|PUBLIC EMPLOYEE DRUG TESTING: A LEGAL GUIDE
TABLE OF CONTENTS
The current law of public employee drug testing began with the Supreme Court’s decisions in Skinner v. Railway Labor Executives’ Assn., 489 U.S. 602 (1989), and National Treasury Employees v. Von Raab, 489 U.S. 656 (1989). In these companion cases, the Court held that the government is allowed to conduct drug tests without individualized suspicion when there is a “special need” that outweighs the individual’s privacy interest. In Skinner, the court found that public safety was such a special need. In Von Raab, the court found a special need in relation to customs agents who carry firearms or are directly involved in drug interdiction.
The federal courts spent the next decade defining which government interests qualified as “special needs” and defining the scope of those that qualified.
It soon became clear that “special need” meant little more than that the nature of the employee’s job was extremely important, and that a great deal of harm could be done if the job was not performed properly. The courts did not require public employers to demonstrate that employees who used drugs were likely to create this harm, nor that there was any special difficulty with preventing the harm through normal supervisory methods. Courts generally resisted, however, attempts to push the Skinner/Van Raab envelope to encompass large sections of the workforce. The result was an unprincipled, but relatively small and well defined exception to normal Fourth Amendment principles.
This picture was complicated by the Supreme Court’s 1995 decision in Vernonia School Dist. v. Acton, 515 U.S. 646 (1995). The Court in Vernonia upheld compulsory random testing of high school student athletes. The fact that the risks of drug use by student athletes do not even approach the levels previously held necessary to establish a special need, and the language of the opinion, suggested that the Court was retreating from Skinner/Van Raab.
Two years later, however, the Court held in Chandler v. Miller, 520 U.S. 305, 117 S. Ct. 1295, 137 L. Ed. 2d 513 (1997), that candidates for state office could not be required to take a drug test. Moreover, it did so in language which indicated continued support for the Skinner/Van Raab standards. While the picture is not yet complete, the best indication is that Vernonia is another example of the reduced rights of students, with little implication for the rights of employees.
Private sector employees, of course, are not protected by these standards because the Constitution does not apply to private corporations. The only protection in the private sector comes from the few state drug testing statutes (discussed in our legislative brief on drug testing which is on the ACLU website) and from the Americans with Disabilities Act (discussed later in this report).
II. THE SKINNER/VAN RAAB STANDARDS
The primary focus of drug testing caselaw has been the development of the Skinner/Van Raab standards, especially their application to specific jobs and specific industries. This has generally consisted of determining whether the government had a “special need” in the context in question. The following is a summary of major decisions organized by industry.
1. Motor Vehicle Operators
The most commonly claimed special need has been public safety. The courts have almost always accepted this argument, especially in the transportation field. Motor vehicle operators, for example, were consistently held to be subject to random testing. This was true not only for interstate truck drivers (International Brotherhood of Teamsters v. Department of Transportation, 932 F.2d 1292 (9th Cir. 1991), subway train drivers (Burka v. New York City Transit Auth., 739 F. Supp. 814 (S.D.N.Y. 1990), and holders of commercial drivers licenses (Keaveney v. Town of Brookline, 937 F. Supp. 975 (D. Mass. 1996), but also for employees who only chauffeur single individuals in passenger cars (National Treasury Employees Union v. Yeutter, 918 F.2d 968 (D.C. Cir. 1990).
Where no passengers are involved, however, the courts have sometimes viewed the issue differently. In AFGE v. Sullivan, 787 F. Supp. 255 (D.D.C. 1992), the court considered the case of mail clerks and messengers who drove motor vehicles, but carried only documents. The court held that such employees were no different from regular highway drivers and that the government’s safety interest was outweighed by the employees’ privacy interest. Other courts, however, have allowed random testing even under these circumstances. In Plane v. United States, 796 F. Supp. 1070 (W.D. Mich. 1992), the court allowed random testing of employees who drove cranes and tractors on public roads. Similarly, random testing was allowed for employees who drove riding lawn mowers on highway medians (Middlebrooks v. Wayne County, 446 Mich. 151 (Mich. 1994)). The best explanation of these divergent results is that the latter two cases involved, not passenger cars, but vehicles which posed special risks (at least in the eyes of the court).
