technique that is in- him or herself from the situation and use trusive is one that requires an individual the machine as intermediary, thereby to reveal facts about his or her thoughts, avoiding the human relationships that act beliefs, or states of mind; to submit sam- as a corrective to overly rigid work envi- ples of body fluids or tissues, or to expose ronments. body parts not ordinarily exposed. A mon- Type.Type refers to the nature of the in- itoring technique is not intrusive if the information gathered through monitoring. formation collected thereby is obtained Information can be either substantive or without probing into the person's mind transactional. Substantive information or body. Note that intrusiveness concerns concerns the actual content or meaning how information is gathered, and not what of communications or documents. Trans- that informatior. is.35 Techniques may be actional information is information about substantive information; the number or *For example, pumping the stomach of a person suspected type of messages sent, to whom, how of "possessing" illegal drugs :s intrusive. See, e.g., Rochin v. California, 342 U.S. 165 (1962) (The fourth amendment guaran- often, in what sequence, etc. Telephone tee* physical security of one's person against procedures that service observation is an example of mon- "shock the consciseme.") itoring substantive information, since it

I1 94 The Electronic Supervisor: New Technology, New Tensions

reveals the content of employees' phone Permanence is important from a privacy stand- conversations. Telephone call accounting, point, since privacy law very often regulates by contrast, reveals only transactional in- what may be done with records that are per- formation, such as the destination called, manent." length of call, cost of call, etc. The distinction between substantive and transac- The way in which these factors interact with tional information can become blurred, electronic monitoring to give rise to problems especially where computers are used to of autonomy, privacy, and dignity can be il- piece together patterns of transactions, lustrated by a brief consideration of the tech- thus allowing inferences regarding the nologies considered in this report. substantive content of those transactions. Computer-based monitoring may be imple- The distinction is important, since both mented in an intensive and invisible manner. societal expectations and the law gener- In other words, the computer can be used in- ally endow substantive information with tensively to chart periods of peak performance greater importance and protection. at a VDT, time spent away from the terminal, Lealdness.Leakiness refers to the abil- time spent idle, and other minutiae of job per- ity of information gathered by monitor- formance. The monitoring may be of extremely ing for one reason to be used for another. low visibilitythe employee may not know Thus, information gathered through tele- how she is doing, but does know that she is phone call-accounting systems tends to be being "watched." The knowledge that one's leaky, because the information gathered every move is being watched, without an abil- can be used to track individuals' extra- ity to watch the watcher, can create feelings work activity, despite the fact that it was that one's privacy is being invaded and that collected for purposes of detecting abuse. one is an object under close scrutiny. Being Computer-based keystroke monitoring, on subject to close scrutiny without an ability to the other hand, tends to be relatively confront the observer may mean the loss of "tight," since the information gathered a feeling of autonomy. This may have subtle often has little use outside of the context yet profound implications for interpersonal of job performance. Like other criteria, power relationships at work. In French philos- leakiness is a factor in determining the le- opher Jr m Paul Sartre's analysis of relation- gality of certain information practices." ships between persons, he observes that: Permanence.Permanence refers to whether the information gathered by mon- ...[w]ith the Other's look the "situation" es- capes me. To use an everyday expression which itoring becomes a record, and how long better expresses our thought, I am no longer that record remains in an employee file. master of the situation." Some information obtained by monitoring is transient, and never becomes a rec- Activities at work that cannot in fact be ob- ord. A computer-based monitoring system served, measured, and thus controlled, are by that determines when the employee has default discretionary activities. In the past, finished a certain job and is ready to move the time an employee spent going to the bath- on to the next (i.e., machine pacing) may room, talking with his or her spouse, pausing generate no records, and is thus transient. between tasks, and so on, were largely discre- Telephone service observation, on the tionary. Obtaining detailed information on other hand, may entail writing comments such activities was either impossible, imprac- on an evaluation sheet, which then be- tical, or not cost-effective. What constituted comes part of the employee's permanent "acceptable" employee performance was in record. "See the discussion of "system of records" under the Privacy Act in part III. Sartre, Being and Nothingness, as reprinted in The Philos- ophy of Jean-Paul Sartre (New York, NY: Vintage Books, 1972), 'See part III of this chapter. p. 203. Emphasis in original. in? Ch. 4Electronic Work Monitoring Law _I-, 1 Policy Considerations 95 part a function of the information a supervi- fashion throughout government and business, sor could collect. There was some domain of may go "wide of the mark," and have inciden- behavior which an employee could call his own, tal impacts on employees' calling decisions, and for which he knew he was unaccountable. and perhaps on the employer-employee rela- But, because the computer dramatically en- tionship, which were not anticipated. While hances the in isiveness of human observa- the past, employers had no choice but to treat tion, the employee may feel powerless and ex- employees as if they were honest,* the abil- posed under the gaze of electronic monitoring. ity to store and process massive amounts of And, since face-to-face exchanges between em- data may reverse this dP facto presumption. ployee and supervisor often involved negotia- Implicit in the installation of call-accounting tions and room for human error and "slippage" systems is the proposition that at least some in the performance of tasks, the employee's employees cannot be trusted in their use of the relationship to a supervisor was on more even employer's property. While the proposition footing. But with systems of evaluation that may in fact be correct, the system neverthe- are invisible to the employee, the transactions less audits the calling activity of all employ- between people that allowed the employee to ees, treating each as a potential abuser of fa- assert his autonomy may be minimized. cilities. Moreover, as the ability to detect abuse is refined through technology, the standard of Different sorts of mcerns about privacy and what constitutes an abr..- may be lowered autonomy are present in telephone call while previous technolc capabilities only al- accounting. Call accounting raises questions lowed an employer to pay attention to extra- about thepermanence and leakiness of records ordinary costs, new telephone call-accounting generated. The legal implications of these fac- systems may allow assessments of calls that tors are dealt with below, but here we call at- are "unnecessarily long" or "redundant." tention to the effect of the existence of records on employee behavior. If records of all calls Customer service observation shares many are being kept, the employee knows he or she of the same characteristics with other moni- may be required to justify those that are "ques- toring systems. Visibility seems an important tionable." Under these circumstances, an em- factor in assessing tne manner and method in ployee may be less inclined to make calls that which customer service observation is carried cannot be easily justified as business calls. This out. The practice of listening in on employee may have an impact on "whistleblowers " telephone conversations with customers is not those seeking to disclose unethical or illegal new, nor is it the result of recent technologi- corporate or government activity. Although cal innovations. It is also not essentially elec- reprisals against the whistleblowing employee tronic monitoring, but instead a variant on hu- may be forbidden by law or company policy, man supervision and observation of employees. the knowledge that all of one's long-distance But, since today's technology permits a super- or e n local calls are being accounted may visor to listen in on an extension at a remote nevertheless act to "chill" such activities. Even location with no audible "click" or diminution calls that can be justified as "business" or in volume, service observation is also a rela- "official" may be subject to supervisors' judg- tively low visibility form of monitoring. These ments regarding propriety or business sense. factors have lead at least one organization, the And, althoug' l the employer does have a right Newspaper Guild, to complain that to pr tect its property by ferreting out non- ..the [employee's] inability to tell under the business calls, the process of identify'' the present equipment whether or not she is be- destination and identity of nonbusiness calls ing monitored has inevitably given rise to feel- may compromise an employee's desire to con- ceal the identity of persons he or she is calling. "The basic tool for measuring abusethe monthly telephone In short, automated telephone call account- billwas inefficient (requiring a human to scan and "flag" ex- ing systems, if implemented in a pervasive pensive calls) and revealed only flagrant abuses. 6 r9c The Electronic Supervisor. New Technology, New Tensions

ings of concern, nervousness and insecurity a year is obviously less frequent than one and has made the job . . . additionally and un- that is conducted daily or weekly. Fre- necessarily burdensome." quency is an important criterion because In addition, telephone service differs from in combination with other criteria, such other forms of monitoring in that it reveals a as continuousness and regularity, it may substantive type of information; i.e., the con- make the difference between monitoring tent of employee conversations. As is discussed as sporadic "spot checks" for efficiency in part III of this chapter, privacy concerns and monitoring as a part of the daily job are most often present where the type of in- environment. formation being gathered is substantive, Continuousness.Continuousness is a rather than transactional. Because of this, measure of how constant a monitoring courtS have held that employers can only lis- technique is. It is closely related to fre- ten into business, and not personal, pbone calls. quency and regularity, but refers to the Recognizing employers' needs to monitor the duration of and intervals between moni- quality of service its representatives offer, no toring. For example, a computer-based court has held service observation to be un- monitoring system that records every lawful per se. transaction, including time spent away from the keyboard, during an 8-hour work- Effects day would be highly contim.Jus. A similar system that recorded only when the Unlike concerns over the purpose, manner, employee logged on and logged off would and method of monitoring, concerns over its be relatively noncontinuous. effects are more tangible, less valua laden, and Regularity.Regularity refers to the pre- are directed at the physical and psychological dictability of intervals between monitor- well-being of the employee. Because of this, ing. Thus, a telephone call-accounting audit most parties opposing monitoring have couched conducted every month is highly regular; 0 sir arguments in terms of observable, objec- a random audit is not. Regularity is an im- tive effects on employee health caused by the portant criterion, because it affects such stress involved in working at a monitored job. issues as actual or constructive knowledge As we saw in chapter 2, however, proving that of being monitored, and it may play a fac- monitoring causes stress can be very difficult, tor in chronic stress (if monitoring is and reliable data hard to find. highly irregular, the employee may have to stay constantly "on guard" to the pos- Electronic monitoring may create new de- sibility of monitoring). mands on employee time, attention, and speed Control.Control refers to the ability of that give rise to concerns about stress. Among the employee to set his or her own pace the factors that cause these concerns are the of work, and to use discretion in organiz- frequency, continuousness, regularity, and con- ing and executing a task. An employee trol involved in the monitoring. Each of these who can determine the pace at which dis- is described and discussed below. crete tasks, such as filling out claim for,, Frequency.Frequency refers to how often are completed has relatively greater con- the act of monitoring takes place. A call- trol than one who doesn't. accounting audit or computer-based key- Electronic monitoring may involve changes stroke monitoring that is conducted once in each of these factors, and may therefore cause greater stress than other forms of ob- "From proceedings of an arbitration of a grievance flied by the Newspaper Guild against the Boston Herald Tra velar, Bos- serving or measuring employee performance. ton Herald Traveler Corp., Case No. 1130-0291.88, arb. award, In computer-based monitoring, for example, Nov. 12,1989; as cited in Alan Westin, and The Educational an employee's control over the pace of work Fund for Individual Rights, Privacy and Quality of Work Life lene3 in Employee Monitoring, cc tractor paper prepared for may be given over to the machine; when one OTA, May 1988, p. A-18. claim form is filled out, another pops up on the Ch. 4Electronic Work Monitoring Law and PolicyConsiderations 97

screen, and delays in processing the second are Table 14.Factors Affecting Electronic Monitoring being recorded. The machine sets the pace. This Factors favoring increasad monitoring: may conceivably cause stress. On the other Economics and increasing sophistication of the technology hand, in customer service observation, continu- Labor market trends Macroeconomic trends ousness and regularity may be the factors caus- Employer liability ing stress. Whether these, or any other, char- Vendor bandwagons acteristics of electronic monitoring factors do Technological Imperatives in fact cause stress is the subject of some de- Factors limiting increased monitoring: Employee backlash, morale, & turnover bate, as will be discussed in part III. Diminishing returns Job deskilling or upgrading Information overload Where Is the Future of Monitoring Management priorities Headed? SOURCE: Office of Technology Assessment, 1987. The full extent of electronic monitoring tech- cated use of the computer to edit and di- niques may have yet to be realized, and we gest this information allows it to be put might see monitoring expand into more and to practical use. different jobs. The only limit, in principle, is In general, the means for obtaining infor- the technology itself. Advances in technology mation about the individual are less phys- may allow a greater range of less routinized ically intrusive than would be possible tasks to be monitored. Sophisticated software without technological methods. design, called expert systems, in combination The storage capacities of modern informa- with the computerization of most office activ- tion systems permits more information ity, may enable tracking the complex trans- about employees to be retained as records. actions of bank loan officers, sales and man- The growing use of computer networks agement personnel, and stock brokers. Profit also permits employee records to be dis- center accounting software, for example, can tributed and shared more easily than pa- keep accurate and timely information on such per folders. items as expense account and investment ac- Technology is not, as : .ctical matter, the tivity, interdepartmental funds transfer, and only limit on electronic mouitoring. A variety business expense structure and account turn- of factors, aside from legisiation, may influ- over. Since many expert systems are applied ence the way in which monitoring technology to assist physicians in diagnosing disease, it is eventually used. The factors can be grouped is conceivable that such systems could also be into those that tend to favor an increasing used to monitor diagnosis and method of treat- amount of monitoring, and those that tend to ment decisions. Depending on the reliability limit monitoring. They are discussed below and record of these expert systems, and their ac- summarized in table 14. ceptance in the medical community, compliance with expert systems' "decisions" may be- Factors Favoring IncreasedMonitoring come prima facie evidence of a standard of due Because purchasing, maintaining, and using care for purposes of determining liability for an electronic work monitoring system often in- negligence. volves cor, siderable expense, monitoring is un- Advances in technology could change mon- likely to be done gratuitously. Beyond achiev- itoring in the following ways: ing the stated goals of enhancing productivity or quality, or detecting and combating waste, More types of information, including in- fraud, and abuse, several factors in combina- formation about employees' behavior out of tion suggest that work monitoring may in- work, may become increasingly available. crease, both in terms of the sheer volume of A greater amount of information about businesses that monitor and the variety of employee performance is now available monitoring techniques and work environments. through the use of technology. Sophisti- Some of these factors include:

111 0

11111...k.a.a...... _. le The Electronic Supervisor. New Technology, New Tensions

The economics and increasing sophistica- diluted," but it may divert the source of tion of the technology. In the past, the employee rights from the provisions of economics of monitoring tended to work labor-management contracts to statutory against intensive mass surveillance; paid or common law; areas which, as we have employees were required to observe other seen, provide a paucity of protection employees. Modern monitoring techniques against monitoring. alleviate this fixed cost, labor-intensive ap- Macroeconomic Trends. Increasingly proach, and substitute an approach with competitive international markets in the a near-zero marginal cost, which is capital- private sector, and decreasing agency intensive. Monitoring systems may be- budgets in the public sector, force em- come cheaper to maintain than the cost ployers to trim expenses, including those of abuse or inefficiency in the labor force. associated with labor. Monitoring is one And, although it is easy to overstate the way of accomplishing this." At the same importance of new technologies," the time economic insecurity within the labor permeation of microelectronics into most market over finding and keeping a job office technology means that monitoring tend to blunt the incentive of employees can be fully integrated into work processes to "rock the boat," particularly if it would without the need for elf borate and costly entail lawsuits against employers. independent measurement devices. Com- Employer Liability. For a variety of rea- puter terminals and PBXs, for example, sons, it is the employer that generally can be monitored through relatively simple suffers economic losses from the wrong- and inexpensive changes in software." doings of its employees. Product liability, negligence, trade secret, and even crimi- Labor Market Trends.Organized labor's nal laws often en.sure this result. Further- share of the work force is currently be- more, plaintiffs in civil suits often look to tween 18 and 19 percent of the nonagricul- the employer's "deep pocket," rather than tural labor force in America 'down from to just the employee, for redress. Jury a high of 35.5 percent in 1945), and it is awards may be very high. Under these cir- expected to decline further over the next cumstances, it is not merely prudent, but 15 years." There is a coneowitant shift in fact mandatory, that the employer ex- in jobs from the manufacturing to serv- ercise a degree of oversight and control ice sector; precisely the sector in which over its employees. Electronic monitoring monitoring is highest and unionization may often be the least expensive and most weakest. It does not necessarily follow Thorough way of facilitating this." from this that employee rights are being Vendor Bandwagons. Vendors of computer-based monitoring and telephone call-accounting software have an obvious "For example, keystroke counters, called "cyclometere," interest in promoting their products. While were available for typewriters in 1913, and Taylorism devel- some vendors are sensitive to privacy and opedvariety of sophisticated techniques for measuring output, sampling piecework, and counting unite of production. See, "Some think that the case is just the contrary: "Some busi- 4, jos Galloway, Office Ahmegeseset: IbPriodpkwandPrac nese leaders think they will getunion-free environment, but tise (New York, NY: Ronald Press, 1919): and William Schulze, what they may get islegalized environment," according to The American Office Its Organisation, Management, and Harvard labor law specialist, Paul Weiler, as quoted in Busi- Beards (London: Key Publishing, 1918). ness Week, July 8, 1985, p. 78. "The Integrated Services Digital Network (ISDN) may do "A press release by OMB in support of its Telephone Call much to accelerate this trend toward integration, since all trans- Reduction Initiative said, for example, that "the PCIE and actional information can be reduced to commensurate, digital GSA/OMB initiatives will address the reduction of the Gov- hem. The first critical steps toward ISDN Lave already been ernment's long distance phone costs." taken, sod will give users access to a broad range of communi- "For example, by using a computer to track the accuracy of cations and data processing services. See: Robert Rosenberg, records maintained by employees to avoid liability for warranty "The Digital Phone Net Finally Starts Taking Off," Electronics, ar negligence in providing information; by sampling the com- Avg. 21, 1988, pp. 5741. munications and memos that go out of the office; or by using 11'Beyond Unions," Business Week, July 8,1986, pp. 72-77. cameras to observe employee conduct. Ch. 4Electronic Work Monitoring Law and Policy Considerations 99

other concerns, some teed to "puff" the for example, may engender a greater num- savings that their systemo offer. ber of unintentional or intentional errors Technological Imperatives.Monitoring (e.g., holding one key to increase total measures many things that employers number). According to a recent work on have always wanted to know. In addition, the subject, greater gains in productivity a manager is concerned with doing all that are often the result of a reorganized work- he or she can to increase efficiency and cut flow and the integration of previously waste. Nor should one rule out an irra- fragmented tasks.° tional, but common response to new tech- Job Deekilling or Upgrading.It is unclear nology: "if it can be done, it should be whether office automation is stripping done." Because of this, the use of moni- relatively skilled jobs of their discretion- toring technologies may become an im- ary and autonomous content (desk:Ming), perative or an accepted way of doing or whether it in fact is taking the drudg- business. ery out of work, leaving the employee with a greater latitude for individual creativity Factors Limiting Increased Monitoring (upgrading). Some studies have suggested Not all factors indicate a headlong drift that both deskilling and upgrading are toward more widespread and intensive moni- occurring, sometimes within the same oc- toring. Indeed, many factors seem to suggest cupation." To the degree that jobs a e that monitoring, if taken to extreme,, may ac- being upgraded by automation, work mon- tually impede some of the goals that it seeks itoring systems that require jobs to be rou- to further (e.g., productivity). Such factors may tinized, and reducible to standardized include: units of production, may become less and less apropos of highly complex, nonstan- Employee Backlash, Morale, and Turn- dardized work environments.° over.Past attempts to drive employees Information Overload.Although elec- to ever-higher levels of production through tronic monitoring offers gains in efficiency close supervision, surveillance, abuse, and over human observation, it very often re- threats of discharge have met with great quires that a human digest the informa- resistance among workers." Employee tion generated by the system and make sabotage and informal collusive "slow- managerial decisions based on that infor- downs," which tended to reduce produc- mation. This in itself may require consid- tion below the average, were often the re- erable investments of time and wages. The sult, even in nonunionized industries. records generated by telephone call ac- During times of economic expansion, job counting systems, for example, can be turnover also increased. To the degree that quite voluminous, and often require a automation is contributing to job upgrad- cadre of auditors to verify and interpret ing (see below), turnover may become an the results." increasingly expensive proposition, because of the time and money involved in "Paul Straussman, Information Payoff (New York, NY: Free Press, 1986). training new employees. "P. Attewell and J. Rule, "Computing and Organisations: Diminishing Returns.A monitoring sys- What We Know and What We Don't Know," Communications tem that emphasizes speed or volume, as of the ACM, December 1934, vol. 27, No. 12, pp. 1184-1192; R. Kling and W. Sacchi, "Computing as Social Action: The So- many computer-based monitoring sys- cial Dynamics of Computing in Complex Organizations," Ad- tems do, may often do so at the price of vances in Computers, vol. 19, 1980. quality or accuracy. A computer-based °Paul Strauseman, In Payoff (New York, NY: Free Prow 1985). monitoring system that counts keystrokes, "See ch. 3 for an example of an expense report generated by telephone call accounting. all. Edwards, "Individual Traits and Organizational Incen- "Sec A. Gouldner, Pattern, of Industrial Bureaucracy tives: What Makes a 'Good' Worker," Journal of Human R. (Gismos, IL: Free Press, 1964 sources, winter 1976.

