The United States Customs Service has adopted a urinalysis drug testing program that will cover all employees having access to certain computer databases. The employees' exclusive bargaining representative, the National Treasury Employees Union ("NTEU"), argues that the agency's program violates the Fourth Amendment's proscription of unreasonable search and seizures and asks us to set aside the district court order granting the Government's motion for summary judgment, 829 F. Supp. 408. Because a compromise of the information in the databases could result in significant harm to an important national interest, we affirm the order.

The United States Customs Service is an agency within the Department of the Treasury. It is charged with the responsibility, among others, of collecting revenue from imports and enforcing customs and related laws. Pursuant to these laws, it is authorized to interdict and seize contraband, including illegal drugs. As such, it plays a critical role in the "war on drugs" that has exacted such a heavy toll on this country: In 1993 alone, Congress authorized some 12.7 billion dollars for drug control efforts. 1992 Sourcebook of Criminal Justice Statistics Table 1.16 at 21 (1993). Illicit drug import operations are "often directed by sophisticated criminal syndicates ... [and] may be effected by violence or its threat." National Treasury Employees Union v. Von Raab, 489 U.S. 656, 660, 109 S. Ct. 1384, 1387, 103 L. Ed. 2d 685 (1989) (citation omitted). Because of the Customs Service's "almost unique mission ... the Government [has] a compelling interest in ensuring that many of [its] employees do not use drugs even off duty, for such use creates risks of bribery and blackmail against which the Government is entitled to guard." Id. at 674, 109 S.Ct. at 1395.

In response to these and other dangers associated with the use of drugs by its personnel, in 1986 the Service adopted a policy of urinalysis testing for persons who applied for or sought promotion to certain positions within the Service. Three years later, in Von Raab, the Supreme Court upheld the policy insofar as it applied to positions directly involved in front line drug interdiction and enforcement activities or which required the carrying of firearms. Id. at 677, 109 S.Ct. at 1396-97. The Court also acknowledged that, in appropriate circumstances, urinalysis may properly be required of personnel who occupy positions where they "are likely to gain access to sensitive information." Id. at 678, 109 S.Ct. at 1397.

In December 1992, the Customs Service proposed expanding the program to include those employees with access to levels 1 or 2 of the Treasury Enforcement Communications System II ("TECS II") database and to two "modules" of the Automated Commercial System ("ACS"). Computers equipped with these databases have proven essential in enabling the Customs Service to process some sixteen million shipments each year. By feeding information regarding incoming cargos into the computers, employees can verify compliance with import laws and identify shipments warranting special scrutiny.

Because Customs has dropped its proposal to test employees having access to one of the ACS modules, we are only concerned with that system's Cargo Selectivity Criteria ("CSC") module. Customs has identified various combinations of these criteria as characteristic of contraband shipments. Certain of them apply nationwide while others are particular to each of the 43 Customs districts.

The process works as follows: When a ship arrives at a port of entry, the relevant information about the shipment (e.g., the type of merchandise, the entity seeking its entry, the manufacturer, the importer, and the country of origin) is entered into the ACS system. Customs employees use this system for a wide variety of routine functions, such as classifying shipments for tariff purposes and determining compliance with quotas, embargoes, and other import restrictions. If the data about the shipment matches any of the criteria requiring a search (for instance, a shipment to a first-time importer), appropriate instructions will appear on the screen, including the intensity of the search that is to be conducted by front-line Customs employees.

The cargo criteria are not the only factors that trigger an inspection, as the CSC module will also order a significant number of random searches. Customs officials may also override the system's recommended level of random inspections and initiate a more intensive search. As of June 1993, of the 2,627 bargaining unit employees covered under the drug testing plan, some 2,194 have access to the CSC module.

The other database of concern here is TECS II, a nationwide system that allows Customs to share anti-smuggling intelligence with its officers in the field and other enforcement agencies. The information contained in TECS II is divided into four levels of confidentiality. The NTEU is challenging the plan on behalf of some 1,090 Customs employees with access to the lowest two: levels 1 and 2. These levels include, inter alia, "suspect subject records," which contain "lookouts" issued by the Customs Service and other agencies "for persons and conveyances suspected of involvement in drug smuggling or other violations of the law." Affidavit of Carol Boyd Hallett, Commissioner of the United States Customs Service, dated January 15, 1993. Joint Appendix ("J.A.") at 289. Lookouts include

the names and/or identifying characteristics of persons targeted for detention and searches at points of entry into the United States; identifying characteristics of aircraft, vessels, vehicles and other means of conveyance targeted for detention and searches at points of entry; specific locations and times targeted for special scrutiny on the basis of intelligence information that drug smuggling or other illegal activity may be likely to take place; and identifying characteristics of suspicious modes of operation.

