Drug Testing in Employment

Author(s): Joseph DesJardins and Ronald Duska

Source: Business & Professional Ethics Journal , Fall, 1987, Vol. 6, No. 3 (Fall, 1987), pp.
3-21
Published by: Philosophy Documentation Center

Stable URL: https://www.jstor.org/stable/27799951

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Philosophy Documentation Center is collaborating with JSTOR to digitize, preserve and extend
access to Business & Professional Ethics Journal

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
BUSINESS & PROFESSIONAL ETHICS JOURNAL, VOL. 6, NO. 3

Drug Testing in Employment

Joseph DesJardins and Ronald Duska

According to one survey, nearly one-half of all Fortune 500 companies were
planning to administer drug tests to employees and prospective employees by
the end of 1987.1 Counter to what seems to be the current trend in favor of
drug testing, we will argue that it is rarely legitimate to override an
employee's or applicant's right to privacy by using such tests or procedures.2

OPENING STIPULATIONS

We take privacy to be an "employee right" by which we mean a presumptive

moral entitlement to receive certain goods or be protected from certain
harms in the workplace.3 Such a right creates a prima facie obligation on the
part of the employer to provide the relevant goods or, as in this case, refrain
from the relevant harmful treatment. These rights prevent employees from
being placed in the fundamentally coercive position where they must choose
between their job and other basic human goods.
Further, we view the employer-employee relationship as essentially
contractual. The employer-employee relationship is an economic one and,
unlike relationships such as those between a government and its citizens or a
parent and a child, exists primarily as a means for satisfying the economic
interests of the contracting parties. The obligations that each party incurs are
only those that it voluntarily takes on. Given such a contractual relationship,
certain areas of the employee's life remain their own private concern and no
employer has a right to invade them. On these presumptions we maintain
that certain information about an employee is rightfully private, i.e. the
employee has a right to privacy.

THE RIGHT TO PRIVACY

According to George Brenkert, a right to privacy involves a three-place

relation between a person A, some information X, and another person B.

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
4 Business & Professional Ethics Journal

The right to privacy is violated only when deliberately comes to possess

information X about A, and no relationship between A and exists which
would justify B's coming to know X about A.4 Thus, for example, the
relationship one has with a mortgage company would justify that company's
coming to know about one's salary, but the relationship one has with a
neighbor does not justify the neighbor's coming to know that information.
Hence, an employee's right to privacy is violated whenever personal
information is requested, collected and/or used by an employer in a way or
for any purpose that is irrelevant to or in violation of the contractual
relationship that exists between employer and employee.
Since drug-testing is a means for obtaining information, the information
sought must be relevant to the contract in order for the drug testing not to
violate privacy. Hence, we must first decide if knowledge of drug use
obtained by drug testing is job-relevant. In cases where the knowledge of
drug use is not relevant, there appears to be no justification for subjecting
employees to drug tests. In cases where information of drug use is job
relevant, we need to consider if, when, and under what conditions using a
means such as drug testing to obtain that knowledge is justified.

IS KNOWLEDGE OF DRUG USE JOB RELEVANT INFORMATION?

There seem to be two arguments used to establish that knowledge of drug

use is job relevant information. The first argument claims that drug use
adversely affects job performance thereby leading to lower productivity,
higher costs, and consequently lower profits. Drug testing is seen as a way of
avoiding these adverse effects. According to some estimates twenty-five
billion ($25,000,000,000) dollars are lost each year in the United States
because of drug use.5 This occurs because of loss in productivity, increase in
costs due to theft, increased rates in health and liability insurance, and such.
Since employers are contracting with an employee for the performance of
specific tasks, employers seem to have a legitimate claim upon whatever
personal information is relevant to an employee's ability to do the job.
The second argument claims that drug use has been and can be
responsible for considerable harm to the employee him/herself, fellow
employees, the employer, and/or third parties, including consumers. In this
case drug testing is defended because it is seen as a way of preventing
possible harm. Further, since employers can be held liable for harms done
both to third parties, e.g. customers, and to the employee or his/her fellow

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 5

employees, knowledge of employee drug use will allow employers to gain

information that can protect themselves from risks such as liability. But how
good are these arguments? We turn to examine the arguments more closely.