One court, however, took a very different position regarding drivers. In National Treasury Employees Union v. Watkins, 722 F. Supp. 766 (D.D.C. 1989), the court was faced with a group of employees whose primary responsibility was to operate motor vehicles which carried passengers. The court recognized the risks of operating any motor vehicle, but pointed out that the risk involved in these employees’ work was no greater than the risk that everyone runs when they drive a car. Sensing that allowing testing of these employees might open the door for testing of all licensed drivers, the court prohibited random testing.
The bottom line appears to be that employees driving vehicles other than passenger cars and/or carrying significant numbers of passengers are consistently held to raise a special need. Other employees who operate motor vehicles may be able to avoid random testing.
2. Other Transportation Employees
Where other transportation employees are concerned, the courts have generally not been willing to blindly accept the mantra of public safety. Instead, they have looked for a nexus between the responsibilities of specific jobs and the safety of the public. Where a position has a direct affect on safety, random testing has been allowed. For example, random testing was consistently allowed for mechanics. (See English v. Talladega County Bd. of Educ., 938 F. Supp. 775 (N.D. Ala. 1996) (school bus mechanic), AFGE v. Skinner, 885 F. 2d 884 (D.C.Cir. 1989), cert. denied, 495 U.S. 923 (1990) (aircraft mechanic)). Random testing was also allowed for air traffic controllers. (Bluestein v. Skinner, 908 F.2d 451 (9th Cir. 1990), cert. denied, 498 U.S. 1083). Where the connection was more remote, the answer was generally different. For example, the District Court for the Eastern District of New York found that railroad mail clerks were not safety sensitive (Laverpool v. New York City Transit Auth., 835 F. Supp. 1440 (E.D.N.Y. 1993), aff’d without opinion, 41 F.3d 1501 (2nd Cir. 1994)). A regional transportation authority’s attempt to classify the custodian at a bus depot as safety sensitive also failed. (Bolden v. Southeastern Pa. Transp. Auth., 953 F.2d 807 (3rd Cir. 1991), cert. denied, 504 U.S. 943 (1992)).
A few creative employers have attempted to justify random testing employees whose jobs were clearly not safety sensitive by arguing that they would have to help out in an emergency. These attempts were generally not successful. In Burka, the court rejected the argument that a turnstile operator was safety sensitive because of the role he might have to play in an emergency. The Coast Guard claimed that all members of a ship’s crew were safety sensitive for the same reason, but this argument was rejected by the District Court for the District of Columbia, striking down random testing for cooks and painters. (Transportation Inst. v. United States Coast Guard, 727 F.Supp. 648 (D.D.C. 1989)).
Perhaps because aviation is perceived as especially risky, the one court which dealt with aviation employees drew the boundaries of what is considered safety sensitive extremely broadly. In Bluestein (section 2 above), the Ninth Circuit allowed random testing not only for flight crews and air traffic controllers, but also for aircraft dispatchers and flight attendants.
B. Medical Professionals
The courts followed a very similar course where medical professionals were concerned. Where employees were involved in patient care, random testing was generally allowed. For example, the Fifth Circuit had no difficulty deciding that an emergency room physician was safety sensitive (Pierce v. Smith, 117 F.3d 866 (5th Cir. 1997)). Similarly, the Northern District of California found that nurses and pharmacists were safety sensitive (AFGE v. Derwinsky, 777 F.Supp. 1493 (N.D. Cal. 1991)). Also found to be safety sensitive were scrub technicians (who assisted during surgery) (Kemp v. Claiborne County Hosp., 763 F.Supp. 1362 (S.D. Miss. 1991)) and emergency medical technicians (Piroglu v. Coleman, 25 F.3d 1098 (D.C. Cir. 1994)).
It may not be necessary that the position involve direct contact with patients. In AFGE v. Barr (794 F. Supp. 1493 (N.D.Ca. 1991)), the court held that physicians and dentists who directly supervise the diagnosis or treatment of patients on a regular basis are subject to random testing.
But the courts rejected the argument that every job in a hospital is safety sensitive. The Derwinsky court rejected random testing of industrial hygienists, air conditioner mechanics, welders, pest control officers, and elevator mechanics. The Kemp court found that purchasing medical supplies was not safety sensitive.
The area where there is no caselaw to guide us is where the position involves regular patient contact, but does not involve providing medical services. Orderlies, social workers, recreational therapists, etc. would fall into this mystery area.
