National Federation of Federal Employees v. Carlucci

680 F. Supp. 416, 1988 WL 17239

• Court: District Court, District of Columbia
• Decided: March 1, 1988
• Docket: Civ. A. 86-0681, 87-1797 and 87-2350
• Status: Published

Opinion

MEMORANDUM OPINION

THOMAS F. HOGAN, District Judge.

Two unions seek to enjoin compulsory random urinalysis drug testing of certain civilian employees of the Department of the Army. These consolidated actions are before the Court on plaintiffs' application for an expanded preliminary injunction and defendants’ motion for summary judgment. As an employer, the government has a compelling safety interest, in some instances, in maintaining a drug-free workplace; it has not demonstrated, however, that urinalysis drug testing is capable of showing whether an individual is impaired by drugs or under the influence of drugs. Under Jones v. McKenzie, 833 F.2d 335, 340-41 (D.C.Cir.1987), urinalysis drug testing lacks the necessary nexus to the employer’s safety concern to satisfy the Fourth Amendment. The government’s nonsafety interests in maintaining a drug-free civilian work force are not sufficiently compelling to justify the substantial intrusion of mandatory, random urinalysis. Thus, the defendants have failed to show that the Army’s civilian testing program is consistent with the Fourth Amendment. Accordingly, the Court shall deny defendants’ motion for summary judgment and grant plaintiffs’ application for an expanded preliminary injunction.

The Court is aware that its decision cannot be squared with Mullholland v. Department of the Army, 660 F.Supp. 1565 (E.D.Va.1987) (upholding random urinalysis drug testing of civilian employees at Army air base), appeal docketed, No. 87-2145 (4th Cir. Aug. 13, 1987), and American Federation of Government Employees v. Dole, 670 F.Supp. 445 (D.D.C.1987) (upholding random urinalysis drug testing of employees of Department of Transportation), appeal docketed, No. 87-5417 (D.C.Cir. Dec. 11, 1987). On the other hand, it is in conformity with Thomson v. Weinberger, No. R-87-393 (D.Md. Feb. 27, 1987) (granting preliminary injunction to two civilian employees of Department of Army), and Railway Labor Executives’ Association v. Burnley, 839 F.2d 575 (9th Cir.1988) (holding unconstitutional federal regulations mandating drug tests of railroad employees after certain train accidents). To date, courts have confronted the difficult issues posed by random drug testing without benefit of direct guidance from the Supreme Court, though the Court recently granted certiorari in National Treasury Employees Union v. Von Raab, 816 F.2d 170 (5th Cir.1987), cert. granted, — U.S.-, 108 S.Ct. 1072, 99 L.Ed.2d 232 (1988) (No. 86-1879). The Court invites defendants to appeal this decision pursuant to 28 U.S.C. § 1292(a)(1) (1982); additionally, the Court respectfully suggests that the Court of Appeals consolidate the appeal in this action with that in American Federation of Government Employees v. Dole and consider scheduling the cases for en banc hearing as presenting a question of exceptional importance. See Fed.R.App.P. 35(a)(2).

I. Procedural Background

The Army’s program of random urinalysis drug testing of civilian employees has been the subject of no less than six court challenges since it was devised in 1986. Three of those suits have been consolidated into the present action. 1

The first of these consolidated cases was filed in this district under the name National Federation of Federal Employees v. Weinberger, Civil Action No. 86-0681. Plaintiffs are the National Federation of Federal Employees (NFFE), a labor organization whose membership includes civilian employees of the Department of the Army; NFFE Local 2058, which represents a bargaining unit of 190 civilian guards employed by the Army at the Aberdeen Proving Ground in Maryland; and Charles W. Jackson, a civilian Aberdeen guard and president of Local 2058. Defendants are the Secretary of Defense, the Secretary of the Army, and the commanding officer of Aberdeen Proving Ground.

This Court on June 23, 1986, denied an application for a preliminary injunction and dismissed the case, holding that it lacked subject matter jurisdiction when the union had not yet pursued its challenge before an appropriate administrative tribunal. National Federation of Federal Employees v. Weinberger, 640 F.Supp. 642 (D.D.C.1986). The Court of Appeals for the District of Columbia Circuit reversed that decision on May 15, 1987. National Federation of Federal Employees v. Weinberger, 818 F.2d 935 (D.C.Cir.1987) (Edwards, J.). The Court of Appeals declined to decide the Fourth Amendment issues, remanding the case for findings of fact with respect to the nature and scope of the drug testing program. Id. at 942. The Court of Appeals did provide guidance on the constitutional issues, however, in the form of “general principles.” Among them is the holding that compulsory urinalysis of public sector employees is a “search and seizure” within the meaning of the Fourth Amendment. Id. at 942-43. The Court of Appeals also provided a framework for determining the reasonableness of compulsory urinalysis, id., which of course will be utilized here.

The second case was filed as American Federation of Government Employees v. Weinberger in the Southern District of Georgia. Plaintiffs are the American Federation of Government Employees (AFGE), a labor organization representing civilian employees of the Department of the Army; AFGE Local 1922, which represents 2,200 civilian employees at Fort Stewart, Georgia; and William Cox, Thomas R. Daniels, Joseph Lane, and James Johnson, civilian police officers employed by the Army at Fort Stewart. Defendants are the Secretary of Defense, the Secretary of the Army, and the commander of operations at Fort Stewart. On December 2,1986, Judge Edenfield issued a preliminary injunction against urinalysis drug testing of any civilian police officer at the Fort Stewart/Hunter Army Airfield military installation absent reasonable suspicion that the employee has engaged in drug use. American Federation of Government Employees v. Weinberger, 651 F.Supp. 726 (S.D.Ga.1986). The court held that urinalysis drug testing without individualized suspicion is unreasonable under the Fourth Amendment. Id. at 733. Plaintiffs had requested a nationwide injunction, but Judge Edenfield saw fit to limit his order geographically so as not to conflict with this Court’s earlier decision in National Federation of Federal Employees v. Weinberger, 640 F.Supp. 642 (D.D.C.1986). On June 29, 1987, after the Court of Appeals for the District of Columbia Circuit ruled in National Federation of Federal Employees v. Weinberger, 818 F.2d 935 (D.C.Cir.1987), the Georgia District Court, sua sponte, ordered American Federation of Government Employees v. Weinberger transferred to this Court. The case was assigned Civil Action No. 87-1797.

The third case was filed as American Federation of Government Employees v. Weinberger in the Eastern District of California. Plaintiffs are the American Federation of Government Employees (AFGE); AFGE Local 1546, which represents 355 civilian employees at Sharpe Army Depot in Lathrop, California; and Rosemarie Alyce Duff, Dewey Lee, Monseis M. Ramos, Ronald George Spooner, and Daniel I. Straight, civilian guards and police officers at Sharpe Army Depot. Defendants are the Secretary of Defense, the Secretary of the Army, and the commander of Sharpe Army Depot. Plaintiffs sought a nationwide injunction against testing of civilian employees. The court declined to address the application for a preliminary injunction, instead granting on August 4, 1987, defendants’ motion to transfer the case to this Court, where it was assigned Civil Action No. 87-2350.

The three cases were consolidated in the present action and the Court received extensive briefing, reviewed voluminous exhibits, and heard oral argument oh plaintiffs’ motion for an expanded preliminary injunction and defendants’ motion for summary judgment on September 16, 1987.

II. Facts

A. Directive 1010.9

On April 8, 1985, the Department of Defense issued Directive 1010.9 authorizing each military department to establish a Civilian Employees Drug Abuse Testing Program. Under the Directive, civilian employees in “critical jobs” and applicants for those jobs may be required to participate (and to sign a form agreeing to participate) in urinalysis drug testing in the following four circumstances: 1) before appointment or selection; 2) periodically thereafter “on the basis of neutral criteria”; 3) when there is probable cause to believe that the employee is “under the influence of a controlled substance 2 while on duty”; and 4) in the course of investigating an accident “for the purpose of accident analysis and the development of countermeasures.”