THE FIRST ARGUMENT: JOB PERFORMANCE AND

KNOWLEDGE OF DRUG USE

The first argument holds that drug use leads to lower productivity and
consequently implies that a knowledge of drug use obtained through drug
testing will allow an employer to increase productivity. It is generally
assumed that people using certain drugs have their performances affected by
such use. Since enhancing productivity is something any employer desires,
any use of drugs that reduces productivity affects the employer in an
undesirable way, and that use is,then, job-relevant. If such production losses
can be eliminated by knowledge of the drug use, then knowledge of that drug
use is job-relevant information. On the surface this argument seems
reasonable. Obviously some drug use in lowering the level of performance
can decrease productivity. Since the employer is entitled to a certain level of
performance and drug use adversely affects performance, knowledge of that
use seems job-relevant.
But this formulation of the argument leaves an important question
unanswered. To what level of performance are employers entitled? Optimal
performance, or some lower level? If some lower level, what? Employers
have a valid claim upon some certain level of performance, such that a failure
to perform up to this level would give the employer a justification for
disciplining, firing or at least finding fault with the employee. But that does
not necessarily mean that the employer has a right to a maximum or optimal
level of performance, a level above and beyond a certain level of accept
ability. It might be nice if the employee gives an employer a maximum effort
or optimal performance, but that is above and beyond the call of the
employee's duty and the employer can hardly claim a right at all times to the
highest level of performance of which an employee is capable.
That there are limits on required levels of performance and productivity
becomes clear if we recognize that job performance is person related. It is
person-related because one person's best efforts at a particular task might
produce results well below the norm, while another person's minimal efforts
might produce results abnormally high when compared to the norm. For ex
ample a professional baseball player's performance on a ball field will be

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
6 Business a Professional Ethics Journal

much higher than the average person's since the average person is unskilled
at baseball. We have all encountered people who work hard with little or no
results, as well as people who work little with phenomenal results. Drug use
by very talented people might diminish their performance or productivity, but
that performance would still be better than the performance of the average
person or someone totally lacking in the skills required. That being said, the
important question now is whether the employer is entitled to an employee's
maximum effort and best results, or merely to an effort sufficient to perform
the task expected.
If the relevant consideration is whether the employee is producing as
expected (according to the normal demands of the position and contract) not
whether he/she is producing as much as possible, then knowledge of drug
use is irrelevant or unnecessary. Let's see why.
If the person is producing what is expected, knowledge of drug use on the
grounds of production is irrelevant since, ex hypothesi the production is
satisfactory. If, on the other hand, the performance suffers, then, to the
extent that it slips below the level justifiably expected, the employer has
prima facie grounds for warning, disciplining or releasing the employee. But
the justification for this is the person's unsatisfactory performance, not the
person's use of drugs. Accordingly, drug use information is either unneces
sary or irrelevant and consequently there are not sufficient grounds to
override the right of privacy. Thus, unless we can argue that an employer is
entitled to optimal performance, the argument fails.
This counter-argument should make it clear that the information which is
job-relevant, and consequently which is not rightfully private, is information
about an employee's level of performance and not information about the
underlying causes of that level. The fallacy of the argument which promotes
drug testing in the name of increased productivity is the assumption that each
employee is obliged to perform at an optimal, or at least quite high, level.
But this is required under few, if any, contracts. What is required con
tractually is meeting the normally expected levels of production or perform
ing the tasks in the job-description adequately (not optimally). If one can do
that under the influence of drugs, then on the grounds of job-performance at
least, drug use is rightfully private. If one cannot perform the task
adequately, then the employee is not fulfilling the contract, and knowledge of
the cause of the failure to perform is irrelevant on the contractual model.
Of course, if the employer suspects drug use or abuse as the cause of the
unsatisfactory performance, then she might choose to help the person with

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 7

counseling or rehabilitation. However, this does not seem to be something

morally required of the employer. Rather, in the case of unsatisfactory
performance, the employer has a prima facie justification for dismissing or
disciplining the employee.
Before turning to the second argument which attempts to justify drug
testing, we should mention a factor about drug use that is usually ignored in
talk of productivity. The entire productivity arugument is irrelevant for those
cases in which employees use performance enhancing drugs. Amphetamines
and steroids, for example, can actually enhance some performances. This
points to the need for care when tying drug testing to job-performance. In
the case of some drugs used by athletes, for example, drug testing is done
because the drug-influenced performance is too good and therefore unfair,
not because it leads to inadequate job-performance. In such a case, where
the testing is done to ensure fair competition, the testing may be justified.
But drug testing in sports is an entirely different matter than drug-testing in
business.
To summarize our argument so far. Drug use may affect performances,
but as long as the performance is at an acceptable level, the knowledge of
drug use is irrelevant. If the performance is unacceptable, then that is
sufficient cause for action to be taken. In this case an employee's failure to
fulfill his/her end of a contract makes knowledge of the drug use unnecessary.

THE SECOND ARGUMENT: HARM AND THE

KNOWLEDGE OF DRUG USE TO PREVENT HARM

Even though the performance argument is inadequate, there is an argument

that seems somewhat stronger. This is an argument based on the potential
for drug use to cause harm. Using a type of Millian argument, one could
argue that drug testing might be justified if such testing led to knowledge that
would enable an employer to prevent harm. Drug use certainly can lead to
harming others. Consequently, if knowledge of such drug use can prevent
harm, then, knowing whether or not one's employee uses drugs might be a
legitimate concern of an employer in certain circumstances. This second
argument claims that knowledge of the employee's drug use is job-relevant
because employees who are under the influence of drugs can pose a threat to
the health and safety of themselves and others, and an employer who knows
of that drug use and the harm it can cause has a responsiblity to prevent it.
Employers have both a general duty to prevent harm and the specific

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
8 Business & Professional Ethics Journal

responsibility for harms done by their employees. Such responsibilities are

sufficient reason for an employer to claim that information about an
employee's drug use is relevant if that knowledge can prevent harm by giving
the employer grounds for dismissing the employee or not allowing him/her
to perform potentially harmful tasks. Employers might even claim a right to
reduce unreasonable risks, in this case the risks involving legal and economic
liability for harms caused by employees under the influence of drugs, as
further justification for knowing about employee drug use.
This second argument differs from the first in which only a lowered job
performance was relevant information. In this case, even to allow the per
formance is problematic, for the performance itself, more than being in
adequate, can hurt people. We cannot be as sanguine about the prevention
of harm as we can about inadequate production. Where drug use can cause
serious harms, knowledge of that use becomes relevant if the knowledge of
such use can lead to the prevention of harm and drug testing becomes justi
fied as a means for obtaining that knowledge.
As we noted, we will begin initially by accepting this argument on roughly
Millian grounds where restrictions on liberty are allowed in order to prevent
harm to others. (The fact that one is harming oneself, if that does not harm
others is not sufficient grounds for interference in another's behavior
according to Mill.) In such a case an employer's obligation to prevent harm
may over-ride the obligation to respect an employee's privacy.
But let us examine this more closely. Upon examination, certain prob
lems arise, so that even if there is a possibility of justifying drug testing to
prevent harm, some caveats have to be observed and some limits set out.

JOBS WITH POTENTIAL TO CAUSE HARM

To say that employers can use drug-testing where that can prevent harm is
not to say that every employer has the right to know about the drug use of
every employee. Not every job poses a serious enough threat to justify an
employer coming to know this information.
In deciding which jobs pose serious enough threats certain guidelines
should be followed. First the potential for harm should be clear and present.
Perhaps all jobs in some extended way pose potential threats to human well
being. We suppose an accountant's error could pose a threat of harm to
someone somewhere. But some jobs like those of airline pilots, school bus
drivers, public transit drivers and surgeons, are jobs in which unsatisfactory

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 9

performance poses a clear and present danger to others. It would be much

harder to make an argument that job performances by auditors, secretaries,
executive vice-presidents for public relations, college teachers, professional
athletes, and the like, could cause harm if those performances were carried
on under the influence of drugs. They would cause harm only in exceptional
cases.6

NOT EVERY PERSON IS TO BE TESTED

But, even if we can make a case that a particular job involves a clear and
present danger for causing harm if performed under the influence of drugs, it
is not appropriate to treat everyone holding such a job the same. Not every
job-holder is equally threatening. There is less reason to investigate an
airline pilot for drug use if that pilot has a twenty-year record of exceptional
service than there is to investigate a pilot whose behavior has become erratic
and unreliable recently, or than one who reports to work smelling of alcohol
and slurring his words. Presuming that every airline pilot is equally
threatening is to deny individuals the respect that they deserve as autono
mous, rational agents. It is to ignore previous history and significant
differences. It is also probably inefficient and leads to the lowering of
morale. It is the likelihood of causing harm, and not the fact of being an
airline pilot perse, that is relevant in deciding which employees in critical jobs
to test.
So, even if knowledge of drug use is justifiable to prevent harm, we must be
careful to limit this justification to a range of jobs and people where the
potential for harm is clear and present. The jobs must be jobs that clearly
can cause harm, and the specific employee should not be someone who is
reliable with a history of such reliability. Finally, the drugs being tested
should be those drugs, the use of which in those jobs is really potentially
harmful.

LIMITATIONS ON DRUG TESTING POLICIES

Even when we identify those jobs and individuals where knowledge of drug
use would be job relevant information, we still need to examine whether
some procedural limitations should not be placed upon the employer's testing
for drugs. We have said that in cases where a real threat of harm exists and
where evidence exists suggesting that a particular employee poses such a