C. Police Officers and Fire Fighters
Many municipalities have attempted to establish random drug testing programs for police officers and fire fighters, arguing that such positions are inherently safety sensitive. Such arguments have generally been quickly accepted for police, largely on the basis of the fact that they carry firearms (a factor specifically mentioned in Van Raab). A typical case of this type is Penny v. Kennedy, 915 F.2d 1065 (6th Cir. 1990), in which the court upheld random testing for police (and firefighters) in the city of Chattanooga. The supreme court of Massachusetts followed the same approach and reached the same result for police cadets in O’Connor v. Police Comm’r of Boston, 408 Mass. 324, 557 N.E.2d 1146 (Mass. 1990).
Other courts, however, applied this rationale in a more honest and discriminating manner. For example, the First Circuit noted that not all police officers carry guns and, while upholding random testing for those who do (or are involved in drug crimes), vacated and remanded for more information concerning employees who meet neither of these criteria. (Guiney v. Roache, 873 F.2d 1557 (1st Cir.), cert. denied, 493 U.S. 963 (1989)). The District Court for the Eastern District of Michigan took the same approach in Brown v. City of Detroit, 715 F. Supp 832 (E.D. Mich. 1989), upholding random testing for officers who use force and make arrests.
The three recorded decisions regarding firefighters reach different conclusions that cannot be reconciled. In Brown v. Winkle, 715 F. Supp. 195 (N.D. Ohio 1989), the court virtually took judicial notice of the “fact” that firefighting is safety sensitive and allowed random testing. The Delaware District Court reached the same conclusion in Wilcher v. City of Wilmington, 891 F. Supp. 993 (D. Del. 1995). The middle district of Florida, however, found that firefighting was not safety sensitive, primarily because there were no accidents that could be attributed to drugs (Beattie v. City of St. Petersburg Beach, 733 F. Supp. 1455 (M.D. Fla. 1990). While this means that we still have the opportunity to argue in future cases that firefighters cannot be constitutionally subjected to random testing, this could be a very hard argument to win. Given the nature of this job, it could easily be true that a firefighter who reported for work under the influence of drugs could pose a safety hazard.
D. Classified/Sensitive Information
The second most common special need claimed by government employers is security. Many agencies have attempted to institute random testing programs for employees who handle sensitive information.
The assertion that drug use creates a risk that sensitive information will be mishandled is anything but obvious. While the risks created by an airline pilot who got high on the job need no explanation, it is not apparent that someone who handles classified information while under the influence will do any harm. Nevertheless, courts have uncritically accepted this argument. The unstated rationale appears to be that if the potential harm from a job being done wrong is great enough, there is a special need that justifies random testing, even if there is no apparent nexus between drug use and poor job performance.
Consistent with this “principle”, the District Court for the District of Columbia allowed random testing for executive department employees with security clearances that gave them access to classified material, but enjoined testing for those employees without security clearances, despite the government’s claim that they had access to sensitive information. (Hartness v. Bush, 794 F. Supp. 15 (D.C.C. 1992), AFGE v. Sullivan, 744 F.Supp. 294 (D.D.C. 1990)). This was consistent with the D.C. Circuit’s earlier decision allowing random testing for holders of both “secret” and “top secret” security clearances, and rejecting claims that only the latter were truly sensitive. (Hartness v. Bush, 919 F.2d 170 (D.C. Cir. 1990), cert. denied, 501 U.S. 1251 (1991)). The D.C. Circuit reached a consistent result in Harmon v. Thornburgh , 878 F.2d 484 (D.C. Cir. 1989), cert. denied, sub nom, Bell v. Thornburgh, 493 U.S. 1056 (1990), allowing random testing of those with top secret clearances, but enjoining testing for other federal prosecutors and other employees with access to grand jury proceedings.
The Ninth Circuit also took this approach, holding in AFGE, Local 1533 v. Cheney, 944 F. 2d 503 (9th Cir. 1991), that navy department civilian employees with top secret security clearances which gave them access to information whose disclosure could cause “exceptionally grave damage to national security” could be randomly tested.
Other federal courts which grappled with this issue reached consistent results, rejecting attempts to expand random testing to public employees generally. For example, the Northern District of Georgia rejected a Georgia state law requiring testing of all applicants for state employment, refusing to accept the argument that maintaining the integrity of government employees was a special need (Georgia Ass’n of Educators v. Harris, 749 F. Supp. 1110 (N.D. Ga. 1990). The Eastern District of Louisiana turned back a slightly less blatant violation of Skinner which tried to label all court house employees either safety or security sensitive, including the plaintiff, a clerk whose job was to swear witnesses and take minutes in civil cases (Romaguera v. Gegenheimer, 1996 U.S. Dist. LEXIS 6272 (E.D. La. 1996).