Jobs could be designated as “critical” only if they fell in a category deemed “sufficiently critical to the DoD mission or protection of public safety that screening to detect the presence of drugs is warranted as a job-related requirement”: 1) jobs in law enforcement; 2) “[pjositions involving the national security or the internal security of the Department of Defense in which drug abuse could cause disruption of operations, destruction of property, threats to the safety of personnel, or the potential for unwarranted disclosure of classified information”; 3) jobs involving protection of property or persons from harm.

The program has three stated purposes:

1.Assist in determining fitness for appointment or assignment to, or retention in, a critical job.

2. Identify drug abusers and notify them of the availability of appropriate counseling, referral, rehabilitation, or other medical treatment.

3. Assist in maintaining the national security and the internal security of the Department of Defense by identifying persons whose drug abuse could cause disruption of operations, destruction of property, threats to the safety of themselves and others, or the potential for unwarranted disclosure of classified information through drug-related blackmail.

The Department of Defense’s initiative is not rooted in the discovery of any particular drug problem among its civilian employees or any group of those employees. Rather, the program is a result of two factors. The first is the success of the Department of Defense in curbing illegal drug use among military personnel through urinalysis drug testing, including random urinalysis. Defendants cite surveys and studies showing a marked decrease in drug use after the uniformed services began widescale urinalysis testing. 3 In 1980, 27 percent of military personnel reported nonmedical drug use within the past 30 days; in 1982 the figure declined to 19 percent and in 1985 the percentage dropped to 9 percent.

The second factor leading to the Department of Defense civilian drug testing program was the general national concern with the problem of illegal drug use. It is beyond dispute that the nation suffers immensely from illegal drug use and that workers impaired by drugs on the job are less safe, reliable, and productive. The defendants do not contend that this problem is particularly prevalent among federal workers 4 or civilian employees of the De partment of Defense. For example, a 1983 report based on a survey of 7,000 randomly selected civilian employees of the Department of Defense showed only 4 percent of those employees used drugs for nonmedical purposes within the past 30 days. Even the anecdotal evidence of on-duty drug abuse by civilian employees is sparse and unconvincing. 5 Additionally, the results of compulsory random urinalysis fail to demonstrate a serious drug abuse problem among civilian employees of the Army. In the six-month period ending March 31, 1987, the Army tested the urine of 5,397 employees covered by the Civilian Employees Drug Abuse Testing Program. A total of 37 positive results were obtained, or .68 percent. 6 Thirty employees tested positive for marijuana metabolites; four tested positive for cocaine metabolites; and three tested positive for both marijuana and cocaine metabolites.

B. Executive Order No. 12,564

The Department of Defense authorized its Civilian Employees Drug Abuse Testing Program, and the Army implemented it, at a time when the nation was focusing increased attention on the problems caused by illegal drugs. On March 3, 1986, the President’s Commission on Organized Crime, citing the failure of law enforcement efforts to reduce the supply of drugs, recommended “suitable drug testing” for all federal employees and employees of federal contractors as part of a renewed effort to reduce the demand for illegal drugs. President’s Commission on Organized Crime, America’s Habit: Drug Abuse, Drug Trafficking, and Organized Crime 450, 483 (1986). On September 14, 1986, the President and First Lady addressed the nation on television, launching a “national crusade” against drug abuse. President’s Address to the Nation: National Campaign Against Drug Abuse, 22 Weekly Comp. Pres. Doc. 1183, 1185 (Sept. 14, 1986).

The next day, September 15, 1986, President Reagan signed Executive Order No. 12,564, “Drug-Free Federal Workplace,” 3 C.F.R. 224 (1987), directing all agency heads to develop programs to test “for the use of illegal drugs by employees in sensitive positions,” id. at 226, and condemning the use of illegal drugs “on or off duty.” Id. at 225. The President stated that drug testing “shall not be conducted pursuant to this Order for the purpose of gathering evidence for use in criminal proceedings.” Id. at 228. He added: “Agencies are not required to report to the Attorney General for investigation or prosecution any information, allegation, or evidence relating to violations of Title 21 of the United States Code received as a result of the operation of drug testing programs established pursuant to this Order.” Id. (emphasis added.) 7

Some members of Congress expressed misgivings about the scope of the testing of federal employees contemplated by Executive Order No. 12,564. As part of a compromise, Congress delayed widespread testing, but exempted a number of agencies, including the Department of the Army, the Department of Transportation, some employees of the Department of Energy, the Central Intelligence Agency, the Bureau of Prisons, the Drug Enforcement Agency, the Federal Bureau of Investigation, the Immigration and Naturalization Service, the Customs Service, the Bureau of Alcohol, Tobacco, and Firearms, and the Secret Service. Supplemental Appropriations Act of 1987, Pub.L. No. 100-71, § 503, 101 Stat. 391, 468-71; see Wash. Post., Sept. 11,1987, at Al, col. 5, A18, col. 6. The Department of Transportation began random testing of half of its 60,000 civilian employees on September 10, 1987. A court challenge to the testing program was rejected. American Federation of Government Employees v. Dole, 660 F.Supp. 445 (D.D.C.1987), appeal docketed, No. 87-5417 (D.C.Cir. Dec. 11, 1987). The Department of Justice is preparing to test up to 60,000 of its employees. 8 After a series of delays, government-wide random drug testing of federal workers in “sensitive” positions could go into effect as early as May. 9

C. Interim Change 111 to Army Regulation 600-85

On February 10,1986, the Department of the Army promulgated regulations implementing Department of Defense Directive 1010.9. Army Regulation 600-85, Interim Change No. Ill (Feb. 10, 1986). Interim Change No. Ill amends the existing Army Regulation 600-85 to include drug testing of civilians. It requires employees in critical jobs, as well as applicants for those jobs, to sign DA Form 5019-R, titled “Condition of Employment for Certain Civilian Positions Identified as Critical Under the Drug Abuse Testing Program.” Form 5019-R states:

SECTION A — REQUIREMENTS

As a prospective or current employee in a position designated by the Department of the Army and approved by the Office of the Secretary of Defense as critical to national or internal security or to the protection of persons or property, you are required to read and sign this statement as a condition of employment. If you are an applicant for a critical job and fail to sign this agreement, you will not be selected for the position. If you are currently in a critical job and refuse to sign the condition of employment, you will be voluntarily or involuntarily reassigned or demoted to a noncritical job or separated from Federal employment. If you sign the condition and later refuse to submit to urinalysis testing, you will be non-selected, reassigned, demoted, or separated according to applicable regulations. To verify that you are not currently using drugs, you will be required, as a condition of your continued employment, to submit a urine sample for testing purposes; (1) periodically, on an unannounced basis, (2) when there is probable cause to believe that you are under the influence of drugs, and/or (3) when there is a mishap or safety investigation being conducted in relation to an accident involving government-owned vehicles, aircraft, or equipment. To assure the validity of these tests, a staff member of the same sex will observe you while you are providing the sample. Detection of drug usage through confirmed positive urinalysis test results may be cause for a determination that you have failed to meet the conditions necessary for your continued employment in the position. Medically prescribed drugs authorized by a physician and confirmed by appropriate evidence are excluded from such determinations. The results of urinalysis will be used only for clinical and necessary administrative purposes. You are entitled to any additional and reasonable information or clarification you desire prior to signing the agreement. A copy of the signed agreement will be given to you and your supervisor. The original will be placed in your Official Personnel Folder.

SECTION B — AGREEMENT

This is to certify that I understand the contents of the policy described above and the reasons therefor, and that I agree to adhere to the terms of this policy as a continuing condition of my employment in positions to which this agreement applies.

Interim Change No. Ill states that at least 90 days before the initial urinalysis test, each employee holding a critical job must be informed of the reasons for the test and the consequences of failing the test or refusing to take it, including “adverse action.” Additionally, employees must be told that they may submit supplemental medical documentation concerning the legitimate use of drugs, and that drug abuse counseling and referral services are available.

The regulation permitted “field testing” of urine samples but specified that positive results from field tests of current employees would be considered preliminary until confirmed by subsequent laboratory testing or an admission of the employee. Only “temporary” personnel action may be taken on the basis of a positive result from a field test. The action must be rescinded if the test result is not confirmed by laboratory testing or an admission of the employee.

