JAMES H. BURNLEY IV, SECRETARY, DEPARTMENT OF TRANSPORTATION, ET AL., PETITIONERS V. RAILWAY LABOR EXECUTIVES' ASSOCIATION, ET AL. No. 87-1555 In the Supreme Court of the United States October Term, 1988 On Writ of Certiorari to the United States Court of Appeals for the Ninth Circuit Brief for the Petitioners PARTIES TO THE PROCEEDINGS The petitioners are James H. Burnley IV, Secretary of the Department of Transportation; and John H. Riley, Administrator of the Federal Railroad Administration. The respondents are the Railway Labor Executives' Association; the United Transportation Union General Committee of Adjustment, the Southern Pacific Company; the Brotherhood of Locomotive Engineers General Committee of Adjustment, the Southern Pacific Company; and the Brotherhood of Railroad Signalmen. TABLE OF CONTENTS Question Presented Parties To The Proceedings Opinions below Jurisdiction Constitutional provision and regulations involved Statement: A. The Federal Railroad Administration regulations B. The present controversy Introduction and summary of argument Argument: The FRA regulations are constitutional under the Fourth Amendment A. A warrantless search may be valid even in the absence of particularized suspicion B. The FRA regulations entail a minimal intrusion on employees' expectation of privacy 1. The employment relationship and the regulated nature of the industry 2. The constraints on official discretion 3. The narrowly-tailored nature of the testing procedures C. The FRA regulations serve compelling governmental interests Conclusion OPINIONS BELOW The opinion of the court of appeals declaring the regulations unconstitutional (Pet. App. 1a-49a) is reported at 839 F.2d 575. The opinion of the district court (Pet. App. 50a-56a) is unreported. JURISDICTION The judgment of the court of appeals was entered on February 11, 1988. The petition for a writ of certiorari was filed on March 17, 1988, and was granted on June 6, 1988. The jurisdiction of this Court rests on 28 U.S.C. 1254(1). CONSTITUTIONAL PROVISION AND REGULATIONS INVOLVED The Fourth Amendment to the United States Constitution provides in pertinent part: The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated * * *. Pertinent excerpts from the regulations involved in this case are reproduced in the appendix to the petition (Pet. App. 57a-78a) and in the Joint Appendix (J.A. 8-16). QUESTION PRESENTED Whether regulations promulgated by the Federal Railroad Administration -- mandating blood and urine tests of railroad employees who are involved in certain train accidents and fatal incidents and authorizing breath and urine tests after certain accidents, incidents, and rule violations -- violate the Fourth Amendment on the ground that they do not require a showing of particularized suspicion of drug or alcohol impairment prior to the testing. STATEMENT A. The Federal Railroad Administration Regulations 1. Section 202(a) of the Federal Railroad Safety Act of 1970, 45 U.S.C. 431(a), authorizes the Secretary of Transportation and by delegation, the Federal Railroad Administration (FRA), to "prescribe, as necessary, appropriate rules, regulations, orders, and standards for all areas of railroad safety." On July 5, 1983, pursuant to that authority (see 48 Fed. Reg. 30723), the FRA published an Advance Notice of Proposed Rulemaking (ANPRM) -- commencing a process that would culminate, two years later, in the promulgation of regulations designed to control the use of alcohol and drugs by railroad employees engaged in railroad operations. The FRA explained that the proposed rulemaking was prompted by pressing safety concerns. "Alcohol impairment and drug impairment," the agency observed, "have been identified as causal or contributing factors in a number of train accidents and employee fatalities over the past ten years." 48 Fed. Reg. 30723 (1983). The agency acknowledged (id. at 30724) that the railroad industry had taken steps to solve these problems on its own, especially by promulgating and enforcing "Rule G," an industry-wide operating rule that prohibits railroad employees from using, possessing, or being under the influence of alcohol or drugs while on duty. /1/ The agency concluded, however, that "past FRA efforts to promote voluntary action by the railroad industry" to curb alcohol and drug abuse had "not met with uniform success" (id. at 30723). For example, a 1979 study prepared by the Railroad Employee Assistance Project (REAP) examining the scope of alcohol abuse on seven major railroads showed that "19% of all employees were 'problem drinkers'"; "23% of operating personnel were 'problem drinkers'"; "(o)nly 4% of problem drinkers were receiving help through an employee assistance program, and even fewer were handled through disciplinary procedures"; "5% of workers reported to work 'very drunk' or got 'very drunk' on duty at least once in the study year (1978)"; "13% of workers reported to work at least 'a little drunk' one or more times during that period"; "13% of operating employees drank while on duty at least once during the study year, averaging about 3 such instances during the year"; and "(a)n estimated one out of every eight railroad workers drank at least once while on duty during the study year." 48 Fed. Reg. 30724 (1983) (footnote ommitted). The costs of alcohol abuse to the seven railroads, moreover, were estimated to be $108 million in 1978, "suggesting industry-wide costs in that year of in excess of $200 million" (ibid.). Drawing on that and other research, the FRA found that from 1972 to 1983 "the nation's railroads experienced at least 21 significant train accidents involving alcohol or drug use as a probable cause or contributing factor" (id. at 30726). "Those accidents," the FRA stated (ibid.), "resulted in 25 fatalities, 61 non-fatal injuries, and property damage estimated at $19 million (approximately $27 million in 1982 dollars)." The FRA also identified (ibid.) "an additional 17 fatalities to operating employees working on or around rail rolling stock that involved alcohol or drugs as a contributing factor." /2/ The agency accordingly solicited comments from interested parties on a variety of possible approaches to the problems of alcohol and drug abuse aboard the nation's railroads. It invited the public to address such disparate options as random drug and alcohol testing, testing on "reasonable suspicion," regularly scheduled "efficiency" testing for train operators, post-accident testing, increased observation of employees by supervisors, the promotion of voluntary programs, and the possibility of no federal regulatory action at all. 48 Fed. Reg. 30726-30731 (1983). 2. Following the publication of the ANPRM, the FRA conducted several days of public hearings and received comments and related materials from a wide spectrum of interested parties. See 49 Fed. Reg. 24252 (1984). The agency carefully reviewed the materials produced in those proceedings and refined the data on which the proposed regulations would be based. On June 12, 1984, the FRA published a Notice of Proposed Rulemaking (NPRM), setting out in detail the factual basis for regulating the on-duty use of alcohol and drugs by railroad employees. 49 Fed. Reg. 24252-24304. On the basis of the evidence submitted, the agency found that "(a) significant minority of railroad employees use alcohol and drugs in connection with railroad operations" and that "alcohol use and drug use are sufficiently common to pose a significant safety problem." 49 Fed. