Drug Testing and Corporate Responsibility:The "Ought Imphes Can" Argument Jennifer Moore

ABSTRACT. Most of the debate about drag testing in theworkplace has focused on the right to privacy. Proponents oftesting have had to tackle difficult questions concerning thenature, extent, and weight of the privacy rights of employees.This paper examines a different kind of argument — theclaim that because corporations are "responsible" for harmscommitted by employees while under the influence ofdrugs, they are entitled to test for drug use. This argumenthas considerable intuitive appeal, because it seems, at least atfirst glance, to bypass the issue of privacy rights altogether.The argument turns, not on rights, but on the nature and.conditions of responsibihty. We may therefore call it an"ought implies can" argument.

In spite of its initial appeal, however, the argument doesnot succeed in circumventing the claims of privacy rights.Even responsibility for the actions of others does not entitleus to do anything at all to control their behavior, we mustlook to rights, among other things, to determine what sortsof controls are morally permissible. In addition, the argu-ment rests on unjustified assumptions about the connectionbetween drug testing and the prevention of drug-relatedharm.

In the past few years, testing for drug use in theworkplace has become an important and contro-versial trend. Approximately 30% of Fortune 500companies now engage in some sort of drug testingor screening, as do many smaller firms. The Reaganadministration has called for mandatory testing of allfederal employees. Several states have already passeddrug testing laws; others will probably consider themin die future. While the Supreme Court has an-

Jennifer Moore is Assistant Professor of Philosophy at the Universityof Delaware. She does teaching and research in business ethicsand business law and is co-editor of the anthology, BusinessEthics: Readings and Cases in Corporate Morality, pub-lished hy McGraw-Hill.

nounced its intention to rule on the testing of federalemployees within the next few months, its decisionwill not settle the permissibiUty of testing privateemployees. Discussion of the issue is likely to remainlively and heated for some time.

Most of the debate about drug testing in theworkplace has focused on the issue of privacy rights.Three key questions have been: Do employees haveprivacy rights? If so, how far do these extend? Whatkinds of considerations outweigh these rights? Ibelieve there are good reasons for supposing thatemployees do have moral privacy rights,' and thatdrug testing usually (though not always) violatesthese, but privacy is not my main concern in thispaper. I wish to examine a different kind of argu-ment, the claim that because corporations are re-sponsible for harms committed by employees whileunder the influence of drugs, they are entitled to testfor drug use.

This argument is rarely stated formally in theliterature, but it can be found informally quiteoften.̂ One of its chief advantages is that it seems, atleast at first glance, to bypass the issue of privacyrights altogether. There seems to be no need todetermine the extent or weight of employees' privacyrights to make the argument work. It turns on adifferent set of principles altogether, diat is, on themeaning and conditions of responsibility. This is animportant asset, since arguments about rights arenotoriously difficult to settle. Rights claims fre-quendy function in ethical discourse as conversa-tion-stoppers or non-negotiable demands.-* Althouighit is widely recognized that rights are not absolute,there is little consensus on how far they extend, whatkinds of considerations should be allowed to over-ride them, or even how to go about settling thesequestions. But it is precisely these thorny problemsthat proponents of drug testing must tackle if they

Journal of Business Ethics 8: 279-287, 1989.© 1989 Kluwer Academic Publishers. Printed in the Netherlands.

280 Jennifer Moore

wish to address the issue on privacy grounds. Facedwith the claim that drug testing violates the moralright to privacy of employees, proponents of testingmust either (l) argue that drug testing does notreally violate the privacy rights of employees;"* (2)acknowledge that drug testing violates privacy rights,hut argue diat there are considerations that overridethose rights, such as public safety, or (3) argue thatemployees have no moral right to privacy at all.̂ It isnot surprising that an argument that seems to movethe debate out of the arena of privacy rights entirelyappears attractive.

In spite of its initial appeal, however, I willmaintain that the argument does not succeed incircumventing the claims of privacy rights. Evenresponsibility for the actions of others, I will argue,does not entitle us to do absolutely anything tocontrol their behavior. We must look to rights,among other things, to determine what sorts ofcontrols are morally permissible. Once this is ac-knowledged, the argument loses much of its force. Inaddition, it requires unjustified assumptions aboutthe connection between drug testing and the pre-vention of drug-related harm.

An "Ought Implies Can" argument

Before we can assess the argument, it must be setout more folly. It seems to turn on the deep-rootedphilosophical connection between responsibility andcontrol. Generally, we believe that agents are notresponsible* for acts or events that they could nothave prevented. People are responsible for theiractions only if, it is often said, they "could have doneotherwise". Responsibility implies some measure ofcontrol, freedom, or autonomy. It is for this reasonthat we do not hold the insane responsible for theiractions. Showing that a person lacked the capacity todo otherwise blocks the normal moves of praise orblame and absolves the agent of responsibility for agiven act.

For similar reasons, we believe that persons cannotbe obligated to do things that they are incapable ofdoing, and that if they fail to do such things, noblame attaches to them. Obligation is empty, evensenseless, without capability. If a person is obligatedto perform an action, it must be within his or herpower. This principle is sometimes summed up by

the phrase "ought implies can". Kant used it as partof a metaphysical argument for fi-ee Avill, claimingthat if persons are to have obligations at all, theymust be autonomous, or capable of acting fireely.'The argument we examine here is narrower in scope,but similar in principle. If corporations are respon-sible for harms caused by employees under theinfluence of drugs, they must have the ability toprevent these harms. They must, therefore, have thefreedom to test for drug use.

But the argument is still quite vague. Whatexactly does it mean to say that corporations are"responsible" for harms caused by employees? Thereare several possible meanings of "responsible". Notall of these are attributable to corporations, and notall of them exemplify the principle that "oughtimplies can". The question of how or whethercorporarions are "responsible" is highly complex, andwe cannot begin to answer it in this paper.' Thereare, however, four distinct senses of "responsible"that appear with some regularity in the argument.They can be characterized, roughly, as follows: (a)legally hable; (b) culpable or guilty; {c) answerable oraccountable; (d) bound by an obligation. The first ispurely legal; the last three have a moral dimensioiL

Legal liability

We do hold corporadons legally liable for the negli-gent acts of employees under the doctrine of re-spondeat superior ("let the master respond"). If anemployee harms a third party in the course ofperforming his or her duties for the firm, it is thecorporation which must compensate the third party.Respondeat superior is an example of what is frequentlycalled "vicarious liability". Since the employee wasacting on behalf of the firm, and the firm was actingthrough the employee when the harmfol act wascommitted, liability is said to "transfer" from theemployee to the firm. But it is not clear that suchliability on the part of the employer implies acapacity to have prevented the harm. Corporationsare held liable for accidents caused by an employee'snegligent driving, for example, even if they couldnot have foreseen or prevented the injury. Whilesome employee accidents can be traced to corporatenegligence,' there need be no fault on the part of thecorporarion for the doctrine of respondeat superior to

Drug Testing and Corporate Responsibility 281

apply. The doctrine of respondeat superior is groundednot in fault, but in concerns of public policy andutility. It is one of several applications of the notionof liability without fault in legal use today.

Because it does not imply fault, and its attendantability to have done otherwise, legal liability orresponsibility a cannot be used successfully as part ofan "ought implies can" argument. Holding corpor-ations legally liable for harms committed byintoxicated employees while at the same time for-bidding drug-testing is not inconsistent. It couldsimply be viewed as yet another instance of liabilitywithout fault. Of course, one could argue that thenotion of liability without fault is itself morallyunacceptable, and that liability ought not to bedetached from moral notions of punishment andblame. This is surely an extremely important claim,but it is beyond the scope of this paper. The mainpoint to be made here is that we must be able toattribute more than legal liability to corporations ifwe are to invoke the principle of "ought impliescan". Corporations must be responsible in sense b , c,or d — that is, morally responsible — if the argumentis to work.

Moral responsibility

Are corporations morally responsible for harmscommitted by intoxicated employees? Perhaps themost frequendy used notion of moral responsibilityis sense b, what I have called "guilt" or "culpability".'"I have in mind here the strongest notion of moralresponsibility, the sense that is prevalent in criminallaw. An agent is responsible for an act in this sense ifthe act can be imputed to him or her. An essentialcondition of imputability is the presence in the agentof an intention to commit the act, or mens rea.^^ Butdoes an employer whose workers use drugs satisfythe mens rea requirement? The requirement probablywould be satisfied if it could be shown that the firmintended the resulting harms, ordered its employeesto work under the influence of drugs, or even,perhaps (though this is less clear) turned a blind eyeto blatant drug abuse in the workplace.'^ But theseare all quite farfetched possibilities. It is reasonableto assume that most corporations do not intend theharms caused by their employees, and that they donot order employees to use drugs on the job. Drug

use is quite likely to be prohibited by companypohcy. If corporations are morally responsible fordrug-related harms committed employees, then, it isnot in sense b .

Corporations might, however, be morally respon-sible for harms committed by employees in anothersense. An organization acts through its employees. Itempowers its employees to act in ways in which theyotherwise would not act by providing them withmoney, power, equipment, and authority. Through aseries of̂ agreements, the corporation delegates itsemployees to act on its behalf. For these reasons, onecould argue that corporations are responsible, in thesense of "answerable" or "accountable" (responsibilityc), for the harmful acts of their employees. Indeed,it could be argued that if corporations are notmorally responsible for these acts, they are notmorally responsible for any acts at all, since corpora-tions can only act through their employees.'' To saythat corporations are responsible for the harms oftheir employees in sense c is to say more than justthat a corporation must "pay up" if an employeecauses harm. It is to assign fault to the corporationby virtue of the ways in which organizational poli-cies and structures facihtate and direct employees'actions.'*

Moreover, corporations presumably have thesame obligations as other agents to avoid harm in theconduct of their business. Since they conduct theirbusiness through their employees, it could plausiblybe argued that corporations have an obligation toanticipate and prevent harms that employees mightcause in the course of their employment If diisreasoning is correct, corporations are morally re-sponsible for the drug-related harms of employees insense d — that is, they are under an obligation toprevent those harms. The "ought implies can" argu-ment, then, may be formulated as follows:

1. If corporations have obligations, they must becapable of carrying them out, on the principleof "ought implies can".

2. Corporations have an obligation to preventharm from occurring in the course of con-ducting their business.

3. Drug use by employees is likely to lead toharm.

4. Corporations must be able to take steps toeliminate (or at least reduce) drug use byemployees.

282 Jennifer Moore

5. Drug testing is an effective way to eliminate/reduce employee drug use.

6. Therefore corporations must be permitted totest for drugs.''

The limits of corporate autonomy

This is surely an important argument, one thatdeserves to be taken seriously. The premise thatcorporations have an obligation to prevent harmfrom occurring in the conduct of their businessseems unexceptionable and consistent with the actualmoral behefs of society. There is not much questionthat drug use by employees, especially regular druguse or drug use on the job, leads to harms of variouskinds. Some of these are less serious than others, butsome are very serious indeed: physical injury toconsumers, the public, and fellow employees — andsometimes even death.'*

Moreover, our convictions about the connectionsbetween responsibility or obligation and capabilityseem unassailable. Like other agents, if corporationsare to have obUgations, they must have the ability tocarry them out. The argument seems to tell us thatcorporations are only able to carry out their obliga-tion to prevent harm if they can free themselves ofdrugs. To prevent corporations from drug testing, itimplies, is to prevent them from discharging theirobligations. It is to cripple corporate autonomy justas we would cripple the autonomy of an individualworker if we refused to allow him to "kick the habit"that prevented him from giving satisfactory job per-formance.

But this analogy between corporate and individualautonomy reveals the initial defect in the argumentUnlike human beings, corporations are never fullyautonomous selves. On the contrary, their actionsare always dependent upon individual selves who areautonomous. Human autonomy means self-deter-mination, self-governance, self-control. Corporateautonomy, at least as it is understood here, meanscontrol over others. Corporate autonomy is essen-tially derivative. But this means that corporate actsare not the simple sorts of acts generated by individ-ual persons. They are complex. Most importantly,the members of a corporation are frequently not theagents, but the objects, of "corporate" action. A gooddeal of corporate action, that is, necessitates doing

something not only through corporate employees,but to those employees." The act of eliminatingdrugs from the workplace is an act of this sort Acorporation's ridding itself of drugs is not like anindividual person's "kicking the habit". Rather, it isone group of persons making another group ofpersons give up drug use.

This fact has important implications for the"ought implies can" argument The argument ispersuasive in situations in which carrying out one'sobligations requires only self-control, and does notinvolve controlling the behavior of others. Presum-ably there are no restrictions on what one may do tooneself in order to carry out an obligation.'^ But acorporation is not a genuine "self", and there aremoral limits on what one person may do to another.Because this is so, we cannot automatically assumethat the obligation to prevent harm justifies em-ployee drug testing. Of course this does not neces-sarily mean that drug testing is unjustified. But it doesmean that before we can determine whether it isjustified, we must ask what is permissible for oneperson or group of persons to do to another toprevent a harm for which they are responsible.

Are there any analogies available diat might helpto resolve this question? It is becoming increasinglycommon to hold a hostess responsible (both legallyand morally) for harm caused by a drunken guest onthe way home from her party. In part, this is becauseshe contributes to the harm by serving her guestalcohol. It is also because she knows that drunkdriving is risky, and has a general obligation toprevent harm. What must she be allowed to do toprevent harms of this kind? Persuade the guest tospend the night on the couch? Surely. Take her carkeys away from her? Perhaps. Knock her out andlock her in the bathroom until morning? Surely not

Universities are occasionally held legally andmorally responsible for harms committed by mem-bers of fraternities — destruction of property, gangrapes, and injuries or death caused by hazing. Whatmay they do to prevent such harms? They maycertainly withdraw institutional recognition andsupport from the fraternity, refusing to let it operateon the campus. But may they expel students wholive together off-campus in fraternity-like arrange-ments? Have university security guards police thesehouses, coverdy or by force? Tliese questions aremore difficult to answer.