.112 100 The Electronic Supervisor. New Technology, New Tensions

Management Priorities.The kinds of indirect input of data by optical scanning formation provided by electronic monitor- technologies, and possibly by speech rec- ing may not assist management in ad- ognition technology; and dressing the workplace inefficiencies that capture of data at the point of origin, in it perceives as most troublesome, and a variety of ways ranging from bar code monitoring may frustrate the human re- readers to consumer use of terminals, e.g., lations goals that many firms see as a key bank automated teller machines (ATMs)." to productivity. Among management's more pressing concerns are employee In conjunction with progress in natural lan- fraud and chronic absenteeism and tardi- guage processing and pattern recognition sys- ness:* conduct that requires no elec- tems," this trend toward automation of tronic monitoring to detect or deter. More highly routine jobs may end up eliminating over, management may have no interest narrow, low-skill clerical positions altogether, in monitoring systems that degrade replacing them with multi-activity skilled po- worker responsibility and morale, since sitions.* Highly complex jobs, requiring mul- commitment to the job is perceived as a tifaceted decisionmaking, interpersonal skills, vital element of employee productivity.* and "common sense" judgment, are unlikely to be as susceptible to electronic monitoring, Wild Cards: Automation and since these jobs are not amenable to merely Artificial Intelligence quantitative measures of performance.° Thus, it is possible that the issue of electronic It is possible that present concerns over work work monitoring is merely a transient phase monitoring may be rendered obsolete by main the automation of office work. There are chines whose functions are to substitute for some indications, however, that data entry re- precisely the type of job that is today the fo- quirements are accelerating faster than the cus of monitoring. As mentioned in chapter ability to automate them. So it may be quite 2, electronic monitoring is most often used in some time before monitored jobs are auto- jobs that require relatively few skills, that are mated out of existence. highly routinized, and that have more or less uniform patterns of input and output. This type of labor is gradually being substituted by automation. Data-entry work (whether nu- meric or textual in nature), for example, can be eliminated by: interorganizational transfer of data, directly from computer to computer; "From U.S. Congress, Office of Technology Assessment, Automation of America's Offices, OTA-CIT-287 (Washington, DC: U.S. Government Printing Office, December 1985), p. 47. "U.S. Congress, Office of Technology Assessment, "inns- Lion Technology R&D: Critical Trends and Issues, OTA-CIT- 268 (Washington, DC: U.S. Government Printing Office, Feb. ruary 1985), see especially ch. 3, "SelectedCUMStudies (Artifi- cial Intelligence). "U.S. Congress, Office of Technology Assessment, Automa- tion of America's Offices, OTA-CIT-287, p. 61. Of course, the "R. Walton, "From Control to Commitment in the Work. low- skilled employee may be eliminated entirely. The conse- place," Harvard Business Review, vol. 36, No. 2,1986, pp. 76 quences of automation for the job market are highly controver- 84; and "The New Industrial Relations," Business Week, May sial, but it is unimportant for present purposes to enter the 11,1981. See also Business Week, "The Hollow Corporation" debate. special issue, Mar. 8,1986; Business Week, "High Tech to the "Thi s . -taut should be qualified by two civvies: any job Rescue," June 16,1986; National Acadscw of Science, Towards performance can in theory be subjected to quantifiable, elec- a Now Era in gal U.S. Manufacturing (Washington, DC: 1988); tronically monitomble criteria; and electronic monitoring of the and Human Resources Practices for Implemeoting Advanced future may be able to build in some sort of assessment of the Manufadming Technology (Washington, DC: 1986). qualitative features of job performance.

413 Ch. 4Electronic Work Monitoring Law and Policy Consideratioi.s 101

PART III: AN OVERVIEW OF APPLICABLE LAW The Framework for the Legal Analysis of employment-at-will and the differing legal status of private and public sector employees. Part II suggested that concerns over the purpose of monitoring can be understood as ob- The Concept of Employment-at-will jections based on notions of fairness; that the manner and method in which monitoring is im- Under the common law tradition in the plemented may irsvolve issues of dignity, au- United States, the relationship between em- tonomy, and privacy; and that issues involv- ployer and employee has been one of "employ - ing the effects of monitoring can be largely ment-at- we'll." Employment-at-will simply means understood as concerns over health and stress. that, in the absence of a specific agreement to The following legal analysis uses this frame- the contrary, an employer has an absolute right work by applying more specific legal concepts to discharge an employee for any reason, and to the purpose/manner and method/effect frame- the employee has a correlative right to resign work. Table 15 shows the relationship of this for any reason." Although subject to consid- framework to applicable law. erable erosion through a variety of judicial and statutory exceptions and qualifications (dis- Each of the major types of monitoring con- cussed below where relevant),4' the employ- sidered in this reportcomputer-based moni- ment-at-w01 doctrine is still law in all 50 States. toring, telephone service observation, and telephone call accountingwill, to the extent that "S. Williston, Contracts 11017 (1967); see e.g., Pearson v. Youngstown Sheet & Tube Co., 332 F.2d 439 (7th ar.). cert de- they raise unique legal issues, be discussed nied, 379 U.S. 91411964). The terminalelity at will doctrine cin separately. Otherwise, the analysis that fol- be modified by a contractual agreement to retain the employee for a specified period of time require that the discharge of en lows is cumulative; what is said of computer- employees be based on a breach of that employee's obligations based moLitoring, for example, applies equally under the terms of his or her contract of employment. to telephonl customer service observation and "The claim of wrongful discharge, for example, has been so- telephone call accounting, unless specifically cepted in a majority of States. "To date, the common law of three-fifths of the states has recopied, albeit to marl mentioned in the text. extents, a cause of action for wrongful discharge in one form or another." Kenneth T. Lopetkay, "The Emerging Law of Before proceeding with the analysis, it is nec- Wrong DischargeA Quadrennial Assessment of the Labor Law Issue of the 80's," 40 Business Law 445 (1984), and see: Wil- essary to discuss two issues common to all liam L. Munk, "Wrongful Discharge The Erosion of 100 Years three types of work monitoring the concept of Fmployer Privilege," 21 Idaho Law Review 201 (1985).

Table 15.A Framework for Addressing Electronic Work Monitoring

Concern Criteria Example of applicable law Purpose of monitoring Fairness Relevance National Lab& Relations Act; Civil Rights Completeness Act; Merit System Principles (as admin- Targeting istered in EEO, OSHA, ERISA, EPA, etc.'); State Law on Privacy; Constitutional Law' Manner and method of monitoring AutonomyIntensiveness State Law on Wrongful Discharge; State Law Dignity Intrusiveness on Privacy; PCIE Guidelines; Title III of Om- Privacy Visibility nibus Crime Control and Safe Streets Act; Type Electronic Communications Privacy Act; Ns- Leakiness tional Labor Relations Act; Privacy Act of Permanence 1974' Effect of monitoring Health Frequency Worker's Compensation Statutes on Stress- Stress ContinuousnessCausing Labor Regularity Control aApplles only to Faders! enhAorlAll- SOURCE Office of Technology Assessment, 1987

1114 102The Electronic Supervisor: New Technology, New Tensions

The significance of the doctrine of employ- fringe a constitutionally protected interest, ment-at-will for electronic work monitoring lies that interest can only be vindicated if the em- in its practical effect on the legal or economic ployer is a local, State, or Federal Government, pressure that an individual employee can bring or if the employer is acting pursuant to or un- to bear against the employer. Unless the con- der authority of a statute or ordinance." Fur- tract of employment includes either substan- thermore, the Privacy Act of 1974, which may tive prohibitions, such as work environment be relevant insofar as electronic monitoring dames that can be construed to extend to work often generates a system of records, applies monitoring, or procedural requirements, such only to records kept by the Federal Govern- as binding arbitration agreements, an em- ment. it is therefore significant primarily to ployee who objects to being monitored has the Federal Government employees." options of accepting the practice, protesting the practice to the employer and facing possi- Notwithstanding this crucial distinction be- ble dismissal, or leaving the job voluntarily. tween private and public sector employers, This is particularly true of Federal employees, there are a number of State and Federal stat- who, though they are represented by the Amer- utes that may be relevant to considerations ican Federation of Government Employees of the purpose, the manner and method, and (AFGE), are forbidden by law to negotiateper- the effect of monitoring by both private and formance standards which are at the heart of public sector employers. The public/private die- many disputes over electronic monitoring." Unction is therefore considered below only where relevant. The Legal Status of Publicv. Private Sector Employees Purpose Computer-Based Monitoring The legal rights of an employee with respect to electronic monitoring depend criticallyon Computer-based monitoring is the computer- whether the employer is a privately owned and ized collection, storage, analysis, and report- operated firm or an agency or subdivision of ing of information about certain employee the local, State, or Federal Government. As work activities. Within this broad definition, a general rule, an employee has no constitu- the chapter focuses on the use of computer- tional rights against private individuals, in- based monitoring to obtain data about employ- cluding private employers." Therefore,even ees directly through their productive use of if some forms of monitoring can be said to in- computer and telecommtmications equipment. In all cases documented by GTA, computer- based monitoring is used by both public and private sector employers for entirely legal pur- "a U.S.C. 143, The Federal Labor Relations Statute. "Phis =mewl is known as "State action." It is a basic prin- poses. As a rule, an employer is not liable for dpie of oonstitutiond law, and provides that rights secured endeavoring to further its legitimate business to bdividuals by the 14th Amendment to the CoConstitution pro- interests, such as enhancing productivity and wzibe only certain actions by the state, state agenciesor subdivision, or individuals acting under color of State law, and do not limit actions between private individuals or private entities. Else e.g., Ma Ikon, Inc. v. Brook., 486 U.S. 14911976). T h e r e i n c e r i r d n n a r r o w e x c e pt i o n s t o t h i s r u b , as where' company assume. ell 6'Am:timed a innoidpality, ihnah v. Ala- . 1.323bona, U.S. 501119461, or when there is substantial State "Even in the latter case, the breadth of State actionmay in bodement with a private entity, e.g., Bart= v. ViIhningtat fact be very narrow. See, e.g., Moose Lodge v. rrvls, 407 U.S. hada g Authority, 10116 U.S. 71511961). Some have argued that 163(1972) (State liquor license for segregated dinerg room was the actins of larep private oandrations that wield greateco- insufficient State involvement). nomic power over individuals should be considered State ac- "A provision that would have made the Privacy Act, 2 don. Sae, e.g., Barb "ConLautional Limitations on Corporate 1552a(1976), appicabb to the private sector was present ActivityProtedion otPersonsl Rights Invasion Through Eco- in the original Senate Man of the Act, S 3418, 93d Cong., mmic Power," 100 UnlveraftydPermsylvanis Law Review 933 2d um f201(a), but was not adopted. The Privacy Act io, how- (1262). ever, applicable to government contractor. 5 U.S.C. 1552a(m).

115 Ch. 4Electronic Work Monitoring Law and Policy Considerations 103

quality. Nor is an employer liable for protect- When used for certain purposes, however, ing its property, or for investigating miscon- computer-based monitoring may become the duct or crime." instrument of illegal ends. It is conceivable, Computer-based monitoring for purposes of for instance, that monitoring could be used to advancing or protecting commercial interests frustrate the rights of employees to organize, and overseeing actions of employees is not by being used as "ptmishment" for individuals merely prudent business practiceit may be seeking to organize." OTA foundno evidence a positive legal requirement." In many in- that monitoring is actually being used in this stances, an employer is held vicariously liable way." for the torts or crimes of its employees, based It is also conceivable that monitoring might in part on the theory that it is in control of be used to discriminate againsta class of em- and responsible for many of the actions of its ployees, by placing stricter scrutiny and stand- employees while in the scope of their employ- ards of job performance on certaingroups. As ment" Furthermore, an employer, as a seller mentioned in chapter 2, the highly specific in- or even gratuitous supplier, may be liable for formation that monitoring generates oftenre- the quality of the goods, services, and perhaps quires a considerable amount of interpretation, even the information produced by its employ- leaving great leeway for (intentionalor =in- ees." And, monitoring the flow of trade tentional) misinterpretation in the guise of "ob- secret information out of a business concern jective," quantitative evidence. OTA foundno may be necessary if an employer is to preserve case where monitoring was intentionally used its rights under trade secrets law." for this purpose, but that does not preclude such a possibility." It is important to point "Based on one sunny of the tap throe reasons for auditing the use of inteMgent desktop terminals were out, however, that the vast majority of employ- To prevent abuse of company PC resources for personal ees whose work is monitored by computer are purposes. To pnwent confidentielitylmanity breeches. To prevent violation of legal/regulatory duties in the use *Such rights are protected, for example, by the National of dknt or employes data. bar Makes Act, 29 U.S.C. 1151 et seq., which secures to em- From Alan Westin, and The Educational Fund for Individual ploye. "the right to ealferganhation, to form, join, or assist Rights, Privacy and Quality of Wort Life hem in Employee labor organizations.." Id., at 1157. The use of monitoring to Monitoring, costeador paper prepared for OTA, May 1986. impose changes in working =Widowe.g., by acalsrating ma- For simple, a Npremntative of American Empress Corp., chine pacingmay be Mega if dons for the purpose of reprisal Inc., informed OTA that American Express is required by the against employee organizational activity. 29 U.S.C. 1158 (a) (3); Fmk Credit Reporting Act to stop dunning a card member who See, e.g., N.L.R.B. v. Sanitary Bag end Burlap Co., 406 F.2d writs. in to say that he or she is having a disputa with a service 750 (3rd Qr. 1969) It is also possible that PC use could be mad- e s t a b l i s h m e s t . B e c a m e the letters are computer generated,the tared to detect =ion communications by *earthing or auditing only way that American Empress can know it is in compthmce PC-user duke or files, although there is no indication that this is through its monitoring system, which aggregates all such activity is widespread today. Monitoring for this purpose, how- transactions and reports on when they are made. ever, would probably not violets labor laws, since employers "See, e.g., Prosser on Tarts, 169 (1971). may observe the activities of employees on its property during "The Uniform Commercial Cods, adopted in 49 States, im- working time. Stone& Webster Barinessing Corp. v. N.L.R.B., poses on the seller of goods three different kinds of warranties. 586 F.2d 461 (1st Cir. 19761; N.L.R.B. v. R.C. Mahon Co., 209 U.C.C. 2413-315 Providers Cl services, such as insurers, may F.2d 44 (tithe Cir. 1969). be liable for breach of either express or implied conditions of "One source told OTA that one practice in monitoring cow the contract. Information providers, such as detainee compa- peter files is to check for human error. Under this circumstance, nies and weather forecesters, may be liable on several theories, some privacy questions may raised, despite the legitimate pur- such as staict Ilability,,wiligsnee, warranty, or defamation. See: poses of the monitoring. e.g., Aetna Casualty & aunty Co. v. Appwrin & Ca, 767 F.2d "For example, the veracity of computer monitoring records 1288 Was Qr. 1965k and Dunn & Bradstreet v. Gretemoes was the subject of an arbitration disputs between The State &Baru inc., 106 &Ct. 2939 (1965). of Oregon Employment Division and the Oregon State Public "A trade secret is a form of intellectual property thatcov- Employees Union (affiliated with the SEIU). on behalf of one ers any confidential fonnula, pattern, device, or compilation of of its members. The employee was fired from her job as word information used in a business, which gives that buskinsan processing specidist for allepdly tampeing with her end others' opportunity to obtain an advantage over competitors who do production statistics generated by a Wang "Machlue Stade not boow or =tc the occra. Ors Cl the foam: far determining Ws System." The arbitrator found the statistics generated by whether a business' secret is a trade secret is "the extort of the computer system ndiable, albeit drcumstantial, widows measures taken by (itl to guard the secrecy of the information." that the employee had tampered with the system co& lel the Restatement Cl Torts, 1757, comment B. States decision to terminate the employee stand. The union ifootems castined on ant NO