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
10 Business & Professional Ethics Journal

threat, an employer could be justified in knowing about drug use in order to

prevent the potential harm. But we need to recognize that as long as the
employer has the discretion for deciding when the potential for harm is clear
and present, and for deciding which employees pose the threat of harm, the
possibility of abuse is great. Thus, some policy limiting the employer's power
is called for.
Just as criminal law places numerous restrictions protecting individual
dignity and liberty on the state's pursuit of its goals, so we should expect that
some restrictions be placed on an employer in order to protect innocent em
ployees from harm (including loss of job and damage to one's personal and
professional reputation). Thus, some system of checks upon an employer's
discretion in these matters seems advisable. Workers covered by collective
bargaining agreements or individual contracts might be protected by clauses
in those agreements that specify which jobs pose a real threat of harm (e.g.
pilots but not cabin attendants) and what constitutes a just cause for in
vestigating drug use. Local, state, and federal legislatures might do the same
for workers not covered by employment contracts. What needs to be set up is
a just employment relationship-one in which an employee's expectations
and responsibilities are specified in advance and in which an employer's
discretionary authority to discipline or dismiss an employee is limited.
Beyond that, any policy should accord with the nature of the employment
relationship. Since that relationship is a contractual one, it should meet the
condition of a morally valid contract, which is informed consent. Thus, in
general, we would argue that only methods that have received the informed
consent of employees can be used in aquiring information about drug use.7
A drug-testing policy that requires all employees to submit to a drug test
or to jeopardize their job would seem coercive and therefore unacceptable.
Being placed in such a fundamentally coercive position of having to choose
between one's job and one's privacy does not provide the conditions for a
truly free consent. Policies that are unilaterally established by employers
would likewise be unacceptable. Working with employees to develop com
pany policy seems the only way to insure that the policy will be fair to both
parties. Prior notice of testing would also be required in order to give
employees the option of freely refraining from drug use. It is morally
preferable to prevent drug use than to punish users after the fact, since this
approach treats employees as capable of making rational and informed
decisions.
Further procedural limitations seem advisable as well. Employees should

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 11

be notified of the results of the test, they should be entitled to appeal the
results (perhaps through further tests by an independent laboratory) and the
information obtained through tests ought to be kept confidential. In
summary, limitations upon employer discretion for administering drug tests
can be derived from the nature of the employment contract and from the
recognition that drug testing is justified by the desire to prevent harm, not the
desire to punish wrong doing.

EFFECTIVENESS OF DRUG-TESTING

Having declared that the employer might have a right to test for drug use in
order to prevent harm, we still need to examine the second argument a little
more closely. One must keep in mind that the justification of drug testing is
the justification of a means to an end, the end of preventing harm, and that
the means are a means which intrude into one's privacy. In this case, before
one allows drug testing as a means, one should be clear that there are not
more effective means available.
If the employer has a legitimate right, perhaps duty, to ascertain know
ledge of drug use to prevent harm, it is important to examine exactly how
effectively, and in what situations, the knowledge of the drug use will prevent
the harm. So far we have just assumed that the knowledge will prevent the
harm. But how?
Let us take an example to pinpoint the difficulty. Suppose a transit driver,
shortly before work, took some cocaine which, in giving him a feeling of in
vulnerability, leads him to take undue risks in his driving. How exactly is
drug-testing going to contribute to the knowlege which will prevent the
potential accident?
It is important to keep in mind that; (1) if the knowledge doesn't help
prevent the harm, the testing is not justified on prevention grounds; (2) if the
testing doesn't provide the relevant knowlege it is not justified either; and
finally,(3) even if it was justified, it would be undesirable if a more effective
means for preventing harm were discovered.
Upon examination, the links between drug-testing, knowledge of drug use,
and prevention of harm are not as clear as they are presumed to be. As we
investigate, it begins to seem that the knowledge of the drug use even though
relevant in some instances is not the most effective means to prevent harm.
Let us turn to this last consideration first. Is drug-testing the most
effective means for preventing harm caused by drug use?

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
12 Business & Professional Ethics Journal

Consider. If someone exhibits obviously drugged or drunken behavior,

then this behavior itself is grounds for preventing the person from continuing
in the job. Administering urine or blood tests, sending the specimens out for
testing and waiting for a response, will not prevent harm in this instance.
Such drug testing because of the time lapse involved, is equally superfluous in
those cases where an employee is in fact under the influence of drugs, but
exhibits no or only subtley impaired behaviour.
Thus, even if one grants that drug testing somehow prevents harm an
argument can be made that there might be much more effective methods of
preventing potential harm such as administering dexterity tests of the type
employed by police in possible drunk-driving cases, or requiring suspect
pilots to pass flight simulator tests.8 Eye-hand coordination, balance,
reflexes, and reasoning ability can all be tested with less intrusive, more easily
administered, reliable technologies which give instant results. Certainly if an
employer has just cause for believing that a specific employee presently poses
a real threat of causing harm, such methods are just more effective in all
ways than are urinalysis and blood testing.
Even were it possible to refine drug tests so that accurate results were
immediately available, that knowledge would only be job relevant if the drug
use was clearly the cause of impaired job performance that could harm
people. Hence, testing behavior still seems more direct and effective in
preventing harm than testing for the presence of drugs perse.
In some cases, drug use might be connected with potential harms not by
being causally connected to motor-function impairment, but by causing per
sonality disorders (e.g. paranoia, delusions, etc.) that affect judgmental
ability. Even though in such cases a prima facie justification for urinalysis or
blood testing might exist, the same problems of effectiveness persist. How is
the knowledge of the drug use attained by urinalysis and/or blood testing
supposed to prevent the harm? Only if there is a causal link between the use
and the potentially harmful behavior, would such knowledge be relevant.
Even if we get the results of the test immediately, there is the necessity to
have an established causal link between specific drug use and anticipated
harmful personality disorders in specific people.
But it cannot be the task of an employer to determine that a specific drug is
causally related to harm-causing personality disorders. Not every controlled
substance is equally likely to cause personality changes in every person in
every case. The establishement of the causal link between the use of certain
drugs and harm-causing personality disorders is not the province of the