Drug Testing and Corporate Responsibility 283

We sometimes hold landlords morally (thoughnot legally) responsible for tenants who are slovenly,play loud music, or otherwise make nuisances ofthemselves. Landlords are surely permitted to cancelthe leases of such tenants, and they are justified inasking for references from previous landlords toprevent future problems of this kind. But it is notclear that a landlord may delve into a tenant's privatelife, search his room, or tap his telephone in order toanticipate trouble before it begins.

Each of these situations is one in which oneperson or group of persons is responsible, to agreater or a lesser degree, for the prevention of harmby others, and needs some measure of control inorder to carry out this responsibility." In each case,there is a fairly wide range of actions which wewould be willing to allow the first party, but thereare some actions which we would rule out Havingan obligation to prevent the harms of others seemsto permit us some forms of control, but not all. Atleast one important consideration in deciding whatkinds of actions are permissible is the rights of thecontrolled parties.^" If these claims are correct, wemust examine the rights of employees in order todetermine whether drug testing is justified. Therelevant right in the case of drug testing is the rightto privacy. The "ought implies can" argument, then,does not circumvent the claims of privacy rights as itoriginally seemed to do.

The agency argument

A proponent of drug testing might argue, however,that the relation between employers and employeesis significantly different from the relation betweenhosts and guests, universities and members of fira-temities, or landlords and tenants. Employees have aspecial relation with the firm that employs them.They are agents, hired and empowered to act onbehalf of die employer. While they act on thebusiness of the firm, it might be argued, they "are"the corporation. The restrictions that apply to whatone independent agent may do to another thus donot apply here.

But surely this argument is incorrect, for a num-ber of reasons. First, if it were correct, it wouldjustify anything a corporation might do to controlthe behavior of an employee — not merely drug

testing, but polygraph testing, tapping of telephones,deception, psychological manipulation, whips andchains, etc.^' There are undoubtedly some peoplewho would argue that some of these procedures arepermissible, but few would argue that all of themare. The fact that even some of them appear not tobe suggests that we believe there are limits to whatcorporations may do to control employees, and thatone consideration in determining these limits is theemployees' rights.

Secondly, the argument implies that employeesgive up their own autonomy completely when theysign on as agents, and become an organ or piece ofthe corporation. But this cannot be true. Agency is amoral and contractual relationship of the kind thatcan only obtain between two independent, autono-mous parties. This relationship could not be sus-tained if the employee ceased to be autonomousupon signing the contract Employees are not slaves,but autonomous agents capable of upholding acontract Moreover, we expect a certain amount ofdiscretion in employees in the course of their agency.Employees are not expected to foUow illegal orimmoral commands of their employen, and we findthem morally and legally blameworthy when theydo so. That we expect such independent judgment ofthem suggests that they do not lose their autonomyentirely.̂ ^

Finally, if the employment contract were one inwhich employees gave up all right to be treated asautonomous human beings, then it would not be alegitimate or morally valid contract Some rights areconsidered "inalienable" — people are forbiddenfrom negotiating them away even if it seems advan-tageous to them to do so. The law grants recognitionto this fact through anti-discrimination statutes,minimum wage legislation, workplace health andsafety standards, etc. Even if I would like to, I maynot trade away, for example, my right not to besexually harassed or my right to know about work-place hazards.

Again, these arguments do not show that drugtesting is unjustified. They do show, however, that ifdrug testing is justified, it is not because the "oughtimplies can" argument bypasses the issue of em-ployee rights, but because drug testing does notimpermissibly violate those rights.^' To think thatobligation, or responsibility for the acts of others,can circumvent rights claims is to misunderstand the

284 Jennifer Moore

import of the "ought imphes can" principle. Theprinciple tells us diat there is a close connectionbetween obligadon or responsibility and capability.But it does not hcense us to disregard the rights ofothers any more than it guarantees us the physicalcondidons that make carrying out our obligadonspossible. It may weU prove that employees' right toprivacy, assuming they have such a right, is secondaryto some more weighty consideradon. I take up thisquesdon briefly below. What has been shown here isthat the issue of the permissibility of drug tesdngwill not and cannot be setded without a close scru-dny of privacy rights. If we are to decide the issue,we must eventually determine whedier employeeshave privacy rights, how far they extend, and whatconsideradons outweigh them — precisely the diffi-cult quesdons the "ought implies can" argumentsought to avoid.