1.1 6 104 The Electronic Supervisor. New Technology, New Tensions female, raising questions about the existence The number of illicit purposes to which mon- of de facto discrimination in working con- itoring can be put is limited only by the imagi- ditions." nation, yet monitoring seems no more or less likely to lend itself to illegal retaliation than Finally, monitoring could be used as a any other form of office technology (the paper method of detecting, preventing or retaliating copier or punch cloc' r, for instance). However, agabst whistleblowers. This might be accom- the employee may c ften know that his or her plished by restricting access to certain com- computer files or phone calls are being "ob- puter files for the purpose of preventing served" by the monitoring system, and this damaging information from being revealed, by knowledge may in itself act as a "chilling" de- tracking the types of files accessed by certain vice to would-be whistleblowers or union or- employees in order to ascertain the source of ganizers. In a case example submitted to OTA "leaks," or by imposing more onerous demands by the AFL-CIO, the chilling effect of video on certain employees for revealing evidence of waste, fraud, or abuse." OTA again found no observation was noted: evidence that monitoring is being used for such . . .employer installed and focused TV moni- pupates toring equipment inside the plant on every work station and worker after organizing ef- destattemaimed from previous pegs) fort began. Monitors were not available for all disputes the arbitrator's fmdings, and suggests that the em- to see, but viewed only by management in ployee'. "work station was frequently used by other employ- management's office. Employer said monitor- ees, pirticularly her supervisor with whom she had a bad working relatiosisidp." From a "case study" submitted to OTA by ing was for safety reasons and would lower the Service Empioyeas International Union, and Westin, Worker Compensation insurance rates. In fact, "Privacy and Quelity-of-Worklife Issues in Employee Monitor no one could determine how that could be. Dur- ing," OTA Contract Report, December 1988. See also,The Wall ing the height of organizing, two workers who Street Journal,June 6,1985. The union thus suggests the possibility that the employee could have been "framed." Regard- left their work stations to go to restroom were less of whether the employee in this case was in fact culpable suspended for leaving their work station, with- for the alleged tampering, it is dear that monitoring systems out permission. Monitoring had chilling effect are eligible of being used for discriminatory or retaliatory rea- on workers attempting to orrnize for pur- mer the objective, mechanically produced measurements of productivity can be "rigged," and undue trust can be phred pones of collective bargaining. Of the 100 in machine printouts. workers in the unit, 89 signed authorization "See Title VII, Civil Rights Act of 1984, 42 U.S.C. 2000c cards calling for a recognition el action. But (MU The de facto feminization of monitoring may give rise to when the final vote came, the union was 12 emits under the Civil Rights Act, since intent to discriminate votes shy of a majority..." There was an un- is not a prerequisite to an action under This VII. spoken fear that Big Brother would catch "Federal employees have been protected by statute against reprisal for disclreng waste, fraud, or abuse in the Federal Gov- them talking for or working for that union." ernment since 1979. As part of "merit system principles," all employees of the executive branch of government, the Admin- Even if the truth of this use of monitoring istrative Office of the U.S. Courts, and the Government Print- for alleged purposes of union-busting are not ing Office: should be protected against reprisal for the lawful disclosure of Wormed= which the employee; noesonably believe evidence. and play a "substantial part" in decision to firer andConnick (A) a violation of fey law, rule, or regulation, or v. Myers,451 U.S. 138 (1983) (whether dissent is a matter of misenuegemset, a gross waste of funds, an abuse of author- public concern is determined by content, form, and context of ity, or a eabetasdel and specific danger to public health or safety communication). In addition. 21 States have enacted whistle- 6U.S.C.1230101191 blowing statutes. "Beyond Unions,"BusinessWeak, July 8, Additional protection may be available under anti-reprisal 1986, p. 73. clauses of various substantive employee protection or public Employees he the private sector may be protected under excep- 'protection laws, EEOpH AF. EsPt Am, and so ont tions to the employment-at-will doctrine. See footnote 49. Other publc pmideae euEnRdIeSrFirst Amendment "From AFLCIO Case Examples, submitted to OTA. The prIndpies, as articulated inPickering v. Board of Education, mausolea allegedly used to monitor these employers was the 1191 U.S. 663 (1988) ("the interests of the [employee], as a citi- video cameraa technology not considered in this report. How- zen, in commenting upon matters of public concern [are to be ever, software for computer-based work monitoring could be weighed against] the interest of the State, as an employer, in used to accomplish much the same purpose as alleged in this promotion the efficiency of the public services it performs case study, by recording the time away from a station, by mon- Wits employees." Id. at 688. See also: Mt.Healthy Cuy itoring internal electronic mail and ecupkyourgersete.ed docu- D1strictBoard el Educative v.Doyle, 429 U.S. 274 (19771 ments, cr by determining who was logged onto a particular work iprefacted whistleblowing most be "a matter of public concern" station.

117 Ch. 4Electronic Work Monitoring Law and Policy Considerations 105 born out, it nevertheless illustrates the height- employer's purpose for monitoring is not a con- ened potential of monitoring used to deter the sidration separate and apart from the man- efforts of whistleblowers or union organizers. ner and method in which the service observa- This potential is also explored below in the con- tion is conducted. text of telephone call accounting. Telephone Call Accounting Telephone Service Observation Chapter 3 discusses telephone call account- Telephone service observation was described ing in detail. Telephone call accounting sys- in chapter 3. It refers to the act of systemati- tems are devices which can be attached to ei- cally intercepting the content of employee tele- ther the central office switch of the local phone calls by listening in on them. This is telephone network or, increasingly, to the pri- often done by a supervisor or quality control vate branch exchanges (PBXs) on the custom- specialist to evaluate courtesy, accuracy, or ers' premises. Call-accounting systems gener- compliance with company guidelines. It is a ate detailed raw data on telephone usage; common practice in a host of businesses which incoming and outgoing call numbers, total sell products or service customers over the tele- number of calls made, total time on the line, phone. As mentioned in chapter 2, service ob- etc (they do not provide information on the con- servation is becoming integrated with various tent of the telephone call). This raw informer forms of computer-based monitoring, the le- tion can be processed by computer to provide gal implications of which were considered summary reports of any type of telephone above. This section will consider service ob- activity that the employer feels is relevant or servation in isolation from other monitoring useful. techniques. Call accounting is often used for purposes The use of service observation for illicit pur- that many might consider legitimate business posese.g., to discourage or listen in on em- functions, such as allocating costs between ployee organizational activities, to discriminate various accounts in a business, billing custom- against certain classes of employees, or to de- ers or clients for particular services, and keep- tect and punish whistleblowersis subject to ing track of abuse or waste of local or long-dis- much the same legal analysis as computer- tance telephone services. The recently enacted based monitoring, and presents few unique Communications Privacy Act of 1986 explicitly problems for the law." Since service observa- recognizes the need for call accounting in the tion is by nature a method of intercepting the course of providing communication services." content of employee communications, legal The extent of personal phone use in the Fed- rights to privacy under State tort law may be eral Government was examined in a call-ac- implicated, and while the employer often en- counting audit conducted by the President's joys a qualified privilege to listen in on em- Council on Integrity and Efficiency (PCIE) in ployee phone calls, that privilege may be viti- conjunction with the General Services Admin- ated by improper motive." Otherwise, the istration (GSA) and Office of Management and Budget (OMB). That audit reported in the fall "A very important qualification to this statement is This HI of the Canmunkations Act of 1934, which prohibits "eaves- (tapping own telephone to protect own interests);Wheeler v. dropping" per se, without regard to the intent of the person Sorenson Mfg.,415 S.W.2d 682 (1967) (publication of wages doing the eavesdropping, with certain crucial exceptions. Thle and deductions of employees to combat union drive); andCity III is considered more fully below under the manner sad method of University Heights v. Conley, 252N.E.2d 198(1969)(spying analYsis- on suspected thief).Prosser on Torte,1117 "(Defenses to common law claims of invasion of privacy in- "The Electronic Communications Privacy Act of 1986, Pub- clude the defense of "privilege") The qualified privilege of the lic Law No. 99-508, 99th Cong., 2d sees., Oct. 21, 1986 amends defendant to protect or further his own legitimate interest has portions of the criminal cods (Title 18) to accommodate digital appeared in a few cases, aswhen, a naphone company has been communications, computer networks, cellular telecommunim psnuittod to monitor calk ; &tor Schmulkor v. Ohio Boil Lions, and other advances in communications. Its import for Telephone. Co.,116 N.E. 2d 81911953) (time and motion studies telephone all accounting and service observation is discussed of employees);Psopk v. Applebaum,97 N.Y.S. 2d 80711960) below where relevant. 118 108 The Electronic Supervisor: New Technology, New Tensions

of 1986 that on the average about 33 percent forceable. If, for example, a Federal employer of long-distance calls made on the Federal Tele- were to discipline or withhold promotion or in- phone System (FTS) were "unofficial," that is, formation from an employee based on that em- made for personal reasons. ployee's contacts with the press, the employee Concerns were raised in Congress over the may find it difficult to prove that the em- implications of the PCIE audit for privacy and ployer's motivations for doing so were the re- whiatleblowing. Problems might also exist, sult of information sbtained through telephone particularly in the private sector, if call call accounting. accounting were be used to frustrate union Although the PCIE study guidelines forbid organizing efforts. As previously discussed, listening to or recording conversations (as does however, legal protections exist to address con- Mae III, discussed below), information on tele- cerns over employee/union rights, and whis- phone transactions can yield a great deal of tleblowers. Moreover, PCIE has adopted guide- inferential knowledge about an employee's per- lines to address some of the concerns over the sonal and life outside of work Knowing that privacy and first amendment implications of an employee contacted a particular newspaper the program." Among the protectionsare: a one day before a damaging article is printed "conservative" approach to classifying calls is sufficient to infer the content of the conver- as "unofficial," prohibitions on invading the sation, regardless of how that call is classified privacy of the persons called from the agency, or whether it is subject to detailed investiga- categorization of "calls possibly made to news tion. Moreover, records of the audit which con- media, congressional offices, public interest nect names and numbers, while protected by groups, and employee unions" as "official,"" the Privacy Act, may nevertheless be subject and a prohibition on using data to single out to disclosure through the Freedom of Infor- individuals or to conduct investigations." It mation Act." remains to be seen, however, whether and how the PCIE initiative will be continued and be- Yet another difficulty with the PCIE study come part of the regular internal auditing Fed- guidelines concerns enforcement and disci- eral agencies. One department indicated that pline. At present the guidelines contemplate in spite of its pilot study resultsindicating disciplinary action, such as removal, suspen- significant unofficial use of the department's sion, demotion, or reprimand only in cases of telephone systemthe agency had no plans for "extreme" cases of FTS abuse. The difficulty further efforts to reduce these misuses, because here is with selective enforcement and uniform- of concerns over privacy implications.7' If the ity of treatment. The PCIE guidelines offer no audit does become a permanent part in intra- guidance on what constitutes "extreme" abuse, agency audits, questions arise over whether and no mention is made of who within each protective guidelines will also become perma- executive agency will be responsible for en- nent, and if so, how such guidelines will be enforcement. This leaves considerable discretion to agencies' Inspectors General in determining who will be disciplined and under what cir- %snare' Services Administration, Office of the Inspector Geared, Office of Audits, "Guide for the PCIE Review of Fed- cumstances. It opens the door to claims of eral Telecommunications System (FTS) utilization.-pert differential treatment between low-ranking see especially app. XIII, pp. 85-74, July 3, 1985. "Ibid., p. 29. Section IX of the "Guide" states that research clerical staff and high-level government execu- activities concerning calls to these destinations should be ter- tives. Since the scope of job responsibility is minated, that the information cannot be released tomamas& often fairly narrow for low-level employees mint, and that the information cannot be used against the person who made the calls. (e.g., claims processing at the Social Security "Ibid., p. 30. The Guide does recommend, however, that "serious or egregious" cases of misuse should be referredto "See Title 5 U.S.C. 1552, intr.,. See also Notice of Proposed the agencies investigative organisation for possible further Privacy Act Guidance for Call Detail Systems (0M13), 51 Fed- action. aid Register 19982,19984 (Friday, May 23,1988), which dis- narA staff taisphons interview with Department of Energy cusses disclosure in the context of a permanent FTS telephone Office of Inspector General representative, fall 1988. call-accounting system.

1 1 Ch. 4Electronic Work Monitoring Law and Policy Considerations 107

Administration), discriminating between "offi- State action (see above), the private sector em- cial" and "unofficial" calls may be relatively ployee can claim no first amendment right to easy. But for high-ranking personnel, whose speak to the public or the press. Of course, communications are more likely to be a mix statutes governing communications between of "business" and "pleasure," such determi- employees and labor organizations, discussed nations may not be so easy. In other words, above, apply with equal force to telephone call the informalities and ambiguities of the PCIE accounting. It should be noted that, unless the guidelines may give greater latitude to high- employer consents to the use of its telephones ranking employees than lower level employ- for labor organizational purposes, the em- ees. Under the proper circumstances, this may ployee probably does not have rights under give rise to a claini of denial of equal protec- statute to protest the use of telephone call tion of the laws under the 14th amendment. accounting to track and squelch union com- Finally, there is a difficulty of administer- munications." ing the telephone call-accounting audit, particularly if it is implemented on a permanent Manner and Method basis. Although OMB has, under the PCIE guidelines, drawn up fairly extensive analyses Computer-Based Monitoring of Privacy Act implications concerning em- The use of computer-based monitoring as a ployee privacy and the disposition of records, means for furthering the legitimate employer the question remains: who will be responsible interests raises few, if any, legal issues. The for overseeing the agencies in the conduct of first hurdle that an attorney challenging the their audits to ensure that the guidelines are practice itself must meet is to identify a "cause followed? A recent OTA report" concluded of action"a legally recognized right that that OMB is not effectively monitoring such forms the basis for a lawsuit. The only right basic areas as: the quality of Privacy Act remotely relevant to monitoring is the right records; the protection of Privacy Act records of privacy." in systems currently or potentially accessible Privacy is a broad value, representing con- by microcomputers; the cost- effectiveness of cerns about autonomy, individuality, personal recordkeeping; and the level of agency resources space, solitude, intimacy, anonymity, and a devoted to Privacy Act limitations. host of related concerns." Since monitoring Such practical difficulties notwithstanding, there appears to be no dearth of legal protec- discharge. See, e.g., Mono v. Beebe Rubber Co, 114 N.H. 130, tion for activities of Federal employees that 316 A.2d 649.82 A.LR.3d 264, 26 EPD P 31,843; and R. Murg ar9 J. Sharman, "Employment at WM: Do Exceptions Over. the law recognizes as legitimate and responsi- whelm the Rule?" 28 Boat= CollegeLaw Review32911982). ble. However, the use of telephone call account- Moreover, employers may be held to their own internal state- ing by private sector employers for illicitpur- manta of policy concerning matters such as privacy and treatment of employees with respect to monitoring. See, e.g.,Wool- poses is not so clearly proscribed by law. In ley v. Hoffman-LaRoche, Inc.,99 NJ 284, 491 A.2d 126711986), fact, the only recourse of the private sectorem- which held that the employer's official statement of policy for ployee against the employer for using call its employees createdcontract of employment for an indefi- nite period. accounting to track whistleblowing activities "Under the National Labor Relations Act., 29 U.S.C. 1168, is the nascent legal right against "wrongful for example, it is not unlawful for an employer to observe em- discharge. ' "6 Because of the principle of ployes activities at the worksight during working hours to see if union activity is being conducted on company time.N.L.R.B. v. R.C. Mahon Co. 269 F.24 44 (8th Cir. 1959). "For example, is a long-distance call to setup"business "Various theories under tort law, such as assault, intentional lunch" with a good friend an "official" or "unofficial" call? infliction of emotional distress, defamation, Wane, and may "U.S. Congress, Office of Technology Assessment, Federal hem were consaiseed, but lack sufficient connection to the types Governunt Information Meltwater: Electimic Record Systems of activities involved in computer-based monitoring. and Individual Privacy, OTA-CTT-296 (Washington, DC: U.S. "See U.S. Congress, Office of 'Technology Assessment, Fed- Government Printing Office, June 1986). eral Government Information Technology: Electronic Record "Retaliatory discharge for whistleblowieg activitiesmay ha Systems andIndividual Privacy,OTA-CIT-298 (Washington, counter to public policy, and may thus constitute a wrongful DC: U.S. Government Printing Office, June 1988). 120 108 The Electronic Supervisor. New Technology, New Tensions is one method of obtaining information about Furthermore, some State courts may hold and control over an employee's activities, some employers to internal statements of policy re- of these concermay be relevant. garding employee privacy, and may award Although monitoring may affect culturally damages for "unjust termination" of employ- held values, there are serious problems in at- ees who seek to withhold information under tempting to stretch the legally enforceable these policy statements.82 This approach has values regarding privacy, whether based on not been widely accepted in the courts, and the common law, statute, or the Constitution, to corporate policy statements seldom address cover the types of monitoring conrdered in this monitoring explicitly. report." Although 34 States have adopted There are several situations in which com- laws regarding employer use of polygraph ma- puter-based monitoring may Implicate certain chines, and 21 have laws addressing the pri- legal rights. The first is where the monitoring, vacy of employee re"ords," none have so far which ordinarily reveals quantitative informa- adopted legislation restricting monitoring, as tion about the amount of work done and the such. One State, Massachusetts, has attempted time spent doing it, reveals "personal" infor- to enact legislation that might prohibit com- mation as a byproduct. For example, if the only puter-based monitoring per se, but the legis- discretionary breaks allowed a monitored lation was found to violate the due process and worker are for trips to the bathroom, the com- equal protection clauses of the U.S. Consti- puter may allow an employer to glean this it tution." formation by the frequency and duration that "The most widely accepted privacy framework under tort the employee is logged off the terminal." In law is that offered by Prosser. See "Privacy," 48alifornia Law this situation, a breach of employee privacy Review,383 (1980). Each of the four distinct tortsintrusion, disclosure, false light, and appropriationrequire a physical in- is arguably present." Another situation con- vasion of the person or his/her property or personality and pub- cerns the monitoring of personel computer use, Radian of the information gained by the invasion. Neither of and the auditing or editing of employee com- these criteria is applicable to monitoring considered in this report. Moreover, consent to monitoring, either explicit or as a puter files. If the employer permits an em- implied condition of the employment contract, would probably ployee to use computer files to store personal vitiate whatever claims an employee might have. Privacy un- information, or electronic mail capabilities for der statutory law, at both the Federal and State level, concerns principally privacy of employee records, and while not relevant personal messages, a breach of privacy may to the act of monitoring itself, may be relevant to records gen- be found under a number of theories if the em- erated and kept by the monitoring system. This is considered ployer subsequently examines or reveals the below where relevant. Privacy under the U.S. Constitution has two main branches; rights under the 14th amendment, designed contents of the files or mail." Finally, to the to protect family relationships,Roe v.Wade, 410 U.S. 113 (1973), "Op cut., Woolley v. Hoffman-LaRoche,99 NJ 284, 491 A.2d &lank, v. Connecticut, 381U.S. 479 (1965); and rights under 1257 (1985). the Fourth Amendment, designed to limit unreasonable searches "See: Karen Nussbaum and Virginia DuRivage, "Computer and seizures.Rata v. United States,389 U.S. 347 (1967), and Monitoring: Mismanagement by Remote Control," 66Business progeny. Both of these branches require state action. Even sup - and Society Review16 (winter 1986); and Nine to Five: The Na- psing that monitoring might be considered a "search" under tiopol Association of Working Women,Computer Monitoring the fourth amendment, it is unlikely that an employee would and Other Dirty Tricks,April 1986. be found to have a "legitimate expectation of privacy" in his "The tort of Intrusion may be applicable, if such monitoring or her performance at a given task.Id. amounts to an invasion of the employee's solitude or seclusion, "These figures are fromCompilation of State and Federal et en if there is not physical intrusion. Prosser on Torte, p. 807. Privacy Laws(1984-85 ed.), Privacy Journal (Washington, DC: If the private activity is publicized, there may also be a tort Privacy Journal), and January, 1986 eupp., but see:Congres- for public disclosure of private facts. sional ResearchReports, Mar. 21, 1986, which reports that 22 "For public sector employees, an action may arise directly States have adopted laws regarding employer use of polygraph under the Constitution. For private sector employees, a tort machhiss, and that 10 have adopted laws addressing employee action may lie. The Electronic Communications Privacy Act access to records. of 1986 is ambiguous as to whether an employer might access "The legislation prohibited "the use of any monitoring de- the contents of its employees' computer files. The prohibitions vice, without the express consent of the employee, by means of the Act speak to an "electronic communication while it is of which the surveillance of employees might be effectuated." in electronic storage." 18 U.S.C. f2701(a) (as amended). While The term "monitoring device" included "any device, electronic, perhaps not intended asa communication when written, all files mechanical, Anti, or photographic" by which "appearance, ac- in a personal computer are potentially communicable. Further, tions, or speech" could be monitored. cite.Re. Opinionof the Act's prohibitions do not apply to "the peree-.4 or entity Justices,8b6 Mese 758, 250 N.E.2d 4481 I- providing a wire or electronic communications service," or to 12i Ch. 4Electronic Work Monitoring Law and Policy Considerations 109 extent that the transactions monitored by com- employee's knowledge of a capability for mon- puter become part of the employee's record of itoring," but must instead be based on a employment, compliance with procedures set knowledge (or imputation of knowledge) that out in the Privacy Act of 1974 (governing only certain types of phones or phone conversations Federal employees) or several State privacy will be listened to." Similarly, the business statutes may be necessary. extension exemption applies only to the interception of particular calls as a part of the en- Telephone Service Observation terprise's ordinary course of business," and Unlike computer-based monitoring, which even at that, one court has held that personal primarily raises serious legal issues only when calls may be intercepted only to determine it is used to tromote ends that are ilk gitimate, their nature, but never their content." the legal difficulties with telephone service ob- Title III and the Electronic Communications servation lie primarily in the manner in which Act of 1986 appear to be the excluzNe, albeit it is carried out. extensive, legal framework for issces that may The principle law governing service obser- emerge from telephone service observation. vation is still Title III of the Omnibus Crime Other legal theories, such as the common law Control and Safe Streets Act of 1968, subject right of privacy and (for governmental employ- to the amendments involved in The Electronic ee:1j the fourth amendment's prohibition on un- Communications Privacy Act of 1986." Title reasonable searches and seizurzas, while possi- III forbids the interception of the contents of bly forming the basis fog a legal action, are telephone calls by government or private per- nproven in the context of service observation. sons, except by judicial authorization." This "A recent case held that, 'although public em- blanket prohibition on "wiretapping." how- ployees are protected by the fourth amend- ever, is subject to two exemptions that per- ment, their expectation of privacy must be mit telephone service observationthe consent balanced against the government's need for and business extension 3xemptions." Both supervision, control, and efficient operation of exemptions have been construed narrowly by the workplace.t9 Moreover, ths government is courts. Consent cannot be implied from the not held to a "probable calm ." standard; in- a "user or that service with respect to a communication of or stead, its actions are assessed under a "reasona- intended for that user " 18 U.S.C. 49701(b) (as amended). bleness under the circumstances" standard. "Title III of the Omnibus Crira Ceitrol and Safe Streets Title III applied only to aural communications, Act, 18 U.S.C. §§ 2510-2520 11976). Public Law No. 90-351, 802, 82 Stat. 212, as amended by The Electronic Communica- but The Electronic Communications Privaci tions Privacy Act of 1986, Public Law No. 99-508, 99th Cong. Act of 1986 extends the coverage of Title 18 2d Bess., Oct. 21, 1986. to address analogous concerns present in the The relevant portion reads: Except as otherwise specifically provided in this chapter any service observation of the content of data com- person who munications." (b) willfully uses, endeavors to use, or procures any other person to the protection of the rights or property of the carrier of such to use of endeavor to use any electronic, mecLanical, or other de- communication . .." 18 U.S.C. c2511(2)(06). vice to intercept any oral communication. . . shall be fined not more than $10,000 orImprisoned not more "Watkins v. L.M. Beny & Co., 704 F.2d 577 (11th Cir. 1983); than five years, or both. Campiti v. Walonis, 811 F.2d 387 (1st Cir. 1979); Croaker v. 18 U.S.C. 12511(10 U.S. Department of Justice, 497 F.Supp. 500 (D.Conn. 1980). The statute also provides for a civil remedy and statutory "Watkins v. L.M. Berry, supra; Jandik v. Village of Brook- damages. 18 U.S.C. 12520. field, 520 F.Supp. 815 (N.D. Ill. 19E). *Section 2511(2)(d) of the law permits interception "where "Watkins v. L.M. Bony & Co., supra; citing Briggs v. Amer- one of the parties to the communication has given prior con- ican Air niter Co., 890 F.2d 414 (5`,.4 Cir. 1980). sent to such interception," and Section 2511( 1Xb) excludes from "Watkins v. L.M. Berry & Co., supra at 589. In essence, the coverage any telephone or telegraph instrument, equipment court held that, once the personal nature of the call is known, or facility, or any component thereof. ..being used 17 the sub- the employer must hang up. scriber or user in the ordinary course of its business; .." In "O'Connor v. Ortega,107S.Ct. 1492, 55 USL vt"40511987). addition, communication.; common carriers may "inkecept, dis- "Title III convened only oral communications, 18 U.S.C. close, or use (an employee's telephone conversations) in the nor- 42510., cf. U.S. v. New York Telephone Co., 434 U.S. 159(1977) mal course of his employment while engaged in any activity holding the a communication -rider Title III must be capable which is a necessary incidentthe rendition of his service or of being overheard. .12e) 110 The Electronic Supervisor. New Technology, New Tensions