Id. Levels 1 and 2 also include intelligence reports that contain "information gathered from informants and other law enforcement agencies on suspected drug smuggling and other illegal activities and anticipated trends." Id. at 290.

The employees affected by the expansion of the Customs Service's drug testing program all work in traditional office environments; none is a front-line employee directly involved in drug interdiction efforts. For example, many of them are import specialists who classify merchandise for tariff purposes, apply quotas, and ensure compliance with commercial tariff regulations. Upon discovering errors or noncompliance, they may recommend sanctions, including fines, penalties, or forfeitures. These employees will be subject to urinalysis tests solely because of their access to the TECS II and ACS databases.

The procedures for testing the employees are set out in the "Mandatory Guidelines for Federal Workplace Drug Testing Programs" issued by the Department of Health and Human Services. See 53 Fed.Reg. 11,970 (Apr. 11, 1988). These regulations have not changed since we last described them:

After arriving at the test site, the employee will present photographic identification and will remove any outer garment such as a coat or jacket. The individual will be supervised by a monitor of the same gender, but will be allowed to urinate within a stall or partitioned area. The toilet water will be tinted with a bluing agent to ensure that the water is not used to adulterate the specimen. After the employee has furnished a specimen, the monitor will inspect the sample to ascertain that a sufficient volume is present, and that the sample is of normal color and temperature.

The laboratory to which specimens are sent will first employ an immunoassay test; any sample identified as positive will then be tested using gas chromatography/mass spectrometry (GC/MS) techniques. If this second test confirms the positive result, a Medical Review Officer shall "review and interpret" the test, "examin[ing] alternate medical explanations for any positive test result." 53 Fed.Reg. 11,985. Before verifying a positive result, the Officer must allow the employee an opportunity to discuss the test. If the Officer verifies the positive result, the employee will be removed from his sensitive position and will be subject to disciplinary proceedings. Possible penalties range from a reprimand to dismissal.

Harmon v. Thornburgh, 878 F.2d 484, 486 (D.C.Cir.1989) (citation and footnotes omitted).

The NTEU filed this action in federal district court in 1992. It seeks a permanent injunction of the proposed testing plan on the ground that it violates the Fourth Amendment's proscription of unreasonable searches. The court granted the Government's motion for summary judgment, which the Union now appeals.

The district court upheld the extension of the Customs Service's drug testing program because it concluded that the information in the ACS and TECS II databases is sufficiently sensitive that the Government's interest in its confidentiality exceeds the employees' privacy interest. On appeal, "we review the record de novo to determine whether it supports" the grant of summary judgment. American Fed'n of Gov't Employees v. Skinner, 885 F.2d 884, 893 (D.C.Cir.1989).

Our analysis is substantially controlled by the Supreme Court's decision in Von Raab. There, the Court concluded that as the production of urine in connection with a drug test constitutes a search for purposes of the Fourth Amendment, "it follows that the Customs Service's drug-testing program must meet the reasonableness requirement of the Fourth Amendment." 489 U.S. at 665, 109 S.Ct. at 1390. The Court's analysis of reasonableness weighed the interest advanced by the Government against the intrusion on the employees' privacy:

Because the testing program adopted by the Customs Service is not designed to serve the ordinary needs of law enforcement, we have balanced the public interest in the Service's testing program against the privacy interests implicated by the tests, without reference to our usual presumption in favor of the procedures specified by the Warrants Clause, to assess whether the tests required by Customs are reasonable.

Id. at 679, 109 S.Ct. at 1398. In keeping with this approach, we must assess two competing interests: the public interest furthered by urinalysis drug testing of employees with access to the two databases and the intrusion on the employees' privacy.

In Von Raab, the Court "readily agree[d] that the Government has a compelling interest in protecting truly sensitive information...." Id. at 677, 109 S.Ct. at 1397. The Court went on to state:

[E]mployees who seek promotions to positions where they would handle sensitive information can be required to submit to a urine test under the Service's screening program, especially if the positions covered under this category require background investigations, medical examinations, or other intrusions that may be expected to diminish their expectations of privacy in respect of a urinalysis test.

Id. The Court declined, however, to decide "whether testing those who apply for promotion to positions where they would handle 'classified' information is reasonable because we find the record inadequate for this purpose," id. at 679, 109 S. Ct. at 1398, and remanded the case for examination of "the criteria used by the Service in determining what materials are classified and in deciding whom to test under this rubric." Id. at 678, 109 S.Ct. at 1397. The Court also cautioned that

[i]n assessing the reasonableness of requiring tests of these employees, the court should ... consider pertinent information bearing upon the employees' privacy expectations, as well as the supervision to which these employees are already subject.

Id.

We had occasion to interpret Von Raab in Harmon. That case involved an appeal from a district court order enjoining the Department of Justice ("DOJ") from enforcing a program for the random urinalysis drug testing of all federal prosecutors and all workers with access to grand jury proceedings, as well as all employees having access to information classified as "top secret." In doing so, we recognized that

[a]pplication of Von Raab to the facts of [Harmon ] presents a delicate task. The Von Raab majority made no effort to articulate an analytical rule by which legitimate drug-testing programs could be distinguished from illegitimate ones. It simply weighed individual privacy interests against the government's policy objectives, enumerating several factors that it deemed relevant in performing this balancing process.

Harmon, 878 F.2d at 488.

In addressing the facts presented in Harmon, we concluded that the inclusion among those to be tested of all prosecutors and all workers with access to grand jury proceedings was overly broad. While we recognized that employees within these categories "[would] have access to information which [they are] duty-bound not to divulge ... we believe[d] that the term ['truly sensitive'] cannot include all information which is confidential or closed to public view." Harmon, 878 F.2d at 492 (emphasis in original). Therefore, while we set aside the injunction as it applied to personnel with access to top secret information, we sustained the injunction with respect to these two broad classes of DOJ employees pending the promulgation of more carefully focused regulations. Id. at 495-96.

In Harmon, we noted that the Supreme Court had failed to "define the contours of 'truly sensitive' information" in Von Raab, id. at 491, and contented ourselves with the safe observation that "[w]hatever 'truly sensitive' information includes, we agree that it encompasses top secret national security information." Id. (footnote omitted). In Von Raab and Harmon, however, there was no need to furnish explicit guidance for determining when the protection of information justifies a suspicionless search because, in each case, the record was found inadequate for the requisite balancing of governmental and private interests where information not classified for national security was concerned. Von Raab, however, provides the necessary guidance when, as here, the record reveals the precise nature of the information sought to be protected, the agency's reasons for wishing to test employees having access to it, and the nature of the work and the work environment of those employees.

In Von Raab, the Supreme Court stated that "[w]here ... the possible harm against which the Government seeks to guard is substantial, the need to prevent its occurrence furnishes an ample justification for reasonable searches calculated to advance the Government's goal." Von Raab, 489 U.S. at 674-75, 109 S.Ct. at 1395 (footnote omitted). As an example, the Court cited the Federal Government's practice of requiring suspicionless searches of tens of thousands of airline passengers and their baggage every day. Id. at 675 n. 3, 109 S. Ct. at 1395 n. 3. The reasonableness of such a search is more than outweighed by the potential harm: the hijacking or pirating of aircraft and the consequent threat to the lives and property of the traveling public. Similarly, when the Government is attempting to prevent the dissemination of "sensitive" information, the extent of the Government's interest is to be gauged not by the stamp the Government sees fit to place on the information, but by the damage that could result from its disclosure.

Here, the Customs Service has issued a directive that describes the information contained in the ACS and TECS II systems as "not 'classified,' but sensitive," and establishes procedures for the background investigation of those having access to them. Customs Directive No. 0991460-010 (Sept. 11, 1992). The Government argues that drug testing is warranted here because compromise of this information could frustrate the Customs Service's drug interdiction efforts in important ways. This is hard to dispute. The TECS II database contains "suspect subject records" and "lookouts" issued by Customs and other law enforcement agencies both for individuals associated with the drug trade and the aircraft, vessels, and vehicles used by them to bring drugs into the United States. It is self-evident that this information is valuable to drug smugglers--preidentified smugglers will be concerned about "suspect subject records," and all drug smugglers will be interested in TECS II's listing of "specific locations and times targeted for special scrutiny." J.A. at 289.

The information in the ACS database is also of obvious benefit to narcotics traffickers because anyone with access to the CSC module would gain important insights into the Service's interdiction strategies. In addition, the ability to run repeated searches would enable traffickers to formulate shipment profiles that minimize the risk of triggering the selectivity criteria. The Customs Service concedes that this information would not provide a smuggler with a 100 percent assurance of avoiding a search because the CSC module still generates random searches. Avoidance of all the criteria, however, would reduce the risk of inspection to the approximately five percent of all incoming shipments that are subject to random searches either by direction of the CSC or on the initiative of Customs inspectors.

The NTEU argues that the information in the CSC is purely commercial and that knowledge of the criteria will only reduce the chances of inspection. Admittedly, we cannot determine the exact value a particular smuggler would place on the information in the CSC, as we do not know the alternate world--the risk of detection inherent in the type of shipment that smuggler would have selected if ignorant of the information in the CSC module. But the NTEU's argument fails to recognize that smugglers seek to reduce the chances of an inspection, and their task will be made easier if they know what characteristics of an incoming shipment will most likely trigger an investigation. Narcotics are often smuggled inside innocuous commercial cargos. A smuggler who must choose between concealing heroin in a crate of endives to be shipped from Antwerp to New York or in a box of brie to be sent from Brest to New Orleans would obviously be interested in knowing which is more likely to be searched. He thus has ample incentive to seek access to the information stored in the CSC module.

The NTEU also argues that in Harmon, we rejected the DOJ's claims that the concededly important interest in maintaining the secrecy of grand jury information was adequate in and of itself to justify random urinalysis tests. We did conclude that the Government was not justified in testing all employees with access to grand jury proceedings because not "all information which is confidential or closed to public view" is sufficiently sensitive to justify drug testing. Harmon, 878 F.2d at 492 (emphasis in original). Thus the NTEU is correct: A general governmental desire for secrecy does not warrant the intrusion inherent in a drug testing program; instead, the Government must demonstrate that sufficient damage could flow from a compromise of that information so as to warrant overriding the affected employee's privacy interest.

In this case, we conclude that smugglers could use the information in both databases to frustrate the Service's interdiction efforts and introduce significant quantities of illicit drugs into the United States. As an analytical matter, however, evaluating the damage that could result from the compromise of the data still leaves us with the task of determining the extent of the governmental interest in protecting information that could benefit drug traffickers. We do not write on a blank slate. The Supreme Court has described illicit drugs as "one of the greatest problems affecting the health and welfare of our population" and has stressed "the veritable national crisis in law enforcement caused by smuggling of illicit narcotics." Von Raab, 489 U.S. at 668, 109 S.Ct. at 1392 (internal quotation marks omitted). In discussing the role of drug interdiction officers in Von Raab, the Court stated that "[a] drug user's indifference to the Service's basic mission, or, even worse, his active complicity with the malefactors, can facilitate importation of sizable drug shipments or block apprehension of dangerous criminals." Id. at 670, 109 S.Ct. at 1392. Moreover, the Court concluded that "the almost unique mission of the [Customs] Service gives the Government a compelling interest in ensuring that many of these covered employees do not use drugs even off duty, for such use creates risks of bribery and blackmail against which the Government is entitled to guard." Id. at 674, 109 S.Ct. at 1395.

Given this backdrop, we conclude that the Government has an obvious and compelling interest in protecting the information at issue here. It is of significant value to narcotics smugglers against whom the Government has been waging a major, costly, and sustained campaign. The question then remaining is whether this compelling interest trumps the employees' privacy interest.

The extent of an employee's privacy interest depends on such factors as the randomness and intrusiveness of drug testing procedures and the extent of the employees' expectations of privacy. Random drug testing represents a greater threat to an employee's privacy interest than does mandatory testing because of the " 'unsettling show of authority' that may be associated with unexpected intrusions on privacy." Von Raab, 489 U.S. at 672 n. 2, 109 S. Ct. at 1394 n. 2 (citation omitted). The employees' actual privacy expectations are also proper grounds for inquiry, as the courts have recognized that employees who occupy positions that require stringent background checks have a lowered privacy interest for purposes of the Fourth Amendment balancing test. See id. at 677, 109 S. Ct. at 1397 (drug tests permissible "especially if the positions ... require background investigations, medical examinations, or other intrusions that may be expected to diminish [the employees'] expectations of privacy in respect of a urinalysis test."); Hartness v. Bush, 919 F.2d 170, 173 (D.C.Cir.1990) ("As for the privacy interest, a number of cases ... have recognized that holders of security clearances have less of an expectation of privacy."). As the Customs Service points out, all employees with access to ACS and TECS II have to undergo comprehensive background checks, must provide detailed information about their finances, and must disclose their history, if any, of past arrests, convictions, and drug use.

The testing procedures themselves may also determine the extent to which a urinalysis test intrudes on an employee's privacy. Here, the employees are subject to testing under Department of Health and Human Services regulations that we have found to " 'significantly minimize' intrusion on the employees' privacy expectations." Hartness, 919 F.2d at 171 (quoting Von Raab, 489 U.S. at 672 n. 2, 109 S. Ct. at 1394 n. 2). As to random testing, although we noted in Harmon that randomness was a relevant factor, we suggested that it "would tip the scales" only "in a particularly close case." Harmon, 878 F.2d at 489.

The NTEU cites an additional factor for our Fourth Amendment calculus: whether there is any warrant here for a suspicionless search, however unintrusive. The NTEU argues that because it is possible "to subject employees and their work product to the kind of day-to-day scrutiny that is the norm in more traditional office environments," Von Raab, 489 U.S. at 674, 109 S. Ct. at 1395, drug testing here is unnecessary and, therefore, unreasonable. This question was addressed in Harmon, where the employees subject to the proposed testing were, as here, office rather than field workers. Harmon, 878 F.2d at 489 ("DOJ employees, by contrast [to the drug interdiction customs officers in Von Raab ], work in 'traditional office environments,' in which drug use is, presumably, more easily detected by means other than urine testing."). We concluded, however, that while the locus of the employee's work "is surely one element to be weighed in the balance, the Von Raab Court gave no indication that it deemed this factor to be one of overriding significance." Id. We accord it little weight here. A use of drugs, including weekend use, that is sufficient to expose an employee to bribery or intimidation by a drug smuggler will not necessarily affect the user's work product, and may thus escape detection through "day-to-day scrutiny" of the employee's performance on the job.

In Harmon and Hartness, as well as in this case, affected employees are subject to random urinalysis testing pursuant to guidelines that minimize the intrusiveness of the procedures consistent with the need for reliable sampling. The employees here, like the holders of the "top secret" and "secret" national security clearances in the other two cases, are subject to intrusive background checks that lower their privacy expectations. Because "the goal of thwarting illegal drug use and trafficking is critically important," Hartness, 919 F.2d at 174 (Edwards, J., dissenting), and because that goal would be adversely affected by the unauthorized disclosure of the information contained in the ACS and TECS II databases, we conclude that the Government's compelling interest in preserving the confidentiality of that information outweighs the privacy expectations of the employees who have access to it.

Over the past decade and more, the United States Government has waged an extraordinary and costly campaign to contain the drug trade. A single shipment that arrives in the United States undetected may have a value in the millions of dollars and can bring misery and death to large numbers of our citizens. The Customs Service has made a plausible case that a drug trafficker in possession of the information contained in the ACS and TECS II databases can significantly improve his chances of successfully smuggling his next shipment of drugs into the United States. That being the case, we have no difficulty in concluding that the Government's interest in protecting this information outweighs the intrusion that random drug testing imposes on employees whose expectations of privacy have been diminished by their subjection to comprehensive background checks. Accordingly, the decision of the district court is

Affirmed.

Whenever individual liberty is at risk of being jettisoned in combatting a national crisis, courts have a special obligation to stand stubborn in requiring that the sacrifice be necessary. Otherwise, our Constitution and those it protects will surely suffer. See Korematsu v. United States, 323 U.S. 214, 65 S. Ct. 193, 89 L. Ed. 194 (1944); Dennis v. United States, 341 U.S. 494, 71 S. Ct. 857, 95 L. Ed. 1137 (1951). For even after the crisis subsides--as it inevitably does--the constitutional fallout lingers on and takes a creeping toll. We cannot count ourselves victorious on any front, in any war, unless our Constitution survives the battle.

Caught in the cross-fire of our nation's "war on drugs," the Fourth Amendment is in a precarious position today. As written, it 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.

U.S. CONST. amend. IV. Although the "probable-cause standard 'is peculiarly related to criminal investigations,' " the Supreme Court has recognized that when the government requires its employees to submit to drug tests, it may do so only within the confines of the Fourth Amendment. National Treasury Employees Union v. Von Raab, 489 U.S. 656, 667, 109 S. Ct. 1384, 1391-92, 103 L. Ed. 2d 685 (1989) (quoting South Dakota v. Opperman, 428 U.S. 364, 370 n. 5, 96 S. Ct. 3092, 3097 n. 5, 49 L. Ed. 2d 1000 (1976)). See id. at 665, 109 S. Ct. at 1390-91. Nonetheless, in Von Raab and its companion case, Skinner v. Railway Labor Executives' Association, 489 U.S. 602, 109 S. Ct. 1402, 103 L. Ed. 2d 639 (1989), the Supreme Court upheld drug testing of certain federal employees despite the absence of either warrant or individualized suspicion. The urinalysis at issue in Skinner was performed on certain railroad employees following all major train accidents and whenever any train incident resulted in the death of an on-duty railroad employee. 489 U.S. at 608-12, 109 S.Ct. at 1408-12. In Von Raab, the United States Customs Service ("Customs" or "Service") had proposed to perform drug tests on all employees prior to their voluntary promotion to positions in which the employee would (1) be directly involved in drug interdiction or enforcement of related laws, (2) carry a firearm, or (3) handle "classified" material. 489 U.S. at 660-61, 109 S.Ct. at 1387-88.

While the Court had no doubt that such drug tests "must be deemed searches under the Fourth Amendment," Skinner, 489 U.S. at 617, 109 S.Ct. at 1413, it nonetheless found the tests passed constitutional muster. The Court reasoned:

[W]here a Fourth Amendment intrusion serves special governmental needs, beyond the normal need for law enforcement, it is necessary to balance the individual's privacy expectations against the Government's interests to determine whether it is impractical to require a warrant or some level of individualized suspicion in the particular context.

Von Raab, 489 U.S. at 665-66, 109 S.Ct. at 1390-91 (citing Skinner, 489 U.S. at 619-20, 109 S.Ct. at 1414-15). In striking this balance in Von Raab, the Court recognized three government interests: protecting the safety of the public from dangers posed by armed Customs officers, protecting the integrity of front-line interdiction personnel, and protecting truly sensitive information. This case involves only the last interest.

The Court recognized that the government had a "compelling interest in protecting truly sensitive information from those who, 'under compulsion of circumstance or for other reasons, ... might compromise [such] information.' " Id. at 677, 109 S.Ct. at 1397 (quoting Dep't of Navy v. Egan, 484 U.S. 518, 528, 108 S. Ct. 818, 824, 98 L. Ed. 2d 918 (1988)). It noted that

employees who seek promotions to positions where they would handle sensitive information can be required to submit to a urine test under the Service's screening program, especially if the positions covered under this category require background investigations, medical examinations, or other intrusions that may be expected to diminish their expectations of privacy in respect of a urinalysis test.

Id. (citing Dep't of Navy, 484 U.S. at 528, 108 S.Ct. at 824). However, because the Court could not balance the privacy expectations against the government's interest based on the record as it existed then, it remanded the case:

Upon remand, the Court of Appeals should examine the criteria used by the Service in determining what materials are classified and in deciding whom to test under this rubric. In assessing the reasonableness of requiring tests of these employees, the court should also consider pertinent information bearing upon the employees' privacy expectations, as well as the supervision to which these employees are already subject.

Id. 489 U.S. at 678, 109 S. Ct. at 1397. The Supreme Court has never struck this balance since.

In our own applications of the principle enunciated in Von Raab and Skinner, this circuit has inched its way ever farther from the core holdings in those two cases, repeatedly mollified by the fact that only a minuscule step extra was needed to arrive at an outcome arguably supported by precedent. See Hartness v. Bush, 919 F.2d 170 (D.C.Cir.1990), cert. denied, 501 U.S. 1251, 111 S. Ct. 2890, 115 L. Ed. 2d 1055 (1991); American Fed'n of Gov't Employees v. Skinner, 885 F.2d 884, 891 (D.C.Cir.1989) ("AFGE "), cert. denied, 495 U.S. 923, 110 S. Ct. 1960, 109 L. Ed. 2d 321 (1990); Harmon v. Thornburgh, 878 F.2d 484 (D.C.Cir.1989), cert. denied, 493 U.S. 1056, 110 S. Ct. 865, 107 L. Ed. 2d 949 (1990). As a prescient Justice Marshall wrote in his Skinner dissent: "The majority purports to limit its decision to postaccident testing of workers in 'safety-sensitive' jobs, [Skinner, 489 U.S.] at 620 [109 S.Ct. at 1414-15], much as it limits its holding in the companion case to the testing of transferees to jobs involving drug interdiction or the use of firearms. [Von Raab, 489 U.S.] at 664[, 109 S. Ct. at 1390]. But the damage done to the Fourth Amendment is not so easily cabined." 489 U.S. at 636, 109 S.Ct. at 1423 (Marshall, J., dissenting).

In Harmon we upheld as constitutional the drug testing of Department of Justice ("DOJ") employees who had access to top secret national security information. The testing involved in that case was performed randomly on incumbent attorneys. Unlike the pre-voluntary-promotion tests in Von Raab, the DOJ attorney with top secret security clearance was not at liberty to remain at her current job and forgo a promotion as a means of avoiding the incursion on her privacy. Thus, the covered employee in Harmon was at risk of losing not only the prospect of a promotion, but her present job as well,1 and the employee would never be put on notice that a drug test was imminent.2 While we recognized the "relevan[ce] " of these distinguishing factors in the overall balance, we ultimately rejected the notion that "this aspect of the program requires us to undertake a fundamentally different analysis from that pursued by the Supreme Court in Von Raab." 878 F.2d at 489 (emphasis in original). Accord, AFGE, 885 F.2d at 891. Applying that analysis, we found that "[w]hatever 'truly sensitive' information includes ... it encompasses top secret national security information," and concluded that under Von Raab the "government may properly make testing a requirement for holding a top secret security clearance." Id. at 491-92.