Is drug testing necessary?

The "ought implies can" argument also has anotherserious flaw. The argument turns on the claim thatforbidding drug tesdng prevents corporadons fromcarrying out their obligadon to prevent harm. Butthis is only true if drug tesdng is necessary for pre-vendng drug-related harm. If it is merely one opdonamong many, then forbidding drug tesdng stillleaves a corporadon free to prevent harm in otherways. For the argument to be sound, in other words,premise 5 would have to be altered to read, "drugtesdng is a necessary element in any plan to rid theworkplace of drugs."

But it is not at all clear that drug tesdng isnecessary to reduce drug use in the workplace. Itsnecessity has been challenged repeatedly. In a recentardcle in the Harvard Business Review, for example,James Wrich draws on his experience in dealingwith alcoholism in the workplace and suggests theuse of broadbrush educadonal and rehabilitadveprograms as alternadves to tesdng. Corporadonsusing such programs to combat alcohol problems,Wrich reports, have achieved tremendous reducdonsin absenteeism, sick leave, and on-the-job accidents.̂ *Others have argued that impaired performancelikely to result in harm could be easily detected byvarious sorts of performance-oriented tests — mentaland physical dexterity tests, alertness tests, flight

simuladon tests, and so on. These sorts of procedureshave the advantage of not being controversial from arights perspecdve.^'

Indeed, many thinkers have argued that drugtesdng is not only unnecessary, but is not even anefFecdve way to attack drug use in the workplace.The commonly used and affordable urinalysis testsare notoriously unreliable. They have a very highrate both of false negadves and of false posidves. Atbest the tests reveal, not impaired performance oreven the presence of a pardcular drug, but thepresence of metabolites of various drugs that canremain in the system long after any effects of thedrug have worn off.̂ * Because they do not measureimpairment, such tests do not seem well-tailored tothe purpose of prevendng harm — which, after all, isthe uldmate goal. As Lewis Maltby, vice president ofa small instrumentadon company and an opponentof drug tesdng, puts it,

. . . [T]he fundamental flaw with drug testing is that ittests for the wrong thing. A realistic program to detectworkers whose condition put the company or otherpeople at risk would test for the condition that actuallycreates the danger. '̂

If these claims are true, there is no real connecdonbetween the obligadon to prevent harm and thepracdce of drug tesdng, and the "ought implies can"argument provides no jusdficadon for drug tesdng

ll==«

Conclusion

I have made no attempt here to determine whetherdrug tesdng does indeed violate employees' privacyrights. The analysis on p. 283 above suggests that wehave reason to believe that employees have somerights. Once we accept the nodon of employee rightsin general, it seems likely that a right to privacywould be among them, since it is an important civilright and central for the protecdon of individualautonomy. There are also reasons, I believe, to thinkthat most drug tesdng violates the right to privacy.These claims need much more defense than diey canbe given here, and even if they are true, this does notnecessarily mean that drug tesdng is unjusdfied. Itdoes, however, create a prima facie case against drugtesdng. If drug tesdng violates the privacy rights of

Drug Testing and Corporate Responsibility 285

employees, it will be jusdfied only under very strictcondidons, if it is jusdfed at all. It is worth taking amoment to see why this is so.

It is generally accepted in both the ethical andlegal spheres that rights are not absolute. But weallow basic rights to be overridden only in specialcases in which some urgent and fundamental good isat stake. In legal discourse, such goods are called"compelling interests".^' While there is room forsome debate about what counts as a "compellinginterest", it is almost always understood to be morethan a merely private interest, however weighty.Public safety might well fall into this category, butprivate monetary loss probably would not Whilemore needs to be done to determine what kinds ofinterests jusdfy drug tesdng, it seems clear that iftesdng does violate the basic rights of employees, it isonly jusdfed in extreme cases — far less often that itis presendy used. Moreover, we believe that over-riding a right is to be avoided wherever possible, andis only jusdfied when doing so is necessary to servethe "compelling interest" in quesdon. If it violatesrights, then drug tesdng is only permissible if it isnecessary for the protecdon of an interest such aspublic safety and if there is no other, morally pre-ferable, way of accomplishing the same goal. As wehave seen above, however, it is by no means clearthat drug tesdng meets these condidons. There maybe better, less controversial ways to prevent the harmcaused by drug use; if so, these must be used inpreference to drug tesdng, and tesdng is unjusdfied.And if the attacks on the effecdveness of drug tesdngare correct, tesdng is not only unnecessary for theprotecdon of pubhc safety, but does not serve any"compelling interest" at all.