Telephone Call Acc.Jung amendment could successfully be brought Many of the legal issues surrounding the use against the practice of telephone call account- of telephone call accounting center on the in- ingeven against its surreptitious use by po- cidental information generated by a call-ac- lice in order to obtain evidence for a criminal counting system. In other words, although the indictment." The Electronic Communica- employer may not purposely set out to infringe tions Privacy Act of 1986, while providing employee rights, many of the by-products of stronger protection than the fourth amend- call-accounting systems may in fact threaten ment by requiring a court order for the appli- employee privacy. In the act of trackingre- cation of pen registers and trap and trace cipients of calls originating from certain phone devices," is applicable to telephone call ac- numbers, employers must, of necessity, obtain counting." However, depending on how the information on the identity of the persons information gleaned from call-accounting sys- called, and the nature of the call (business or tems is used and whether it is disclosed, all nonbusiness). Depending on how the audit is employees may have rights under common law conducted, and how closely focused on indi- theories of privacy or defamation. viduals it is, a "picture" of extra-employment activity may be obtained merely from the iden- Effects tity of the destination phone numbers,even if the intent of the audit is to identify non- Aside from the abusive purposes and meth- business-related calls." Once the information ods of electronic monitoring discussed above, is collected, it may be intentionally or acciden- the most salient legal issue presented by mon- tally disclosed to people whom the employee itoring concern its health-related effects on par- would prefer remain unaware. Although a call- ticular workers. Other, less tangible, effects accounting audit may disclose misuse, such "Pen registers, which are devices that attach to a telephone misuse may not be the fault of the employee line to record dial pulses, may be used in law enforcement to (especially when others have access to the em- obtain information oa suspects without the need of a search warrant. Smith v. Maryland, A2 U.S. 786 (1980). By extension, ployee's phone)a claim that may be hard to the use of call accounthog systems (that achieve very much the prove. same resultalbeit, in a more detailed fashion), which often form an integral part of modern PBXs, would seem to raise no unique Federal employees are the most protected fourth amendment problems, especially when they are used on segment of the labor force. If the records gen- the employer's premises and it is known by employees that they exist (thus raising no "subjective expectation of privacy"). erated by the telephone call-accounting sys- "18 U.S.C. Ch. 206 "Pen Registers and Trap and Trace tem form part of a "system of records" per- Devices." sonally identifiable to particular individuals, "litle 18 has been amended so as to specifically exclude a then, under the Privacy Act of 1974, the Fed- "provider of electronic communication service to record the fact that a wire or electronic communication was initiated or com- eral employee is subject to a number of proce- pleted in order to protect such provider, another provider fur- dural safeguards concerning notice that such nishing service toward the completion of the wire or electronic communication, or a user of that service, from fraudulent, un- records are being collected, the subsequent use lawful, or abuse use of sue_ service" 18 U.S.C. 2611(8)(h)(ii) to which they can be put, the right of the em- (emphasis added) ployee to correct or amend the records, the ne- Call-accounting software might arguably be a "pen register" cessity, and the acquisition for lawful purposes for purposes of The Electronic Communications Privacy Act of 1988, since, like SMDR PBX equipment, it is defined as "a of those records." device which records or decodes electronic or other impulses which identify the numbers dialed or otherwise transmitted on For public employees in general, it is unlikely the telephone line to which such devices is attached... ." How- that a constitutional claim under the fourth ever, SMDR equipment and software is excepted from the prohibitions of the act "(pen register) doss not include any device used by a provider or customer of a wire or electronic communi- "Calls to collection agencies may re. sal debt trouble; calls cation service for billing. or recording as an incident to bil ling, to counselor may reveal psychological or marital trouble; calls for communications services provided by such provider or any to employers in similar businesses may reveal an jutted to change device used by a provider or customer of a wire communication jobs; and calls to the news media may reveal the source of service for cost accounting or other like purposes in the ordi- " kaki." nary course of its business .." 18 U.S.C. P128 (emphasis U.S.C. 1552(a), infra. added)

123 Ch. 4Electronic Work Monitoring Law and Policy Considerations 111

are sociological in nature, and concern the hn- Many of those opposed to work monitoring pact of monitoring on the overall climate of focus on its health, and particularly stress- work in the United States. related, effects. Moreover, many of the "case studies" submitted to OTA by a variety of The literature on stress and work monitor- unions emphasize the deleterious effects that ing is not broken down cleanly into the three monitoring has on employees' health. The ap- categories of monitoring dealt with in this re- proach of this section will therefore be to take port. Computer-based irinitoring, telephone the assertion that monitoring causes stress and service observation, and telephone call account- health-related problems as a given, and ask: ing may each entail widely different work envi- how might present law address concerns over ronment factors (e.g., different organization these effects of monitoring? factors, different physical relationships between the employee and the technology on All State jurisdictions recognize stress as which the employee is working, and different a compensable injury, either under their tort expectations). Thus, any particular findings on or Worker's Compensation laws."' However, stress are likely to vary widely between types "stress" is subject to a wide variation in defi- of work monitoring. Nevertheless,,at of the nition in the way it is manifested, and the man- legal analysis that follows will hold true so long ner and context in which it is inflicted. Stand- as stress can be shown to be caused by or asso- ards of proof for its existence, and the degree ciated with the particular type of monitoring of injury necessary for compensation, are de- in question. terminative of whether monitoring-induced stress (if it can be shown to exist) will rise to This section relies on the analysis developed the level of a legally recognized claim. in chapter 2 examining the evidence for computer monitoring as a cause of stress, and ap- Worker's Compensation, which was estab- plies relevant law in light of this evidence. The lished in all 50 States to provide compensa- principle conclusions of chapter 2 are: 1) that tion on a "no-fault" basis for the loss of abil- evidence that computer monitoring, per se, ity to earn wages, is a substitute for employer causes stress is Ex igestive, but not conclusive; tort liability. Most Worker's Compensation and that 2) many other aspects of job design statutes require that the injury be accidental, and work enviornmente.g., computer pacing, and that it arise out of or in the curse of em- heightened work pressure, routinized activi- ployment. Courts have read these require- ties, variable workloads, lack of control over ments expansively in recent years, so that even the task, lack of decision latitude, lack of peer "accidents" that are slow in manifestation and and supervisory support, and fear of job loss which rise out of employment-related risk are may also cause stress among VDT office work- compensable.102 This means that, as a thresh- ers. Research to date has not succeeded in sep- old matter, chronic stress caused as a result arating the effects of computer monitoring of monitoring may be compensable. from effects of these other workplace factors, insofar as stress is concerned.m not separate computer-based monitoring from customer service observation and call - accounting systems, as is done in this "%tress due to work monitoring should be considered sepa- report. However, the way in which monitoring was described, rately from that due to the use of Video Display Terminals as a system of electronic supervision or feedback in work orga- (VDTs) per se. A recent OTA report concluded that "evidence nization, would include the first two forms of monitoring con- for a relationship between stress-related disease and VDT work sidered in this report. Moreover, machine pacing the use of is still sparse." U.S. Congress, Office of Technology Assess- a computer to control when and how fast a task is performed ment, Automation of America's Offices, OTA-CIT-287 (Wash- can lead to anxiety, depression, boredom, dissatisfaction, fre- instal, DC: U.S. Government Printing Office, December 1985), quent health complaints, and decreases in productivity with p. 150. In part, this is due to the methodological problems in increases in error rates. Id. at 128-29; citing Selveady and Smith attributing stress and stress-related ailments to any one factor fads.), Machine Pacing and Occupational Stress (London: Tay- or combinations of factors in the workplace. The OTA report lor & Francis, Ltd., 1981). did conclude, however, that electronic monitoring in general is 10118 Larson, The Law of Workmen's Compensation. 142 in- associstad with the symptoms of chronic arousal, and can lead fra (1982), and Prosser, Law of Torts, 112. to increased anxiety, fatigue, psychosomatic complaints, and '0l1A Larson, The Law of Workmen's Compensation, job dissatisfaction. Id. at p. 180. In this regard, the report did 1137.00.39.00 and 1 p.m.®, infra ,124 112 The Electronic Supetvisor. New Technology, New Tensions

Even if stress meets these threshold require- ceptance as an ordinary part of the work envi- ments, not all States recognize psychological ronment in which it is deployed, any stress that effects as compensable injuries caused by it causes (if any) is arguably "not unusual." stress.'" Only a handful of States would al- In order to be recognized in most States, there- low recovery for monitoring-induced stress, if fore, electronic monitoring-induced stress must that stress can be characterized as "not un- manifest itself in a physiological symptom to usual," or "not in excess of everyday life or be compensable. employment."101 Otherwise, in order to be a compensable injury under Worker's Com- Finally, many States today recognize the pensation laws, stress must be "unusual, "106 tort known as "intentional infliction of emo- or even "sudden," "frightening," or "shock- tional distress." Although compensation for ing."10e As electronic monitoring gains ac- emotional distress previously required some sort of physical invasion or injury, such as a battery or assault, this is no longer the law in "Florida, Georgia, Kansas, Louisiana, Minnesota, Montana, a substantial number of jurisdictions. This not- Nebraska. Ohio, and Oklahoma do not recognize purely mental or emotional injuries as the result of stress.Emotional Stress withstanding, monitoring-induced stress is un- in the Workplace,op. cit. All jurisdictions recognize stress- likely to be actionable under tort law. First, related ailments, whether "mental" or "physical," that have the distress-producing act must often be of an an antecedent physical injury. And, with the exception of Ohio, all States recognize "physical" disabilities resulting from stress. "extreme and outrageous" naturea charac- 1A Larson, 142.72 at seq. terization that is probably not fitting to elec- "California, Hawaii, Kentucky, Michigan, New Jersey, Ore- tronic work monitoring. Secondly, as with the gon, a 4 West Virginia. See generally "Emotional Stress in the WorkplaceNew Legal Rights in the Eighties," National Coun- tort of invasion of privacy, consent (found in dl on Compensation Insurance, 1985. the implied or explicit terms of an employment 'Arizona, Arkansas, Maine, Massachusetts. New Mexico. contract) will probably vitiate an employee's New York, Pennsylvania, Rhode Island, Washington, Wiscon- sin, and Wyoming.Id claim. Finally, many States still require that "Illinois, Maryland, Missouri, Masissippi, South Carolina, physical illness or some other nonmental ef- Tennessee, Texas, Virginia.Id. fect be present before allovr;ng recovery.

