Business Article Critique IIShaun Webbs
Employee Testing: An Ethical Perspective Dan Davidson
ABSTRACT. This paper deals with the conflict between the desire of an employer to test employees for honesty and chemical dependency, and the right of the employee to privacy. Not only is the physical privacy of the employee infringed upon, but the psychic privacy of the individual as well. It is the conclusion of the paper that such an invasion of privacy is not justified without serious and compelling reason, and not die mere chance that testing will reveal problems among some percentage of the tested persons.
The United States Supreme Court officially recog- nized that each individual has a consdtudonal right to privacy in its opinion in Griswold v. Connecticut [381 U.S. 479 (1965)] rendered in 1965. In diis opinion the Court recognized that the right to privacy is penumbral, being based in the First Third, Fourth, and Fifth Amendments. It also foimd that the right apphed to the states through the Fourteenth Amendment However, the opinion did not actually define privacy, it merely recognized that the right to privacy existed and could not be infringed upon without a compelling state interest Since that opinion, the scope of one's privacy has been ques- doned in a number of cases (see. e.g. Roe v. Wade and Stanley v. Ceorgia), and the courts have been forced to "fine-tune" the right usually by balancing the right of the individual to privacy against the pubUc policy consideradons which led to the alleged infiingement upon the right
Dan Davidson is Professor of Business Law at Radford University in Virginia. He has Jive teaching awards, including "Razorback Outstanding Business Faculty' Award, University of Arkansas. He is the author of four textbooks published by Kent Publishing Co., and his articles have been published in the Education Forum, Journal of Insurance Issues and Practices, and the Business Law Review, among others.
Privacy was deemed an important enough issue that Congress enacted the Privacy Act of 1974 [5 U.S.C. 552a (1974)] and then created die Privacy Protecdon Study Commission to invesdgate the issue of privacy. This Commission spent two years (1975—77) studying the issue without ever reaching a consensus on the appropriate definidon of privacy. The Commission submitted a report to Congress in 1977 which induded a chapter dealing with an employee's right to privacy in employment This chapter, 'Personal Privacy in an Informadon Society', offers numerous recommendadons to ensure em- ployee privacy.
The Commission also recognized three major pubhc pohcy objecdves that should be served in an effecdve privacy protecdon pohcy:
(1) the minimizing of obtrusiveness; (2) the maximizing of faimess; and, (3) legidmizing expectadons of privacy (Desjar-
dins&McCall, 1985). At the dme of the report, the Commission felt that these objecdves and the overall goal of privacy protecdon would best be served by voluntary methods in the employment area. It was felt that any governmental intrusions or reguladons would be unnecessary and undesirable. However, the former chairman of the Commission, David Linowes, sub- sequendy wrote that the volimtary compUance in employee privacy by the business sector had failed and that the time for "new incendves" to encourage comphance m^ht be at hand (Desjardins and McCall, 1985).
Today a number of employees — and a number of commentators — are arguing that the right to privacy in the workplace is too often being denied. These people also feel that the new tesdng pro- cedures so often utilized in the employment setdng provide the opportunity for a further decline in
Journal of Business Ethics 7 (1988) 211-217. © 1988 by D. Reidel Publishing Company.
212 Dan Davidson
employee privacy. A number of challenges to employee tesdng procedures are being asserted, often raising consdtudonal arguments, and nearly always raising ediical challenges as well. Several of these tests, and several of the challenges to these tests, will be explored below.
One of the major problems in any attempt to analyze the right of privacy is the development of an acceptable defijiidon of the relevant area. The most common atdtude toward the topic is the view that privacy means the individual has the ri^ht to be "let alone" pesjardins and McCall, 1985). This right to be "let alone" is seen as creating a zone of privacy aroxmd the individual, a zone which should not be infiinged upon by the government without compelling cause. Such a definidon presents several problems for advocates of an employee's right to privacy. First how can one properly assert a zone of privacy in what is, essendally, a sodal environment such as the workplace? Second, even if such an asserdon can be made, how can one object to non- governmental intrusions into this zone of privacy?