What do these conclusions tell us about theresponsibility of employers for prevendng harmscaused by employees? If it is decided that drugtesdng is morally impermissible, then there can beno duty to use it to andcipate and prevent harms.Corporadons who fail to use it cannot be blamed fordoing so. They cannot have a moral obligadon to dosomething morally impermissible. Moreover, if itturns out that there is no other effecdve way toprevent the harms caused by drug use, then it seemsto me we may not hold employers morally respon-sible for those harms. This seems to me unlikely tobe the case — there probably are other effecdvemeasures to control drug abuse in the workplace.

But corporadons can be held responsible only to theextent that they are permitted to act It would not beinconsistent, however, to hold corporadons legallyliable for the harms caused by intoxicated employeesunder the doctrine of respondeat suphior, even if drugtesdi^ is forbidden, for this kind of liability does notimply an ability to have done otherwise.

Notes

' Employees do not, of course, have legal privacy' rights,although the courts seem to be moving slowly in thisdirection. Opponents of testing usually claim that employeeshave moral rights to privacy, even if these have not beengiven legal recognition. See, for example, Joseph Des Jardinsand Ronald Duska, "Drug Testing in Employment", inBusiness Ethics: Readings and Cases in Corporate Morality, 2ndEd, ed. W. M. Hoffman and J. M. Moore (McGraw-Hill,forthcoming).

^ See, for example, "Work-Place Privacy Issues and Em-ployer Screening Policies," Richard Lehr and David Middle-brooks, Employee Relations Law journal 11, 407. Lehr andMiddlebrooks cite the argument as one of the chief justi-fications for drug testing used by employers. I have alsoencountered the argument frequently in discussion withstudents, colleagues, and managers.•* Ronald Dworkin has referred to rights as moral "trumps".This kind of language tends to suggest that rights overwhelmall other considerations, so that when they are flourished, allthat opponents can dp is subside in silence. Rights arefrequently asserted this way in everyday discourse, and inthis sense rights claims tend to close, rather than open, thedoor to fruitful ethical dialogue.

* In his article "Privacy, Polygraphs, and Work," Business andProfessional Ethics journal 1, Fall, 1981, 19, George Brenkerthas developed the idea that my privacy is violated whensome one acquires information about me that they are notentitled, by virtue of their relationship to me, to have. Mymortgage company, for example, is entitled to know mycredit history, a prospective sexual parmer is entitled toknow if I have any sexually transmitted diseases. Thus theirknowledge of this information does not violate my privacy.One could argue that employers af'e similarly entitled to theinformation obtained by drug tests, and that drug testingdoes not violate privacy for this reason. A somewhat dif-ferent move would be to argue that testing does not violateprivacy because employees give their "consent" to it drugtesting as part of the employment contract. For a sustainedattack on these and other Type 1 arguments, see Joseph DesJardins and Ronald Duska, "Drug Testing in Employment".^ One might defend this position on the ground that the

286 Jennifer Moore

employer "owns" the job and is therefore entitled to placeany conditions he wishes on obtaining or keeping i t Theproblem with this argument is that it seems to rule out allemployee rights, i n d u d i i ^ such basic ones as the right toorganize and bargain collectively, or the right not to bediscriminated against, which have solid legal as well asethical grounding. It also implies that ownership overridesall other considerations, and it is not at all clear that this istrue. One might take the position that by accepting a job, anemployee has agreed to give up all his rights save thoseactually specified in the employment contract. But thismakes the employment contract look like an agreement inwhich employees sell themselves and accept the status ofthings without rights. And it overlooks the fact that webelieve there are some things ("inalienable" rights) thatpersons ought not to be permitted to bargain away. AlexMichalos has discussed some of the limitations of theemployment contract in "The Loyal Agent's Argument", inEthical Theory and Business, 2nd edition, ed. Tom L. Beau-champ and Norman E. Bowie (Englewood Cliffs, NJ: Pren-tice-Hall, 1983), p. 247.