At the same time, however, Harmon struck down the DOJ's related attempt to test all federal prosecutors or all federal employees with access to grand jury proceedings. We held that truly sensitive information

cannot include all information which is confidential or closed to public view. A very wide range of government employees--including clerks, typists, or messengers--will potentially have access to information of this sort.... The fact that the employees covered by the Department's drug testing regulations deal with confidential information therefore does not distinguish them from private attorneys, or from government employees generally.

878 F.2d at 492 (emphasis in original). The court warned that "caution should be used in approving this [information protection] justification for testing." Id.

Next, we held that Department of Transportation mail van operators with "secret or top secret security clearance and regular access to classified information" could constitutionally be subjected to random drug testing, especially in light of the fact that tested employees "work in settings other than the 'more traditional office environments' where 'drug use is, presumably, more easily detected by means other than urine testing.' " AFGE, 885 F.2d at 893 (quoting Von Raab, 489 U.S. at 674, 109 S. Ct. at 1395, and Harmon, 878 F.2d at 488). Finally, a divided panel of this court held more generally that the difference between top secret and secret security clearance holders "is not significant enough to tip the constitutional scales against [conducting random urinalysis drug] testing" on Executive Office of the President employees with "secret" national security clearances. Hartness v. Bush, 919 F.2d 170, 174 (D.C.Cir.1990) (Mikva, J., concurring). See also id. ("I do not think it necessary ... to go beyond this narrow holding."). While we have thus taken significant strides toward upholding suspicionless drug tests aimed at protecting sensitive information, my review of the caselaw has revealed no decision by a federal court of appeals--and the government has pointed to none--upholding urinalysis drug tests for the purpose of protecting information that is not classified as "secret" or "top secret" pursuant to Executive Order No. 12,356, 47 Fed.Reg. 14,874 (Apr. 6, 1982).3

In the instant appeal the government's need to test Customs employees is far less compelling than in prior cases. Basically, the majority seems to say that access to any information that might prove useful to a drug dealer justifies random urinalysis. The information that Customs seeks to protect with the help of urinalysis is not classified under Executive Order No. 12,356 at all, nor do the Customs employees here enjoy secret or top secret security clearances. The employees at issue here work in traditional office environments, presumably permitting the kind of day-to-day monitoring that would reveal a pattern of drug abuse constituting a security threat, and the government readily admits that "a substantial population of Customs officers require access to TECS II for reasons unrelated to drug enforcement." Joint Appendix ("J.A.") 419.

Moreover, Customs seeks to test not only job applicants, but incumbent job holders, and it proposes to test them randomly without notice. But nothing in the obvious ethical or physical demands of these employees' jobs would indicate to them that they should expect random intrusions on their privacy. Cf. Von Raab, 489 U.S. at 679, 109 S. Ct. at 1398 (noting that intrusions into employee's privacy were to be expected due to "the special, and obvious, physical and ethical demands of those positions"). Indeed, the sole factor reducing these employees' privacy expectations is a one-time background check. By relying on such a background check alone as reducing an employee's expectation of privacy to the point of permitting the intrusive urinalysis, the majority effectively makes background checks simply an added cost for the government to insure that random drug testing will not be rejected as too invasive by courts.

Finally, I do not believe that the record permits us to decide that, under these circumstances, it would be impracticable to require Customs to develop individualized suspicion prior to testing any given employee. True, the Supreme Court noted that " '[t]he reasonableness of any particular government activity does not necessarily or invariably turn on the existence of alternative "less intrusive" means.' " Skinner, 489 U.S. at 629 n. 9, 109 S. Ct. at 1419-20 n. 9 (quoting Illinois v. Lafayette, 462 U.S. 640, 647, 103 S. Ct. 2605, 2610, 77 L. Ed. 2d 65 (1983)). But while we have followed Skinner's rejection of a least intrusive means requirement, we have at the same time recognized that under certain circumstances, such as employment in a traditional office environment, drug use might be easily detected without intrusive random urinalysis. Compare National Fed'n of Fed. Employees v. Cheney, 884 F.2d 603, 610-11 (D.C.Cir.1989), cert. denied, 493 U.S. 1056, 110 S. Ct. 864, 107 L. Ed. 2d 948 (1990), with id. at 614. Indeed, this criterion of alternative means of supervision was recognized explicitly by the Court in Von Raab. See 489 U.S. at 674, 109 S. Ct. at 1395 ("it is not feasible to subject [front-line drug interdiction] employees and their work product to the kind of day-to-day scrutiny that is the norm in more traditional office environments"); id. at 678, 109 S. Ct. at 1397 (remand with instruction that "court should also consider pertinent information bearing upon the employees' privacy expectations, as well as the supervision to which these employees are already subject"). Nothing in the record before us indicates that the government is incapable of effectively monitoring these employees at the workplace to detect serious drug problems.