The regulation gives broad discretion to unit commanders to decide whom to test:

The decision to require an individual covered by the biochemical testing program to undergo such testing to detect drug abuse is the commander’s prerogative. The management of available quotas, both for field and laboratory tests, is the commander’s responsibility. He or she must decide which segments of the total population, civilian and military, are most at risk and allocate quotas accordingly. Beyond the pre-accession test for civilian employees in critical positions, subsequent testing is left to the commander’s discretion.

Interim Change No. Ill listed five categories of jobs designated as “critical” and subject to urinalysis testing:

1) Aviation positions: air traffic controller, pilot, aircraft engine mechanic, aircraft overhaul specialist, prop and motor mechanic, aircraft mechanic, and aircraft servicer.

2) Guard and police positions: guard, police officer, criminal investigator, and correctional officer.

3) Personnel Reliability Program personnel: chemical and nuclear surety positions.

4) Alcohol and Drug Abuse Prevention and Control Program direct service staff.

5) All employees at Army forensic drug testing laboratories.

Defendants estimate a total of 9,400 of the Army’s nearly 450,000 civilian employees are subject to random drug testing, as follows: 2,800 in aviation positions; 3,700 in guard and police positions; 2,250 in the Personnel Reliability Program; and 650 in drug abuse and drug testing positions.

According to Raymond J. Sumser, the Army’s Director of Civilian Personnel, a total of 65 job classes covering 20,000 employees were nominated by unit commanders as “critical” for purposes of drug testing. The Department of De fense reviewed the nominations and approved 16 job classes covering 9,400 employees as “critical” within the meaning of Department of Defense Directive 1010.9.

Sumser’s declaration of August 20, 1987, offered explanations of why each position was selected for the Civilian Employees Drug Abuse Testing Program:

Air traffic controllers — Air traffic controllers provide information to pilots concerning weather, air navigation, and airfield conditions before and during flight. They provide information to pilots in distress, orient lost pilots, and initiate search and rescue action to locate overdue aircraft; control aircraft en route and near airfields to assure proper separation between aircraft and minimize delays caused by congestion. According to Sumser, air traffic controllers “must have the ability to act decisively under stressful situations and maintain alertness over sustained periods of pressure. Controllers must be able to coordinate plans and actions and direct aircraft operating at very high speeds and exercise sound judgment to select and take the safest and most effective courses of action among several available choices.”

Pilots — Pilots, of course, fly aircraft; civilian pilots employed by the Army are responsible for transporting passengers and cargo, including weapons, munitions, and other hazardous materials. “Performance of duties requires extreme precision in taxiing, taking off, climbing, hovering, approaching, and landing using visual and instrument controls. In flight, pilots must closely monitor sophisticated hydraulic, electrical, pressurization, navigation, and communications systems, space patterns, and fuel consumption.”

Aircraft mechanics — Aircraft mechanics inspect, adjust, and repair aircraft systems, assemblies, and surfaces. “Duties require visual checks of components, parts, and systems to determine the need for repair or replacement of parts and the extent of adjustment and alignment required, as well as skills in the use of fixtures, templates, precision dial and feeder gauges, test stands, and cockpit instruments.”

Aircraft attendants — Aircraft attendants meet and guide aircraft to parking areas, secure aircraft with wheel chocks, ground wires, lock pins, and engine covers, refuel aircraft, operate ground support equipment such as auxiliary power units, and service aircraft with oil, hydraulic fluid, compressed air, nitrogen, and liquid oxygen. “Aircraft attendants must be skilled in safe work procedures involving hazardous operations such as engine starts and shutdowns and refueling aircraft while engineers are operating (‘hot’ refuels). Attendants must exercise care to ensure that fuel, hydraulic fluid, engine oil, and other vital fluids replenished during servicing procedures meet the exact technical specifications for the various types of aircraft being serviced.”

Sumser stated that the duties of aviation personnel require “keen sensory perception and neuromuscular coordination. Drug use could easily have an adverse impact on these faculties and increase the potential for undue distractions, carelessness, and accidents____ Even slight impairment could reduce the ability of an employee occupying one of these positions to recognize and react to potentially threatening circumstances and materially lessen the chance of preventing accidents.” He concluded that drug use by these employees could result in “disruption of operations, damage or destruction of extremely high-value equipment, personal injury, and loss of life.”

It is apparent from the Army’s rationale for testing these aviation employees that the government’s interest is safety and its safety concern is identifying employees who are impaired by drugs on duty.

The next category of “critical” positions is law enforcement. Sumser stated that civilian guards and police employed by the Department of the Army perform typical law enforcement duties and are responsible as well for the protection of Army installations from the hazards of espionage, sabotage, fire, theft, and accidental or willful damage or destruction. Frequently this property includes very dangerous weapons and facilities, including nuclear and chemical, munitions, and electronics and communications systems of vital importance. Additionally, these employees perform courier duties or otherwise have access to classified information. Most guards are required to possess at least a “secret” security clearance. On-duty security personnel are armed, some with automatic weapons, and are authorized to use deadly force in certain circumstances. Their duties include enforcement of drug laws. “Obviously, great doubt is cast on the ability of a law enforcement official who personally uses illegal substances to perform required drug suppression duties.”

Some law enforcement officials must perform their duties in relative isolation, “without close and continuous supervision,” Sumser stated. “These employees must, at all times, be able to think rationally and respond intelligently despite adverse circumstances and deal effectively in interpersonal confrontations. The inability to fully perform their duties due to drug use could readily create the possibility of theft, security breakdown, damage to or destruction of sensitive government property, personal injury to self and others, and loss of life.”

From this rationale for testing civilian law enforcement employees, it is possible to identify three government interests. The first is safety, and again the government’s safety concern is identifying employees who are impaired by drugs on duty. The second concern is security, and the government's security concern is identifying employees who use drugs on duty, or off duty because of the possibility of “drug-related blackmail.” The third concern might be labeled integrity, and this concern is identifying employees who use drugs either on duty or off duty.

Third, employees who receive special access certifications as part of the Army’s Personnel Reliability Program (PRP) are automatically subject to drug testing. These jobs include supervisory personnel who administer the Personnel Reliability Program, nuclear reactor operators, nuclear weapons technicians, chemical ammunition maintenance specialists, quality assurance personnel, materiel handlers, laboratory workers, and intrusion detection system maintenance personnel. Additionally, “other employees who require routine access to surety material or require routine entry into chemical or nuclear exclusion areas” are included; for example, a secretary might be included in the Personnel Reliability Program not because of his or her duties but because the workplace was within a secured area.

“Personnel occupying PRP positions are responsible for the custody, transportation, storage, maintenance, demilitarization, and security of nuclear and chemical surety materials. Failure to perform these responsibilities increases the risk of untoward incidents, unnecessary costs, and unsafe work environments. This could lead to the definite potential for unthinkable danger to the public, destruction of property, mission degradation, and major security problems. Drug abuse by personnel in these jobs poses a serious threat to the individual, fellow workers, and the surrounding community. Disruption of operations critical to national defense readiness, destruction of property, injury to persons, and disclosure of highly classified information could occur.”

The first rationale for drug testing of PRP employees is safety, and the government’s safety interest is identifying employees impaired by drugs on duty. The second interest is security, which is served by identifying employees using drugs on duty or off duty.

Finally, civilian employees involved in drug and alcohol counseling, employees in the chain of custody of samples taken for testing, and employees of Army drug testing laboratories are subject to urinalysis drug testing. “Obviously, testing of these personnel is essential to the continued credibility and integrity of the ADAPCP [Alcohol and Drug Abuse Pre vention and Control Program] and the Army’s biochemical testing program. If ADAPCP staff are to function properly in assisting with identification and treatment of military and other civilian personnel with substance abuse problems, they must be certifiably drug free.”

The rationale for testing these employees is integrity, which is served by identifying those who use drugs on duty or off duty.

Sumser’s declaration makes clear that jobs were designated as “critical” (and therefore subject to drug testing) solely on the basis of their duties and the safety, security, and integrity interests implicated. The designations were not based on drug-related problems that had arisen among particular employees or any suspicion concerning drug use by a group of employees. For example, the defendants cite no examples of “drug-related blackmail,” though reference is made to it in Directive 1010.9 and the defendants’ pleadings.

