Full opinion text
MEMORANDUM OPINION AND ORDER GETZENDANNER, District Judge: This is a civil rights class action, on behalf of the class of correctional officers and supervisors at the Cook County Department of Corrections (“DOC”), challenging the constitutionality of a compulsory urine testing program. The program compels all correctional employees, on pain of termination, to submit urine specimens for the purpose of drug testing. An employee must submit a urine sample in the absence of a reasonable suspicion that the employee is illegally using drugs. The named plaintiffs, Yvonne L. Taylor, Charles Edwards, and Jean S. Jackson, are currently employed as correctional officers with the Department of Corrections. Also named as a plaintiff is Frankie McNeal, a correctional sergeant and supervisor. Defendant James O’Grady is the Sheriff of Cook County. Within his office the Illinois General Assembly has established the Cook County Department of Corrections. Ill. Rev.Stat. ch. 125, 11202. Among the Sheriff’s express powers is the appointment of the Executive Director of the Department of Corrections to act as the chief executive. Id. ¶ 212. The current Executive Director, Spencer Leak (replacing former Executive Director Philip Hardiman), is also a defendant in this suit. Acting pursuant to their responsibilities, defendants have ratified and implemented General Order 9.8A, the subject of this lawsuit. The complaint seeks certification of the class of correctional employees and supervisors at the Department of Corrections (“DOC”). In a previous order, I certified the class to include all correctional employees and supervisors at the DOC, except for the defendants and those employees of the DOC who testified for the defendants. The complaint also seeks a declaration that the annual, compulsory urine testing of correctional employees authorized by General Order 9.8A are unconstitutional searches and seizures in violation of the fourth amendment. Further, the complaint seeks a declaration that the follow-up compulsory tests (for those who test positive initially) be declared violative of the fourth amendment. Relatedly, the complaint seeks a declaration that the continued employment of correctional employees may not be conditioned on their willingness to consent to mandatory urine testing. Finally, the complaint requests that defendants be permanently enjoined from implementing and enforcing the urine testing program described by General Order 9.8A and from conditioning plaintiffs’ continued employment with DOC upon their willingness to consent to compulsory urine testing. Under the current standards for evaluating fourth amendment challenges to state action, to be discussed below, numerous factual questions must be resolved. See National Federation of Federal Employees v. Weinberger, 818 F.2d 935, 942 (D.C.Cir.1987) (the legal judgments in a fourth amendment drug testing case “will hinge on findings of fact ... with respect to the nature and scope of the drug testing program.”). Accordingly, I held a full trial on the merits over several days. Having heard testimony from several expert and non-expert witnesses on the scope of and need for the program of compulsory urinalysis of all employees, I now make my findings of fact and conclusions of law in accordance with Fed.R.Civ.P. 52(a). For the reasons stated below, I grant the plaintiffs’ request for declaratory and injunctive relief and accordingly enjoin defendants from implementing and enforcing the urine testing program described by General Order 9.8A. I. Facts A. The Urine Testing Program The urine testing program at the DOC is described in General Order 9.8A, the stipulation of facts entered into evidence, and in part of the testimony of Robert E. Glotz. It may be summarized as follows. Once annually every correctional employee, no matter of what rank, will be compelled to submit a urine specimen for chemical urinalysis. The correctional officer to be tested will not be advised of the pendency of the test until roll call on the day on which the specimen is to be submitted. There are two situations in which an officer may be compelled to submit a urine specimen more than once annually. First, if the officer has already submitted a urine specimen which tested positive, and the officer has elected to enter a drug counseling and treatment program, the officer will be tested twice monthly for a period of six months. Second, if, in the view of a correctional superintendent or higher officer, there is a reasonable and articulable suspicion that an officer is using drugs illegally, is in possession of illegal drugs, or is bringing drugs into the institution, he or she will be tested. The actual procedure for eliciting urine from the employees and chemically analyzing the urine samples is as follows. The employee is notified at roll call at the beginning of her or his shift that the test is to take place. The employee is then taken to the holding area for urinating and is required to wait there until able to urinate. The employee has the balance of the shift to provide her or his urine specimen. Once the employee indicates that she or he is able to urinate, the employee is then requested, but not required, to specify medical information relating to the employee’s substance use. The purpose of this inquiry is informational. In the event that there is a positive test result, the employee will not appear to have recently fabricated a defense to a prospective charge of drug use. The parties have indicated plans to enter a consent decree on the provision of medical information, so the legality of this particular procedure is no longer at issue. After the employee has provided the medical information, or indicated that she or he declines to provide it, the employee will be taken to the appropriate washroom. Prior to the entry of the employee into the washroom, the investigator administering the test will inspect the washroom to assure that there are not articles present which could impair the test. After doing so, the investigator will then conduct a “pat down” search of the employee for the same purpose. (The standard “pat down” search consists of the visual observation of the person searched and lightly patting or running the hands over the body extremities and torso to assure that there are no objects present which may be used to corrupt the specimen.) The correctional officer urinates in a cup provided for that purpose. The procedure is observed generally by the assigned investigator from a “discreet” distance. The actual physical observation of the flow of urine from the body into the cup is not required or recommended, unless the investigator is possessed of facts which indicate that the person giving the sample is going to attempt to impugn the veracity of the test by switching samples, substituting water for urine, or the like. Under ordinary circumstances, therefore, women can be in a stall with the door closed, and men at a urinal with their back to the investigator. After the urine specimen has been obtained, it is placed in two containers and sealed. One container is opened and tested at the Department of Corrections utilizing the DAU/EMIT urinalysis test. If the EMIT test shows a positive result for the presence of marijuana, cocaine, heroin, or opiates (the only drugs being tested for), the opened sample is resealed and, together with the unopened sample, is sent to the Met-Path Laboratories for a confirmation