One recent case which could be read to cast doubt on these guidelines is Stigile v. Clinton 110 F. 3d 801 (D. C. Cir. 1997) cert. denied, ____U.S. ___,1998 U.S. LEXIS 1458 (1998). Here, the Court of Appeals for the D.C. Circuit allowed random testing of economists working for the Office of Management and Budget even when they did not have security clearances. These employees did, however, have access to the areas in the Old Executive Office Building frequented by the President and Vice President. Thus, the case is better understood as a sui generis response to the unique security needs of the President than as a sign that the D.C. Circuit is retreating from its position in Hartness, Harmon, and Cheney.
E. Drug Interdiction
One aspect of security that has received a great deal of judicial attention is drug interdiction. The Court in Van Raab found that customs officials involved in drug interdiction represented a special need that outweighed the privacy interests against random searches. The Court found that, while there was no history of drug abuse among customs agents, there was a history of attempts to bribe them. The official rationale is that agents who use drugs are more susceptible to bribery, presumably because they need money to buy drugs. There are obvious difficulties with this argument, especially where marijuana is concerned, which the Court did not address. A more likely explanation is that it seems facially inappropriate to have the fox guarding the henhouse.
While the lower courts have generally been willing to look at the specific requirements of the job where transportation safety is involved, or even national security, in this area all willingness to examine and weigh the facts has been thrown to the winds. In each of the four major post-Van Raab cases, the government was completely victorious. Not only did the government prevail in cases where the employees involved in drug interdiction had security clearances (National Treasury Employees Union v. Hallett, 756 F. Supp. 947 (E.D. La. 1991)), a result consistent with the national security decisions, but it prevailed in cases where the employees had no access to classified information (National Treasury Employees Union v. Hallett, 776 F. Supp. 680 (E.D.N.Y. 1991). In the latter case, the court held that all 14 of the jobs in question were sensitive. The D.C. Circuit also found that employees with access to information that was “sensitive” but not classified were subject to random urine testing (National Treasury Employees Union v. United States Customs Serv., 27 F.3d 623 (D.C. Cir. 1994)).
One might assume that this attitude would spill over into decisions concerning employees involved in drug testing. This, however, has not been the case. In the only recorded federal case on this issue, the D.C. Circuit held that lab workers and employees involved in the chain of custody are not subject to random testing. (National Fed’n of Fed. Employees v. Cheney, 884 F.2d 603 (D.C. Cir. 1989), cert. denied, 493 U.S. 1056 (1990)).
F. Prison Employees
Prisons have a well deserved reputation as a place where the rules don’t apply, and the manner in which the courts have handled testing issues for correctional employees does nothing to disturb that reputation.
The courts uniformly upheld random testing of prison guards, despite the absence of any clear rationale for this exception to normal Fourth Amendment principles. (AFGE v. Roberts, 9 F.3d 1464 (9th Cir. 1993), Taylor v. O’Grady, 888 F.2d 1189 (7th Cir. 1989), Seeling v. Koeher, 76 N.Y. 2d 87, 556 N.E.2d 125, 556 N.Y.S.2d 832 (N.Y.), cert. denied, 498 U.S. 847 (1990)). Even employees who are not guards could be randomly tested, so long as they came into contact with prisoners. In Roberts, the Ninth Circuit allowed random testing of dietitians, telephone operators, plumbers, and other employees having nothing to do with law enforcement or prison security. The Georgia Supreme Court reached the same conclusion in Georgia Dep’t of Corrections v. Colbert (260 Ga. 255, 391 S.E.2d 759 (Ga. 1990)), in which it held that the prison fire chief was subject to random testing.
Attempts to push the envelope even further and test prison employees without contact with prisoners have not been successful. In Taylor, the Seventh Circuit, while allowing testing of guards, struck down the portions of the testing program dealing with clerical workers and other employees without contact with prisoners. The court found that the smuggling issue, while compelling, did not apply to such employees, and that public perception was not a sufficient interest to allow random testing. Efforts by prison authorities to extend testing to employees who worked outside prisons got even shorter shrift. In Roberts, the testing program originally included all employees of the department of corrections, including those who did not work in the prison. Testing of these employees was enjoined by the district court, and the Federal Bureau of Prisons did not appeal.