PART CONCERNS NOT ADDRESSED BY LAW

Table 16.A Framework for Addressing Electronic Work Monitoring

Concern Criteria Example of applicable law Possible "gaps" in law Purpose 3f monitoring Fairness Relevance National Labor Relations Act; Due processtype Completeness Civil Rights Act; Merit System guidelines for private Targeting Principles (as administered In employees EEO, OSHA,ERISA, EPA, etc.'); State Law on Privacy; Constitu- tional Law' Manner and method of monitoring Autonomyintensiveness State Law on Wrongful Dis- Privacy In transactional Dignity Intrusiveness charge; State Law on Privacy; information; "Human Privacy Visibility PCIE Guidelines; Title III of Om Rights"type law; Type nibus Crime Control and Safe laws requiring notice Leakiness Strests Act; Electronic Commu- of monitoring Permanence nications Privacy Ac;; National Labor Relations Act; Privacy Act of 1974' Effect of monitoring Health Frequency Worker's Compensation StatutesGuidelines/regulations Stress Continuousness on Stress-Causing Labor on stressinducing Regularity labor practices Control 'Applies only to Federal employees. SOURCE Of of Technology Assessment, 1987 Ch. 4Electronic Work Monitoring Law and Policy Consijerations 113

If we look at the types of concerns raised transactional data becomes part of a per- by electronic monitoring, and at those ad- sonally identifiable record, then the subse- dressed by law, several broad conclusions quent use and disposition of that record is follow: regulated by both Federal and State law. With some exceptions, no law prevents an PurPose employer from using the monitoring systems considered in this report in a secretive, In general, public employees are better pro- low visibility manner. For example, an em- tected against "unfair" monitoring practices ployer is not under a positive duty to reveal than private sector employees. Constitu- to its employees the fact that their keystrokes tional due process protections afford pub- outgoing lic sector employees the opportunity to chal- are being counted, or that their lenge &missals, demotions, or other actions long-distance calls are being documented. based on monitoring that is irrelevant, un- Unless the employee has an expectation of privacy in the activity or location while at fairly targeted, or incomplete. Although the work, the employer is free to collect as much doctrine of employment-at-will is gradually being eroded in State courts, a suit for un- information on the employee's performance just dismissal of private sector employees as it sees fit. based on unfair monitoring is unlikely to Although employees may regard some meth- succeed. ods of monitoring as an assault on their dig- Aside from provisions made in union con- nity or autonomy, there is no legal right to tracts, no law compels an employer to im- be treated with dignity or as an autonomous plement monitoring with fairness, unless it person. Unless the monitoring technique is taken intrusiveinvading either the bodily or can be shown that the employer has mental integrity of the person (as, perhaps, actions against certain employee(s) based on that in drug testing or brain wave analysis) race, sex, or religion or for motivations there are no legal protections against moni- are against narrow public policy exceptions toring because it is "dehumanizing." Al- to the employment-at-will doctrine. though monitoring may affect interpersonal Electronically monitoring formerly unmoni- and power relationships at work, no law pre- tored tasks may change the very nature of vents employers from using intense, low vis- that task by accommodating the task to the ibility monitoring. For example, using com- system of measurement. While some em- puters to set the pace at which tasks are ployees may object to this as an unbargained accomplished, to measure the employees' for change in job description, no legal pro- performance, or to document time away tections, aside from employment contracts, from a terminal, are not prohibited by law. exist. Effects Manner and Method Although some forms of monitoring may Monitoring most often involves the collec- cause stress, and may therefore havehealth tion of transactional, rather than substan- effects, no law currently protects workers tive, information about employees' perform- against stressful environments, whether the ance. No privacy protections exist against stress is caused by monitoring or by other the collection of transactional information aspects of the work environment. Lawmak- on employees' activities while at work. For ing with respect to stress in the work envi- example, no law prevents the collection of ronment is not unprecedented, however, and telephone usage data in a call-accounting several foreign countries have adopted leg- system, or of performance data in a com- islation that attempts to address stress in puter-based monitoring system. If, however, the work environment.

1 2 114 The Electronic Supervisor: New Technology, New Tensions

In some cases, stressmay be a compensa- vats realms may vitiate claims that the indi- hie *jury under Worker Compensationstat- vidual's expectation of privacywas a reason- utes, but stress-related health effectsare dif- able one.'" ficult to prove, andare not accepted in a majority of State courts. Work Environment Changes Much of employee behavior in the past went unobserved or undocumented simply because What Does the Future Hold? the technical facility for monitoring it didnot exist, or was too cumbersome to employ. As Depending on the influence of a variety of noted in chapter 2, however, the business, economic, and social factors (see part use of mod- II), the next 10 to 15 years may em information technology enables employers see substan- to keep track of more informationon employee tial changes in monitoring technologies and performance in much greater detail. Given this settings in which they are conducted. These new ability, much of the "looseness" of previ- changes may raise a whole new set ofconcerns warranting continued congressional ous work environments may be reduced or sc.. utiny. eliminated. What was in the pasta de facto perquisite of the job, suchas a limited ability to make nonwork-relatad plnue calls,or an oc- Incremental Changes casional break from a given task,may in the future become grounds for disciplineor dis- Today's monitoring techniques, whichare missal. In such a case, the question isnot in and of themselves neither illegalnor clearly whether the employer is "within his rights," in conflict with employer-employee custom, but whether the work environment should be- =cease.* form a precedent for future moni- come so rigorously controlled as to eliminate toring techniques. As these techniques become all discretionary employee activity. more sophisticated and permeate the work environment, law and lawmakers may havea Qualitative Changes difficult time distinguishing between eachnew innovation and the one that preceded it. The As discussed elsewhere in chapter 1,a clear law and practice that grows up arounda par- distinction can be made between work moni- ticular form of monitoring may easily assimi- toring and worker testing; the former isan late a new, incremental change in the technol- evaluation of the performanceor behavior of ogy or application. The cumulative changes in an employee, while the latter is an evaluation work environment may be great, despite their of an employee's physicalor mental state. In flue: and hence imperceptable nature. The theory, it may be possible for legal rules to be ork for analyzing claims to privacy, framed in accordance with this distinction. which relies on an assessment ofan individ- However, while the distinctionmay be rela- ual's "reasonable expectations," 1°7 can easily tively easy to make in theory, it is breaking become teinple descriptive statements of what down in fact. Research in the field ofpsy- the monitoring milieu is, rather than prescrip- chophysiology, discussed elsewhere in thisre- tive statements of what ought to be. An indi- port, may be able to correlate behavior with vidual's knowledge that certain technologies psychophysiological states; blurring the bound- are capable of intruding into previously pri- aries between monitoring work and monitoring the worker. "For example, inCalifornia v.Moab, a recent case in which 'elate v. United States, 889U.B. 347 (1967) announced thepolice used an aircraft and camera to obtain evidence of mari- constitutional "reesonable expectation of privacy" standard that juana growing in a suspect's backyard, the Court concluded her guided the Swum Court ever since. It has twocomp- that Itjhe Fourth Amendment simply does not require thepo- mate whether the individual actually expected that his or her lice traveling in the public airways. . . to obtain a warrant in activity remain Ovate, and whether that expectation isone order to observe what is visible to the naked eye." The stand- tied sodety is premed to accept as reasonable. This standard ard for determining what expectations of privacy are reason- has gradually been applied in nonconstitutional, tart-privacy able is, in part, dependant on the state of technology for intrud- cases. ingon that privacy. x.27 Ch. 4Electronic Work Monitoring Law and Policy Considerations 115

PART V: POLICY OPTIONS Before addressing the problem of how Con- than national policy, are the best way of ap- gress might act, it is first necessary to con- proaching what appears to be situation-specific sider whether and when action may be appro- problems (see part III). Under these circum- priate. Some factors suggest that a "wait and stances, Congress may want to avoid legislat- see" posture may be appropriate; uncertainty ing on the issue of monitoring per se, and in- about whether monitoring causes stress, the stead make monitoring an item for compulsory lack of judicial precedent, the possibility of pri- arbitration or collective bargaining under Fed- vately negotiated restraints on monitoring, eral labor law. and marketplace checks on monitoring are among these. Other factors indicate that Con- This, of course, does not necessarily ensure gress may want to act now to alleviate grow- an outcome that is satisfactory for the majority ing concern about monitoring in the workplace. of monitored workers, who are not unionized These include the lack of union representation and are therefore powerless to negotiate a fair in the bulk of the monitored work force, inade- monitoring practice, or any other aspects of quacy of current law to address concerns over the quality of work life, through the collective health, privacy, and dignity, difficulties of leg- bargaining process. Furthermore, a growing islating against powerful economic interests segment of the work force are temporary work- at the State level, and increasing sophistica- ers, who, since they come and go on a weekly tion of the technology itself. Several possible or monthly basis, have little ability to improve directions of Federal policy are described the quality of worklife. below. There is the argument that natural "market krces" may tend to limit unfair monitor inr and preclude the need for congressional ac- Option I: Take no Federal action concern- tion even on behalf of nonunionized workers: ing work monitoring at this time. employee backlash, low morale, cud high turn- Questions of the fairness of work monitor- over should dissuade employers from monitoring practice would be left, as they are at ing practices that their workers find onerous. present, in the hands of stakeholders, em- If monitoring is indeed stress-producing, then ployers and employees. In industries where la- employers who use it will inevitably see the bor unions are active, collective bargaining effects of stress on diminished quality and out- with regard to technology change, monitoring, put of its product or service. The response to and methods of evaluation should continue un- this is that many monitored jobs comprise rou- der current rules. tine work subject to and indifferent to a high turnover rate. And, in many instances, high Although many unions have adopted posi- attrition works to the employer's benefit (by tions opposing electronic work monitoring (see lowering the costs of pension, salary increases, table 11), their bargaining strength with re- etc.). Thus it is not clear that "natural" checks spect to it, whether by informal negotiations will be sufficient to ensure that monitoring is or by formal collective bargaining or arbitra- not abused. tion, is probably not great. However, some forms of monitoring take place within specific If natural checks are not sufficient, politi- industries or companies. An argument can cal action is still available. Unions and other therefore be made that, pending the develop- interest groups have worked to pass State level ment of a longer history of negotiations be- legislation on monitoring, service observation, tween labor and management on this issue, or VDT health and safety. These activities will monitoring is beat addressed at the union level; probably continue. Some of these attempts the parties concerned are most familiar with may be successful, giving rise to a variety of the specific problems, and contracts, rather legislative or regulatory approaches to deal-

128 116 The Electronic Supervisor. New Technology, New Tensions

ing with issues related to electronic monitor- tant as or more important than monitoring in ing. Some may serve as models for Federal ac- contributing to stress-related illness, and that tion at some later time, should the need for the these should also be co-iered by protective leg- harmonizing effect of national legislation be islation or regulation. seen more dearly in the future. Option 3: Consider Federal legislation aimed Option 2: Establish whether stress effects at gaps in current law. This could be in of electronic monitoring are an occupa- two postalle directions: general legislation tional health hazard; if they are, consider aimed at establishing certain rights for creating Federal legislation or regulations employees within the workplace or surgi- governing the use of electronic moni- cal legislation aimed at specific monitor- toring. ing practices. The effect of monitoring on stress and healthissues which might vovide the policy- There have been no court cases challenging maker with the most direct and least value- the types of monitoring considered in this re- laden approach to acting on monitoringis in port. Two conclusions can be drawn from this. a state of scientific uncertainty. There exist The first is that, until the judiciary acts, Con- few authoritative studies on the effects of elec- gress has very little clue (aside from analyses tronic monitoring on health. Many studies and of tie sort found in part III of this chapter) informal polls of workers suggest that moni- as to the type of legal inadequacies it should toring has stressful effects, and there is a cer- address, and ought therefore to wait to legis- tain common sense appeal to the idea that late on work monitoring. The second is that working in fast paced, highly monitored envi- current law is inadequate to even form the ba- ronments may be very stressful. However, not sis for a lawsuit, and that Congress must take much is known about the types of monitoring the lead in "roviding rights to monitored em- that are stressful, how stress might be reduced, ployees, should it decide that certain forms of or how stress due to monitoring manifests it- monitoring are pernicious. self (if at all) in physiological symptoms. Un- til more is known about the effects of moni- Current worker protection legislation gives toring on health, policy action under a "stress" workers a variety of rights, such as the right rationale may be premature. The policymaker to organize, to bargain collectively, to mini- may consider it appropriate, therefore, to ini- mum wage, and increasingly, the right to know tiate studies on stress in the workplace, and about health and safety hazards that form part on the role that monitoring plays in such stress. of the working environment. However, U.S. law has not heretofore involved itself deeply The National Institute of Occupational Safety in quality of worklife issues nor in issues of and Health would seem to be the logical agency personal privacy or dignity in the workplace. to supervise or carry out studies of stress as There is no legal right to be treated with dig- a workplace hazard. Specific studies of moni- nity or as an autonomous person. There is no tored workers would have to be done with an legal right to a well-designed, interesting job, eye to understanding the effects of monitor- nor is there law that compels employers to con- ing independent of other workplace stressors, sider employee input in decisions about new a major deficiency in existing studies. In addi- technology or new monitoring procedures. To tion, however, it would be useful to understand the extent the law treats privacy in the work- more about the phenomenon of workplace place, it looks to r standard of what an em- stress in general, given the rising number of ployee might reasonably expect to remain pri- worker compensation claims and other evi- vate; as mentioned earlier in this chapter, this dence of the growing importance of stress in standard may fail as a guide for action in the occupational health. Research may reveal that face of employer's increasing use of monitor- other factors in the workplace are as impor- ing, surveillance, or testing technologies.

129 Ch. 4Electronic Work Monitoring Law and Policy Considerations 117

That these issues are not currently addressed present law, such as the Privacy Act and the in law does not mean they could no be. As Fair Credit Reporting Act, regulates what can is discussed in appendix A, a number of other be done with transactional information once countries have quality of worklife legislation. collected, it does not forbid its collection as Such legislation could give guidelines on the such As we have seen, however, the collection rights to health, safety, privacy, constitutional of transactional information, particularly if protections, or information that employees can done on an intensive basis (see part II) can expect to enjoy in the workplace. As indicated arouse feelings of having one's privacy, dig- earlier in this chapter, the erosion of the doc- nity, and autonomy invaded. Moreover, be- trine of "employment-at-will" through anti- cause of the power of computers to generate discrimination, health and safety legislation, profiles and crossmatch many transactions, and public interest concerns, has already transactional information can yield informed marked some involvement of the U.S. Govern- estimates of the substantive content of com- meat in regulating the work environment. The munications or patterns of behaviorit can be, issue of electronic monitoring in offices is too in other words, a "back door" for getting at narrow to serve as a basis for comprehensive personal information that existing law reg- work environment legislation. It should be just ulates. one factor of many to be considered in determining what rights U.S. citizens have in the Certainly, to forbid or regulate the collection workplace, both as employers and employees. of all transactional information would be unreasonable. Much transactional data collected However, assuming that blanket legislation by electronic monitoring software is used to on worklife quality is neither wise nor dogra- monitor equipment utilization, to track totals ble, Congress might address concerns over spe- of tranbaetions made, and to Bete: mine whether cific issues through the use of specific amen- security systems are working properly. The col- datory legislation. If, for example, telephone lection of transactional data becomes most call accounting is an area of particular concern, subject to controversy when it is collected Congress might address the problem specifi- about the performance of an individual worker. cally by amending the Electronic Communi- It may be that Congress would choose to treat cations Privacy Act to comport with what it electronic monitoring as a "right to know" is- considers "fair" monitoring practice. The sue for workers; that is, employers could have President's Council on Integrity and Efficiency the right to collect whatever kind of transac- guideline may form a template for such legis- tional data they wish about employee perform- lation, or instead, Congress may mandate a:ter- ance, but would be required to give employees natives to telephone call accounting discussed access to, and if need be, correct, this infor- in chapter 3 of this report. mation. Another example of an area of the law not As this report has indicated throughout, currently addressed, and on which Congress however, the issue of work monitoring cannot may wish to act, is what might be called trans- be adequately understood, nor appropriately actional pri^ cy, or the collection of "informa- addressed, in isolation from larger labor-man- tion about information." For example, the agement, privacy, and the health and safety number of keystrokes, the number of visits to context in which it is embedded. Nor will spe- the bathroom, the destination of calls, etc., are cific policy actions taken with respect to par- all type of information about transactions, ticular forms of monitoring neitessanly end the rather than about the content of communica- controversies arising out of the application of tions or activities (see part II).'" Although new technology forms in the workplace. The policymaker should therefore be aware that an exclusive focus on the forms of monitoring con- 'N'hansactiosal information, it will be recalled, differs from sidered in this report will at best form the ba- =bawdw information, in that the letter reveals the content or meaning of coo nos or documents. Transactional in- sis for a series of patchwork solutions to what formation, In contrast., reveals facte about oomnumications or has been a perennial issue between workers and documents. employers. 130 Appendixes

131 Appendix A NotesonComputer Work Monitoring in Other Countries

Introduction issues, including introduction of computer technology and use of work monitoring. In addition many Because the American economy is so tied to the of our competitor nations have a tradition of gov- global economic structure, incoming attention hes ernmental involvement in labor relations and a been given to approaches taken by other industrial more developed tripartite government-labor-man- nations in their efforts to maintain a competitive agement approach to industrial policy, economic ed. and adapt to microelectronic technology. It development, and growth. This approach has no is kWul to look at the experience of other coun- real equivalent in this country, but there are in- tries in evaluating what lessons might be learned creasing calls for some efforts in this direction. with respect to work monitoring. There are a num- A traditiot of worker participation, including ber of ways in which the different legal structure employee involvement in applying new technology and the institutional structure of labor relations in factories and offices, is also more developed in has resulted in different policies toward moni- some other industrial countries. In some countries, toring.' this participation depends almost entirely on the The American labor relations system and labor collective bargaining process, as for example in law model differ importantly from many other in- England, Australia, and Canada, where the adver- dustrial nations. Most other industrial democra- sarial labor relations model is closer to that of the cies have a higher level of unionisation. Tablo 17 United States. In others, however, worker partici- &Iwo the approximate percentage of the work pation or "co-determination" is required by law, .force that is unionized in the United States and as in Sweden, Norway, Germany, andHolland.In in 15 other developed countries. Given the higher some of these countries as well, work environment level of union participation in the work force in laws define certain workplace rights and give some of these countries, unions have naturally had guidelines for job design. Tables 18 and 19 list some a greater influence in a wide variety of workplace work environment legislation. The Norwegian Work Environment Act of 1977 reads, in part: General requireneste.Technology, organisa- 'Porta d die section draw heevily item Stever Deutsch, "I'be In- teredioned Corset at Lebor-Managenrat Reindeer Implication for tion of the work, working hours and wage systems Woripince Mesiteting," pew pegrared for OTA, shall be set up so that the employees are not exposed to undesirable physical or mental strain and so that

Table 17.Percent of Unionized Waters by Country Table 13. European Work Environmen! Acts Providing Goals for Changing Worldns; United States 22% Conditions During Office Automation France 23 Japan 30 Federal Republic of Canada 35 Germany The Works Constitution Act of Switzerland 35 West Germany 1972 40 The Netherlands Netherlands Working Environment Act of 1980 40 Norway Work Environment Act of 19776 Italy 50 Sweden Ireland Working Environment Act of 1974 50 Denmark Act Respecting the Working Britain 52 Austria Environment 60 German Democratic Belgium 65 Republic Norway ..65 Labour Code as Amended, June 1977 Denntwk 70 These Acts dell with of working conditions, thus we specifically appecable Finland 86 _tooleosautomation. Sweden 90 °Also known as Act Reepecting Wows Protection and the Working En. SOURCE: Alm mean and Musee11 Pips, "Wectfonic kinfiNenna In Min MR* Wronment. Pleased Couneise," contrador report mewed for OM, INS. SOURCE: Office of Technology Assessment, 1094.