examination utilizing the Gas Chrometogra-phy-Mass Spectroscopy Method (“GCMS”). In the event that the results of the EMIT test are confirmed, the employee will be informed of her or his option to have a second GCMS test performed upon the given urine sample by a reputable laboratory. If this test fails to confirm the positive test results already obtained, no action will be taken against the employee. If the employee has tested positive on the initial EMIT/GCMS tests and fails to have a second confirmatory test done, or if the second confirmatory test is positive, the employee has the option of entering a drug treatment program. If the employee chooses not to do so, the Sheriff will file a complaint with the Cook County Police and Corrections Merit Board seeking termination of the employee. If the employee chooses to enter the drug treatment program, she or he will be compelled to submit two urine samples per month for six months to assure that the employee is not continuing to ingest drugs. Positive test results during the treatment program, or positive test results at a later date after completion of the program, will result in the Sheriff seeking the termination of the employee before the Cook County Police and Corrections Merit Board. At any time, the failure to provide a urine sample when directed to do so is grounds for termination. Additionally, an employee who does not or is not able to provide the required sample within the allotted time will be presumed to have refused to give same and the Sheriff will then file a complaint with the Cook County Police and Corrections Merit Board seeking the employee’s termination. B. The Interests the Government Seeks to Protect The defendants advanced three interests which they believe will be promoted through this program of urine testing: (1) the interest in preventing correctional officers from smuggling contraband into the prison (which includes selling drugs to inmates and demanding money from inmates in return for drugs); (2) the interest in having an unimpaired work force (which includes protecting the employees and maintaining the security of the institution as a whole); and (3) the interest in maintaining the public’s perception of the integrity of correctional officers as law-abiding law enforcement officials. The defendants instituted the current drug testing program primarily out of their first interest in countering what they perceived to be the problem of correctional officers smuggling contraband into the prison facility. The other two interests, however, also factored into their decision to implement and ratify the testing program. C. Evidence of Substance Abuse and Drug Trafficking The evidence presented at trial on the nature of substance abuse (alcohol and drug, legal and illegal) and the extent of the problem came in two forms. First was the testimony of Dr. Sidney Schnoll, M.D., Ph.D., an expert in the research and treatment of substance abuse and the field of pharmacology. Schnoll testified, among other things, to the extent of the problem of drug and alcohol abuse in society. The second form of substance abuse testimony came from witnesses Hardiman, O’Grady, and Glotz, who testified to their belief that there was a drug problem specifically at the DOC. Also presented to the court were Defendants’ Group Exhibits A and B. These exhibits contain 147 reports of various types of drug related problems at the DOC from 1981 through April, 1987. Fifty-six of these reports involve allegations that correctional officers were handling illegal drugs. (Twenty-seven involve arrests for suspected drug possession off duty; three involve cases where the officer was found in suspected possession inside the DOC; twenty-six involve inmate or anonymous allegations that correctional officers were trafficking drugs in the DOC.) The remaining two-thirds of the reports describe cases where inmates were found in possession of drugs, drugs were found somewhere in the institution (such as in clothing intended for inmates or mail intended for inmates), or drugs were found on visitors of inmates. Throughout the trial it was agreed that these reports were not admitted for the purpose of the truth of the allegations asserted therein. See Tr. 248-49, 508, 466-67. Both parties acknowledged that the purpose of admitting these reports was only to establish, confirm or clarify the state of mind of the witnesses who read some or all of these reports— Glotz and Hardiman. The acknowledgment that these reports were admitted for this limited purpose is well founded. The substance of every single report was hearsay. Of the fifty-six reports involving correctional officers, only two of them indicate that a criminal judgment of conviction was entered against the officer. These two convictions were for possession of drugs off duty. The remaining cases either do not indicate the criminal disposition, or indicate that the charges are still pending or were dismissed. Many of these reports also indicate how the DOC investigator (who conducted an internal investigation) disposed of the charges. Some charges were sustained, some dismissed, and some suspended. Nevertheless, as all parties and I expressly agreed at trial, these reports were not admitted for their truthfulness. Thus, as indicated before, I regard them as merely providing evidence for and clarification of the subjective state of mind of those who read them. Since Glotz and Hardiman read Groups A and B, respectively, I find that, consistent with their testimony, they in fact believed there to be a drug problem involving correctional officers at the DOC. I do not and cannot regard the reports themselves as tending to show that a drug problem involving correctional officers in fact exists at the DOC, or what the nature of that problem is, since, as just explained, those reports were not admitted for their truth. However, I do find that the testimony of Glotz, Hardi-man and O’Grady as to their personal belief of the existence of the drug problem is probative of its actual existence. I also find that Dr. Schnoll’s testimony is probative of the nature and existence of the drug problem. While his testimony pertained to the drug problem in society generally, there was no evidence suggesting that the subpopulation of correctional officers was worse than the overall population. Indeed, there was testimony provided by O’Grady that the drug problem at the DOC was, in fact, similar to the overall population. See O’Grady Dep. at 50. With this in mind, I now state my findings as to the nature of drug problems at the DOC. Dr. Schnoll testified that the commonly accepted percentage of persons in the workplace with substance abuse problems is 10%. Of that 10%, over 70% are abusers of alcohol, 30% are abusers of non-alcohol drugs. Schnoll also testified that the drugs most commonly abused, aside from alcohol, are legal drugs such as prescription drugs, sleeping pills, and sedatives. Thus, less than 3% of the work force abuses the sort of drugs to which General Order 9.8A relates. Finally, Schnoll explained that in his opinion, after reviewing reports from the National Institute on Drug Abuse and his own observations and work, under one-tenth of those who use illegal drugs are “chronic” drug users. In his testimony, Dr. Schnoll defined a chronic drug user as one who uses drugs on a regular basis, usually three or four times a week, and displays addictive and sometimes desperate behavior. Thus, based on