III. POST ACCIDENT TESTING
Post accident testing presents different issues than random testing. At least in theory, such testing is based on individualized suspicion. One would think that courts which were willing to allow random testing with such slender justification would automatically bless testing whenever there was any individualized suspicion. This, however, was not always the case. For example, in Plane v. United States, the District Court for the Western District of Michigan struck down the Defense Logistics Agency’s post accident testing program. The agency tested employees who were “involved in” an accident. The court held that such testing was overbroad because it included all employees who were involved, even those employees whom the facts indicated could not have caused the accident.
In Connelly v. Newman, 753 F. Supp. 293 (N.D. Cal. 1990), the Northern District of California struck down the post-accident testing program of the Office of Personnel Management. Under this program, employees who were involved in accidents which produced $1,000 or more in property damage were subject to testing. The court held that being involved in an accident of this magnitude was insufficient to provide probable cause for testing. This holding is especially interesting in light of the fact that the employees were in positions that were arguably safety sensitive.
IV. REASONABLE EXPECTATION OF PRIVACY
Many courts have used the concept of “reasonable expectation of privacy” in justifying their decisions. The idea is that if, under the circumstances, employees have no reasonable expectation of privacy, a government search is less likely to be unreasonable. For example, the court in Skinner stated that the heavily regulated nature of the railroad industry reduced the amount of privacy employees could expect. Many other decisions cited a diminished expectation of privacy as a factor in their decision to allow random testing.
In practice, however, this factor had little, if any, impact on the courts’ decisions. Courts which found that the government had a special need virtually always allowed random testing, with or without a diminished expectation of privacy. Where courts found no special need, they did not allow random testing, even where there was a diminished expectation of privacy.
There are even a number of cases in which the court explicitly stated that the presence of a special need in and of itself creates a diminished expectation of privacy. For example, in Pierce, 117 F. 3d 866 (5th Cir. 1997), the court found that state’s interest in medical safety created a diminished expectation of privacy (in addition to a special need), even in the absence of any other privacy diminishing factors. The court Burka also explicitly collapsed the concept of expectation of privacy into special need.
This does not mean that we should ignore this concept in our litigation efforts. We should look for factors which support the argument that employees in particular cases have a reasonable expectation of privacy and urge them as a reason to prohibit random testing. But we should be aware of the limited weight this factor carries when we make strategic and case selection decisions.
V. THE VERNONIA DECISION
The continued validity of the entire line of cases beginning with Skinner and Van Raab was thrown into doubt by the Court’s decision in Vernonia School Dist. v. Acton, 515 U.S. 646 (1995). In Vernonia, the Supreme Court reversed the Ninth Circuit and held that it was constitutional for a school district to conduct mandatory random urine testing of high school student athletes.
This decision was obviously inconsistent with Skinner and Van Raab. The bedrock principle of these cases was that random testing is impermissible in the absence of some special need. The court paid lip service to this principle by pointing to the risks involved in some forms of athletics, especially football. But some degree of risk is inherent in almost all human activity, and the mere existence of some risk does not meet the Skinner/Van Raab test. Skinner and Van Raab (and the federal caselaw flowing from them) clearly required an elevated level of risk to justify abandoning the usual requirement of individualized suspicion. Perhaps football rises to the level of risk required by previous caselaw, but many (probably most) sports do not.
The language in Vernonia also raised concerns about the continued vitality of Skinner. The Court explicitly rejected the language from that case requiring the government to have a “compelling” interest in order to justify random testing. Instead, the court held that there is no minimum standard which the state’s interest must meet; any state interest will suffice if it is “important enough to justify the particular search at hand.”
While the Vernonia court obviously failed to follow Skinner, the significance of that failure was not immediately clear. Students have traditionally had much fewer rights than adults, and the Court’s opinion placed great emphasis on the fact that the school stands in loco parentis to the students. The opinion stated, in its lengthy discussion of this point, “the subjects of the Policy are children who have been committed to the temporary custody of the state as schoolmaster; in that capacity, the State may exercise a degree of supervision and control greater than it could over free adults.” The Court also pointed out that it had already identified schools as a special need for Fourth Amendment purposes in New Jersey v. T.L.O., 469 U.S. 325 (1985). It was thus possible that Vernonia represented “merely” the refusal to apply Skinner to students, not a general retreat from Skinner.