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 13

employer, but the province of experts studying the effects of drugs. The
burden of proof is on the employer to establish that the substance being
investigated has been independently connected with the relevant psycho
logical impairment and then, predict on that basis that the specific
employee's psychological judgment has been or will soon be impaired in such
a way as to cause harm.
But even when this link is established, it would seem that less intrusive
means could be used to detect the potential problems, rather than relying
upon the assumption of a causal link. Psychological tests of judgment,
perception and memory, for example, would be a less intrusive and more
direct means for acquiring the relevant information which is, after all, the
likelihood of causing harm and not the presence of drugs per se. In short,
drug testing even in these cases doesn't seem to be very effective in
preventing harm on the spot.
Still, this does not mean it is not effective at all. Where it is most effective
in preventing harm is in its getting people to stop using drugs or in identifying
serious drug addiction. Or to put it another way, urinalysis and blood tests
for drug use are most effective in preventing potential harm when they serve
as a deterrent to drug use before it occurs, since it is very difficult to prevent
harm by diagnosing drug use after it has occured but before the potentially
harmful behavior takes place.
Drug testing can be an effective deterrent when there is regular or random
testing of all employees. This will prevent harm by inhibiting (because of the
fear of detection) drug use by those who are occasional users and those who
do not wish to be detected.
It will probably not inhibit or stop the use by the chronic addicted user,
but it will allow an employer to discover the chronic user or addict, assuming
that the tests are accurately administered and reliably evaluated. If the
chronic user's addiction would probably lead to harmful behavior of others,
the harm is prevented by taking that user off the job. Thus regular or
random testing will prevent harms done by deterring the occasional user and
by detecting the chronic user.
There are six possiblities for such testing:
1. regularly scheduled testing of all employees;
2. regularly scheduled testing of randomly selected employees;
3. randomly scheduled testing of all employees;
4. randomly scheduled testing of randomly selected employees;

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
14 Business & Professional Ethics Journal

5. regularly scheduled testing of employees selected for

probable cause; or finally,
6. randomly scheduled testing of employees selected for
probable cause.
Only the last two seem morally acceptable as well as effective.
Obviously, randomly scheduled testing will be more effective than regu
larly scheduled testing in detecting the occasional user, because the occa
sional users can control their use to pass the tests, unless of course tests were
given so often (a practice economically unfeasible) that they needed to stop
altogether. Regular scheduling probably will detect the habitual or addicted
user. Randomly selecting people to test is probably cheaper, as is random
scheduling, but it is not nearly as effective as testing all. Besides, the random
might miss some of the addicted altogether, and will not deter the risk takers
as much as the risk aversive persons. It is, ironically, the former who are
probably potentially more harmful.
But these are merely considerations of efficiency. We have said that
testing without probable cause is unacceptable. Any type of regular testing of
all employees is unacceptable. We have argued that testing employees
without first establishing probable cause is an unjustifiable violation of
employee privacy. Given this, and given the expense of general and regular
testing of all employees (especially if this is done by responsible
laboratories), it is more likely that random testing will be employed as the
means of deterrence. But surely testing of randomly selected innocent
employees is as intrusive to those tested as is regular testing. The argument
that there will be fewer tests is correct on quantitative grounds, but
qualitatively the intrusion and unacceptability are the same. The claim that
employers should be allowed to sacrifice the well-being of (some few)
innocent employees to deter (some equally few) potentially harmful em
ployees seems, on the face of it, unfair. Just as we do not allow the state
randomly to tap the telephones of just any citizen in order to prevent crime,
so we ought not allow employers to drug test all employees randomly to
prevent harm. To do so is again to treat innocent employees solely as a
means to the end of preventing potential harm.
This leaves only the use of regular or random drug-testing as a deterrent
in those cases where probable cause exists for believing that a particular
employee poses a threat of harm. It would seem that in this case, the drug
testing is acceptable. In such cases only the question of effectiveness
remains: Are the standard techniques of urinalysis and blood-testing more