Schnoll’s numbers, only 3% of the work force abuses non-alcohol drugs; most of this 3% abuse legal and prescription drugs, and of the remainder who abuse illegal drugs, only one-tenth of that remainder are chronic users of illegal drugs. There are approximately 1700 correctional officers and supervisors at the DOC. For purposes of making a calculation, I can assume that Dr. Schnoll’s estimate (that “most” of the 3% who abuse non-alcohol drugs abuse legal and prescription drugs) means that at least half (50%) of that 3% abuse legal and prescription drugs. “50%” would be an estimate most favorable to defendants and still consistent with Schnoll’s testimony. With these numbers, I therefore can estimate that only about three DOC employees are chronic, illegal drug abusers (1700 X 3% X 50% X 10%), around 23 employees occasionally use illegal drugs (1700 X 3% X 50% X 90%), around 26 employees abuse legal and prescription non-alcohol drugs (1700 X 3% x 50%), and around 120 employees abuse alcohol (1700 x 7%). Of course, the urine testing program, testing as it does for cocaine, heroin or opiates, and marijuana, would not detect members of this last and obviously largest group. There is no evidence contrary to these numbers. The testimony of Hardiman as to the extent of the drug abuse problems at the DOC actually confirms Schnoll’s estimate. Hardiman estimated that between 1981 and 1986 about 300 to 500 correctional employees had substance abuse problems. He also explained how he arrived at this figure. As just stated, there are about 1700 correctional employees at any one time. Hardiman took this number and multiplied it by the accepted 10% estimate for overall substance abuse in the workplace that Schnoll cited. This is 170 people. He then explained that he had to factor in an additional number of substance abuse cases because of attrition and new hires over the six year period. Thus, he revised his estimate upward from 170 to 300 to 500 employees. This does not mean that Hardi-man differed with the estimate that at any one time 10% of the work force has substance abuse problems. Rather, Hardiman merely had to adjust that one-time estimate upward to determine the number of employees with abuse problems over a six year period. Hardiman’s reliance on Group Exhibit B (covering the six year period of 1981-1986) also does not undercut Schnoll’s estimate of the drug abuse problem. As explained earlier, that exhibit was not admitted for its truth; it only shows Hardiman’s state of mind in believing that the DOC has a drug problem. I therefore examined those reports for the purpose of clarifying Hardi-man’s state of mind. Three of the ninety-four reports in Exhibit B indicate that a correctional officer was found in possession of a suspected illegal drug while on duty. Twenty-three of the reports involve allegations (some dismissed) of off-duty possession. Thirty involve allegations by an inmate or an anonymous person that a correctional officer was trafficking in drugs. However, only in four of these trafficking cases did the DOC investigator actually sustain the charges. (In two others, the officer resigned prior to a determination of the charge; in the remaining investigations for which a disposition is indicated (in seven none is indicated), and which were completed (four were closed or suspended prior to completion), the charges were either not sustained or held to be false (thirteen such cases)). As Hardiman explained at trial, “[w]e get a lot of reports from inmates on officers ... we take [them] with a grain of salt ... Inmates are likely to report officers for no reason at all_” Tr. at 469. The remaining thirty-eight cases in Exhibit B involve problems with inmates and drugs, drugs in inmate mail, and visitors with drugs. Therefore, I regard Group Exhibit B as evidencing Hardiman’s belief that, as to correctional officers, the drug problem at the DOC consists of off-duty drug possession, and, to a somewhat lesser extent, drug trafficking and on-duty possession. I also view this belief on the nature of the DOC drug problem as fully consistent with the numbers derived from Schnoll’s testimony. The testimony of Sheriff O’Grady also failed to contradict Schnoll’s figures. Indeed, as indicated above, the Sheriff willingly assumed that the drug problems at the DOC were similar to the population at large. While he labeled the DOC drug problem “serious,” he did not quantify this and expressly refrained from calling it “extremely serious.” O’Grady Dep. at 52-53. O’Grady also explained that his evaluation of the drug problem at the DOC was based on just four reports he had read. All of them involved allegations of drug smuggling into the jail, and they were all currently under investigation. O’Grady Dep. at 50-51. He knew that two of the reports involved only individuals; he did not know how many were involved in the other two investigations. On the basis of this hearsay (not included in Group Exhibits A or B), O’Grady reached his conclusion that the problem was “serious,” as serious as in the general population. See id. at 53. Because O’Grady’s testimony that the drug problem is “serious” was based on these four hearsay reports of drug sales, I find that O’Grady believes that there is a “serious” drug trafficking problem at the DOC. This label, of course, is not a quantification. However, O’Grady did expressly testify that a drug problem qualifies as “serious” in his mind if just one employee has broken the law. O’Grady Dep. at 57. Thus, O’Grady’s testimony about the “seriousness” of the drug problem does not strike me as inconsistent with Schnoll’s. I therefore affirm my conclusion that Schnoll’s figures have not been contradicted. While failing to quantify the actual extent of the DOC drug problem, except to confirm that Schnoll’s numbers applied to the DOC, O’Grady did offer testimony that those who sell drugs also use drugs, as a general rule. Interestingly, O’Grady did not state that those who use drugs generally also sell. There is a difference. O’Grady’s testimony, which I accept, suggests that if you have evidence on a number of sellers at the DOC, you also know the number (approximately) of users. But since O’Grady did not state that most users also sell, I cannot infer from his testimony alone the number of sellers at DOC just by knowing the number of users. However, his testimony does tell me that, as a general proposition, if I can identify all the drug users, I will have detected the sellers as well. Finally, Acting Director Glotz also testified that there was a drug problem at the DOC. He too did not quantify his opinion nor contradict Schnoll’s numbers. He did explain that his opinion was based on Group Exhibit A. This exhibit contains no cases of allegations of drug trafficking and no cases of allegations of a correctional officer possessing illegal drugs while on duty. While the exhibit does contain eleven reports of off-duty possession, none of these has been criminally disposed of yet, except that one was dismissed. The remaining forty-two cases in this exhibit involve drug problems with inmates or visitors or, in a few cases, something else unrelated to correctional officers. I therefore regard Glotz’s opinion about the drug problem as pertaining only to off-duty possession and not drug sales. I also discount his opinion somewhat because it is based on pending criminal cases. In any event, I have no reason to view Glotz’s opinion as inconsistent with