The meaning of Vernonia was revealed by Chandler v. Miller, 520 U.S. 305, 117 S. Ct. 1295, 137 L. Ed. 2d 513 (1997). Chandler involved a challenge to a Georgia state law requiring candidates for designated state offices to pass a drug test. The Eleventh Circuit, reading Vernonia as a retreat from Skinner, upheld the requirement. If the Court had really abandoned Skinner’s requirement of a special need to legitimize random testing, it could easily have agreed with the Court of Appeals and found this search to be reasonable. In fact, it would have been the perfect vehicle to continue to undermine Skinner without formally overturning it. The Court did not take this course, however. Instead, by a vote of 8-1, it struck down the state law. Even the state attorney general, with responsibility for enforcing drug laws, was held to be ineligible for random testing.
The language in Chandler was also consistent with Skinner. The opinion did not echo the balancing test language of Vernonia, despite the urging of Chief Justice Rehnquist (who filed the lone dissenting opinion). Instead Justice Ginsburg’s opinion states that “the proffered special need for drug testing must be substantial–important enough to override the individual’s acknowledged privacy interest, sufficiently vital to suppress the Fourth Amendment’s normal requirement of individualized suspicion.” The opinion faulted the state’s presentation for failing to provide an “indication of a concrete danger demanding departure from the Fourth Amendment’s main rule.”
Ginsburg’s opinion also described the Court’s holdings in Skinner, Van Raab, and Vernonia, in the narrowest possible terms. She described Skinner as allowing random testing because of compelling public safety concerns and found the basis of Van Raab in the “almost unique mission” of the INS. Vernonia is described in terms of the school’s “large responsibilities—as guardian and tutor of children entrusted to its care.”
Thus, public employees still have the protection of the Skinner/Van Raab standards. Vernonia, while it represents still another failure of the Court to enforce the Fourth Amendment, is apparently confined to students.
VI. OTHER OPEN ISSUES
There are other critical issues where the law is not clear, either because the issue has not been sufficiently litigated, or because Vernonia has unsettled a point that was formerly clear.
A. Direct Observation
The most offensive aspect of urine testing is direct observation, which can amount to a virtual strip search. Fortunately, few testing programs have included this feature, or did so only when there was evidence that an individual was cheating.
For years it was relatively clear that direct observation was illegal unless there was evidence that the individual being observed had given prior indications that he or she was attempting to tamper with the test. Both the Ninth Circuit (in Piroglu v. Coleman) and the D.C. Circuit (in International Brotherhood of Teamsters v. Department of Transportation) held that direct observation in the absence of evidence of fraud was unconstitutional.
Once again, however, Vernonia has muddied the law. The Court in Vernonia allowed direct observation of randomly chosen student athletes with no evidence of cheating (all students tested were directly observed).
As previously discussed, the meaning of this decision for testing in employment and other contexts is anything but clear. At best, it is based on the traditionally diminished rights of students, and means nothing in other contexts. This is probably at least close to the truth. Even at worst, however, the decision may not be as bad as might appear. The boys in Vernonia gave their sample at a wall urinal, fully clothed. The observer stood somewhere behind them, and observed them, if at all, from the back at some distance. Girls gave their sample in a closed stall with the monitor standing outside, listening for signs of tampering. While this is more intrusive than having the testee go into a bathroom (without running water) alone while the monitor waits outside, it is still a far cry from the nightmare scenario in which the person has to undress and urinate in full view of the observer who stares at their genitals.
There have been two post-Vernonia direct observation cases, with different results. In Wilcher v. City of Wilmington, the Third Circuit allowed direct observation for firefighters (in the absence of fraud), citing Vernonia. There was no direct genital observation, but observers did go inside the stalls with both men and women.
The Southern District of New York took a different approach in Kennedy v. City of New York, 1995 U.S. Dist. LEXIS 7437 (S.D.N.Y. 1995). Here, a police officer was selected for random testing while in the hospital recovering from an injury received in the line of duty. He was forced to give the sample in his hospital room in full view of five officers who apparently had full view of his genitals. The court had no hesitation in finding that this was unnecessarily degrading and an illegal invasion of privacy. It is possible, however, that the court was influenced by the particularly egregious facts of this case.
It is possible that the common law right of privacy might have some application here. This right exists in virtually every state, and most courts which have considered the issue have held that it applies to employers. The basic elements of this tort are that the intrusion be highly offensive to the average person and that there is no legitimate justification for the intrusion. The first of these elements is clearly met by direct observation (at least in any but the mildest cases). While the employer would appear to have a strong argument that the intrusion is necessary where there is evidence of tampering, it might be hard pressed to justify observation in the absence of such evidence. There is little caselaw on this subject, but the argument is well worth making, especially for private sector employees who do not have the benefit of the Fourth Amendment.