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 15

effective means for preventing harms than alternatives such as dexterity

tests? It seems they are effective in different ways. The dexterity tests show
immediately if someone is incapable of performing a task, or will perform
one in such a way as to cause harm to others. The urinalysis and blood
testing will prevent harm indirectly by getting the occasional user to curtail
their use, and by detecting the habitual or addictive user, which will allow the
employer to either give treatment to the addictive personality or remove
them from the job. Thus we can conclude that drug testing is effective in a
limited way, but aside from inhibiting occasional users because of fear of
detection, and discovering habitual users, it seems problematic that it does
much to prevent harm that couldn't be achieved by other means.
Consider one final issue in the case of the occasional user. They are the
drug users who do weigh the risks and benefits and who are physically and
psychologically free to decide. The question in their case is not simply "will
the likelihood of getting caught by urinalysis or blood-testing deter this
individual from using drugs?" Given the benefits of psychological tests and
dexterity tests described above, the question is "will the rational user be more
deterred by urinalysis or blood-testing than by random psychological or dex
terity tests?" And, if this is so, is this increase in the effectiveness of a
deterrent sufficient to offset the increased expense and time required by drug
tests?9 We see no reason to believe that behavioral or judgment tests are
not, or cannot be made to be, as effective in determining what an employer
needs to know (i.e., that a particular employee may presently be a potential
cause of harm). If the behavioral, dexterity and judgment tests can be as
effective in determining a potential for harm, we see no reason to believe that
they cannot be as effective a deterrent as drug tests. Finally, even if a case
can be made for an increase in deterrent effect of drug testing, we are
skeptical that this increased effectiveness will outweigh the increased
inefficiencies.
In summary, we have seen that deterrence is effective at times and under
certain conditions allows the sacrificing of the privacy rights of innocent
employees to the future and speculative good of preventing harms to others.
However, there are many ways to deter drug use when that deterrence is
legitimate and desirable to prevent harm. But random testing, which seems
the only practicable means which has an impact in preventing harm is the one
which most offends workers rights to privacy and which is most intrusive of
the rights of the innocent. Even when effective, drug testing as a deterrent
must be checked by the rights of employees.

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 16 Business & Professional Ethics Journal

THE ILLEGALITY CONTENTION

At this point critics might note that the behavior which testing would try to
deter is, after all, illegal. Surely this excuses any responsible employer from
being overly protective of an employee's rights. The fact that an employee is
doing something illegal should give the employer a right to that information
about his private life. Thus, it is not simply that drug use might pose a threat
of harm to others, but that it is an illegal activity that threatens others. But
again, we would argue that illegal activity itself is irrelevant to job
performance. At best conviction records might be relevant, but, of course,
since drug tests are administered by private employers we are not only
ignoring the question of conviction, we are also ignoring the fact that the
employee has not even been arrested for the alleged illegal activity.
Further, even if the due process protections and the establishment of guilt
are acknowledged, it still does not follow that employers have a claim to
know about all illegal activity on the part of their employees.
Consider the following example: Suppose you were hiring an auditor
whose job required certifying the integrity of your firm's tax and financial
records. Certainly, the personal integrity of this employee is vital to the
adequate job performance. Would we allow the employer to conduct, with or
without the employee's consent, an audit of the employee's own personal tax
return? Certainly if we discover that this person has cheated on his/her own
tax return we will have evidence of illegal activity that is relevant to this
person's ability to do the job. Given one's own legal liability for filing
falsified statements, the employee's illegal activity also poses a threat to
others. But surely, allowing private individuals to audit an employee's tax
returns is too intrusive a means for discovering information about that
employee's integrity. The government certainly would never allow this vio
lation of an employee's privacy. It ought not to allow drug testing on the
same grounds. Why tax returns should be protected in ways that urine, for
example, is not, raises interesting questions of fairness. Unfortunately, this
question would take us beyond the scope of this paper.

VOL UNTARINESS

A final problem that we also leave undeveloped concerns the voluntariness

of employee consent. For most employees, being given the choice between
submitting to a drug test and risking one's job by refusing an employer's