Schnoll’s. In summary then, I find, in accordance with Schnoll’s uncontradicted testimony, that 7% of the workforce abuses alcohol (roughly 120 employees) and that 3% of the work force abuses non-alcohol drugs (around 50 employees). I also find that most of this 3% abuses legal and prescription drugs (around 26), and that of the remainder who abuse illegal drugs, only one-tenth are chronic abusers (around 3). I find further, in accordance with O’Grady’s testimony, that these figures apply to the DOC. I also find that Glotz, O’Grady, and Hardiman believe that there is a serious problem with correctional officers abusing drugs, but they do not persuade me that the abuse problem is any more pervasive than Schnoll has indicated. As to drug trafficking at the DOC, I rely on O’Grady’s, Hardiman’s, and Schnoll’s testimony. (Glotz’s testimony applied only to drug abuse.) I interpret the testimony of O’Grady and Hardiman as opining that there is a serious problem of drug sales at the DOC. O’Grady’s opinion is based on four investigations which are not yet completed. For O’Grady, if there is just one verified case, he would consider there to be a serious drug problem. Hardiman’s testimony is based on four cases compiled over a six year period where charges of trafficking were sustained, and two cases where the officer involved resigned. (Four of these six cases occurred in 1983 or before.) From this, I find that O’Grady and Hardi-man believe there to be a serious drug sale problem. However, given the foundation for their belief (which was not even admitted for its truth but merely as evidence of their belief), I am not persuaded based on the opinions alone that the problem of correctional officers trafficking drugs into the DOC is in fact “serious”; rather, I can say only that it exists. Schnoll’s testimony suggests to me that chronic illegal drug abusers are the ones most likely to smuggle illegal drugs into the DOC. He described their addiction behavior in terms that suggest desperation and deceit, conditions one would normally expect to exist before a sworn law enforcement officer would undertake to smuggle illegal drugs. Accordingly, I believe the number of drug traffickers among correctional officers is closely tied to the number of chronic, illegal drug abusers. Thus, it is on the order of three or so employees. This is consistent with my understanding of Hardi-man’s and O’Grady’s testimony. (In particular, O’Grady’s opinion — that just one case constitutes a “serious” drug problem, and that there is indeed a “serious” problem— is consistent with my finding that the number of traffickers is around three.) There was evidence presented at trial as to how illegal drugs entered the DOC other than through correctional officer drug trafficking. This evidence came through plaintiff’s witness Charles Edwards, a correctional officer at the DOC with fourteen years of experience at the facility. He explained that illegal drugs can enter the institution and reach inmates by way of clothing brought for an inmate, or actual contact between certain civilians and inmates. Drugs may be concealed within the stitching of clothing or within the heels of shoes. (This testimony was confirmed by plaintiff’s witness Frankie McNeal.) When civilians enter the institution, normally there is no opportunity for actual contact between them and the inmates. However, Edwards explained that this is not always true. Civilians who temporarily work in the institution and members of church groups can come into actual physical contact with inmates. Glotz also testified on cross-examination that about eighty teachers from the Chicago Board of Education, medical personnel from the Cermak Memorial Hospital, and church group members can and do have physical contact with the inmates. These civilians are put through a pat down search before the possibility of contact, but Edwards added that this is not a thorough search, so cleverly hidden drugs could go undetected. Edwards testified that he knew of one case where illegal drugs entered the institution because a church group member exchanged a bible containing drugs with another bible held by an inmate. (However, Glotz could not recall any such incident, even though he was certain that such an incident would have stuck in his mind.) Given this testimony, I find that drugs can and in fact do enter the institution through clothing brought by visitors destined for inmates, and through brief contacts between some civilians (such as church members and outside workers) and the inmates. I cannot make a finding on the degree to which the drug trafficking problem can be attributed to these sources, but I do find that these sources exist. D. Detecting Substance Abusers and Traffickers There was much testimony on the methods available for detecting substance abusers. By contrast, there was little evidence at trial as to the methods for detecting drug traffickers, other than the traditional means of relying on allegations from inmates or anonymous callers and following those up with an investigation. However, O'Grady’s testimony that most drug sellers are users is helpful here. If defendants can identify all users of drugs, they are bound to catch the sellers since, according to O’Grady, the sellers are generally also users of drugs. Thus, detecting drug users will indirectly help detect drug sellers. In fact, I can be more precise about this. I have already found that the number of drug sellers is closely related to the number of chronic, illegal drug abusers. Indeed, there was no evidence presented to suggest that non-chronic drug abusers will sell drugs. Because of all this, I find that detecting the chronic, illegal drug abusers will indirectly detect virtually all, if not all, drug traffickers. Of course, drug abuse itself (apart from drug trafficking) was a separate problem which the defendants asserted an interest in eliminating. Specifically, eliminating drug abuse at the DOC was necessary to maintain an unimpaired work force and to maintain the integrity of officers in the public eye. Therefore, given these interests and given that detecting abusers will substantially detect sellers, I will now focus on the methods for detecting substance abusers. 1. Urinalysis The urinalysis program which defendants have implemented, consisting of the EMIT/GCMS chemical analysis procedure, is a highly accurate and reliable procedure for detecting the presence of metabolites of cocaine, marijuana and heroin or opiates in a sample of urine. Plaintiffs’ attorney indicated in conference that he was not challenging the reliability or accuracy of the EMIT/GCMS results insofar as they purported to identify the presence of metabolites of cocaine, marijuana or heroin in a urine specimen. Moreover, courts have uniformly approved EMIT and GCMS tests as accurate and reliable predictors of these metabolites. As to EMIT testing, see Spence v. Farrier, 807 F.2d 753 (8th Cir.1986); Harmon v. Auger, 768 F.2d 270, 276 (8th Cir.1985); Jensen v. Lick, 589 F.Supp.35, 38-39 (D.N.D.1984); as to GCMS testing, see Wykoff v. Resig, 613 F.Supp. 1504, 1512 (N.D.Ind.1985); Lovvorn v. City of Chattanooga, 647 F.Supp. 875, 878 n. 4 (E.D.Tenn.1986). On the basis of all this, I conclude that the EMIT/GCMS testing procedure is an accurate and reliable procedure for detecting the presence of metabolites of cocaine, marijuana and heroin in a urine specimen. This finding, however, does not end the inquiry. The relevant