Many commentators have avoided these problems by arguing that the right to privacy is inadequately defined in Griswold v. Connecticut, supra, and that the definidon of privacy must be expanded in order to properly treat the issue. It has been urged that privacy must be classified into at least three distinct areas, based on the situadon involved and the interests of the individual in the pardcular situadon (Wasserstrom, 1985). The first area involves the unauthorized use of a person's name, image, or reputadon in order to gain a benefit for the user. This is the type of privacy addressed by Brandeis and Warren in 1890 (Desjardins and McCall, 1985). The second type, grounded in the opinion in Criswold, supra, relates to zones of privacy and protecdon from unwarranted government intrusions into these zones. The third type of privacy deals with the accumuladon of data and informadon about an individual without the consent — or even the knowledge, in many circumstances — of the indi- vidual (Desjardins and McCall, 1985). This diird type is pardcularly troublesome to many people
today because of the ease with which data can be stored, and transferred, by computers.
The first type of privacy has been dealt with legally by the tort laws governing invasions of privacy, and it does not deal with the employee and tbe issue of employee privacy. The second type of privacy has also been dealt with legally in such cases as Griswold, supra, and again, this type of privacy does not deal with the employee or the issue of employee privacy. It is the third type of privacy, the accumula- don and potendal disseminadon of informadon about the individual without the consent of that individual that relates to employee privacy. This third type of privacy has not been legally defined, nor its parameters established by judicial opinions. And it is this tbird type of privacy that is most likely to be involved in alleged infiingements of the privacy of an employee in the workplace.
The fact that informadon about an individual is accumulated is not in and of itself, an infrigement of the individual's right to privacy. There are many ways in which informadon can be acquired, and many types of informadon that can be acquired in a legitimate manner without any infringement To further clarify the problem, there must be some defining of the scope of the right of privacy. Thomas Garrett has isolated three "spheres of privacy" that can be readily identified (Garrett 1966). These three spheres of privacy are: psychic privacy, which is the privacy of the inner self, such as one's thoughts, ideals, and ambidons; physical privacy, which in- volves seclusion from the view or contact of other people; and sodal privacy, which involves the role each individual is called upon to play in his or her daily life. Each sphere of privacy carries with it a presumpdon of privacy that is distinct within the sphere. The area of psychic privacy demands special respect and protecdon because it is personal and intemal. Psychic privacy deals with the very core of one's idendty. Physical privacy is also important and personal, but it is also more easily idendfied and protected. Each of us needs to get off by ourselves occasionally, to be alone. Such privacy is necessary to retain our individualism. Social privacy places certain demands on each of us, since we are expected to perform in certain manners at certain times due to our roles in society. These roles are largely selected by us, and the responsibilides we are obligated to meet are part and parcel of the roles we
Employee Testing 213
have assumed. The accumulation of information about an individual based on that individual's actions in the performance of his or her sodal roles is not necessarily an infringement of the individual's right to social privacy. However, the accumulation of information about the same individual based on the surreptitious reading of die individual's diary is quite likely to be an infringement of the individual's right to psychic and/or physical privacy.
physical privacy is, to a great extent, surrendered in the workplace. The employee knows that he or she will be "forced" to interact with other employees, and that complete physical privacy or isolation is impractical at best and impossible in most circum- stances. Similarly, sodal privacy is circumscribed to a great extent by the employment setting. Since social privacy involves acting within the role one assumes, and since one of the social roles one assumes is die employment role, one surrenders a certain amount of social privacy when one accepts an employment position and the role that goes with that employ- ment. Of course the individual will still retain some physical privacy in the employment setting, as well as the normally expected physical privacy in the non-employment phases of life. Similarly, the fact that the employment setting carries with it a sur- render of some sodal privacy at least in the workplace does not deprive the individual of social privacy in the other sodal roles the individual chooses to perform, such as spouse, parent, and so forth. However, the psychic privacy of the individual is not normally expected to be surrendered in the employment setting. It is this phase, the invasion of the psychic privacy of the employee or the prospec- tive employee, that first brought the issue of the right of employee privacy to the fore. Before the issue of protection of an employee's psychic privacy had been fully resolved, the issue of protecting an employee's right to physical privacy again became timely. Both issues are dealt with below.
Every year American business loses literally billions of dollars due to employee theft, and additional billions of dollars in lost person-hours of work due to illnesses and injuries that are, at least to some extent, drug-related. One estimate of the loss due to
employee theft set the total at $10 bilUon for 1985 (Brenken, 1985). It has also been estimated that alcohol abuse alone reduced American productivity by some $65.6 biUion in 1983 (Takas, 1986), with die total believed to have grown since that time. It is believed that productivity losses from other forms of chemical abuses and dependendes add significandy to the losses related to alcohol abuse. In an effort to combat these losses, which exceed $100 billion per year, many employers have turned to employee testing. The apparent purpose of this new emphasis on testing is to detect those employees or potential employees who are likely to steal from the employer, and to detect those employees or potential em- ployees who have a chemical dependency problem, and to separate them from their employment before the losses begin to accumulate.