D.Changes in Interim Change No. Ill

The Army made two significant changes in its civilian drug testing program after the issuance of Interim Change No. Ill on February 10, 1986. The first followed Executive Order No. 12,564’s requirement that “[procedures for providing urine specimens must allow individual privacy, unless the agency has reason to believe that a particular individual may alter or substitute the specimen to be provided.” 3 C.F. R. 224, 226-27 (1987). On October 10, 1986, the Pentagon established procedures conforming to the presidential order and notified unit commanders as follows:

A. Biochemical testing will be done with maximum respect and concern for human dignity.

B. The term “biochemical test monitor (BTM) will be used instead of “observer” for the purpose of the civilian urinalysis program. The BTM has the responsibility to ensure that the sample is not contaminated or altered in any way.

C. Care must be taken to ensure that the individual cannot access a container previously filled with a foreign substance or a prepared sample. For example, employees/applicants should not be allowed to carry such items as briefcases, carry bags, or purses into the collection area, coats for all employees/applicants should be left outside, sleeves should be rolled up, and a visual check of potential hiding places in the stall or collection area should be made prior to giving the employees/applicants the collection bottle.

D. Individual privacy shall be assured by permitting employees or applicants to produce urine specimens alone in closed stalls or other similar structures or enclosures. There will be no observation of any employee’s or applicant’s private anatomy. The BTM will maintain control of the integrity of the specimen by listening for normal sounds associated with urination and by oversight of the setting.

E. When the bottle is returned to the BTM, he or she will verify the normal warmth and appearance of the sample.

F. If at any time during the testing procedure the BTM has reason to believe or suspect that an employee or applicant is attempting to adulterate or substitute the sample, he or she should stop the procedure and inform the installation biochemical testing coordinator (IBTC) of that individual’s conduct. The IBTC will then determine if direct observation is required____

On March 19, 1987, the Pentagon revoked the authorization for field testing of urine samples of employees. It ordered that “[a]ll samples of civilian employees will now be forwarded directly to the forensic toxicology drug testing laboratory (FTDTL) for testing.” The sole exception permitting continued use of field testing as a pre-screen is for persons selected for critical positions; a positive result would delay hiring until the laboratory test was conducted. The field testing used by the Army was the Enzyme Multiplied Immunoassay Technique (EMIT) System, marketed by Syva Company. The Army laboratories use a radioimmunoassay (RIA) test and a gas chromatography/mass spectro metry (GC/MS) test, which is considered the most reliable and acceptable method of testing for drugs or drug metabolites.

E. Uses of Drug Test Results

The Army considers its civilian drug testing program “non-punitive and non-disciplinary in nature,” in the words of Samuel S. Horn, Deputy Chief of the Labor and Civilian Personnel Law Office, Office of the Judge Advocate General, Department of the Army. Nonetheless, failure to take the test or a positive result could result in dismissal.

Horn explained in an August 11, 1987, declaration that a current employee in a critical position who does not want to be subject to drug testing could request reassignment to a noncritical position for which he or she was qualified. “If there is no noncritical job available at the same grade level, the employee may be offered a job at a lower grade level. In the event there is no alternative noncritical position available into which the employee may be assigned, the employee may be separated from the service.”

Nor does a positive urinalysis result in automatic dismissal, although dismissal may result. An employee will be offered drug counseling or treatment and will be “reassigned to an available noncritical position for which he or she is qualified, with no loss of pay or benefits. If no noncritical positions are available at the same grade level, the employee will be offered a position at a lower grade if such a position is available. Only if no alternative noncritical positions are available will the employee be separated for failure to meet a condition of employment.”

Horn also stated that employees who successfully complete a treatment and rehabilitation program are eligible to return to their former jobs. The report for the six-month period ending March 31, 1987, attached to Horn’s declaration, shows that of the 37 employees who tested positive, four resigned; five employees were fired, two because of a “second positive,” two for “other reasons,” and one who was a temporary employee; the other 28 employees were detailed or reassigned, with 14 still in rehabilitation, eight refusing rehabilitation, and six successfully rehabilitated and returned to a critical job.

F. Army Test Procedures

The Army has abandoned EMIT field testing of urine samples of current employees, and instead relies on a two-tier laboratory testing procedure. Lieutenant Colonel John S. Jewell, who holds a Ph.D. in organic chemistry and serves as consultant to the Surgeon General of the Army for drug laboratory operations, said in an August 10, 1987, declaration that the initial test used by all military laboratories is the radioimmunoassay (RIA) test, which is based on immunological principles. If the RIA test is positive, a gas chromatography/mass spectrometry (GS/MS) test is conducted. A positive result is issued only after both tests have confirmed the presence of the drug or drug metabolite above an administratively defined cutoff level. The administrative cutoff level is designed to screen out positive results that could be associated with passive inhalation or “non-participative ingestion” of drugs.

Jewell offered this explanation of how the tests work:

The RIA procedure (Abuscreen) is based on a physical reaction between antibodies (substances manufactured in the bodies of animals or humans to assist the bodies’ elimination and/or detoxification of these substances) and antigens (drugs) that causes physical separation of these substances from urine. The end point consists of measuring a decrease in the radioactivity found as a result of competitive binding reactions between naturally occurring drugs and drugs modified with radioactive Iodine-125. The RIA procedure correctly identifies the presence of drug or drug metabolites greater than 99% of the time.

[] The GC/MS procedure combines both the physical and chemical properties of the molecule to be identified with the analytical capabilities of the instrument. The first part of the procedure relies on the physical and chemical properties of the drug or drug metabolite to separate them from the urine matrix. Then the drug or drug metabolite is usually taken through a chemical derivitization that makes use of the chemical properties of the compound to separate it from other molecules of different chemical structure. Then the derivatized drug or drug metabolite of interest is separated from all other molecules by its physical properties using gas chromatography and an absolute identification is made through a mass spectrometer fragmentation pattern. This procedure is recognized as the state-of-the-art method and is 100% accurate for chemical, drug or drug metabolite identification. The GC/MS identification has been accepted by the scientific community as being the most reliable and acceptable method for drug or drug metabolite identification.

The tests detect the presence of drugs in some cases, drug metabolites in others. A drug metabolite is an inactive product resulting from the body’s breakdown of a drug. Cocaine is generally detected by its metabolite, benzoylecgonine; the active ingredient of marijuana, delta-9-tetrahydrocannabinol (THC), is also detected by its metabolites, principally THC-9-carboxylic acid (THC-acid). Morphine, phencyclidine (PCP), amphetamines, barbiturates, and LSD are excreted directly into the urine. Declaration of Robert E. Willette, Ph.D., June 3, 1987.

Even “state-of-the-art” tests have severe limitations; for example, defendants’ counsel informed the Court that the Army no longer tests civilians for opiates because consumption of a quantity of poppy seeds before testing produces a “false positive” test result for opiates. In other words, the most technologically advanced urine test cannot distinguish between a heroin user and a person who eats poppy seed strudel.

More importantly, it is beyond dispute that urinalysis drug testing, whether by the EMIT method, the RIA method, or the GS/MS method, is not capable of determining whether a person is impaired or intoxicated by drugs. Courts focusing on the issue have agreed that urinalysis cannot show impairment, 10 and government counsel have conceded the point at oral argument and in their pleadings and submissions. Willette provides one reason for this incontrovertible fact: “Unfortunately, the relationship between concentrations of certain drug metabolites in urine and their parent drugs in the body is obscured by fluctuations in urine flow, sometimes in differences in metabolic rates, and other individual characteristics.” Only a blood test can show a correlation between the effects of drugs and their concentrations in bodily fluids. 11 As the testing technology now stands, tests for only one substance can show impairment; that substance is alcohol.