121 63-982 0 - 87 - QL: 3 4 132 122 The Electronic Supervisor: New Technology, New Tensions

Table 19.European Acts Requiring Employers To Provide Information About New Technology and Worker Representation in the Decisions About the Quality of Work life

Country Information provision Work representation Federal Republic of Germany Works Constitution Act of Works Constitution Act of 1972 1972° The Netherlands Works Council Act of 1979' Works Council Act of 1979 United Kingdom Employment Protection Act of Employment Protection Act of 975° 1975 France Act No. 82-915 of 28 October Act No. 82.689 of 4 August 1982 1982 Sweden Act Representing Co- Act Representing Co- Determination of Work of Determination of Work of 1976 1976° ainfonnation is provided to a Work Council siddi can be a croaseection of all interested pollee la:iodine persimmon, Industry, and labor. binformation is provided to the worker or worker representative. SOURCE: International Labour Organization, Automation: Work Organization and Occupational Straus (Geneva, Switzerland: 1964).

their possibilities of displaying caution and observ- perience in job redesign, improvement of working ing safety measures are not impaired.... life quality and participative management. There Arrangement of walk. The individual employ- is evidence of a substantial growth of a more par- ee's opportunity for self-determination and professional responsibility shall be taken into considera- ticipative and cooperative trend in the American tion when planning and arranging the work. labor-management relations community,' and the Efforts shall be made to avoid tmdiversified, Federal Government has started an initiative on repetitive work and work that is governed by ma- labor-management cooperation in the U.S. Depart- chine or conveyor belt in such a manner that the ment of Labor.' employees themselves are prevented frog.. varying Since collective bargaining covers only one-fifth the speed of the work. Otherwise efforts shall be of the American work force, other initiatives have made to arrange the work so as to provide possibil- been explored by advocates and policymakers, in- ities for variation and for contact with others, for cluding the use of occupational safety and health conneution between individual job assignments, and laws and agencies (which somewhat parallel work for employees to keep themselves informed about production requirements and results. environment laws in other nations that have ad- Control and Oneida' systems.The employees dressed job stress, work organization and technol- and their elected union representatives shall be kept ogy concerns). Concern over health effects, stress- informed about the systems employed for planning related worker compensation claims, are among the and effecting the work and about planned changes driving forces for reform here. Thus far, only mod- in such systems. They shall be given the training est changes in Federal legislation and standards, necessary to enable them to learn these systems, such as the OSHA hazard communication stand- and they shall take part in planning them. ard, have been seen. More activity seems focused This law, and the Swedish Work Environment at the State and local level in efforts to pass legis- Act of 1978, were both based on evidence that lation to I Id employees against possible health machine-paced, monotonous work, done in social and safety hazards associated with video display isolation and involving shift work, leads to un- terminals (VDTs). European developments have healthy outcomes in both emotional and physical been a significant stimulus for some of this effort. terms' There is a well-established international communications system across management and union circles which has shared research results and ex- 'Sea, for example Baldness Week, "Th. Hollow Corporation," Mar. 3,1966; Steven Deutsch and Sandra Albrecht, "Worker Partidpstion 'Bjorn Gustavus and Gerry Honshu, Na,' Patterns of Work Re- in the United States: Efforts to Democratise Industry and the Econ- im& The cow at Norway (Odds Oslo University Press, 1961); Barth omy," Labour and Society, vol. S. July-September 1963, pp. 246-269. Oardeil and Bjorn Gustayson, "Work Envirooment Research and Sc- 'See U.S. Department of Labor. Bore= of Luber-Management Rela- old Change Curtest Devdopanots in Scandinavia," Jourealot °veget- tions and Cooperative Programs, U.& Labor Lew and the Attars at ated Ikbeekr. vol. 1, January HMO. Labor-Management Cooperation, 81.116 104 (Washington. DC: 1966).

. 133 App. ANotes on Computer Work Monitoring in Other Countries 123

Policy Approaches to Monitoring in in other matters. According to the Act on Works Selected Countries' Constitution, which created works councils, . the works council will, if no statutory rules or West Germany collective bargaining agreement exist, codetermine: . . . the introduction and use of technical installa- Both the West German Trade Union Federation tions that are intended to monitor conduct or per- (DGB) and a number of individual unions have been formance of employees. active in trying to set ergonomic and work environment standards for VDT users. In 1979-80 Ger- Norway man unions opened a campaign to write "model codes" for VDT work into both industrywide and There is an understanding in Norway, among plant-level agreements. These models inlluded a both unions and employers, that the work meas- ban on individual monitoring, for tr..ample: "It shall urement capabilities of new office technologies not be permitted to monitor the performance of have groat potential for increasing productivity workers, for the purposes of messnrement, control, and helping in planning and management of work. or comparison, by use of the installed [VDT] equip- However, there is a strong aversion to using the ment."' A study of about 50 actual contracts con- information for individual performance meas- cluded during this period found that work moni- urement. toring clauses hie the above were often included. The general trend is to use the work monitoring Part of the concern about monitoring arises at least system to collect information, but to only use ag- partly from the fear of social isolation of monitored gregate data. The social security administration, workers. A study of bargaining agreements found for example, recently introduced a computerized that such clauses are often included in actual la- case handling system. The daily work statistics on bor management agreements as well. individuals are collected by the system, but the Another example of a clause preventing moni- reports issued are aggregated by work groups. At toring comes from the 1984 contract between the the present time the data are available to both C-o___merzbank AG and the bank employees' union and management as part of an experiment union: designed to test the productivity of two different The performance or behavior of employees shall work organization approaches. not be eff ected by means of existing or planned EDP. Similarly, the bank union has included in its tech- Data and programs which serve to verify perform- nology agreements with employers that data on ance or behavior shall be erased. ..; A guarantee work volume and speed be used only at the work shall be given that personal data on the employees group level: which are a by-product of the working process or . . . local regulations laid down under the collective which can be deduced from work process data will agreement are designed to ensure that such infor- not be such as can be used or interpreted as a check mation .. . is not used to evaluate employees. The on personnel behavior or performance. union points out that the only way to assess theim- An interesting feature of the West German la- pact on employment and working conditions of new bor relations scene is the Works Council, an elected technologies.... is by using such work measurement group which management must consult in all mat- devices. However the union stresses the importance ters of "internal order" in an enterprise; its func- of cor 'rolling the use to which the information is put. tions are independent of the collective bargaining process. Works councils are active in the process Sweden of "co-determination," that is, they represent the employees' voice in the selection of technology and The Swedish Codetermination Act of 1978 requires that employee representatives participate in decisions about computer system design, includ ing the possible use of electronic monitoring. Fur- ther, The Work Environment Act of 1978, which guarantees workers the right to a "satisfactory" 'Much of the information in this section comes from Alan F. Westin work environment, is generally interpreted to mean and Russell Pipo, "Employee Monitoring in Othar Industrialized De- mocracies," contractor report preps:ad kr OT 1966. that jobs should be designed to avoid machine pee ."Viest Germany: Workplace AgreementsanNA,ew Technology," ing or individual computer monitoring, if possible. Haropeas Industrial Retie= Review, No. 106, November 1962, pp. 74. Cited in Alan F. Weals and Russell Pipe, "Employee Monitoring Electronic monitoring systems, as negotiated be. in Odom Industrialised Demomacies," contractor report prepaid for tween employers and unions, are generally only OTA, August 1906, p. M. used to measure group performance, as in Norway.

134

1111.1MM,.. 124 The Electronic Supervisor: New Technology, New Tensions

However, in some cases, where the union has It is recognized thatvoidemeasurementmay agreed to the necessity, individual monitoring can be necessary to obtain objective evaluation of and has been used. This has been the case, for ex- the level of production of a group, a section or an ample, where computers are used to keep track of office. However, there shall be no individual work inventory in order to prevent theft. In addition, measurement. nommionized temporary workers are sometimes Such language has been adopted in a few contracts. paid by piece rates, so their work is sometimes The Postal Workers, for example, negotiated for monitored electronically.' work measurement by group in 1981 when new equipment was installed. The Telecommunications Canada Workers got a commitment from British Colum- bia Telephone Co. that data collected on computer- Canada has a voluntaristic labor union system ized cash registers at Phone Marts would be Used similar to that in the United States; although the only for inventory purposes, rather than individ- level of unionization is higher. There are no spe- ual performimce. cific provisions dealing with work monitoring in A task force appointed by the Canadian Minis- any of Canada's national or provincial labor codes, try of Labour also addressed the question of elec- nor have there been any regulations on monitor- tronic monitoring in its 1982 report In the Chips: ing issued by national or provincial regulatory au- Opportunities, People, Partnerships." The task thorities, although several have been considered. force considered monitoring "the most serious What limitations on employer conduct that have manifestation" of the introduction of new office taken place have been the result of (a few) collec- technology, pointing to the stress, performance tive agreement clauses negotiated on that topic, pressure, and lack of autonomy suffered by moni- or arbitrator rulings interpreting rights of employ- tored workers. ees under contract. The Task Force regards close monitoring of work Electronic monitoring appears to be fairly com- as an employment practice based on mistrust and mon in Canada, in the same sorts of work it is ap- lack of respect for basic human dignity. It is an in- plied to in the United States. Service observation fringement on the rights of the individual, and un- is also an established practice in the telephone in- desirable precedent that might be extended to other dustry and in other types of telephone customer environments unless restrictions are put in place now. We strongly recommend that this practice be service. prohibited by law. A survey conducted for the Canadian Labour The Task Force recommendations were in gen- Congress' (CLC) 1980 study of VDT health and eral considered too sweeping, too costly, and too safety issues found that monitored workers experi- "pro-labor" and were not endorsed by the Minis- enced stress-related illnesses (general tiredness, try of Labour. However, there have been attempts irritability, headaches, and sleeplessness) to a greater to pass legislation at the provincial level to create extent than nonmonitered workers. As a result, the ergonomic standards for VDTs and to ban individ- report recommended that "direct electronic moni- ual monitoring. Legislation was introduced in On- toring of individual worker's activities and pro- tario, British Columbia, and Saskatchewan be- ductivity be discontinued." Where productivity tween 1981 and 1985, but none of the bills passed. monitoring was deemed necessary, the report suggested that indirect or aggregate monitoring tech- Japan niques be used. Canadum unions began mobilizing around VDT Interviews and published reports indicate that related issues, including work monitoring, in the individual electronic monitoring is not common in early 1980s. Model contracts proposed by the Japan. This may be due in part to the still low Canadian Union of Public Employees and the penetration of automated equipment into offices Canadian Labour Congress, among others, con- there, but is most strongly related to cultural tained language banning monitoring of individuals. values toward work. According to an executive of For example: one computer firm: Individual work monitoring is not an issue in Ja- pan. Employers do not measure individual output and make individual judgements on that basis. If

Canadion Union of Public Employees, Model contract on Tecimologi. /Minim with Elisabeth Lagerloi, Labor Specialist, Swedish Em- cal Champ, 1962. Mad in Westin & Pipe, op. cit. hussy, Mar. 26, 1906. 'Westin and Pips, op. cit.

1 :3 5 App. ANotes on Computer Work Monitoring in Other Countries 125

they tried to do that, unions would complain, be- Table 20 outlines the main provisions of legisla- cause it would violate the union-company attitude tion of 11 nations that use the Western European toward worker productivity. The climate inour workplaces is for employees to work hard, and for model of privacy protection. These featuresgen- the whole work groupemployees andmanagers erally include a requirement that the data be col- to strengthen the norm of Lard work. We would not lected for legitimate reasons and used only for the measure each person. specified purpose, and that the individual have the In one documented case, individual monitoring right to inspect the data In some countries, has been used, not to increase worker performance, citizens can withhold sensitive information; any- but to insure that employees did not hurt them- one storing it in a database without their consent selves by working too hard. In the 1960s,a wave would be in violation of the data protection law. of repetitive strain injuries among keypunchoper- While the words "work monitoring" do notap- ators was attributed to the high number of key- pear in the privacy legislation of any European strokes they performed daily. Several companies country, computerized work monitoring is covered and unions proposed voluntary limitson keystrokes by data protection legislation to the extent that: and in 1964 the Ministry of Labor issueda guide- 1) information is collected and stored by computer line of 12,000 keystrokes per hour. The guideline and 2) information can be related to a specificem- also called for breaks of 10 to 15 minutesper hour ployee. Table 21 indicates coverage of computer of work, a total of 300 minutes keypunchingper usage monitoring, telephone call accounting, and day, and regular medical exams. To enforce these work measurement records under the laws of 11 guidelines, employers have monitored individual nations. performance, at least on a test basis. These Minis- Although records created by work monitoring try of Labor guidelines are still in use, andare the are covered by privacy laws in these nations, it ap- model for updated guidelines now being considered pears that these laws have not been used exten- for VDT work." sively to prevent or modify any monit-Tbig practices that unions or individual workers might have Privacy Legislation found objectionable. Data protection authorities have been preoccupied with their first-linerespon- Most Western European nations have privacy sibility of bringing major government and private legislation intended to ensure that personal privacy databases into compliance with the law, andre- is not eroded as a result of data processing appli- cently have had to deal with increasingly frustrat- cations. Most of these laws were passed in the ing problems of enforcing the laws in anera of 1970s, beginning with Sweden's Data Act of 1972; proliferating personal computers. There appear to this was the same epoch as the work of the U.S. be no jurisdictional problems involved with data Privacy Protection Commission (final report in protection authorities becoming involved in the 1977), the passage of the Privacy Act, and several case of some inappropriate use of work moigtor- other privacy-related laws in the United States. ing records; data privacy officials in somecoun- Unlike the approach of the United States, how- tries have expressed increased interest and true- ever, European laws do not leave it to the injured trativn with their inability to regulate such aspects individual to complain or sue if heor she believe. of the new information technologies. Itappears that personal data has been misused. Rather, in- that the unions, works councils and labor courts dependent government agencies (data protection have not requested their involvement. Thus, trade authorities) were created to supervise and enforce unions and labor law will continue to be the focal pres..-ribed data handling practices. All databases point of actions to deal with monitoring. containing personal information must be with the data protection authority so that use Privacy Legislation and Telephone Call can be supervised. In most cases, this applies to both government and private databases. One prin- Accounting cipal focus of the data protection authorities has In West Germany, telephone call accounting has been to prevent linking of databases to buildup been the topic of at least two courtcases and the "profiles" or "dossiers" of citizens. abject of a continuing disagreement between tiv privacy commission and the Federal government. °Westin and PM(, op. nit. Inkematlon complied by Alan F. Westin. The privacy commission objects to collectingrec- from interviewswithIndustry, hp!, and academic mmorta, Tokyo, May ords of personal telephone calls by government 1965. employees and urges utmost restrainteven in keep- 1 3 6

4 44 1, 128 The Electronic Supervisor: New Technology, New Tensions

Table 20.Main Provisions of Foreign Personal Data Protection Legislation Relevant to Coverage of Employee Monitoring

National legislation

CC LL >. c >. C es 0 O co v -c a c 0 c as is' E vi 11 ... 2 2 O as c. D II 0 0 3 Provisions a 0 u. 0 0 z (.0 D Scope of application: Central government Y Y Y N Y Y Y Y Y Y Y Y Provinces/states Y N Y N Y YY`f Y Y Y Y Private sector Y N N Y Y Y Y Y Y Y Y Y Covers all Information traceable to identifiable individuals Y Y Y Y`f Y filiNby 'if y 'if Information collected and/or processed using computers Y Y Y Y Y Y Y Y Y Y Y Y Limits placed on personal data collection Y Y YcY Yrly 'if y 'if 'if Ng Personal Information must be collected for specified, legitimate purposes Y Y Y Y Y Y Y Y Y Y Y Y individuals have right of access to inspect personal information Y Y V Yf YaY Y Y Y Y Y Sensitive personal details specified (collection only with data subject's knowledge and consent) N N Y Y Y N Y N Y YYf KEY:1' Yee; N No ,11Covers Information concerning private affairs, such as financial situation of individuals. °Covers information on an Individual's personal statue, intimate affairs, economic position and vocational qualifications. *Collection, of person:- data limited unless It Is "natural part of the normal operations of an enterprise." dpersonal Information collection Is permissible If It eaves the purpose of a corractual relationship or there Is a legitimate Interest In (a business) storing It. *stow law, may be enacted that for personal data maintained by the public sector. SOURCE: Russell Pipe and Ala, F. Westin, "Employee Monitoring in Other Industrialized Democracies," contractor report prepared for OTA, late.