B. De Minimis Performance of Special Needs Responsibilities
Several cases have raised the issue of whether an employee who performs a job which presents a special need, but does so only on rare occasions, can be subjected to random testing.
Some courts have held that a job is not truly sensitive when the sensitive tasks are not a regular part of the employee’s job. For example, in Bannister v. Board of City Comm’rs, 829 F. Supp. 1249 (D. Kan. 1993), the court struck down random testing for a secretary who delivered meals to the elderly at most twice a week when volunteers failed to show up. The court held that such incidental performance of a safety sensitive function would not support random testing. The District Court for the Southern District of New York reached the same conclusion in the case of a municipal sanitation inspector whose driving was only occasional. (Watson v. Sexton, 755 F. Supp. 583 (S.D.N.Y. 1991)).
Other courts, however, have reached the opposite conclusion. In Derwinsky, the court held not only that nursing is a sensitive position where random testing is appropriate, but also that the 5% of her time a nursing instructor spent with patients was enough to consider her job sensitive. The court in Cheney held that the frequency with which employees were exposed to top secret documents did not determine whether their job was sensitive, any exposure was considered enough.
These decisions cannot be reconciled, and the few cases on point are equally divided, so there is no majority view. Only when more cases are decided will the law in this area be clear.
C. Applicant Testing
A few courts have held that applicants can be tested without cause even where testing of incumbent employees would be illegal under a Skinner special needs analysis. In Willner v. Thornburgh, 928 F.2d 1185 (D.C. Cir.), cert. denied sub nom., Willner v. Barr 502 U.S. 1020 (1991), the D.C. Circuit held that applicants for staff lawyer positions in the Justice Department Court could be subjected to urine tests. The court did not find that the position involved a special need, and would have been hard pressed to do so, since it did not involve drug enforcement. But the position was obviously somewhat sensitive, which makes the implications of this case for truly rank and file applicants unclear.
There is no lack of clarity, however, in Loder v. City of Glendale, 14 Cal. 4th 846, 927 P. 2d 1200, 59 Cal. Rptr. 2d 696 (Cal.), cert denied, ___U.S.___, 118 S. Ct. 44 (1997), in which the California Supreme Court held that it was constitutional to conduct drug tests of all applicants for municipal employment, regardless of the nature of the position. The court’s rationale was that employers cannot judge applicants on job performance, the way they can employees. The opinion also cited the prior existence of a drug abuse problem in the city’s workforce and the fact that the drug test took place in the context of a pre-employment medical exam as factors which made the test reasonable. But there is no reason to believe that the absence of those factors would have changed the decision.
There is a germ of truth to this argument. It is obviously more difficult to get information about applicants. But the court was wrong in its assumption that information about an applicant’s previous job performance is unavailable to employers. Experienced human resource managers and search firms routinely get candid reference information from prior employers. There is no need to subject applicants to drug tests to avoid hiring problem employees.
Other courts have rejected the Loder rationale. The Northern District of Georgia struck down across the board testing for all applicants for state jobs in Georgia Association of Educators v. Harris. The Southern District of New York rejected testing of applicants for non-sensitive jobs in Burka.
This issue is still very much up for grabs and we should seize opportunities to challenge applicant testing. The Washington state affiliate is currently challenging a policy identical to the one in Loder in Robinson v. City of Seattle.
D. Hair Testing
As bad as random urine testing is, hair testing is even worse. While it is not as intrusive as urine testing, it is not reliable. Every reputable scientific organization which has examined hair testing, including the FDA and the Society of Forensic Toxicologists, has concluded that hair testing is not reliable enough to be the basis of employment decisions. Moreover, the link between a positive test and on duty impairment, already strained in urine testing, is completely erased with hair testing because drug metabolites remain in the hair permanently. (These problems are discussed in more detail in our affiliate memo of June 29, 1990–Item DT8 in the workplace rights document bank.)
For years, this was not a great concern because hair testing was not used by a significant number of employers. Unfortunately, this has now changed. Wayne Huizenga, the Florida centimillionaire who owned the Florida Marlins and Blockbuster Video, acquired a major stake in Psychemedics and has provided them with financial resources and political connections to beef up their marketing efforts. The result is a dramatic increase in their commercial success which shows no signs of abating.