The National Institute on Drug Abuse of the U.S. Department of Health and Human Services states flatly that urinalysis testing can detect only prior drug use, not present impairment. “Although urine screening technology is effective in determining prior drug use, positive results of a urine screen do not prove intoxication or impaired performance. Drugs or their metabolites may appear in urine for several days, even weeks (depending upon the drug), without apparent impairment.” National Institute on Drug Abuse, Consensus Summary: Interdisciplinary Approaches to the Problem of Drug Abuse in the Workplace 6 (1986). 12

G. Alternatives to Urinalysis

Defendants have presented no evidence that alternative methods of detecting drug use have been seriously considered or attempted. Defendants merely note that the nature of some jobs makes close supervision difficult because the employees often work alone. They also point out that drug tests result in some positive results when a trained observer would detect no impairment; in light of the foregoing that is both not surprising and not supportive of the government’s position. Executive Order No. 12,564 requires that each agency’s drug testing plan include “[sjupervisory training to assist in identifying and addressing illegal drug use by agency employees.” 3 C.F.R. 224, 226 (1987). However, this requirement is not incorporated in Department of Defense Directive 1010.9 or Interim Change 111 to Army Regulation 600-85.

Alternative methods of detecting drug use by employees are suggested by the plaintiffs, and criticized by the defendants, in the pleadings before the Court. The Court also finds helpful Judge Getzendanner’s discussion of alternative methods of detection in Taylor v. O’Grady, 669 F.Supp. 1422, 1431-33 (N.D.Ill.1987) (enjoining compulsory urine testing of county correctional officers).

The first alternative is observation by trained supervisors, which can detect chronic drug use and some on-duty drug use; it cannot detect off-duty use that does not manifest itself on the job. The second alternative is neurobehavioral testing, or testing the behavioral output of the brain to determine whether a worker’s thinking or behavioral processes are impaired. There is no evidence in the record before’ the Court that defendants have seriously considered either alternative.

III. Conclusions of Law

Plaintiffs raise a number of constitutional and statutory issues in addition to their main challenge under the Fourth Amendment. The Court finds it unnecessary to address those issues at this juncture because Fourth Amendment considerations entitle plaintiffs to injunctive relief.

A. General Fourth Amendment Principles

The Fourth Amendment provides:

The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.

It restricts the activities of civilian authorities as well as law enforcement officers, New Jersey v. T.L.O., 469 U.S. 325, 335-37, 105 S.Ct. 733, 740-41, 83 L.Ed.2d 720 (1985), and applies to the conduct of governmental employers. O’Connor v. Ortega, — U.S.-, 107 S.Ct. 1492, 1497, 94 L.Ed.2d 714 (1987) (plurality opinion).

Fourth Amendment does not restrict all searches and seizures by the government, but it does prohibit all unreasonable searches and seizures. Fourth Amendment analysis proceeds in two steps. First, the Court must determine whether the government’s conduct constitutes a search or seizure by infringing upon a reasonable expectation of privacy. Compulsory urinalysis of public employees is a “search and seizure” in this Circuit. Jones v. McKenzie, 833 F.2d 335, 338 (D.C.Cir.1987); National Federation of Federal Employees v. Weinberger, 818 F.2d 935, 942 (D.C.Cir.1987) (NFFE). Second, if a search or seizure occurred, the Court must determine whether it was reasonable.

The interest served by the Fourth Amendment is the protection of the “privacy and security of individuals against arbitrary invasions by government officials.” Camara v. Municipal Court, 387 U.S. 523, 528, 87 S.Ct. 1727, 1730, 18 L.Ed.2d 930 (1967). To safeguard this interest, “one governing principle, justified by history and by current experience, has consistently been followed: except in certain carefully defined classes of cases, a search of private property without proper consent is ‘unreasonable’ unless it has been authorized by a valid search warrant.” Id. at 528-29, 87 S.Ct. at 1730-31. Even where warrantless searches are permitted, they ordinarily “must be based upon ‘probable cause’ to believe that a violation of the law has occurred.” T.L.O., 469 U.S. at 340, 105 S.Ct. at 743.

However, neither a warrant nor probable cause is an irreducible requirement of a valid search and seizure. “The fundamental command of the Fourth Amendment is that the searches and seizures be reasonable, and although ‘both the concept of probable cause and the require ment of a warrant bear on the reasonableness of a search, ... in certain limited circumstances neither is required.’ ” Id. at 340, 105 S.Ct. at 743 (quoting Almeida-Sanchez v. United States, 413 U.S. 266, 277, 93 S.Ct. 2535, 2541, 37 L.Ed.2d 596 (1973) (Powell, J., concurring)).

There are even some rare circumstances where individualized suspicion is not necessary to validate a search. T.L.O., 469 U.S. 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, 49 L.Ed.2d 1116 (1976)), and “[ejxceptions to the requirement of individualized suspicion are generally appropriate only where the privacy interests implicated by a search are minimal____” T.L.O., 469 U.S. at 342 n. 8, 105 S.Ct. at 744 n. 8.

Because the privacy intrusion represented by urinalysis is not “minimal,” Jones v. McKenzie, 833 F.2d at 339 (“strong privacy interests are involved”), the Fourth Amendment requires a heavy presumption against searches not based on individualized suspicion. 13 Only a compelling need of the government as employer could justify dispensing with the requirement of individualized suspicion. Random testing, of course, is based on no suspicion whatsoever, and there exist no grounds at all for suspecting that the search represented by urinalysis will turn up evidence. Indeed, the government’s goal is to obtain no positive results.

The Supreme Court has been guided by a twofold inquiry in determining whether a search without a warrant or probable cause is reasonable: “first, one must consider ‘whether the ... action was justified at its inception.’ Terry v. Ohio, 392 U.S. [1], 20 [88 S.Ct. 1868, 1879, 20 L.Ed.2d 889] [ (1968) ]; second, one must determine whether the search as actually conducted ‘was reasonably related in scope to the circumstances which justified the interference in the first place,’ ibid.” T.L.O., 469 U.S. at 341, 105 S.Ct. at 744.

This requirement that both the inception and scope of the intrusion be reasonable applies to the government workplace. O’Connor v. Ortega, 107 S.Ct. at 1502-03. Determining the “standard of reasonableness applicable to a particular class of searches requires ‘balancing] the nature and quality of the intrusion on the individual’s Fourth Amendment interests against the importance of the governmental interests alleged to justify the intrusion.’ ” Id. at 1499 (quoting United States v. Place, 462 U.S. 696, 703, 103 S.Ct. 2637, 2642, 77 L.Ed.2d 110 (1983)).

The D.C. Circuit elaborated on this inquiry in the context of compulsory urinalysis in NFFE:

On the one side of this balance, in the matter at hand, are the employees’ reasonable expectations of privacy — those expectations which society is “prepared to recognize as legitimate.” On the other side of the balance is “[t]he governmental interest [in] the efficient and proper operation of the workplace.” This balancing inquiry has two reference points: the court must determine first “whether the [search] was justified at its inception” — ie., whether “reasonable grounds [exist] for suspecting that the search will turn up evidence” of work-related drug use; and second, “whether the search as actually conducted ‘was reasonably related in scope to the circumstances which justified the interference in the first place,’ ” — ie., whether “the measures adopted are reasonably related to the objectives of the search and not excessively intrusive.”

NFFE, 818 F.2d at 942-43 (citations omitted). 14

B. Jones v. McKenzie

The D.C. Circuit, again speaking through Judge Edwards, revisited drug testing of public employees recently in Jones v. McKenzie, 833 F.2d 335 (D.C.Cir.1987). While the factual situation was somewhat different, the case has elements in common with the case sub judice. Jones v. McKenzie involved drug testing in the context of a regular medical examination, not random testing, of school bus drivers, mechanics, and attendants employed by the District of Columbia Public School System. The testing program arose from a perceived “drug culture” among the employees, and the record contained repeated incidents of “bizarre or dangerous drug-related behavior by drivers and attendants while omduty.” Id. at 336. Syringes and bloody needles were found in restrooms, and the head of the Transportation Branch estimated 60 percent of the employees were using narcotics. In response to this perceived problem, the School System decided to include drug testing in its medical examinations, and so notified the employees. Id. at 336— 37.