Table 21.Applicability of Foreign Personal Data Protection Legislation to Employee Monitoring

National legislation

cc U- '2 >. 2, c as c .0 >. 5 as c E 03 . as -6 a 7, X 'S = i i O 0 V = 0 3 he Type of monitoring a C sr. 0 0 so -J z 0 D Employee computerusage monitoring (by IDs, terminals, and passwords) Y Y Y Yf rUbYcy Yay Teiephone-call accounting YYY`fVDU Ycy'fay 'if Work measurement YYffYfaf Y flof KEY: 1, Yes; U Uncertain. °Personal data covered by this law must be organized or filed so as to be retrievable automatically using identifiers that can be linked to aparticular person. binfonnallon must be related to a person's private affairs, such as financial situation. OM official translation (ch. I, pars 2) states that It Is an Infringement of privacy to: 1) spy on or trail a person in amanner likely to harass him, or any other harassment; and 2) 'listen In." SOURCE: Russell Pipe and Alan F. Westin, "Employs Monitoring In Other Industrialized Democracies," contractor report prepared for OTA, 19(18 fIN

App. ANotes on Computer Work Monitoring in Other Countries 127 ing records of official call destinations. Of particu- ing with the new technology at the workplace, in- lar concern are calls by union representatives and cluding electronic monitoring. It is a model which others (counselors, medical services) in sensitive is different in important ways from the more cen- positions. In one agency, on the basis of an unwrit- tralized, regulatory, legislated model of emr ' ver- ten understanding in effect since 1984, calls by the employee relations in many other industrial na- "personnel council" have been channeled through tions. While voluntary and recommended styles of a separate telephone line to bypass the call-account- participative management and other approaches ing system. In general, however the Federal gov- towards dealing with microelectronic technology ernment finds value in call accounting and has not developed abroad may have growing influence in complied with the privacy commissioner's request the United States, there will continue to be a dis- to terminate the practice and erase previous rec- tinctly different approach taken in this county in ords. A similar controversy is going on at the state the foreseeable future. level Conclusion The force of law, tradition of labor-management relations, and the current economic and political milieu have shaped the American approach to deal- Appendix B Privacy and Civil Liberties Implications of Testing Employees in the Workplace

Introduction by urinalysis, and genetic screening. In addition, the appendix will review some research in the field The focus in this section shifts from the moni- of brain wave analysis that could give rise to new toring of work to the testing of employees, It looks forms of worker testing in the next few years. at some of the tests that can be given by employers Some of the technology used for testing is not in hopes of predicting some aspects of an employ- new. The polygraph, for example, has had limited ee's future work quality. Can technology help em- use in law enforcement for 60 years. Now, however, ployers predict how good an employee will be be- its dominant use is in personnel screening; of 2 mil- fore the person is hired? Could there be some lion polygraph tests given annually, about 98 per- corporate equivalent of a carnival "weight and cent are given by employers to job applicants and fate" machine, capable of succinct predictions: employees.' Medical screening for drug or alcohol "This woman will try to embezzle money: don't use, formerly used primarily as a diagnostic tool trust her." "This man will have a heart attack at in clinical settings, began to be used by the De- 45; don't invest in training him." "This one uses partment of Defame in the 1970s to identify return- drugs: don't hire him." "This is is a healthy, honest ing military personnel with drug problems. Now, worker; hire heel" nearly all military personnel, millions of private em- Clearly, such a single tool for personnel selection ployees, and a growing number of government em- does not exist, but a growing number of einnktvP-s ployees find that their jobs depend on passing the are relying on tests for empkorees and joke appli- drug test. cants to try to predict behavior or personal char- Genetic screening. an emerging technology for acteristics that may affect their job perforzoGzez. predicting a person's likelihood of developing dis- Limited evidence suggests that over the past few eases, is now used only in a few workplaces, usually years workplace testing has been Amidergoing a re- to identify workers who may be hypersusceptible vival that may surpass the heyday of psychologi- to chemicals found in those workplaces. However, cal testing in the 1950s. The future trend may well tests for many common diseases will be commer- be in the direction of more testing. If this is the cially available within the next 5 years, and em- case, then the controversy about worker rights, ployers may want to include them in preemploy- workplace privacy, and privacy of personnel rec- ment physicals. ords, may intensify. Still in the research stage are tests based on brain Issues such as those explored in earlier chapters waves. Currently under study is the possible use will con' inue to arise and become more prominent of brain wave analysis in monitoring concentration, concerning what information is reasonable and nec- detecting lies, and predicting certain illnesses. A essary for an employer to have, rho shculd have computer-based system to detect drug use by access to it, and how it ought to be d.rie and where measuring brain waves is already on the market. the line should be drawn between lotion that Controversy about worker testing focuses on the is personal and private and that s -1-2a is noz. accuracy and predictive value of the tests. Em- As noted earlier, the past pattern nas been for ployers who test employees for drug use or hon- these new capabilities to come into use in a piece. esty may believe that the tests work or at least meal fashion over time. Sometimes, however, they that the fear of testing discourages the unwanted are put to use almost immediately, before adequate behavior. Washington Area Metropolitan Transit research car be conducted, as with polygraph test- Authority, for example, noted a decrease in drug ing, and before consideration can be given to the use and in accidents sham beginning its drug test long-term consequences for society. When this hap pens, issues arise that have no established legal, 1Harrieon Donnelly. "Privacy in the Workplace," Editorial Rowan* Report.. Mar. 21,1966, p. 214, dting figure. from the American Poly- ethical, or other useful framework for bvaltation. graph Amodation. This appendix focuses on three technologies that *Kenneth F. Englade, "rim Business of the Polygraph,~ Across tbo are already in use today: polygraphs, drug tests Board. Oetobw 1962, pp. 20-27.

128 App. 9Privacy and Civil Liberties implications of Testing Employees in the Workplace 129

program. Officials of the Eckerd Drug Co. believe At present 3,000 polygraphers belong to the Amer- that requiring a test of all applicants and periodic ican Polygraph Association, but it is estimated tests of employees, is the beat way to deter theft that there are 8,000 to 9,000 full-time polygraphers and "keep basically honest people basically honest." nationally. Some are employed by law enforcement But there have been no systematic studies of em- agencies, some by detective agencies, and some are ployee theft to support this claim' part of in-house security departments of large On the other hand, there are concerns about test- firms.' ing, and many argue that these gains, if they ex- Paper-and-penal honesty testing has gained ist, are achieved at a heavy cost: undue intrusion popularity in the past few years, partly in response into p.-ivate lives of employe* creation of an to criticism of polygraphs and partly as a lower atmosohere of fear and intimidation in the work- cost alternative. Compared to a cost of $40 to $50 place; and false accusation and denial of job op- per test for polygraphy, paper-and-penal testa can portunities for many innocent people. be administered and scored for $8 to $15 a piece.' This chapter outlines some trends in worker At the present time, about a dozen firms nation- testingwho uses it, what can be learned, and wide are dominant producers of these tests, two directions of research. The chapter also explores of the largest being John E. Reid & Associates in why worker testing is controversial, and looks at Chicago and Stanton Corp. in Charlotte, NC. At some of the ethical and legal questions raised by least 2 131111i081 of these tests are given annually in its use. preemployment screening. Thirty-two States have legislation limiting the Polygraph Testing use of polygraphs in employment, including 12 with an outright ban on employers' requiring or Extent of Honesty Testing requesting that employees take a polygraph test. Nine States require licensing for the polygraph Employee theft is a major business problem, re- operator. Four States have legislation regardiug sulting in losses estimated at $5 billion to $10 bil- the types of questions that may be asked, prohibit- lion annually.' To counteract it, employers are in- ing questions on such topics as sexual preference, creasingly using "honesty testing" on workers, religion, union affiliation, or politics. Twenty-one either using the polygraph ("lie detector') or paper- States have laws providing that polygraph tests and-pencil honesty tests. One major object is to be voluntary? (see table 22). cut down on employee theft by screening out po- At the present time there seem to be few State tential thieves before they are hired. In addition, laws dealing directly with paper-and-pencil honesty a number of employers use polygraphs as part of tests. One 1986 Massachusetts law outlaws hon- paternal investigations of theft and other wrong- esty tests that amount to paper-and-pencil poly- doing, and some administer polygraph tests on a graph tests. There have been several attempts to regular or random basis as a deterrent to pass polygraph legislation at the national level, in- wrorgdoing cluding bills in the 99th Congress" A testing industry has grown in response to this Part of the reason for the growing use of hon- demand, and perhaps has helped to fuel the de- esty testing is the increasing difficulty and high mand. While some large firms have in-house poly- cost of doing good background checks. A thorough graphers (one Florida drug firm has a staff of 40), most rely on detective or personnel security firms check might cost as much as $250, and many employers hesitate to give detailed information about who provide polygraph services on a contract ba- former employees, partly due to fear of Bei sis. There is no good estimate of the total number suits.' of such firma throughout the country, but a gauge of their growth can be found by looking at selected 'Joseph Buckley, III, President, John E. Reid & Antedates.person& cities. In 1970, for example, there were only three communication, Aug. 19, 1986. such firms listed in the Atlanta yellow pages. By "Susan Deoshr, et aL, "Can You Pass The Job Test,"Norwalk May 5. 1966. pp. 4663; Kenneth F. Engirds, "The %ahem of the Poly- 1975 there were 20, and in 1985 there were 33.' graph," Across the Board. October 1982, pp. 20-27. ViRiam E. HartafiekL "Polygraphs." Labor Law Journal,vol. 36, November 1:45, pp. 817-834. 'Employee I. Aygraph Protection Act of 19861H.R. 1624)paused the 'Susan Gardner, "Wiretapping the Mind: A Call to Regulate Truth House; Polygrap, Protection Act of 198618. 1815)was reported out Verification in Employment." San Di110 Law Review, von 21 No. 2, by the Committee s. Labor and Human Resources but not actedon March 1963, pp. 296-323. by the Senate. 'Kenneth F. Epode, 'Ilse Business of the Polygraph," Across the High Cost of Employes 'Theft," Dun's Basinger Month, Os- Board. October 1962, pp. 2047; Atlanta Yellow Pages, 1985416. Vb., 1682. 140 1X The Electronic Supervisor: New Technology, New Tensions

Table 22.State Legislation on Polygraph Testing

Employermay Employee must Certainpersonal not test or Employer may not be told test is questions License for requesttest require test voluntary prohibited poiygraphers Alabama x Alaska X Arizona x Arkansas x California Colorado Connecticut X Delaware D;strIct of Columbia X Florida x Georgia x Hawaii Idaho Illinois X Indiana Iowa Kansas Kentucky X Louisiana X Maine X Maryland X Massachusetts X Michigan X Minnesota X Mississippi X Missouri Montana x Nebraska x Nevada New Hampshire New Jersey X New Mexico X New York X North Carolina x North Dakota x Ohio Oklahoma X Oregon X Pennsylvania X Rhode Island X South Carolina South Dakota Tennessee Texas Utah x Vermont Virginia' Washington x West Virginia X Wisconsin x Wyoming Applicant or hployem must give illtten consent and be told all questions In advance. SOURCE:Adapted from VAIllam E t(artsfleld, "Polygraphs," Labor Law Journal, vol 36, November 1962, PP 617534

141 App. BPrivacy and Civil Liberties implications ofTesting Employees in the Workplace 131

Technological Considerations Table 23.Accuracy of Polygraphs for Specific incident Criminal investigations' Polygraph technology has not changedsubstan- tially since it came intouse in 1921.10 The poly- Field studies graph measures and records Range Average a number of phy- Six prior reviews 6498% siological responses, including skinresistance, respiration, and blood OTA review of 10 individual field studies: pressure, while the subject Correct guilty 70.8-98.8% 86.3% answers a series of questions posed byan inter- Correct innocent 12.5.94.1 76.0 viewer. The most commonly accepted theoriesof False positive° 0-75.0 19.1 pokgaphy hold that the subject's guiltand ner- Fasie negative° 0-29.4 10.2 vousness will produce measurable physiologicalre- OTA review of 14 individual analog studies: Correct guilty actions when hear she is lying. Thesereactions must 35.4-100.0% 63.7% be interpreted by a trained polygraph Correct Innocent 32.0-91.0 57.9 interviewer. False positive 2.0.50.7 14.1 Critics argue that the physiologicalchangedur- False negative 0-28.7 10.4 ing lying is still not well understood,and even des sail. bFeles Innocent afaradaItsisceptIve. proponents of polygraphy agree thata unique set False negellsoduay persons foundno, deceptive. of physiological reactions to lyinghas not been SOURCE:Adopted from U.B. Caldron. Once of Technology Assessment.scien- found. The reactions often attributedto lying can tific Validity of Polygraph Testing: A Reetrach fieview and Evoke. donA TedmimeMemorandum, OTA.TM44-I5 (Wardengton. DC: U.S. also be caused by anxiety,anger, or humiliation. Government Printing Office.November 1113),p. 97. Being required to takea polygraph test elicits pre- Validity in screening situations, where questions cisely these feelings inmany people. are of a more general or hypothetical nature ("Have In determining theaccuracy and reliability of you ever taken something that didn't belong to polygraph tests, it is important to considerwhether you?") and responses cannot be easily compared the test is used for investigatinga specific crimi- to control questions, is more problematic. Some nal incident or for screwing. Thequestioning tech- critics suggest that screening polygraph tests nique is also important. are as discos/led in a previous strongly biased agohast honest people. In this view, OTA report.11 The polygraphappears to be most the basically honest person, theone who feels reliable when the interviewer isasking relevant guilty about small past wrongdoingsor angry at questions about a specific event (e.g.,"Did you a challenge to his or her integrity, is likely to do take 3200 out of the cash draweryesterday after- worse on a polygraph test than a person with less noon?") and comparing the reactionsto a list of developed conscience." The American Psychologi- irrelevant control questions (e.g.,"except for what cal Association, for its part, charges that poly- you told me already, did you even steal anything graph tests produce "an unacceptable number of before the age of 21?")." Subjects that react more false positives. "" The OTA report foundno strongly to relevant questions thancontrol ques- studies evaluating whether polygraph testing is tions are believed to be deceptive. OTA'sprevious valid in personnel security situations." OTA is review of research on polygraph validityfound that currently reviewing the Defense Department's most research has focusedon use in specific inci- polygraph test and research programs." dents. Methodological problems anddifferences The validity of pencil-and-paper honesty tests made it difficult to draw overall conclusions about has also been called into question, and thereap- validity. In the studies reviewed,accuracy rates pears to be a dearth of independent research on ranged from 0 to 100 percent, and innocent people their validity. Some critics note thatmany "cor- were more likely to be assessed deceptive than vice rect" test answers are based on values and defini- versa. A summary of the findings of the OTAre- tions of honesty that may not be shared by alltest view is shown in table 23. takers."

"David T. Lykkan. "Detecting Deceptionin 1984," American Br hawked Solentiet. March/April 1984. pp. 481.499. ""Cen You Pass the Job Test," Newsweek.May 6.1968, pp. 48-63. "For a desaiption of the various polygraph "Kenneth F. Engler* "The Businessof the Polygraph." Across the techniques and an evalu- Board, October 1984, pp. 20.27. ation of scientific research on polygraphs,see, U.S. Congress, Office of Technology Assimement, Scientific Validity of "For a description of the various polygraphtechniques and en evalu- Polygraph Testing: ation of scientific resew& on polygraphs, A Research Review and Evaluation A Technical MeinrandomOTA- see, U.B. Congress, Office TMH15(Washington, DC: U.& Government Printing Office, Novem- of Technology Aseememeh, Scientific Validityof Polygraph Testing: A Ramada &view add EnkletlonA Technkal ber 19831. Memorandum OTA- "OTA, Review al the Wawa Department'sPolygraph Tact and Re- TM-H-16 (Washington, DC: U.& Government PrintingOffice,Novem- ber1983). search Program forthcoming (1987). "Quoting Michael Mecham In "Can YouPose the Job Test," Newsweek. May 8. 1988. p. 49.

142 132 The Electronic Supervisor. New Technology, New Tensions

Legal and Ethical Considerations Fortune 500 firms now do some testing of employees, as compared with 10 percent in 1982." Asof The privacy issue pits the employees' interest mid-1986 nine Federal agencies had drug-testing in being left alone, and in keeping certain informa- programs in place, and at least eight were plan- tion private, against the employers' interest in pro- ning to begin testing in the immediate future." tecting their businesses and selecting employees An Executive Order of September 1986 established by their standards. a "drug free workplace" as the policy of theU.S. However, questions addressed to workers in on- Government. the-job polygraph tests sometimes go beyond job- A number of firms that have instituted drug- related topics and probe into sensitive areas of per- testing programs have reported dramatic de- sonal life. Several critics have charged that the pur- creases in on-the-job accidents and injuries-90 pose of such questioning is not only to intimidate percent in the case of Georgia Power and 70 per- workers but to screen out minority group members cent at the Southern Pacific Railway." However, or those who may show dissident tendencies or an there are some who dispute that testing is respon- inclination to join unions. At least four States have sible for these results; for example, in the case of legislation prohibiting employers from asking Georgia Power, the accident rate began its decline questions about politics, religion, union affiliation, before testing started.* In addition, the tests or sex life." have also led to charges that they are invasive of privacy, that tests are sometimes inaccurate, and Drug Abuse Testing that testing programs are sometimes used to in- Rationale and Extent of Drug Testing timidate, discriminate against, or harass certain groups of workers. The problem of drug abuse on the job has gained Public concern with the substance-abuse prob- increased attention in the past few years, but it lem, on the one hand, and concern about the ad- is not a new problem. In fact, overall abuse of most visability of on-the-job drug testing, on the other, drugs in the United States has held steady or de- readied a hem b 1986. March of that year, the clined since 1979." Substance abuse is stall a ma- President's COMMifiSiC 1 on Organized Crime rec- jor problem, however, and the current public aware- ommended that Federal employees and contractors ness and concern should help to reduce it further. be subjected to "suitable" drug testing as an es- Alcohol abuse is estimated to cost the U.S. econ- sential step in reducing the demand for drugs. Sev- omy a total of $89.5 billion per year in lost employ- eral months later, President Reagan held "volun- ment, illness, reduced productivity, and death and tary" testing for the White House staff and injury due to automobile accidents. Other types suggested that government agencies and private of drug abuse are estimated to cost society about industry follow suit. Executive Order 12564 of Sep- $46.9 billion per year." The U.S. Chamber of timber 1986 directed the Office of Personnel Man- Commerce estimates the direct costs of drug and agement to develop governmentwide guidelines on alcohol use among workers costs employers $60 testing. Objections to widespread use of testing billion per year in reduced productivity, increased have been voiced in the press and in a report by medical claims, and absenteeism." Workers who subcommittees of the House Committee on Civil are dependent on alcohol or other drugs tend to Service." have more accidents, which may increase their employers' insurance costs. Addicts are also more likely to steal money or property from the employer or from co-workers in order to support their habits . "Mark A Rothstein, ning Workers for Drew A Legal and Ethi- In order to ensure a drug-free workplace, some cal Framework,"Employs. Rolstacas Law Journal, winter1986/1966, pp. 422437. employers have resorted to testing their worker "U.S. Congress, House Committee on Post Office and Civil Service, for drug and alcohol use. About 25 percent of the Subcommittee on Civil Service, "Drug Testing in the Federal Govern- ment," Staff Report, June 20. 1966, pp. 7-10. "Stephen R. Deja*,"An Unhadthy Swarm: Drug Testa Air Un- "William E. Hartaddd. "Polygraphs,"Lobar Law Journal,vol. 36. oonstitutionsl sod Elansthoes Wrong,"Washington Post,Aug. 17.1906. November 1966. pp. 317-634. "Testimony of Mt Leslie Price bateau the Subcommittee on Human "Ramer& ?MingleInstitute. Econotnic Costs to Soday of Akohol Resources, House Committee on Poet Office and CivilSwimSept sod Drug Abuts and Mortal Moms,Research Triangle Park, 1964. The 16 1986. continuations of this trend is supported by 'teddies of the National "U.S. Congress, House Committee on Post Office and Civil Service, Instituteof Drug Abuse. Subconnuittee on Human Resources. "Drug Testing Federal Employ- "Ibid. ees," Hearings, Mar. 16 1966. fluid No. 9944: Subcommittee on Civil "Sum Dander. at al, "Can You Pau the Job Test?" Newsweek. Service, "Drug Testing in the Federal Government," Staff Report, June May 6, 1966. pp. 4643. 20. 1966.