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 17

request is not much of a decision at all. We believe that such decisions are
less than voluntary and thereby would hold that employers cannot escape our
criticisms simply by including within the employment contract a drug-testing
clause.10 Furthermore, there is reason to believe that those most in need of
job security will be those most likely to be subjected to drug testing. Highly
skilled, professional employees with high job mobility and security will be in a
stronger position to resist such intrusions than will less skilled, easily replaced
workers. This is why we should not anticipate surgeons and airline pilots
being tested, and should not be surprised when public transit and factory
workers are. A serious question of fairness arises here as well.
Drug use and drug testing seem to be our most recent social "crises."
Politicians, the media, and employers expend a great deal of time and effort
addressing this crisis. Yet, unquestionably, more Uves, health, and money
are lost each year to alcohol abuse than to marijuana, cocaine and other
controlled substances. We are well-advised to be careful in considering
issues that arise due to such selective social concern. We will let other social
commentators speculate on the reasons why drug use has received scrutiny
while other white-collar crimes and alcohol abuse are ignored. Our only
concern at this point is that such selective prosecution suggests an
arbitrariness that should alert us to questions of fairness and justice.
In summary, then, we have seen that drug use is not always job-relevant,
and if drug use is not job relevant, information about it is certainly not job
relevant. In the case of performance it may be a cause of some decreased
peformance, but it is the performance itself that is relevant to an employee's
position, not what prohibits or enables him to do the job. In the case of
potential harm being done by an employee under the influence of drugs, the
drug use seems job relevant, and in this case drug testing to prevent harm
might be legitimate. But how this is practicable is another question. It would
seem that standard motor dexterity or mental dexterity tests, immediately
prior to job performance, are more efficacious ways of preventing harm,
unless one concludes that drug use invariably and necessarily leads to harm.
One must trust the individuals in any system in order for that system to
work. One cannot police everything. It might work to randomly test people,
to find drug users, and to weed out the few to forestall possible future harm,
but are the harms prevented sufficient to over-ride the rights of privacy of the
people who are innocent and to overcome the possible abuses we have
mentioned? It seems not.
Clearly, a better method is to develop safety checks immediately prior to

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
18 Business & Professional Ethics Journal

the performance of a job. Have a surgeon or a pilot or a bus driver pass a few
reasoning and motor-skill tests before work. The cause of the lack of a skill,
which lack might lead to harm, is really a secondary issue.

DRUG TESTING FOR PROSPECTIVE EMPLOYEES

Let's turn finally to drug testing during a pre-employment interview. Assum

ing the job description and responsibilities have been made clear, we can say
that an employer is entitled to expect from a prospective employee whatever
performance is agreed to in the employment contract. Of course, this will
always involve risks, since the employer must make a judgment about future
performances. To lower this risk, employers have a legitimate claim to some
information about the employee. Previous work experience, training, educa
tion, and the like are obvious candidates since they indicate the person's
ability to do the job. Except in rare circumstances drug use itself is irrelevant
for determining an employee's ability to perform. (Besides, most people who
are interviewing know enough to get their systems clean if the prospective
employee is going to test them.)
We suggest that an employer can claim to have an interest in knowing (a)
whether or not the prospective employee can do the job and (b) whether
there is reason to believe that once hired the employee will do the job. The
first can be determined in fairly straightforward ways: past work experience,
training, education, etc. Presumably past drug use is thought more relevant
to the second question. But there are straightforward and less intrusive
means than drug-testing for resolving this issue. Asking the employee Ts
there anything that might prevent you from doing this job?" comes first to
mind. Hiring the employee on a probationary period is another way. But to
inquire about drug use here is to claim a right to know too much. It is to
claim a right to know not only information about what an employee can do,
but also a right to inquire into whatever background information might be
(but not necessarily is) causally related to what an employee will do. But the
range of factors that could be relevant here, from medical history to
psychological dispositions to family plans, is surely too open-ended for an
employee to claim as a right to know.
It might be responded that what an employee is entitled to expect is not a
certain level of output, but a certain level of effort. The claim here would be
that while drug use is only contingently related to what an employee can do,
it is directly related to an employee's motivation to do the job. Drug use then

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 19

is defacto relevant to the personal information that an employee is entitled to

know.
But this involves an assumption mentioned above. The discussion so far
has assumed that drugs will adversely affect job performance. However,
some drugs are performance enhancing whether they are concerned with
actual output or effort. The widespread use of steroids, pain-killers, and
dexadrine among professional athletes are perhaps only the most publicized
instances of performance enhancing drugs. (A teacher's use of caffeine
before an early-morning class is perhaps a more common example.) More to
the point, knowledge of drug use tells little about motivation. There are too
many other variables to be considered. Some users are motivated and some
are not. Thus the motivational argument is faulty.
We can conclude, then, that whether the relevant consideration for pro
spective employees is output or effort, knowledge of drug use will be largely
irrelevant for predicting. Employers ought to be positivistic in their
approach. They should restrict their information gathering to measurable
behavior and valid predictions, (What has the prospect done? What can the
prospect do? What has the prospect promised to do?) and not speculate
about the underlying causes of this behavior. With a probationary work
period always an option, there are sufficient non-intrusive means for limiting
risks available to employers without having to rely on investigations into drug
use.