The purpose of the tests was to enforce a long-standing directive prohibiting employees “to possess, use or be under the influence of intoxicating liquors, narcotics, or other drugs such as LSD, marijuana and the like, while on school premises.” Id. at 337. Employees were required as part of medical examinations to provide urine specimens in the privacy of a restroom. The samples were tested with the EMIT Cannabinoid Urine Assay. An employee who was fired after a positive test for THC metabolites filed suit in district court. The court held the testing was an unreasonable search “in the absence of particularized probable cause.” Jones v. McKenzie, 628 F.Supp. 1500 (D.D.C.1986) (order).

The Court of Appeals reversed as to the requirement of “probable cause,” holding that testing could be required when safety interests were high, the testing was part of a routine, employment-related medical examination, and “the test employed is one that has a nexus to the employer’s legitimate safety concern.” Jones v. McKenzie, 833 F.2d at 340.

The court found, or rather found the School System had conceded, that the EMIT test used “is not a valid measure of whether the subject is in possession of, is using, or is under the influence of illicit drugs at the time of the test. As this test therefore lacks a sufficient nexus to the appellant’s legitimate concern that its employees not possess, use or be under the influence of drugs while on duty, it is clear that the School System could not constitutionally test its employees for drugs in the manner Jones was tested____” Id. at 339.

The more sophisticated RIA and GS/MS tests used by the Department of the Army are perhaps more reliable than the EMIT test, but they do not reveal any more infor mation about whether a person is impaired by or under the influence of drugs while on duty. Like the EMIT test, they detect only relatively recent drug use.

C. Application of the Law to This Case

1. Justified at Inception

The first prong of the reasonableness test is whether the search is justified at its inception. The government has advanced three interests in compulsory random urinalysis. Only “strong governmental concerns” can outweigh the privacy interests involved here, Jones v. McKenzie, 335 F.2d at 340. The first interest is safety, of the workers, the workplace, and the public. This is an important and serious interest. “The case law on this point is clear that a governmental concern is particularly compelling when it involves the physical safety of the employees themselves or others.” Id. This interest is served by detecting on-duty impairment. At this stage of the scientific development of drug testing, urinalysis testing does not advance this interest because there is no convincing evidence that a positive result can ipso facto be equated to on-duty impairment. Off-duty drug use does not automatically mean on-duty impairment any more than off-duty consumption of alcoholic beverages means on-duty impairment.

The second interest is security. This too can be a compelling interest in the context of the nation’s defense. The government argues that on-duty drug use could lead to breaches of security directly, and that off-duty drug use could lead to breaches of security because of bribery or blackmail possibilities. No evidence correlating off-duty drug use to bribery and blackmail has been produced, and on the facts presented to date the Court considers this relationship speculative. 15

The third interest is integrity. The government argues that those responsible for enforcing the law and conducting the testing program must be “certifiably drug free” if they are to have the respect and confidence of those they police and test, and of the public. Under this rationale all Cabinet officers and the President should undergo random drug testing. The Court finds this interest less compelling than safety and security. The government has presented only supposition but no evidence of any harm that would occur if the personnel were not “certifiably drug free.” This interest, of course, would be advanced by testing for any drug use, off-duty as well as on-duty. 16

On the issue of whether the search represented by compulsory random urinalysis is “justified at its inception,” the Court concludes as follows:

The government’s safety interest, where compelling, justifies testing for on-duty impairment. Safety does not justify testing for off-duty drug use because there is insufficient demonstrable effect on workplace safety.

The government’s security interest, where compelling, justifies testing for on-duty impairment. Security does not justify testing for off-duty drug use because there is insufficient showing that off-duty drug use is related to security.

The government’s interest in integrity as presently articulated as to the job categories under consideration does not justify testing for off-duty drug use because the interest is not sufficiently compelling in the face of the intrusion represented by random urinalysis and because there is insufficient evidence suggesting that employees must be “certifiably drug-free” to perform their duties.

To summarize, compulsory random urinalysis is justified at its inception by compelling safety and security concerns that would be jeopardized if employees were impaired or under the influence of drugs while on duty. Testing for off-duty drug use is not justified at its inception because the government’s interest in integrity, as presented in this record, is not sufficiently compelling and the government’s safety and security concerns are not jeopardized by off-duty drug use with not on-duty consequences.

2. Related in Scope

The next step in the analysis is to determine “whether the search as actually conducted ‘was reasonably related in scope to the circumstances which justified the interference in the first place.’ ” T.L.O., 469 U.S. at 341, 105 S.Ct. at 744 (quoting Terry v. Ohio, 392 U.S. at 20, 88 S.Ct. at 1879).

Here, the search as actually conducted is not reasonably related to the justification, in two ways: first, urinalysis is not “reasonably related” to the justification because, as the technology now stands, it fails to show whether an employee is impaired on the job. This is the same conclusion reached by the Ninth Circuit recently in Railway Labor Executives’ Association v. Burnley, 839 F.2d 575, 588 (9th Cir.1988). In that case, federal regulations mandating drug tests of employees after certain train accidents were struck down as unreasonable under the Fourth Amendment. The court found suspicionless testing constitutionally infirm because the tests do not detect on-the-job drug use:

We see one flaw in the reasonableness of this approach to the problem. Blood and urine tests intended to establish drug use other than alcohol are not reasonably related to the stated purpose of the tests because the tests cannot measure current drug intoxication or degree of impairment.

Id.

Second, urinalysis is “excessively intrusive” because it reveals information in which the employee has a legitimate expectation of privacy concerning his or her personal life while off-duty. 17 In short, the test employed fails to validly detect the activity with which the government is legitimately concerned.

Whether random urinalysis is “excessively intrusive” turns on whether less intrusive but effective alternatives are available. Both trained supervision and neurobehavioral testing can usually detect those who are impaired at work because of drug use, something urinalysis drug- testing cannot do. Thus, to the extent that the legitimate government interest is served by detecting on-duty drug use, less intrusive alternatives make random urinalysis excessively intrusive.

It is worth noting that the random urinalysis testing at issue here is more intrusive than the regular or announced testing as part of an employment-related medical examination reviewed in Jones v. McKenzie. The only “notice” given of the Army testing program is that an employee is henceforth subject to random, unannounced testing; the employee’s only alternative to undergoing testing is giving up the “critical” job. The testing also is more intrusive than that in Jones v. McKenzie because it occurs in the work setting, not in a physician’s office or other medical facility. While the Army’s procedures were changed to provide more privacy in that no visual observation is permitted absent suspicion of tampering, the intrusion can only be minimized, not eliminated. The act of urinating is still “observed” by senses other than sight.

Accordingly, as it now stands absent some scientific advances, urinalysis is both not “reasonably related” to the government’s safety interest and is “excessively intrusive,” and therefore necessarily violates the Fourth Amendment. Illegal drugs are an enormous and dangerous problem in this country. It is with some regret that the Court removes what might be a powerful weapon from the nation’s arsenal in the campaign against illegal drug use, but the values represented by the Constitution and the Fourth Amendment are transcendent.

The Court’s analysis has been guided by the teachings of the Supreme Court on general Fourth Amendment principles and the Court of Appeals for the District of Columbia Circuit on drug testing of public employees. It has striven to be faithful to controlling precedent, even where the Court might have reached a different result were it writing on a blank slate. No attempt has been made to line up the many cases decided by federal and state trial and appellate courts on this difficult subject, first because they are not relevant to the extent that they depart from cases decided by this Circuit, and second because they simply cannot be lined up. Judges across the nation have struggled to balance the public and private interests at stake, often reaching irreconcilable results. The prospect of Supreme Court guidance is a welcome one.

IV. Standards for a Preliminary Injunction

Four factors are to be considered in determining whether a preliminary injunction is warranted: 1) the likelihood that the party seeking the preliminary injunction will prevail on the merits; 2) the likelihood that the moving party will be irreparably harmed absent such relief; 3) the prospect that others will be harmed if the court grants a preliminary injunction; and 4) the public interest in granting the relief sought. Washington Metropolitan Area Transit Commission v. Holiday Tours, Inc., 559 F.2d 841, 842-42 & n. 1 (D.C.Cir.1977); Virginia Petroleum Jobbers v. FPC, 259 F.2d 921, 925 (D.C.Cir.1958).