In 1895 the Italian criminologist Cesare Lombroso advanced the idea that involuntary physiological responses — blood pressure, in this case — could be used to detect the veracity of people (Dujack, 1986). In 1930, William Marston developed a machine diat could measure blood pressure, perspiration, and respiration and graph die results of these three physiological responses. He dubbed his new machine the polygraph, and declared that it was the first effective "lie detector" (Dujack, 1986) However, the courts of the United States determined that the claims advanced by Marston were not scientifically supported, and refused to allow polygraph results as evidence in trials. Marston was driven out of the very profession he started, but his polygraph lived on — and is enjoying an acceptance today greater than at any time in the past
T h e federal court system still refuses to enter into evidence test results from Marston's wonder machine, but nearly half of the state courts permit polygraph tests under some circumstances, and its use is on the rise among private employen and in the federal government Desperate to do something about the steady flow of classified information to the press and beyond our borders, the Reagan adminis- tration has repeatedly pushed the polygraph as a technological watchdog over the government's vital secrets. And private employers, facing billions of
214 Dan Davidson
dollars worth of losses from employee theft and other misconduct, order perhaps as many as one million tests a year" (Dujack, 1986).
It is estimated diat as many as twenty percent of the nation's employers routinely include polygraph testing as part of the recruiting process and/or routine employee evaluations (Barry, 1986). Hiring and firii^ decisions are regularly based, at least in part, on the results of these polygraph examinations. Therein lies the rub. A polygraph measures stress — which may well occur simply because the person is being subjected to a test that may well result in the denial or loss of employment — and not truth (Dujack, 1986). Yet many people treat the test results as a measure of honesty and as an indicator of trutL To further compound the problem, these examina- tions are not accurate from 10% of the time (accord- it^ to the polygraph industry) to 40% of the time (according to some critics of polygraphs) ('Machines Can Lie', 1983). Even if one assumes the best practical results, a 95% accuracy rate in the test results, what would occur if an employer tested 1000 employees for suspected theft? The test would classify 950 employees accurately and 50 employees inaccurately as far as the "truthfulness" of each employee. If any of the employee thieves were among the 950 properly categorized as truthful, these employees would be found out, and appro- priate disciplinary action could then be taken. But what of the other 50 employees? Some of these employees might well be honest, but would have been classified as thieves because they "failed" die test These employees would be unfairly disdpUned or fired, and could be stigmatized as a result of the polygraph test Others among this group would be classified as non-thieves even though they were, in fact, guilty of thefr from the employer, and would have no action taken against them. This "sample" of 50 misclassified employees was based on a best practical result from the test In reality, the figure could be closer to 100 employees misclassified, if one accepts the polygraph industry estimates of accuracy, or 400, if one accepts the estimates of the critics of the industry. Of course, many people claim that inaccurate results are due to improper training of the test administrators or to improper techniques in administering the test, and not to any inherent flaws with the machine itself (Dujack, 1986). To these proponents of the tests, the "problem" of inaccurate
test results could be solved with tighter standards for the conduct of examiners, and with better training of the examiners. However, the Office of Technological Assessment released a report in 1983 that concluded that while the polygraph may be valuable in the area of criminal investigation, it is virtually worthless as a screening device (Holden, 1986).
Given the apparent problems with the adminis- tration of "He detector" tests, why are they so widely used? One reason frequendy given is the mobiHty of the modem work force. Since workers today are not likely to be long-term residents of a community, it is difficult to safely and thoroughly check the refer- ences and background of applicants for employment Another reason is the relatively inexpensive nature of the test A polygraph examination will normally cost between $50 and $100 to administer. Still another reason is the faith diat many employers place in this machine, coupled with the ease of accepting die test results rather than making the hard choice of whether to hire — or fire — the individual without any "technological" basis for so deciding. If the machine deddes the issue for the supervisor, he or she can then follow the machine's "decision" with a clean consdence.