In summary, we believe that drug use is information that is rightfully

private and that only in exceptional cases can an employer claim a right t
know about such use. Typically, these are cases in which knowledge of dr
use could be used to prevent harm. However, even in those cases we believ
that there are less intrusive and more effective means available than dru
testing for gaining the information that would be necessary to prevent t
harm. Thus, we conclude that drug testing of employees is rarely justified
and mostly inefficacious.

NOTES

* Versions of this paper were read to the Department of Philosophy at

Southern Connecticut State University and to the Society of Business Ethics.
The authors would like to thank those people, as well as Robert Baum and
Norman Bowie, the editors of Business and Professional Ethics Journal for

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
20 Business & Professional Ethics Journal

their many helpful comments. Professor Duska wishes to thank the Pew
Memorial Trust for a grant providing released time to work on this paper.

1. The New Republic, March 31,1986.

2. This trend primarily involves screening employees for such drugs as
marijuana, cocaine, amphetamines, barbituates, and opiates (e.g., heroin,
methadone and morphine). While alcohol is also a drug that can be abused
in the workplace, it seldom is among the drugs mentioned in conjunction with
employee testing. We believe that testing which proves justified for
controlled substances will, a fortiori, be justified for alcohol as well.
3. "A Defense of Employee Rights," Joseph Des Jardins and John
McCall, Journal of Business Ethics 4, (1985). We should emphasize that our
concern is with the moral rights of privacy for employees and not with any
specific or prospective legal rights. Readers interested in pursuing the legal
aspects of employee drug testing should consult: "Workplace Privacy Issues
and Employee Screening Policies" by Richard Lehe and David Middlebrooks
in Employee Relations Law Journal (Vol.11, no.3) pp. 407-21; and "Screening
Workers for Drugs: A Legal and Ethical Framework" by Mark Rothstein, in
Employee Relations Law Journal (vol. 11, no. 3) pp. 422-36.
4. "Privacy, Polygraphs, and Work," George Brenkert, Business and
Professional Ethics Journal vol. 1, #1 (Fall 1981). For a more general
discussion of privacy in the workplace see "Privacy in Employment" by Joseph
DesJardins, in Moral Rights in the Workplace edited by Gertrude Ezorsky,
(SUNY Press, 1987). A good resource for philosophical work on privacy can
be found in "Recent Work on the Concept of Privacy" by W. A. Parent, in
American Philosophical Quarterly (Vol. 20, Oct. 1983) pp. 341-56.
5. U.S. News and World Report Aug. 1983; Newsweek May 1983.
6. Obviously we are speaking here of harms that go beyond the simple
economic harm which results from unsatisfactory job performance. These
economic harms were discussed in the first argument above. Further, we
ignore such "harms" as providing bad role-models for adolescents, harms
often used to justify drug tests for professional athletes. We think it
unreasonable to hold an individual responsible for the image he/she provides
to others.
7. The philosophical literature on informed consent is often concerned
with "informed consent" in a medical context. For an interesting discussion of
informed consent in the workplace, see Mary Gibson, Worker's Rights
(Rowman and Allanheld, 1983) especially pp. 13-14 and 74-75.

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms
 Drug Testing in Employment 21

8. For a reiteration of this point and a concise argument against drug

testing see Lewis L. Maltby, "Why Drug Testing is a Bad Idea.", Inc. June,
1987, pp.152-153. "But the fundamental flaw with drug testing is that it tests
for the wrong thing. A realistic program to detect workers whose condition
puts the company or other people at risk would test for the condition that
actually creates the danger. The reason drunk or stoned airline pilots and
truck drivers are dangerous is their reflexes, coordination, and timing are
deficient. This impairment could come from many situations-drugs, alcohol,
emotional problems-the list is almost endless. A serious program would
recognize that the real problem is workers' impairment, and test for that.
Pilots can be tested in flight simulators. People in other jobs can be tested by
a trained technician in about 20 minutes-at the job site." p. 152.
9. This argument is structurally similar to the argument against the
effectiveness of capital punishment as a deterrent offered by Justice Brennen
in the Supreme Court's decision in Furman Georgia.
10. It might be argued that since we base our critique upon the contractual
relationship between employers and employees, our entire position can be
underminded by a clever employer who places within the contract a privacy
waiver for drug tests. A full answer to this would require an account of the
free and rational subject that the contract model presupposes. While
acknowledging that we need such an account to prevent just any contract
from being morally legitimate, we will have to leave this debate to another
time. Interested readers might find "The Moral Contract between Employers
and Employees" by Norman Bowie, in The Work Ethic in Business, edited by
Hoffman and Wyly (Oelgeschlager and Gunn, 1981) pp. 195-202, helpful
here.

This content downloaded from

140.182.176.13 on Tue, 07 Jan 2025 16:19:45 UTC
All use subject to https://about.jstor.org/terms