Here, the Court finds it likely the plaintiffs will prevail on the merits of their Fourth Amendment claim. The law is well settled that the burden is on the government to show that a warrantless search is reasonable under the Fourth Amendment. “Those 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); see also 4 W. LaFave, Search and Seizure § 11.2(b), at 218 (2d ed. 1987) (burden on defendant when attacking search pursuant to warrant; burden on government when warrantless search is attacked). Here, the government has failed to demonstrate that compulsory, random urinalysis is justified at its inception and reasonably related to the objectives and not excessively intrusive.

Without the relief requested, plaintiffs must undergo an unconstitutional search or face transfer, demotion, or even dismissal. No substantial harm to other parties will result from a preliminary injunction because urinalysis drug testing does not detect drug-related impairment on the job and other methods of detecting such impairment are available. Finally, the Court concludes that the public interest lies in enjoining unconstitutional searches.

V. Conclusion

For these reasons, the Court shall preliminarily enjoin the Department of the Army from compulsory random urinalysis of its civilian employees under Army Regulation 600-85. An order consistent with the foregoing accompanies this memorandum opinion.

ORDER

Pursuant to Fed.R.Civ.P. 25(d)(1), it is this 1st day of March, 1988,

ORDERED that Frank C. Carlucci, Secretary of Defense, is substituted as defendant in this consolidated action for Casper Weinberger, who has resigned as Secretary of Defense.

Upon consideration of plaintiffs motion for an expanded preliminary injunction, defendants’ motion for summary judgment, the oppositions and replies thereto, supporting memoranda, oral argument, and the entire record herein, and for the reasons set forth in the accompanying Memorandum Opinion, it is this 1st day of March, 1988,

ORDERED that defendants are hereby restrained from implementing or enforcing Department of Defense Directive 1010.9 and Interim Change No. Ill, amending Army Regulation 660-85 as they pertain to suspicionless drug testing of civilian employees of the Department of the Army. Defendants are specifically enjoined from conducting any urinalysis or other drug test unless the test is based upon a reasonable, articulable, and individualized suspicion that a specific employee is under the influence of drugs or alcohol while on duty; and it is

FURTHER ORDERED that security pursuant to Fed.R.Civ.P. 65(c) is set in the amount of $0.00 and no bond need be posted.

1 . One of the other three cases, filed by a union seeking a nationwide injunction, was dismissed without prejudice on grounds of comity and judicial efficiency. American Federation of Government Employees v. Weinberger, No. C86242T (W.D.Wash. Aug. 5, 1986) [Available on WESTLAW, 1986 WL 15495]. Two cases were filed by individuals: Mullholland v. Department of the Army, 660 F.Supp. 1565 (E.D.Va.1987), appeal docketed, No. 87-2145 (4th Cir. Aug. 13, 1987) (denying injunction sought for 50 employees at Army air base) and Thomson v. Weinberger, 682 F.Supp. 829 (D.Md.1987) (granting preliminary injunction to pipefitter and research biologist).

2 . The Directive defines "controlled substance[s]” as those listed in 21 U.S.C. § 812 (1982 & Supp. Ill 1985).

3 . Compulsory urinalysis of military personnel withstood constitutional challenge in Murray v. Haldeman, 16 M.J. 74 (C.M.A.1983), and Committee for GI Rights v. Callaway, 518 F.2d 466 (D.C.Cir.1975). The Court of Appeals noted “the differences between military and civilian life and the constitutional standards to be applied to each,” 518 F.2d at 474, and said the increased incidence of drug abuse in the Armed Forces “poses a substantial threat to the readiness and efficiency of our military forces." Id. at 476.

4 ."There is no evidence that federal employees as a group are any more or less likely to abuse illegal substances than the population as a whole." Declaration of J. Michael Walsh, Ph.D., Director of Workplace Initiatives, National Institute on Drug Abuse, August 11, 1987.

5 . For example, the government submitted the declaration of Captain Bernard Leavitt of the Seneca Army Depot in Romulus, New York. Leavitt related an incident that occurred on July 4, 1985, when a security guard reported that another guard had twice used cocaine while on duty. An investigation revealed that other guards had observed the drug use on earlier occasions but had not reported it. Other “evidence” cited by the government consists primarily of positive results from urinalysis testing. This testing does not show on-duty impairment, however, but merely recent ingestion, as will be seen.

6 . Similar results were obtained in the first few months of drug testing of civilian employees of the Department of Transportation. A total of 890 workers were subjected to random urinalysis; seven positive test results were obtained, or .79 percent. Results Show Federal Drug Testing Not Fail-Safe, Wash. Post, Feb. 21, 1988, at Al, col. 2. That figure does not include the positive test result reported for Malcolm Tindall, an air traffic controller at Santa Rosa Tower in northern California; that positive result, Department of Transportation officials later acknowledged, was a mistake. They attributed the error to the private testing laboratory under contract to the Department of Transportation. Id.

7 .Congress limited the use that can be made of drug test results. For example, no disclosure to law enforcement officials is permitted without the employee’s consent. Supplemental Appropriations Act of 1987, Pub.L. 100-71, § 503(e), 101 Stat. 391, 471.

8 . Time Magazine, Nov. 16, 1987, at 18. Random urinalysis testing in the Department of the Army has been limited to "critical” employees under the terms of Department of Defense Directive 1010.9. About 9,400 of the 450,000 civilian employees of the Department of the Army hold “critical” jobs. It seems likely, in light of the Department of Transportation and Department of Justice testing programs, that more civilian employees of the Department of the Army would have been tested had the program been designed pursuant to Executive Order No. 12,564, which calls for testing of employees in "sensitive" positions, 3 C.F.R. 224, 226 (1987). Executive Order No. 12,564 leaves it to the head of each agency to determine which positions are "sensitive." Id. Common sense suggests that more federal employees hold "sensitive” positions than "critical” positions, and the Justice Department program appears to confirm this. For example, the Justice Department plans to test "people who have authority of one sort or another to make settlements involving large amounts of money, and those with security clearances.” Marcus, Justice Department vs. Its Lawyers? Group of Government Attorneys Considers Suing Over Drug Tests, Wash. Post, November 13, 1987, at A21, col. 3 (statement of Justice Department spokesman Patrick S. Korten).

9 . Results Show Federal Drug Testing Not Fail-Safe, Wash. Post, Feb. 21, 1988, at A1, col. 2.

10 . E.g., Railway Labor Executives’Association v. Burnley, 839 F.2d 575, 588 (9th Cir.1988) ("tests cannot measure current drug intoxication or degree of impairment"); Jones v. McKenzie, 833 F.2d 335, 339 & n. 10 (D.C.Cir.1987) (trial court found, and defendant conceded, tests did not show whether user is under influence at time of test); Schaill v. Tippecanoe County School Corp., 679 F.Supp. 833, 857 (N.D.Ind.1988) (witnesses testify without contradiction, and court concludes, that "little, if anything, can be inferred relative to current impairment" from urinalysis testing) (1988 WL 7127); Taylor v. O’Grady, 669 F.Supp. 1422, 1431 (N.D.Ill.1987) ("I find that there is no substantial relationship between a positive urine test and an impairment at the time of testing”); Artable v. Ford, 663 F.Supp. 149, 153 (W.D.Ark.1985) ("The utmost that the most precise of urine tests can reveal is that a student has ingested marijuana at some time in the preceding days or weeks.”). See also Dubowski, Drug-Use Testing: Scientific Perspectives, 11 Nova L.Rev. 415 (1987); Hudner, Urine Testing for Drugs, 11 Nova L.Rev. 553, 556-57 (1987); Joseph, Fourth Amendment Implications of Public Sector Work Place Drug Testing, 11 Nova L.Rev. 605, 632 (1987). But see American Federation of Government Employees v. Dole, 670 F.Supp. 445, 448 (D.D.C.1987) (discussed infra note 12), appeal docketed, No. 87-5417 (D.C.Cir. Dec. 11, 1987).

11 . The experts are in considerable disarray on the so-called “hangover effects” of various drugs. However, in the face of criticism from plaintiffs’ experts, the defendants’ experts appear to be in retreat on the subject. For example, in a supplemental declaration dated August 28, 1987, J. Michael Walsh, Ph.D., Director of Workplace Initiatives, National Institute on Drug Abuse, stated that "the fact is that there are very few studies that look beyond the acute affects of drugs, i.e., the direct pharmacological effects of drugs in the first three to four hours after the drug is ingested.”