'' The term "responsibility" is deliberately left ambiguoushere. Several different meanings of it are examined below.' See Immanuel Kant, Critique of Practical Reason, trans.Lewis White Beck (Indianapolis: Bobbs-Merril, 1956), p. 30.* In this paper I have tried to avoid getting embroiled in thequestion of whether or not corporations are themselves"moral agents", which has been the question to dominate thecorporate responsibility debate. The argument I offer heredoes, I believe, have important implications for the problemof corporate agency, but does not require me to take a standon it here. I am content to have those who reject the notionof corporations as moral agents read my references tocorporate responsibility as shorthand for some complexform of individual or group responsibility.' One example would be negligent hiring, which is anincreasingly frequent cause of action against an employer.Employers can also be held negligent if they give orders thatlead to harms that they ought to have foreseen. Domino'sPizza is now under suit because it encouraged its drivers todeliver pizzas as fast as possible, a policy that accidentvictims claim should have been expected to cause accidents.'" This understanding of moral responsibility often seemsto overshadow other notions. In an article on corporateresponsibility, for example, Manuel Velasquez concludesthat because corporations are not responsible in this sense,they are "not responsible for anything they do". "WhyCorporations Are Not Responsible For Anything They Do",Business and Professional Ethiajoumal 2, Spring, 1983, 1.'' There is also an actus reus requirement for this type ofresponsibility — that is, the act must be traceable to thevoluntary bodily movements of the agent. Obviously, cor-porations do not have bodies, but the people who work forthem do. The question, then, has become when may we call

an act by one member of the corporation a "corporate act". Ifit is possible to do so at all, the decisive feature is probablythe presence of some sort of corporate "intention." This iswhy I focus on intention here, and why intention has beencentral to the discussion of corporate responsibility.'- There are some, like Velasquez, who hold that a cor-poration can never satisfy the mens rea requirement becausethis would require a collective mind. If this were true, theargument would collapse at the outset Others believe that amens rea can be attributed to corporations metaphorically, ifit can be shown that company policy includes an "intention"to harm, and it is this model I follow here." There are, of course, those who take precisely thisposition. See Velasquez, "Why Corporations Are Not Re-sponsible For Anything They Do".

'* See, for example, Peter French, Collective and CorporateResponsibility (New York Columbia University Press, 1984)." It is tempting to conclude from this argument that drugtesting is not only permissible, but obligatory, but this is notthe case. The reason why it is not provides a clue to one ofthe major weaknesses of the argument Drug testing wouldbe obligatory only if it were necessary for the prevention ofharm due to drug use, but it is not clear that this is so. Butalso means that it is not clear that corporations are deflectedfrom their duty to prevent harm by a prohibition againstdrug testing. See below for a fuller discussion of thisproblem.

"" For example, it has been claimed that employees who usedrugs cause four times as many work-related accidents as doother employees. The highly publicized Conrail crash in1987 was determined to be drug-related. Of course there areharms to the company itself as well, in the form of higherabsenteeism, lowered productivity, higher insurance costs,etc. But since these types of harm raise the question of whata company may do to preserve its self-interest, rather thanwhat it may do to prevent harms to others for which theyare responsible, I focus here on harm to employees, con-sumers, and the pubhc.

" In our eagerness to assign "corporate responsibihty", thisfact has frequently been overlooked. This in tum has led, Ibelieve, to an oversimplified view of corporate action. Idiscuss this problem more fully in a paper in progressentitled "The Paradox of Corporate Autonomy"." It is an interesting question whether there are hmitationson what individuals can do to themselves to control theirown behavior. What about individuals who undergo hyp-nosis, or who have their jaws wired shut in order to loseweight? Are they violating their own rights? Underminingtheir own autonomy? It could be argued plausibly that thesekinds of things are not permissible, on the Kantian groundthat we have a duty not to treat ourselves as merely as meansto an end. Of course, if there are such restrictions, it makesthe "ought imphes can" argument as applied to corporationseven weaker.

Drug Testing and Corporate Responsibility 287

'̂ None of these analogies is perfect. In the case of thehostess and guest, for example, the guest is clearly intoxi-cated. This is rarely true of employees who are tested fordrugs; if the employee were visibly intoxicated, there wouldbe no need to test. Moreover, in the hostess/guest case thehostess contributes directly to the intoxication. There areimportant parallels, however. In each case one party is heldmorally (and in two of the cases, legally) responsible forharms caused by others. Moreover, the first parties areresponsible in close to the same way that employen areresponsible for the acts of their employees: they in somesense "facihtate" the harmful acts, they have some capacity toprevent those acts, and they are thus viewed as having anobhgation to prevent them. One main difference, of course,is that employees are "agents" of their employers. See p. 283.-" There are other, utihty-related considerations, as well —for example, harm to employees who are unjustly dismissed,a demoralized workforce, the costs of testing, etc. I concen-trate here on rights because they have been the primary focalpoint in the drug testing debate.