question is not so much whether urinalysis can detect drug metabolites in a urine specimen, but whether urinalysis can identify those employees who are using drugs, and therefore possibly using drugs on the job (producing an impairment), or using drugs off duty (thereby affecting workforce integrity), or selling drugs, or a combination of all these. As to on-the-job impairments from drug abuse, the uncontradicted testimony, which I accept, came from Dr. Schnoll. He explained that a positive urine test does not indicate impairment at the time the test was taken — during the employee’s work shift. He further explained that drug metabolites can remain in the urine long after the impairing effects of the drug have ceased, and the particular metabolites that the urinalysis identifies may be inactive. In the case of marijuana, in particular, a person’s urine may test positive several days after the person was exposed to marijuana and after there is no longer an impairment. Based on this testimony, I find that there is no substantial relationship between a positive urine test and an impairment at the time of testing. However, Schnoll’s testimony indicates that a positive urine test indicates that the employee may have been impaired either at some earlier point or at the time of testing. I therefore conclude that a urine test indicates only that a person was exposed to a drug at some point. It does not necessarily indicate impairment at the time of testing or at some other on-duty time. To the extent that the positive urine test may indicate some on-the-job impairment (either at the time of testing or on a previous day), I have been presented with no evidence how likely a positive test will suggest an on-the-job impairment. Dr. Schnoll also provided testimony which leads me to believe that the procedure for eliciting urine from a correctional employee will result in some specimen-switching or tampering and will therefore impugn the veracity of the test results. After receiving an explanation of the stipulated procedures for eliciting urine (specifically, inspection of the washroom, patting down the employee, observing the employee from a discreet distance unless facts already indicate tampering may occur), Schnoll stated that “any person who wants to could easily give a specimen that was not their own.” Tr. at 45. He then indicated that a chronic abuser (but not an occasional abuser) would, in behavior characteristic of an addict, likely falsify her or his specimen. Id. at 45-46. From this I believe that some chronic drug abusers will go undetected. After providing the above testimony, Schnoll was then asked on direct examination to consider the possibility that the procedure for eliciting urine would include the provision that the employee to be tested would have her or his entire shift to provide a urine specimen. He testified that this would make urine-switching even easier. Id. 46. On cross-examination, it became apparent that the primary reason Schnoll believed there could be urine-switching or tampering was that he thought employees could go to a locker or some such place unsupervised, and then provide a false specimen. This premise is incorrect because the actual procedures provide that an employee must wait, under supervision, in a designated area until able to urinate. Therefore, I take Schnoll’s entire testimony as establishing only a small likelihood, but a likelihood nevertheless, that a chronic drug abuser will falsify her or his specimen and therefore go undetected. This finding is confirmed by Schnoll’s additional testimony, given under cross-examination, that the stipulated procedures “could minimize” the possibility for tampering. He does not state that the procedures would definitely minimize the possibility of tampering, and he certainly never suggested that the procedures could or would eliminate that possibility entirely. Accordingly, I find that, under the urine-eliciting procedures of General Order 9.8A, a small fraction of the chronic drug abusers will go undetected even after urinalysis. Conversely, given Schnoll’s testimony that occasional drug abusers will generally not attempt to impugn the urine test’s veracity, I find that the procedures will likely detect the occasional, non-addictive abusers. 2. Trained Supervision An alternative method for detecting substance abuse is trained supervision. Dr. Schnoll testified that chronic drug use is generally detectable by easily identifiable changes in behavior, i.e., tardiness, decrease in ability to perform tasks, increased absences, and changes in relationships with supervisors and coworkers. These are often the first signs of chronic drug abuse and can readily be detected by untrained lay personnel. Tr. 31-32. One way for employers to enhance their ability to detect chronic drug abuse in the workplace, according to Schnoll, is by a basic educational program of short duration that discusses the behavioral changes reflective of impairment and chronic drug abuse. The defendants put in place about one year ago (see Tr. 384) a comprehensive supervisor evaluation form which provides an overall assessment of the performance of the correctional officers. See PI. Ex. 7. Upon reviewing a copy of the new evaluation forms, Dr. Schnoll testified, without impeachment or contradiction, that the DOC’s evaluation process was sufficiently detailed to ensure that a correctional officer who is impaired on the job can be identified. Tr. 118-122. Acting Director Glotz testified that this form was an important part of the employee evaluation process. Tr. 431. He believed, however, that in the year the form has been used he has not noticed a diminution in the problem of the correctional staff abusing drugs. Notwithstanding this belief, Glotz also stated that the absenteeism rate at the DOC has dropped from about 10% per day prior to six months ago, to 5%-7% in the last six months. Tr. 441. Yet, Glotz still believes that the drug problem at the DOC has not improved, and has possibly worsened in the last six months. On the basis of this evidence, I conclude that trained supervision, including the sort contemplated by properly completing the evaluation form of PI. Ex. 7, can detect those who chronically abuse any type of drug and who are impaired at work. It cannot detect the occasional drug abusers who abuse drugs only off-duty. Finally, I find that the evaluation process has not in itself eliminated or significantly reduced the drug abuse problem among correctional officers in the year since the process was implemented. However, judging from the reduced absenteeism rate and recalling that Glotz’s testimony is based in part on hearsay which did not include cases of employee drug trafficking or on-duty employee possession, I conclude that the evaluation process has had some effect. I have little difficulty concluding that the evaluation process has not in itself reduced the drug abuse problem at the DOC. For that to happen, there must, of course, be some productive method for disciplining or at least responding to those employees whose actions indicate an impairment. As to this, it is instructive to observe the rather shocking testimony I heard regarding how the Merit Board handles information which, according to Schnoll, is suggestive of impairment. According to Glotz, only one out of ten employees who exhibit an attendance problem can be terminated. Tr. 438. The problem is so extreme that, as Glotz conceded, only one out of ten correctional officers who are absent without excuse on twenty days over a six month period (i.e., a full month of work days) can be terminated. So it seems to me that, if the drug problem at the DOC has not improved, the supervisory evaluation form cannot be blamed. Rather, the blame lies with how the DOC and the Merit Board respond to information contained in those forms. Finally, I note that the uncontradicted testimony at trial was that observation of behavioral changes at the workplace simply identifies those who are either chronic abusers of drugs or otherwise impaired on the job (by observing behavioral changes). It does not identify the cause of the impairment or behavioral changes. And, of course, supervision cannot identify those who use drugs only off-duty and are unimpaired on the job. This contrasts with urinalysis which indicates the nature of the drug that has been injected and who has injected it, but cannot tell when the injection occurred or when the person tested may have been impaired. 