Marian W. Fischman, Ph.D., Associate Professor in the Department of Psychiatry and Behavioral Sciences at Johns Hopkins University School of Medicine, filed a supplemental declaration dated August 27, 1987, to respond to criticism of one of his earlier assertions. "The quoted sentence, ‘Hangover effects after marijuana smoking may represent an additional, heretofore unappreciated aspect of marijuana toxicity,’ was carefully worded to include may," Fischman stated. "The data collected were suggestive of such residual effects, not conclusive, as no single study of this sort can be conclusive. The point of carrying out these kinds of research studies is to accumulate sufficient evidence pointing in one direction to be able to draw a conclusion____ [A] substantial body of literature in support of residual drug effects does not currently exist....”

The Court, of course, cannot resolve this dispute among scientists. To the extent that a "clear” relationship, see Jones v. McKenzie, 833 F.2d 335, 336 (D.C.Cir.1987), must be established between testing and impairment, however, the Court finds that the government has failed at this juncture in the case to make this showing. Even the government’s experts conceded the verdict is not in on hangover effects.

12 . The Court takes note of Judge Gesell’s statement that the Department of Transportation "presented proof that drug use, at the level sought by testing generally impairs the normal functioning of employees.” American Federation of Government Employees v. Dole, 670 F.Supp. 445, 448 (D.D.C.1987), appeal docketed, No. 87-5417 (D.C.Cir. Dec. 11, 1987). The Court has examined the defendant’s papers in that case without finding evidence that a positive test shows on-duty impairment. For example, the defendant states, "A confirmed positive urinalysis test result that has been verified by the agency’s' medical review officer (MRO) demonstrates that the employee is an illegal drug user, and presents the government with the substantial risk that the employee is currently impaired or will in the future be impaired while on duty.” Defendant’s Memorandum of Law in Opposition to Plaintiffs' Motion for a Preliminary Injunction, and Reply Memorandum in Support of Defendant's Motion for Summary Judgment at 14-15, American Federation of Government Employees v. Dole, 670 F.Supp. 447 (D.D.C.1987), notice of appeal filed, No. 87 — 1815 (Nov. 25, 1987). The passage quoted above contains a footnote, which states: "Plaintiffs repeatedly complain that drug tests do not measure impairment and are therefore worthless. See McBay Dec. at ¶ 12. Defendants agree that drug tests do not measure impairment in the sense that, for example, a breathalyzer test measures alcohol impairment; differences in individual metabolic rates make it difficult to draw definite conclusions about the exact relationship between the presence of drug metabolites in the urine and the presence of such metabolites in the body. However, a drug test that uses a high cut-off level like the DOT plan does demonstrate that the person tested ingested an illegal drug very recently. Willette Dec. at ¶ 12; Jewell Dec. at ¶ 3.” Id. at 15 n. 8.

The two declarations cited fail to support the contention that the cut-off levels used by the government demonstrate "very recent[]” drug use. One states that marijuana metabolites fall below cut-off values within two to five days for infrequent users, and adds, "The quantity of cannabinoid metabolites in urine depends primarily on the quantity of liquids ingested prior to the time of sampling.” The other states that smoking a single marijuana cigarette can be detected "for two or three days.”

Another government expert, J. Michael Walsh, Ph.D., Director of Workplace Initiatives for the National Institute on Drug Abuse, reported in an August, 11, 1987, declaration in the case sub judice that studies show that any measurable effects caused by marijuana end long before the ability of testing to detect metabolites of the drug in urine.

13 . The Court does not believe the “heavily regulated industries" exception to the warrant requirement, New York v. Burger, — U.S.-, 107 S.Ct. 2636, 96 L.Ed.2d 601 (1987); Donovan v. Dewey, 452 U.S. 594, 101 S.Ct. 2534, 69 L.Ed. 2d 262 (1981); United States v. Biswell, 406 U.S. 311 (1972); Colonnade Catering Corp. v. United States, 397 U.S. 72, 90 S.Ct. 774, 25 L.Ed.2d 60 (1970), is applicable to this case. The Third Circuit’s decision in Shoemaker v. Handel, 795 F.2d 1136 (3d Cir.), cert. denied, — U.S.-, 107 S.Ct. 577, 93 L.Ed.2d 580 (1986), which held the exception applies to warrantless breath and urine testing of jockeys “in the heavily regulated horse-racing industry,” 795 F.2d at 1142, is readily distinguishable. Unlike jockeys, the federal employees in this case are not members of a “highly regulated industry” subject to licensing requirements and a comprehensive regulatory scheme diminishing their privacy expectations. The Eighth Circuit’s extension of the heavily regulated industries exception in McDonell v. Hunter, 809 F.2d 1302, 1308 (8th Cir.1987), is likewise distinguishable. The Eighth Circuit limited its holding to prison guards "who have regular contact with the prisoners on a day-today basis in medium or maximum security prisons.” The court was unwilling to extend across-the-board testing even to other prison employees not falling within this select group. Id. The narrowness of the holding in McDonel v. Hunter distinguishes it from this one.

14 . The Fifth Circuit applied a somewhat different analysis in National Treasury Employees Union v. Von Raab, 816 F.2d 170 (5th Cir.1987), cert. granted, — U.S.-, 108 S.Ct. 1072, 99 L.Ed.2d 232 (1988). The Fifth Circuit upheld a testing program for some Customs Service employees tentatively selected for transfer into a sensitive position. The testing is not random, but is a regular element of the required procedure for transferring into certain jobs. 816 F.2d at 173. Additionally, the integrity interest articulated by the government is more compelling in the case of the Customs Service, which is charged with a key role in interdicting illegal drug smuggling, than in the case of civilian employees of the Army. The Von Raab decision also has been criticized for failing to adequately identify employee interests to be balanced against the government interest. E.g., Feliciano v. City of Cleveland, 661 F.Supp. 578, 592 (N.D.Ohio 1987).

15 . In Berry v. District of Columbia, 833 F.2d 1031, 1035 (D.C.Cir.1987), the Court of Appeals, again speaking through Judge Edwards, said the trial court could not merely accept the argument that there is a positive correlation between drug use and the likelihood that an arrestee released on bail will commit crimes or fail to appear for court dates. "On remand, the District must proffer reliable evidence, statistical or otherwise, from which the trial court can reasonably conclude that drug use makes it significantly more likely that an arrestee will commit crimes or fail to appear for scheduled court dates.” Id.

16 . The government also argues that drug use, on duty or off, is "simply incompatible" with federal employment. Concededly, no one should engage in illegal activities, but the government’s argument has several flaws. First, it fails to articulate a compelling interest that would justify the substantial intrusion represented by compulsory random urinalysis. Second, if accepted the argument would justify testing of all federal employees, not only those holding “critical” jobs. Third, as Judge Edenfield noted, the question "is not whether drug use, off-duty or on-duty is compatible with federal employment. Rather, the question is by what means is it permissible to come by evidence of such drug use.” American Federation of Government Employees v. Weinberger, 651 F.Supp. 726, 735 (S.D.Ga.1986).

17 . Tests detecting off-duty drug use "provide Government officials with a periscope through which they can peer into an individual’s behavior in her private life, even in her own home.” Jones v. McKenzie, 833 F.2d at 339. This . amounts to a form of surveillance of a person's off-duty activities. Id. (quoting Capua v. City of Plainfield, 643 F.Supp. 1507, 1511 (D.N.J.1986)); cf. American Federation of Government Employees v. Weinberger, 651 F.Supp. 726, 733 (S.D.Ga.1986) (urinalysis highly intrusive into "private affairs of an individual”). The information available to government officials from urinalysis is not limited to illegal drug use. For example, secret pregnancy tests were conducted on the urine samples of women who applied to be District of Columbia police officers from November, 1985, until the practice was disclosed recently. Applicants for D.C. Police Secretly Tested for Pregnancy, Wash. Post, Nov. 5, 1987, at A1, col 1.