While polygraph tests are used by some 20% of the nation's businesses, drug testing of one type or another is currendy restricted to a more elite group. At the present time some 25% of the Fortune 500 companies are testing employees and applicants for drug usage, with an additional 20% of the fijms planning to implement mandatory drug test- ing within the next two years (The Yellow Peril', 1986).
Among those firms that currendy test employees, 25% report that any employee who fails the test is fired, with another 64% reporting that they either require any employee who fails the test to submit to treatment, or that they take disdplinary action against the employee (Tlie Yellow Peril', 1986). To compound the problem, many employees are not informed that they are being subjected to a drug test These employees are submitting urine samples as a part of a routine physical examination, never know-
Employee Testing 215
ing that the sample will be analyzed for the presence of drugs until such time as they are informed of the action the employer is taking against them for their drug usage ("The Yellow Peril', 1986).
Most people have no objection to an employer who disciplines any employees who are impaired by drugs while on the job. However, the drug tests currendy being administered do not relate exclu- sively to drug impairment on the job. Rather, the tests deal with any use of prohibited substances at any time and at any place — both on and off the job. To further compound the problem, the Joumal of the American Medical Association reported in 1985 that labs with government contracts to administer drug tests were wrong as much as 100% of the time (Press, 1986). Not only that, but many of the tests are not overly sophisticated. These tests have reported drug use by persons who drink herbal teas, or who are taking prescription medications ('The Yellow Peril', 1986). Finally, a forensic chemist in Adanta has discovered evidence that many dark-skinned individuals, espedally blacks, naturally test positive for marijuana due to the natural pigmentation of their skin (Jet, 1986). Dark-skinned individuals tend to possess high levels of the pigment melanin, which is chemically similar to the THC in marijuana. Unless the drug test is rather sophisticated, these people will test positive for marijuana usage even if they have never used this dmg.
One of the major flaws of several of the most widely used dmg tests is that the tests do not measure the degree of impairment, if any, at the time of the test (The Yellow Peril', 1986). Instead, the tests show only a positive or a negative result as to the presence of drug traces in the urine. Cocaine may show up in such a test as much as three days after consumption. Marijuana may still register for as much as three weeks after its consumption ('The Yellow Peril', 1986). Employers often claim to be testing for the use of such drugs as cocaine, heroin, or PCP when they administer urine tests. However, these dmgs vanish from the bloodstream in 48 to 72 hours, and the tests are often scheduled at least a week in advance. Thus an employee can test "drug- free" by staying off of the dmg in question until after the scheduled examination CThe Yellow Peril', 1986). Meanwhile, the tests detect the presence of marijuana, and may also give "false-positive" read- ings for dmgs due to the use by the employee of
prescription dmgs, Advil, Nuprin, herbal teas, or even poppy seed bagels! (The Yellow Peril', 1986).
Judge Irvir^ Kaufhian of the Second Circuit Court of Appeals serves as the chairman of the President's Commission on Organized Crime. The Committee recommended that all federal employees be screened for dmg use, that all private employers who have government contracts should screen their employees for dmg use, and that any private employers who refused to initiate dmg testing for all employees be denied any further government contracts until such time as they comply with the testing "suggestion" (Hentoff, 1986). The Commis- sion also urged all other private employers to begin testing soon. It is the opinion of the Commission that such wide-spread testing in the workplace will help to diminish the demand for illegal dmgs, thus alleviating the dmg problem to a significant extent (Hentoff, 1986). Both Judge Kaufman and Attorney General Edwin Meese feel that such testing is not an unlawful search and seizure, primarily because the emiployee "consents" to the search before it is undertaken as a condition of the employment They also analogize the mandatory dmg testing to manda- tory searches via x-rays and metal detectors at airports (Hentoff, 1986).
An employer who requires a polygraph examination as a part of the employment screening process or one who uses a polygraph examination in routine employee reviews is probably invading the psychic privacy of that employee. In so doing, the employer is on shaky grounds, at best The accuracy and utility of polygraph tests is so questionable that the FBI forbids polygraph dragnets, and the American Psy- chological Association prohibits its members from admitiistering such tests (Dujack, 1986). Even those employers who only administer polygraph tests to those employees who voluntarily consent to such tests are facing potential problems. "Polygraphers claim that their subjects voluntarily waive their rights, but there's nodiing voluntary about waiving basic rights to keep or to get a job. Moreover, no waiver can be voluntary if it's made without full knowledge of the risks, and few polygraph subjects know what they are getting into" (Pyle, 1983). Added
216 Dan Davidson
to this fact is the low accuracy rate, and general doubts about the validity of the tests. Dr. David T. Lykken, a psychology professor at the University of Minnesota says of polygraphs, "They're a fake, they have at best a 50% accuracy rate" (Press, 1986). David C. Raskin, a psychology professor at the University of Utah and a Ucensed polygrapher, feels that while the polygraph may have valid uses in the area of law enforcement, they "should be banned in screening and periodic testing of employees. The level of examiners is so low that it's just easier to eliminate (testing) in the private sector" (Press, 1986).