3. Neurobehavioral Testing A final alternative for detecting impairments such as those caused by substance abuse is neurobehavioral testing. Plaintiffs expert, Dr. David Hartman, Ph.D., testified about this technique. Dr. Hartman, supervisor of neurobehavioral testing at Cook County Hospital, described this form of testing as testing of the behavioral output of the brain. The test could detect whether a worker’s thinking and behavioral processes were impaired. Although, like supervision, the test does not identify the source of impairment, it would detect impairment from any source, including heroin, cocaine, and marijuana, as well as other drugs and toxins. Hartman recommended administering the tests in a battery of six to eight tests. This would include a finger tapping test, the “digit symbol” test, the Strube Color test, the Wexler Memory Scale or Benton visual retention test, the block design test, the “trail” test, and the Shipley test. The testing takes about one hour per person and can be administered by a properly trained lay person, although the results must be evaluated by an expert. No elaborate or expensive equipment is needed. Hartman estimated the equipment would cost about $150, and a supply of scoring sheets would cost about $15 per hundred. In Hartman’s opinion, neurobehavioral testing was an optimal method for detecting workplace impairments. Test results from an individual employee could be measured against a preestablished norm, or against previous test results from that individual. Hartman conceded that a person could cheat on the test by strategically underperforming on one occasion (when unimpaired), so that on a later date when she or he was impaired, the impairment might go undetected. Given Schnoll’s testimony that only the chronic drug abusers tend to devise elaborate strategies to avoid detection, I believe that neurobehavioral testing can accurately detect current impairments in occasional drug users. It seems that the detection rate might be less with chronic abusers, but since Hartman regards this testing as optimal, I conclude that neurobe-havioral testing is fairly good as to detecting the chronic drug abusers. However, given the time it takes to administer each test, and the size of the DOC work force, I doubt that this test could easily be administered to every member of the entire work force one time each year. E. Intrusiveness of Compulsory Urine Testing I received testimony from Frankie McNeal, a correctional sergeant at the DOC. As a supervisor, she was required to and did submit to a compulsory urine examination under an earlier incarnation of General Order 9.8A. The procedures to which McNeal was subjected in that earlier incarnation were similar to those now authorized by the order. One possible exception is that when McNeal was compelled to urinate in the washroom stall, she was instructed to keep the stall door open. McNeal then had to hold the stall door open and hold her pants off the floor with her one hand while she attempted to urinate into a cup held in her other hand. To effectively urinate into the cup, she had to do all this while standing. While urinating, McNeal was directly and constantly observed by the officer administering the test. The current procedures still permit direct visual inspection of the tested employee. Although direct inspection is not recommended unless facts indicate that there might be tampering, McNeal’s testimony is instructive because it indicates how she reacted to one way in which the urine program can now be administered. However, it is unclear why McNeal had to hold the door to her stall open while urinating, rather than having the observer do so. The current procedures do not call for this. Another aspect of the procedure applied to McNeal and deserving comment is the person who observed McNeal. It was a subordinate correctional officer. The current procedures specify that the employee being tested is to be observed by an “investigator or other officer.” It appears that this could include a subordinate, but ordinarily the observer would be a DOC investigator. McNeal testified that her experience was humiliating and hateful. I found it painful to listen to her testimony when it was so obvious how embarrassing it was for her, both at the time of the test and when relating it publicly in the courtroom before me. Based on her testimony, I am convinced that compelling an individual to urinate at a particular time and place into a cup for subsequent chemical analysis can be and was for McNeal an extreme intrusion and invasion into her most narrow sphere of personal privacy. Beyond the intrusion of compelling an act of urination, this intrusion was magnified even further by the officer’s visual inspection of McNeal’s act of urination. While not all people may react as negatively as McNeal did, I received no other experiential evidence on intrusiveness. Moreover, as discussed later in this opinion, I conclude that McNeal’s aversion to her experience was, as a matter of law, entirely reasonable under the circumstances. II. Conclusions of Law A. General Fourth Amendment Principles The current controversy over urine testing has spawned much litigation over its constitutional validity. Specifically, this litigation has addressed the question, presented in this case, of the constitutional validity of compelling a person to submit a urine specimen for the purpose of drug testing in the absence of a reasonable suspicion of illegal drug use. I therefore do not write on a clean slate but instead am guided by a body of relevant legal principles. This body of law was recently summarized by the Court of Appeals for the District of Columbia Circuit in National Federation of Federal Employees v. Weinberger, 818 F.2d 935 (D.C.Cir.1987) (with Judge Fairchild of the Seventh Circuit sitting by designation). Although the court in National Federation remanded the case for the development of a factual record on the fourth amendment issues, the governing legal principles set forth there apply here as well. First, as the court wrote: Compulsory urinalysis of public sector employees, we hold in common with our sister circuits, qualifies as a “search and seizure” within the meaning of the Fourth Amendment. See National Treasury Employees Union v. Von Raab, 816 F.2d 170, 176 (5th Cir.1987); McDonell v. Hunter, 809 F.2d 1302, 1307 (8th Cir.1987); Division 241 Amalgamated Transit Union v. Suscy, 538 