-' The assumption here is that persons are entitled to dowhatever they wish to themselves. See Note 18." See Michalos, "The Loyal Agent's Argument"." Some violations of right, of course, are permissible. Seep. 285.•^* James T. Wrich, "Beyond Testing: Coping with Drugs atWork",Hfln'<jr(/B»«i«essRew«*'Jan-Feb 1988,120.-' See Des Jardins and Duska, "Drug Testing in Employ-ment", and Lev«s Maltby, "Why Drug Testing is a Bad Idea",Inc., June 1987. While other sorts of tests also have thepotential to be abused, they are at least a direct measurementof something that an employer is entitled to know — per-formance capability. Des Jardins and Duska offer anextended defense of this sort of test-<> See Edward J. Imwinkelreid, "False Positive", The Sciences,Sept.-Oct. 1987, 22. Also David Bearman, "The MedicalCase Against Drug Testing", Harvard Business Review Jan-Feb. 1988, 123.

-' Maltby, "Why Drug Testing is a Bad Idea', pp. 152-153.- ' It could still be argued that drug testing deters drug use,and thus has a connection with preventing harm, eventhough it doesn't directly provide any information thatenables companies to prevent harm. This is an importantpoint, but it is still subject to the restrictions discussed in theprevious section. Not everything that has a deterrent value ispermissible. It is possible that a penalty of capital punishmentwould provide a deterrent for rapists, or having one's handremoved deter shoplifting, but there are very few advocatesfor these penalties. Effectiveness is not the only issue here;rights and justice are also relevant.

-'' The principle that fundamental rights may not be over-ridden by the state unless doing so is necessary to serve a"compelling state interest" is a principle of constitutionallaw, but it also reflects our moral intuitions about when it isappropriate to override rights. The legal principle would notapply to all cases of drug testing in the workplace becausemany of these involve private, rather than state, employees.But the principle does provide us with useful guidelines inthe ethical sphere. Interestingly, Federal District JudgeGeorge Revercomb recently issued an injunction blockingthe random drug testing of Justice Department employeeson the ground that it did not serve a compelling stateinterest. Since there was no evidence of a drug problemamong the Department's employees, the Judge concluded,there is no threat that would give rise to a compellinginterest. See "Judge Blocks Drug Testing of Justice Depart-ment Employees", New York Times July 30, 1988,7.

Department ofPhilosophy,University of Delaware,

Newark,Delaware 19716,

U.SA.

288 Review

(continuedfom p. 270)phy, management studies and finance-orientedhusiness. It reinforces an approach as old as Aris-totle and as recent as Hahermas and Gadamer. Itdemonstrates that a broad socially interpretiveand critical philosophy has an empirical andpractical potential for hoth husiness and ethics.

Secondly, Toffler's work indicates that thebureaucratic organization, by diffusing responsi-bility and occluding accountability, transformsethical decisions into tough choices. It producesorganization men, not autonomous managers. Inthem, ethics means whistle-blowing, disloyalty,and exit The primary difficulty is to gain legi-dmacy (voice) for ethics.̂ Obversely ethics impliesa practical and structural critique of the collec-tivism and rigidity of the bureaucratic organiza-don. Top-level CEO commitments and formalcodes and committees, while steps forward, mayonly reinforce its hierarchical culture. They arenot enough.

Toffler does not, then, just report the needfor change from the standpoint of individualmanagers; she also shows how ethics arises withinthe firm and how it consdtutes a dynamic pres-sure for organizadonai transformadon. Standingstill in patiological bureaucradc rigidity withone's eyes firmly on the past will only mean thatchange is imposed externally and, usually, viasome 'sudden' — but foreseeable and preventable

— crisis. An ethically responsive business. ToughChoices suggests, would be more open, morehuman, more flexible and dynamic. All are char-acterisdcs of organizadons capable of foresight,organizadons with a future.

Tough Choices is a work of significance, then;for it shows that the ethical change process inorganizadons is one of the major implicadons ofbusiness ethics.

References

' See H.-G Gadamer, Truth and Method (Sheed andWard: London, 1978), 285fF, and J. Habermas, Knowledgeand Human Interests (Beacon, Boston: 1971).^ In Aristotle's Nicomachean Ethics, Hegel's Philosophy ofRight, and A. Maclntyre's After Virtue (University ofNotre Dame, 1981).' Respectively, Moral Mazes: The World of CorporateManagers (Oxford, New Y o r k 1988), and Images of theOrganization (Sage, San Francisco: 1986).* See Gadamer's comments on Aristotle in Truth andMethod (285f).' See Albert O. Hirschmann, Exit, Voice and Loyalty(Harvard UP, Cambridge, Mass: 1970).

V I N C E N T D I N O R C I A

University ofSudhury,Sudbury, Ontario,Canada P3E2C6.