Similarly, Customs has not presented any evidence concerning alternative methods of storing and securing the information so as to minimize the number of employees it must test. Even putting the general question of a Fourth Amendment "less intrusive means" inquiry to one side, cf. Nadine Strossen, The Fourth Amendment in the Balance: Accurately Setting the Scales through the Least Intrusive Alternative Analysis, 63 N.Y.U.L.Rev. 1173 (1988), the instant situation merits more than a nod in the direction of attempting to minimize the user pool of truly sensitive information so as to cut down on unnecessary searches. This criterion, too, has been implicitly recognized as significant by the Von Raab Court. See 489 U.S. at 678, 109 S. Ct. at 1397 (remanding with instructions to "examine the criteria used by the Service in determining what materials are classified and in deciding whom to test under this rubric"). While perhaps in the case of information classified as "top secret" we may presume that access to such information is not freely given without cause, where nonclassified information is at stake, we must ask more incisively whether a pattern of broad access does not undermine the government's claim to the information's sensitivity. See Harmon, 878 F.2d at 492. Although this suit concerns only the 2,194 NTEU members who are covered by the drug testing plan, nationwide approximately 12,000 Customs employees, including secretaries, clerks, and typists, with access to the database in question will be subject to testing. Yet the present record contains only a bald assertion by Customs that "it would be extremely difficult, cumbersome and expensive to attempt to segregate computer access to information specifically related to drug interdiction from access to other enforcement related data in the system." J.A. 402. To be sure, in today's information age, where access to data is often only a few keystrokes away, heightened vigilance is required in the protection of the government's secrets. However, we must equally ensure that the burdens of secrecy are carried in the first instance by the government itself, e.g., by appropriately restricting employees' access to sensitive information they do not need or by policing the data requests that are ultimately made. The government may not simply throw up its hands and err on the side of liberally granting its employees access to a wide range of data with the effect of losing the Fourth Amendment somewhere in cyberspace.

I am aware that Von Raab approved a drug testing program without any documentation that drug using Customs officials were the source of information leaks. See Von Raab, 489 U.S. at 683, 109 S.Ct. at 1400 (Scalia, J., dissenting). So it may be argued that we cannot dismiss the government's asserted need simply on the basis that it has not offered a single documented instance of an integrity violation of the kind feared here. But apart from the inevitable doubts this lack of evidence raises about the need for such widespread testing, we have been instructed by the Supreme Court that we must consider "the criteria used by the Service in determining what materials are classified [or considered truly sensitive] and in deciding whom to test under this rubric." Von Raab, 489 U.S. at 678, 109 S. Ct. at 1397. We are further charged with "assessing the reasonableness of requiring tests of these employees, [by] ... consider[ing] pertinent information bearing upon the employees' privacy expectations, as well as the supervision to which these employees are already subject." Id. Ultimately, the task we face is to "balance the individual's privacy expectations against the Government's interests to determine whether it is impractical to require a warrant or some level of individualized suspicion in the particular context." Von Raab, 489 U.S. at 665-66, 109 S.Ct. at 1390-91 (citing Skinner, 489 U.S. at 619-20, 109 S.Ct. at 1414-15).

In applying the Von Raab framework to this case, it is immediately apparent that the drug testing program here is a far cry indeed from the one-time, pre-voluntary-transfer test at issue in Von Raab. The urinalysis to be conducted here is performed randomly on a large pool of incumbent employees who work in traditional office environments in a wide range of positions with no expectation of repeated intrusions into their privacy. Without an ounce of evidence of a prior loss of integrity akin to the one it now seeks to prevent, Customs sets out to test its employees in order to protect information that it does not deem sensitive enough to merit inclusion in any category of classified material. Certainly when dealing with information to which the government has granted such wide access, we may not, as the majority does, simply presume that any approach short of drug testing to safeguard the information at hand is too cumbersome, that any further restriction on access to sensitive information is impracticable, or that day-to-day monitoring in a traditional office environment to detect drug use is unfeasible. If balancing the individual's expectations of privacy against the interests of the government is to have any meaning at all as a way of assuming the Fourth Amendment remains alive and well in this field and if the Supreme Court's mandated judicial determination of the impracticality of requiring individualized suspicion is to retain any semblance of reality, the government must bear the burden of demonstrating that other less intrusive efforts at protecting nonclassified information in an office environment have been explored and found inadequate. The government has not done so in this case. Without any such assurances, we should not blindly permit our constitutional shield against an ever-inquisitive government to be further compromised by a single-minded dedication to fighting the drug war. Because the record as it currently stands does not indicate that the Fourth Amendment balance which we must so carefully strike tips in favor of the government, I dissent.