Fortunately, there are strong legal arguments that can be made against hair testing, at least for public employees. Courts have generally held that drug tests of public employees must be reasonably accurate and reliable. In Jones v. McKenzie, 628 F. Supp. 1500 (D.D.C. 1986), rev’d on other grounds, 833 F.2d 335 (D.C. Cir. 1987), vacated sub nom, Jenkins v. Jones, 490 U.S. 1001 (1989), the District Court for the District of Columbia held that it was unconstitutional for the city to fire a bus driver on the basis of an unconfirmed urine screening test. The District of New Jersey reached the same conclusion in Capua v. City of Plainfield, 643 F. Supp. 1507 (D.N.J. 1986). The Supreme Court’s opinion in Skinner indicated that their approval of the federal testing program was predicated on the care which had been taken to make the tests reliable.
In light of the scientific community’s consensus that hair testing is not reliable, we have a good argument that terminations based solely on its results are also arbitrary and capricious and a violation of due process.
E. Americans with Disabilities Act (ADA)
Unfortunately, the limited degree of protection afforded by the Fourth Amendment protects only public employees. It does nothing to protect those who work in the private sector. Traditionally, the only source of protection for these employees is the handful of state statutes that either outlaw random testing or restrict it to safety sensitive jobs.
The ADA, however, may offer some protection. The ADA prohibits discrimination against those who are disabled and those who are perceived to be disabled. While individuals who are currently using illegal drugs are explicitly denied protection (even if their drug use causes them to meet the statutory definition of disability), no such exclusion applies to those who are erroneously perceived to be disabled because of drug use. Thus, those who are denied employment because of a false positive test result may be able to claim that the employer discriminated against them because it perceived them to be disabled. The EEOC has recognied the legitimacy of this argument in a complaint brought by the Tennessee affiliate and the workplace rights taskforce.
VII. FUTURE CHALLENGES AND OPPORTUNITIES
The infamous Skinner/Van Raab standards have now been with us for almost a decade and there is no indication that the Court has any intention of reconsidering them. There remains, however, a great deal that the ACLU can do in this area.
The most obvious challenge is to keep these standards from being stretched beyond their legitimate narrow scope. Employers have constantly tried to argue that large segments of their workforce have critical responsibilities that justify random testing. These arguments are not entirely far fetched. Many jobs have the potential to cause great harm, even injury and death, if they are done wrong. Skinner’s distinction between critical and non-critical jobs is at best a matter of degree. Currently, relatively few public employees are required to submit to random testing. We must make every effort to keep it that way, or the exception will swallow the rule..
We need to pay special attention to applicant testing. Unlike random testing for incumbent employees, where the law is on our side and we have “only” to protect it, the law on applicants is an open question. If it is decided wrongly, all employees will be subject to testing without cause whenever they change jobs.
We also need to focus on direct observation. Vernonia has at least created the impression that this practice is now legitimate. We need to bring well selected cases to reestablish that this odious practice is off-limits in the absence of tampering.
There are also opportunities to expand protection that we can do more to develop. The de minimis argument (VIB above) might be successfully used to eliminate random testing for people whose special need responsibilities are only a small part of their job. And the ADA might do much to help those who lose employment opportunities because of laboratory errors.
One requirement of much of this litigation will be greater use of expert witnesses. Many of the current issues are factual, such as the extent to which it is possible for an employer to get the information it needs about applicants without drug testing and whether hair testing is reliable. The taskforce is working on identifying and recruiting the experts on these issues so that they will be available when they are needed.
Another area where we can do more is to present courts with alternatives to random urine testing. Many of the bad decisions are the result of judges deliberately (and possibly consciously) bending the law because of the perceived urgency of the drug problem. While there is little we can do in the short run to eliminate this hysteria, we can reduce the problem by showing that there are other ways to respond to drug abuse which are both more effective and less intrusive, such as impairment testing. The taskforce has a wealth of materials on this subject, which we are happy to share.
Finally, there may be circumstances in which state constitutional law can be made to provide greater protection than the federal constitution. For example, after the First Circuit had held that most police officers could be subjected to random testing (see page 4), the Massachusetts Supreme Court relied on the state constitution to rule that they could not. Many other states constitutions include privacy protection provisions which might support similar rulings.
The national workplace rights taskforce will be happy to work with affiliates to evaluate drug testing complaints and help litigate those that will help maintain and develop good law in this area.
VIII. TABLE OF CASES