143

121...... -... ! 134 The Electronic Supervisor New Technology, New Tensions

fractions of the body) for a byproduct of THC (the On the other hand, some employees thatare using active ingredient of marijuana), thus producing a legal drug for legitimate medical reasons may be higher false positive rates among black people." impaired using it. There is also controversy over the cutoff points Is there a socially acceptable need foran em- to be used for tests. While some expertssuggest ployer to mandate periodic or random substance- that tests for msrijnana should be consideredposi- abuse testing for the workforce,or should testing tive only if they have a concentration ofover 50 be limited to specific instances wherean employ- nanograms per mil liliter (nglml) a more typically ee's conduct raises "reasonable suspicion" of sub- used cutoff point is 20 nikial andsome organize- stance abuse? Random or periodic testingseems thins use even lower cutoff points. Inaccuracies in contrary to the principle that people are innocent all tests are much more Nicely at lowconcentra- until proven guilty; such a testing schemerequires tions.° Even the use of confirmatorytests has everyone to prove his or her innocence on a regu- not stalled controversy about accuracy, especially lar basis. In a number ofcases to date, the courts .since some commercial labs relyon lower cost tests have required that employees be tested only when for confirmation as well as screening." there is reasonable suspicion of druguse. Random testing of Iowa prison guards. and compulsorytest- Ethical and Legal Issues ing of New York City teachers applying fortenure Privacy. There are several reasons why employ- were struck down by the courts." In bothcases, ees have objected to drug testing on privacy however, the employers were government entities, grounds. For one thing, it can reveal details of whose employees are protected by theFourth per- Amendment to the Constitution against sonal life outside of work. Current methods ofdrug "un- screening reveal recent use of a drug, not current reasonable search by their employer." A number knpairment; people who use drugs outside of work of private employers makeuse of random testing, test positive, even th3ugh they might be sober and and some of their programsare also being chal- fit for duty during work hours. Is after-hoursuse kngs& of legal or illegal drugs an employer's concern? Another imports_ t point is what drugs should In addition, producing a urine sample beforewit- be covered by the tests. Asurvey of drug testing nesses violates most people's reasonable expecta- programs in the Federal Government found wide tion of privacy for bodily functions. Both tIlese ob- variation in the types of drugs being screened." jections apply to deficiencies of urinalysis, notto While many employers focuson illegal substances the concept of drug testing in general. like cocaine and marijuana,a number of legal sub- Methods and Conditions of Testing.The first stances are also abused. However, testing for these drugs will also reveal their use by individuals who question is whether testing is the appropriatere- sponse to ed drug use in workplaces. In use them legitimately, thus raising another privacy issue. many pop a testing program may cast too broad a net, subjecting many workersto testing One interesting point is the lack of interestof in order to identify a very few drugusers. The next many employers at the present time in testing for alcohol abuse in the workplace, despite all question is who should be tested? Should allem- the evidence that alcohol is responsible for far ployees be tested or only those with jobs thatpose more work- medal problems of safety or security if thereis place accidents and absenteeism than theillegal drug abuse? If only some employeesare tested, drugs. (Alcohol intoxication is usually tested by does this create equal protection problems ifthose breath or blood tests, rather than urinalysis.) tested are disproi Aionately racial minoritiesor blue collar workers? What drugs should betested for? Is it an employer's job to identify onlyusers of illegal drugs? Many legal drugsare also abused. *For simple, cuss of New York artyOpeceers, New Jenny govern. meet workers, Josue Rangel, "Government BarsDrug Testing of "Janes Woodford. Ph.D., personal Timbers," New York Times, Aug. 12, 1966,1).131;Alfonso A. Novae, oommunicatiou Ana. 26, 1986, "U.S. Judge Blocks Urine Drug Tees," New heeed mi stades conducted st the U.S. Army FaroukLab, Wkebaden, p. Al. York Thom Sept. 19. Nee. Ciormsey. See also 'The Melanie Defense, Debated by Woodford and "See U.S. Cowen. Hour Zommittaat Post Ofike and Civil Serv- Bobetesno Mere Report, Dee1, 1986, p. 3. ice. Subcommittee ma MA Service, "Drag Toting ""The Mobs Defeo Debated byWoo.d ard and McBay," Sub. in the Federal Gov- seam Abase Report, Dec. 1, 1988, p. ernment," Staff Report, June 90,1986,pp. 7-10; also Lawrence Mae 9. "Amery and Reliability at Urine Dreg Testa," in "Tubb on "Modified Drug Stow"from conference material House Committee an Post Office and Civil Service, Subcommitteeon Human Resources, predated by Smithillise Mika Laboratory,Mosphis. TN. Hearings of Sept. 16, 1966.

145 App. 8Privacy and Civil Liberties Implications of Testing Employees in the Workplace 135

Use of Test Results dation and death if not treated early, is called for in statutes or regulations of 46 States and is cus- Another question germane to any testing pro- tomary in the others.* gram is what to do with the drug abusers when Genetic screening in the workplace is not yet a they are discovered. In the case of pre-employment widespread practice. In its 1983 report on genetic screening, the answer of most firms is not to hire screening in the workplace, OTA found that few them, although a few do tell the applicants why large U.S. firms were using genetic screening in they have been turned down and invite them to their personnel selection practices.* Out of 366 reapply once they are drug free. respondents, only 8 said they were currently do- In some firms, a current employee who tests ing any type of genetic test; 17 had done so in the "positive" might be referred to an employee &ask,- past; but 59 had plans to do so in the future. While tame program (EAP) for rehabilitation. Some ob- these r3sults indicate that genetic screening in the servers have expressed concern that existing EAPs, workploce was not widespread at the time, many which heretofore have assisted employees who vol- researchers in the field assert that it is difficult to untaril sought help for drug problems, will be se- get good information about firms that do genetic verely weakened by an influx of clients who have studies. Such firms may not answer questionnaires been "sentenced" to rehabilitation after a positive or talk about their policies because of the con- urine test. An alternative followed by many clients troversies, including charges of discrimination, is the dismissed or discipline of drug users. Between that have arisen when other firms have publicly 1971 and 3980 the Armed Forces tended to reha- discussed screening programs." bilitate returning Vietnam veterans whose drug Despite the bad publicity that might attach it- problems were detected through urinalysis. In re- self to the concept of genetic testing, screening cent years, however, the emphasis has turned to might offer some advantages to employers. It is discipline and dismissal. The cmrent guidelines is- known that human beings have varying suscepti- sued by the Office of Personnel Management for bilities to illness, including illnesses related to ex- Federal agency drug programs directs each agency posure to toxic substances at work. Not everyone plan to include an EAP, but also lists disciplinary who mines coal gets black lung disease, for exam- actions including reprimands, suspension, and dis- ple, just as not everyone who smokes cigarettes missal. According to the guidelines, dismissal from gets cancer. If firms could determine that certain Federal service would be mandatory upon a sec- people are especially susceptible to a toxin found ond confirmed finding." in the workplace they could decline to hire them A forthcoming OM technical memorandum will or otherwise avoid assigning them to work near look in greater detail at some of the technical, administrative, and legal problems involved in the hazardous substance. Thus the employer helps drug testing.* the employee stay healthy while also avoiding the possible costs of a future illness that the employer might have to bear. Employers often pay part or Genetic Screening all of their employees health or life %sure= costs. Group insurance rates depend on the health experi- Extent and Rationale for Genetic Testing ence of the group. Thus it would be to an em- Genetic screening is not new, though in many ployer's advantage to identify and eliminate un- ways it is still an emerging technology. Tests for healthy or potentially unhealthy employees in diagnosing or predicting some genetically based order to keep rates low. diseases have been available for some time. For example, screening tests for sickle cell trait, a condition especially common among those of African an- cestry, were available in the early 1970s. Blood tests of newborns to screen for phenylketonuria (PKU), a genetic ailment that causes mental retar- slAmwicen Bar Foundation, State Laws and Regulations Govern* Nowiorn Sam*, compiled by Lori B. Andrews, Chicago, 1966. *U.S. Congress, Office of Technology Asseennont. The Bak el Genetic Teeth,: in the Prevention el Oceapetiosal Mee" OTA-BA- 194 MalgIbIltak DC: US. Oorwmoust Printing Moe, April 19831- 870f6ce if Personnel Management, PPM Letter 792-, "Establishing 41Intarriew with Mine Draper, Ph.D., School of Public Health, a Dreg-Free Federal Workplace," Dec. 27, 1955. Univenity of California, Berkeley, September 1986. See also Thomas seOffloe of Technology Aasunwant, Chine Drug Testier, Technical K Murray, "The Social Contast of Workplace fieneatag." The Hast- Mamoresion forticosaing 19871. ings Center Report October 1964.

146 r

, App. B Privacy and Civil Liberties Implications of Testing Employees in the Workplace 137 these tests to screen applicants for insurance cov- Research on electrical and magnetic recordings erage, and some have withdrawn from the market of human brain activity is being conducted in a in areas where such screening is forbidden by law. number of government, government-supported, and academic laboratories. Government funding Brain Wave Research for research is provided by a variety of Federal agencies, including the National Institutes of State of the Technology Health, the National Science Foundation, and the Department of Defense. Brain activity that underlies psychological proc- Five areas of focus characterize much of the cur- esses can be recorded from the surface of the head rent research: and body in the form of weak electrical and mag- 1. assessment of neurological function and neu- netic signals, or "brain waves" as they will be called rological disorders; in this section. Many of these signals are not well 2. assessment of mental disorders; understood by scientists despite decades of re- 3. analysis of normal cognitive processes, includ- search. However, some are known to reflect cogni- ing perception, memory, language, and deci- tive (memory, language, learning) or sensory (vi- sionmaking; sion, hearing, touch) processes stimulated by 4. analysis of cognitive disorders; and external stimuli. These signals, often called "event- 5. human factors applications. related brain potentials" or ERPs, are extremely Table 24 shows some of the recent research in this weak (typically on the order of a few microvolts, area or millionths of a volt), but they can be monitored The potential benefits to society appear to be in through sensors attached along particular regions several main areasas a tool for medical diagnos- of the scalp." Magnetic recordings of brain activ- tics and cures, for developing optimal learning and ity are also possible by using a specialized mag- educational techniques, and for enhancing man- netic detector termed SQUID (superconducting machine interfaces. This technology is already in quantum interference device). use as an aid to diagnosing brain tumors. multiple sclerosis, dyslexia, epilepsy, and strokes. It has also been used to test for mental retardation, coma, "Cavities processes are identifiable through ERPs that occur between about 100 and 700 thousandths of a second or more following and autism.° each *ironies; Mowry Foamed Are reflected in ERN that occur within "Carol Tumid, "Watching the Brain at Work,"IEEESpedewm, 100 thousandths of a second following the etimuluu Marc, 1983, pp. 62-67.

Table 24. Examples of Research on Brain Waves

Research center Areas of investigation National Institutes of Mental Health Predict risk of psychiatric disorders; tendency toward behavioral problems (sbstance abuse, antisocial personalities). University of California Analysis of sensory and cognitive processes diagriosis, e.g., deafness. La Jolla, CA Man-machine interface: Analysis of mental workload, e.g., attention and (ONR, NIMH, NSF funds) concentration

Veterans Administration Medical Center Detect neurological disorders, multiple sclerosis and make more precise Westhaven, CT neurological diagnoses. Understanding brain structures and processes responsible for surface electrical activity. University of Illinois Cognitive processes, e.g., memory, learning, decision-making. Champagne-Urbana, IL Man-machine interface Harvard Medical School Brain electrical activity mapping for detecting mental disorders, e.g., Boston, MA dyslexia, Alzheimer's disease Advanced Research and Development, Inc. Man-machine Interfaces, e.g., aircraft pilots Columbia, MD (NASA funds) AIr Force Aerospace Medical Research Lab Man-machine interfaces; analysis of mental workload Dayton, OH University of Florida at GaInesvile Lie detector test (ended 1906) (CIA funds) SOURCE: Office of Technology Ameament, leST. 148 The Electronic Supervisor. New Technology, New Tensions

Workplace Testing Applications refers to commercial applications of scientific dis- coveries before the underlying principles are The examples above of brain wave research at thoroughly understood: least raise the possibility that brain wave analy- Polygraphization occurs when the commercial de- sis could lead to usable technologies with possible velopment is done without an anchor in the scien- in the workplace. If developed as prac- tific community. Actions are taken to assure the systems, they could be used to gather exten- profitability of the product, and caution and con- sive information about a subject's psychological trol become less critical.. I emphasize that all this state, genetic propensities, or honesty; they might is done well within the law. But, it remainsthecase be useful in new means of measuring or pacing that it is quite possible to have what appears to be an work. impressive instrument that is essentially worthless." Some predictive tests that might be of interest The danger of using such a device in the work- in the area of work monitoring or worker testing place, of course, is that decisions affecting people's could be derived from the above avenues of re- lives will be made based on flawed technology or search. These could include the following: flawed principles. Due to the complexity of the predicting whether a person is at risk of cer- nervous system, it is likely that only very general tain diseases, such as Alzheimer's disease or links will be drawn between physiological processes alcoholism; like brain waves, and psychological ones like lying determining whether a person is concentrat- or concentration. Even these tests may only be ing and predicting the speed of mental re- valid in a very structured environment, such as in sponse to stimuli; a controlled laboratory setting. A workplace set- determining recognition of persons, places, ting would introduce too many uncontrollable and objects; variables. testing for knowledge of a specific subject; Other researchers, however, are more optimis- detecting lies. tic about the possibility of developing practical sys- Brain waves are also being explored as a possi- tems. Researchers at Westinghouse Research & ble means of improving man-mschice interfaces. Development Center have, for a number of years, Future systems are envisioned that would moni- been exploring the use of brain waves, in particu- tor the operator's ability to cope with information lar a wave called the P300, to determine an indi- flows and to make decisions. On the basis of the vidual's level of attention and cognitive process- information about his performance, the system ing. A Westinghouse researcher has predicted that could either adjust the rate of information flow to within the next 10 years, Westinghouse could mar- the operator or automatically take on some of the ket "a complete system capable of monitoring the operger's tasks to optimize his performance. Some mental processing effort of employees as they future applications could include pilots, air traffic worked."' controllers, and other computer-band work. In a slightly different direction, a system for If practical brain wave systems could be devel- using brain wave analysis for determining whether oped, the implications for privacy would be tremen- a person is intoxicated on alcohol or drugs is al- dous. In the case of workload measurement, for ex- ready on the market. Called the Veritas 100 ample, the distinction between monitoring the Analyzer, it is marketed by National Patent Ana- work and monitoring the worker completely dis- lytical Systems. The Analyzer is small, about the appears. In the case of something like an improved size of a personal computer, and is designed to be lie detector, such technology might actually give used at the workplace. A disposable headband is the ability to "read the mind," removing all possi- placed on the subject's head, and the analyzer ex- bility of a person's keeping information private. amines the cm neal-mtinal potential transmitted Whether practical systems can be developed, along the ve*.tibular nerve. According to the man- however, is another question. There are serious ufacturer, the system re.mgnizes the characteris- limitations on our of brain waves, tic brain waves that this nerve group produces and at least one researcher worries about the "poly- when the subject is under the influence of particu- graphization" of brain wave research.' By this he lar substances. The signal is unique because each drug produces a specific "fingerprint," a waveform --end Doodah% Psychophysialogioal Monitoring Possibilities known as a "drug-evoked potential," according to sad Prospects, contractor report prepared for DTA, September 1986. Also, Dr. Charles Wood, Department of Neuropsychology, VA Medical °Emanuel Donchin, Psychophysiological Monitoring: Possibilities Quaff, Weethaves, CT, sad Dr. Steve Hillywd, Department of New and Prospects, contractor report prepared for DTA, September 1988. remcbace, University of C4dffenda at La Jolla, mewed with this view allichad &Iirege, "Technology Could Let Boma Read Minds." is telephone interviews with OTA staff. September 1988. Washington Post, June 3, 1984, p. Cl.

14D App. 9Privacy and Civil Liberties Implications of Testing Employees in the Kir *place 139 the inventor.* Results of the test are available ing pits the interests of the employer in reducing within a few minutes. The analyzer shows a report costs, increasing workplace safety, limiting liabil- on the screen and also prints out a report and stores ity and exercising managerial control against em- a record of the test. ployee interests in maintaining personal dignity The manufacturer claims accuracy in the 99 per- and privacy. Some of the legal questions involved cent range. The device is currently undergoing in- in testing are discussed in chapter 4 of this report. dependent testing, but the results were not avail- In addition, listed below ara some OTA analyses able to OTA at this writing. The Veritas Analyzer dealing in detail with the topics of polygraph test- has already been used by several police depart- ing, drug screening, genetic screening, and with ments and in some workplaces. the constitutional issues involved in workplace testing: Conclusion Review of Deftzse Department's Polygraph Test and Research ProgramsHealth Staff While somewhat different issues are raised by Paper (March 1987); each type of employee testing discussed above, Tests for Human Genetic Disorders (forthcom- there are some common themes. In general, testing, 1988); Urine Drug TestsHealth Testimony (June 10, 1987); " s. Thomas Wrstermau, at el., "Qualitative Measuremest of Drugs," Science, Technology, and the Constitution Laryngoscope, vd. 94, No. 2, February 1994. (forthcoming, 1988).

k,OVERNMENT PRINTING OFFICE : QL: 3 1987 0 63-982 Superintendent of Documents Publication Order Form Charge your order.is rior' sasy! 3111

YES,please send me the following indicated publications:

The Electronic Supervisor: New Technology, New Tensions, GPO stock num'052-003-01082-8; price $6.30.

1. The total cost of my order is $ (International customers please add an additional 25%.) All prices include regular domestic postage and handling and are good through 3/88. After this date, please call Order and Information Desk at 202-783-3238 to verify prices. Please Type or Print 2. 3. Please choose method of payment: (Company or person .; name) Check payable to the Superintendent of Documents (Additional address/attention line) GPO Deposit Account VISA, CHOICE or MasterCard Account (Street address) {11111_1[1 11 I I (City, State, ZIP Code) Thank you for your order! (Credit card expiration date) ( ) (Daytime phone including area code) (Signature) 9/87 4. Mail To: Superintendent of Documents, Government Printing Office, Washington, D.C. 20402-9325

15j Office of TechnologyAssessment

The Office of Technology Assessment(OTA) was created in 1972 as an analytical arm of Congress. OTA's basic functionis to help legislative policymakers anticipate and plan forthe consequences of technological changes and to examine themany ways, expected and unexpected, in which technology affectspeople's lives. The assessment of technology calls for exploration ofthe physical, biological, economic, social, and political impacts thatcan result from applications of scientific knowledge. OTA provides Congresswith independent and timely information about the potentialeffectsboth beneficial and harmfulof technological applications. Requests for studies are made by chairmen ofstanding committees of the House of Representativesor Senate; by the Technology Assess- ment Board, the governing body of OTA;or by the Director of OTA in consultation with the Board. The Technology Assessment Board iscomposed of six members of the House, six members of the Senate,and the OTA Director, who is a non-voting member. OTA has studies under way in nineprogram areas: energy and materials; industry, technology, andemployment; international security and commerce; biological applications;food and renewable resources; health; communication and informationtechnologies; oceans and environment; and science,education, and transportation. AMIN

156