F.2d 1264, 1266-67 (7th Cir.), cert. denied, 429 U.S. 1029 [97 S.Ct. 653, 50 L.Ed.2d 632] (1976); cf. Shoemaker v. Handel, 795 F.2d 1136, 1142 (3d Cir.1986). This is true whether or not the urinalysis procedures employed involve direct visual inspection of urination. See American Fed’n of Gov’t Employees v. Weinberger, 651 F.Supp. 726, 732-34 (S.D.Ga.1986); Allen v. City of Marietta, 601 F.Supp. 482, 488-89 (N.D.Ga.1985). National Federation, 818 F.2d at 942. The principle that compulsory urinalysis is a “search and seizure” comports with the evidence I heard in this case and with reality. Employees have a reasonable expectation of privacy in the act, place, and decision of urination: There are few activities in our society more personal or private than the passing of urine. Most people describe it in euphemisms if they talk about it at all. It is a public function traditionally performed without public observation; indeed, its performance in public is generally prohibited by law as well as social custom. National Treasury Employees Union v. Von Raab, 816 F.2d 170,175 (5th Cir.1987). Not only does compulsory urination invade a reasonable expectation of privacy, but so does the actual urine testing, both of which are involved here. Urine testing may disclose not only the presence of drug traces but much additional personal information about an employee — whether the employee is under treatment for depression or epilepsy, suffering from diabetes, or, in the case of a female, pregnant. Even tests limited to the detection of controlled substances will reveal the use of medications prescribed for relief of pain or other medications. Id. at 175-76. That compulsory urinalysis constitutes a “search and seizure” does not, of course, end the fourth amendment inquiry. “[T]he Fourth Amendment safeguards persons only against ‘unreasonable searches and seizures.’ ” National Federation, 818 F.2d at 942. The determination of whether a search is reasonable requires a balancing of the nature and quality of the intrusion against the importance of the government interests alleged to justify the intrusion. O’Connor v. Ortega, — U.S. -, 107 S.Ct. 1492, 1499, 94 L.Ed.2d 714 (1987); National Federation, 818 F.2d at 942. On the one side of this balance is the strength of “the employee’s reasonable expectations of privacy — those expectations which society is ‘prepared to recognize as legitimate.’” Id. (citing New Jersey v. T.L.O., 469 U.S. 325, 338, 105 S.Ct. 733, 742, 83 L.Ed.2d 720 (1985); O’Connor, 107 S.Ct. 1497-99. In this case, I must assess the strength of an employee’s interest in not having to urinate at the order of an employer and in not having her or his urine chemically analyzed at the employer’s insistence. In O’Connor v. Ortega, — U.S. -, 107 S.Ct. 1492, 94 L.Ed.2d 714 (1987), eight justices explicitly agreed that a public employee generally has a reasonable expectation of privacy in his desk and file cabinets. See O’Connor, 107 S.Ct. at 1499 (plurality opinion written by O'Connor, J., and joined by Rehnquist, C.J., White, J., and Powell, J.); id. at 1510 (dissenting opinion of Blackmun, J., joined by Brennan, J., Marshall, J., and Stevens, J.). The plurality concluded that the employee of that case enjoyed a reasonable expectation of privacy in his desk and files because he did not share these items with other employees and often used his desk for the storage of personal items. Id. at 1499. However, the plurality also observed that the intrusion was not particularly great because the case really only involved office space, not the home, and the employee could avoid exposing personal belongings at work to a search “by simply leaving them at home.” Id. at 1502. Given this reasoning, it follows that an employee must be deemed to have a fairly strong expectation of privacy in the act of urination and in the urine itself. In contrast to O’Connor, this case involves the search and seizure of a person, not an office. “Persons” is one of the categories explicitly guaranteed protection in the fourth amendment. This case also involves the search and seizure of urine. Virtually everyone disposes of urine in a fashion to preclude the public from witnessing the urine itself, thus making the urine a personal item at least during the act of urination. The very purpose of the urine testing program of this case is to make sure an employee’s urine is seized at work. Thus, unlike O’Connor, one cannot avoid the seizure and search by “leaving at home” one’s urine. In view of all this, I must conclude the correctional employees of this case have a fairly strong expectation of privacy in the place, act, and decision of urination. Having concluded this, I must then consider the other side of the balance in the reasonableness inquiry: “The governmental interest [in] the efficient and proper operation of the workplace.” O’Connor, 107 S.Ct. at 1501; National Federation, 818 F.2d at 942; McDonell v. Hunter, 809 F.2d 1302, 1308 (8th Cir.1987). Defendants contend this interest is of overriding significance in this case. Accordingly, they seek to compel urination of an employee even in the absence of some individualized, reasonable suspicion that the employee may be abusing drugs. Ordinarily, the government must obtain a warrant before a valid search may be executed. The warrant requirement, however, is not appropriate in a case like this when “the burden of obtaining a warrant is likely to frustrate the governmental purpose behind the search.” New Jersey v. T.L.O., 469 U.S. at 340, 105 S.Ct. at 743 (quoting Camara v. Municipal Court, 387 U.S. 523, 533, 87 S.Ct. 1727, 1733, 18 L.Ed.2d 930 (1967)). Also ordinarily, “a search — even one that may permissibly be carried out without a warrant — must be based upon ‘probable cause’ to believe that a violation of the law has occurred.” Id. (citing Almeida-Sanchez v. United States, 413 U.S. 266, 273, 93 S.Ct. 2535, 2539, 37 L.Ed.2d 596 (1973); Sibron v. New York, 392 U.S. 40, 62-66, 88 S.Ct. 1889, 1902-1904, 20 L.Ed.2d 917 (1968)). “However, ‘probable cause’ is not an irreducible requirement of a valid search.” Id. Thus, in certain cases the Supreme Court has recognized the legality of searches and seizures based on suspicions that, although “reasonable,” do not rise to the level of probable cause. Id. 469 U.S. at 341,105 S.Ct. at 743 (collecting cases). “Where a careful balancing of governmental and private interests suggest that the public interest is best served by a Fourth Amendment standard of reasonableness that stops short of probable cause,” such a standard should be adopted. Id. There are even some rare circumstances where no individualized suspicion is necessary to validate a search. See id. at 342 n. 8, 105 S.Ct. at 744 n. 8. Nevertheless, “some quantum of individualized suspicion is usually a prerequisite to a constitutional search or seizure,” id. (quoting United States v. Martinez-Fuerte, 428 U.S. 543, 560-61, 96 S.Ct. 3074, 3084-85, 49 L.Ed.2d 1116 (1986)), and exceptions to the requirement of individualized suspicion are generally appropriate only where the privacy interests are “minimal.” Id. Since I have already held that the privacy interests in this case are not minimal, it follows that there should be a heavy presumption against searches based on no individualized, reasonable suspicion. However, in a compelling case, the needs of the government as an employer may be so strong as to justify dispensing with the individualized suspicion requirement in a urine testing program. Cf. O’Connor, 107 S.Ct. at 1501-03. Thus, it is crucial that I assess the documented strength of the government interest in this case, and determine whether it is sufficiently strong to justify compulsory urination in the absence of a reasonable suspicion of substance abuse. This will depend on two factors. First, whether “reasonable grounds” exist for suspecting that the urinalysis will turn up evidence of “work-related drug use.” National Federation, 818 F.2d at 943 (citing New Jersey v. T.L.O., 469 U.S. at 342, 105 S.Ct. at 744). This is another way of saying that the search must be justified “at its inception.” Id. (citing T.L.O., 469 U.S. at 341-42, 105 S.Ct. at 743-44). I must also determine whether the search as it is actually conducted is “reasonably related in scope to the circumstances which justified the interference in the first place.” Id. (quoting Terry v. Ohio, 392 U.S. 1, 20, 88 S.Ct. 1868, 1879, 20 L.Ed.2d 889 (1968)). This means that “ ‘the measures adopted [must be] reasonably related to the objectives of the search and not excessively intrusive.’ ” Id. (quoting T.L.O., 469 U.S. at 342, 105 S.Ct. at 744). Finally, “[t]hose seeking to invoke an exception to the warrant requirement bear the burden of establishing that circumstances required dispensing with that requirement.” United States v. Longmire, 761 F.2d 411, 417 (7th Cir.1985) (collecting cases). Similarly, whenever the government seeks to avoid the requirement of probable cause and reasonable suspicion prior to a search, the government bears the burden of demonstrating the necessity for doing so. See O’Connor, 107 S.Ct. at 1503-04. As just discussed, this must be especially true when the privacy interests to be intruded upon are not minimal. With these general principles in mind, I now apply them to the evidence of this case. B. The Law Applied to This Case The government interest in this case has been advanced in three parts: first, and primarily, the interest in preventing correctional officers from smuggling contraband into the prison; second, the interest in an unimpaired work force; third, the interest in the integrity of a law-abiding work force. I have already discussed the evidence on the extent to which these interests are in fact undermined. Specifically, 3% (fifty employees or so) of the DOC abuse non-alcohol drugs. Only one-tenth of a small part of this 3% (about three employees) are chronic abusers of illegal substances. The other nine-tenths of this “small part” (about twenty-three employees) are occasional abusers, and the remainder of the 3% (about twenty-six) abuse legal and prescription drugs. As to the first interest asserted — drug trafficking — there was no direct evidence on the extent to which the number of drug abusers indicates the number of drug traffickers. One court has asserted that it is “logical” to assume “that employees who use the drugs, and who come into regular contact with the prisoners, are more likely to supply drugs to the inmates.” McDonell v. Hunter, 809 F.2d 1302, 1308 (8th Cir.1987) (disagreeing with the trial court’s finding on this point and doing so without following the “clearly erroneous” standard for appellate review of facts); but see id. at 1311-12 (Lay, J., dissenting and chastising majority for ignoring lower court finding that no evidence connected drug users to traffickers). In this case, based on Schnoll’s evidence about the behavioral attributes of a chronic drug abuser versus those of the occasional abuser, I can conclude that generally the chronic abusers of illegal drugs and generally not the occasional users are likely to be involved in drug trafficking. Thus, only one-tenth of a small part of 3% of the work force (around three employees) is involved in drug trafficking at one time. This figure would be consistent with the hearsay evidence on trafficking, had that evidence been admitted for its truth. This confirms my earlier factual finding that the problem of correctional officers trafficking drugs at the DOC “exists” but is not serious. I also found that those who chronically abuse drugs (i.e., those who usually go to work impaired and are most likely the drug traffickers) can generally be detected through the “trained supervision” about which Schnoll testified. Further, I found that in the short time the DOC has implemented such supervision, there has been some reduction in the drug problem — as reflected through decreased absenteeism— albeit to an undefined extent. Finally, I observed that the DOC has often failed to act on information which Schnoll indicated could be indicative of chronic drug abuse. As to impairments on the job, there was no direct evidence on the extent to which occasional, non-alcohol drug abusers were drug-impaired on duty. Again, I know that this group comprises slightly under 3% (around fifty employees) of the work force. Again, given Schnoll’s testimony regarding the nature of addictive and non-addictive behavior, I believe that most occasional users will not use drugs or become impaired on-duty. However, for those that do, I found that trained supervision could detect the impairments. Next, as to maintaining the integrity of the work force as law-abiding officers, the 3% figure is directly relevant. Since 3% of the work force either chronically or occasionally abuse illegal drugs, that number reflects the extent to which the integrity of the work force is now corrupted. Finally, I found that urinalysis would not identify those who are impaired at the time of the urine tests, but would probably identify chronic drug abusers. It would also identify any type of user who has ingested drugs within the relevant time frame and who were impaired at some previous time, perhaps including some on-duty time. Therefore, urinalysis would likely identify those officers who might be trafficking drugs (because they are chronic abusers) and those officers who are impugning the integrity of the work force (because they abuse drugs at least occasionally). As just explained, I also found that trained supervision would likely detect chronic abusers (and therefore probably drug traffickers) and those who were impaired on duty. It would not detect occasional abusers who are never impaired during work but nevertheless use illegal drugs. These people, however, are unlikely to be traffickers. With this set of facts, I conclude as follows. There are reasons for suspecting that the defendants’ urinalysis program will successfully turn up evidence of work-related drug use. In this sense, a urinalysis search is indeed “justified at its inception,” the first requirement for the reasonableness of a government search. The urinalysis program fails, however, at the second step: “whether the measures adopted are reasonably related to the objectives of the search and not excessively intrusive.” T.L.O., 469 U.S. at 342, 105 S.Ct. at 744. Trained supervision can usually detect the chronic abusers and therefore most, if not all, of the traffickers as well as those who usually go to work impaired. Trained supervision also can detect those of the occasional drug abusers who go to work impaired. Thus, trained supervision — an unintrusive mode of analysis — will successfully promote the first two government interests (the third is discussed below). Indeed, urinalysis will not substantially promote the second interest of detecting on-the-job impairments since there is no necessary relationship between th