If the polygraph is still to be used despite these concerns, a number of considerations should be added to the tests. First, the information being sought should be clearly and significandy related to the job. The tests should not be fishing expeditions, nor should the test results be the sole factor upon which job-based dedsions are based. Second, since the use of polygraphs involves an invasion of psychic privacy, the persons dedding to use these tests should consider whether they have compellii^ reasons to resort to the test, or whether they are merely using the tests as the easiest available alterna- tive. Third, die information gathered from polygraph exams that are given despite these considerations should be handled carefully, with access to the infonnation being restricted to the greatest extent possible.
Dmg testing presents two possible invasion of employee privacy issues. Obviously a dmg test involves the invasion of physical privacy, and it may also involve the invasion of psychic privacy as well. Dragnet dmg tests of all employees seems to imply that the employee is guilty until proven innocent This attitude direcdy contradicts our basic legal model. In addition, state or federal employees seemingly cannot be subjected to such testing unless there is a reasonable suspidon that the particular individual being tested is guilty of using dmgs (Hentoff, 1986). In addition, many of the most popular tests in use today have a tremendous inaccu- racy factor. The single most common test, EMIT (Enzyme Multiplied Immunoassay Test), has a 25 percent false positive rate and a 50 percent false negative rate (Dentzer et al., 1986). Until a substan- tially higher accuracy rate can be developed, the invasion of employee privacy does not appear to be
offset by any adequate state interests or by adequate safeguards of the employee's rights.
In the end, we come down to a simple question: "which set of rights is paramount, those of com- panies seeking a productive and safe workforce — or those of employees trying to protect their privacy?" (Dentzer et al, 1986). At the present time, die set of rights that is paramount are those of the employees, as much by default as anything else. Until the employer can vahdate the testing and also validate its application to the rights of the employer, there should be no contest Employee privacy must win out
Criswoldv. Connecticut, 381 \5S. 479,85 S.Ct. 1678 (1965). Stanley v. Ceorgia, 394 U.S. 557,89 S.Ct. 1243 (1969). Roe V. Wade, 410 US. 113,93 S.Ct. 705 (1973).
Privacy Act of 1974, 5 U.S.C 552a (1974).
Books and articles
Garrett, Business Ethics, 64—67 (1966). 'Machines Can Lie, If Truth Be Known', USA Today,
February 17,1983 at lOA. Pyle, These Tests Are Meant to Scare People', USA Today,
February 17, 1983 at lOA. Brenkert, 'Privacy, Polygraphs, and Work', Contemporary
Issues in Business Ethics, 227 (1985). Desjardins and McCall, Contemporary Issues in Business Ethics,
200-203(1985). Wasserstrom, 'Privacy', Contemporary Issues in Business Ethics,
204 (1985). Barry, Moral Issues in Business, 283 (2d ed., 1986). Dentzer et ai, 'Can You Pass the Job Test?', Newsweek, May
5,1986 at 46-53. Dujack, 'Polygraph Fever', The New Republic, August 4, 1986
at 10. Hentoff, 'Presumption of Guilt', The Progressive, May, 1986 at
Employee Testing 217
Holden, 'Days May Be Numbered for Polygraphs in the The Yellow Peril', T/ieNeH/ilep«W('c, March 31,1986 at 7. Private Sector', Science, May 9,1986 at 705.
'Sciendst Reveals Blacks Who Fail Marijuana Tests Could Department of Aaounting and Finance, Prove Drug Free', Jet, May 19,1986, at 10— 11. pQ gp^ 5 y^$
Press, 'First the lie Detector, Then the Chemicals', Newsweek, Radford University January 27,1986 at 56—57. Radfo d VA 24141
Takas,'They Want Your Body: Can Your Boss Test You for -̂ ' r j c / Drug Use?', Vogue, April, 1986 at 156. " "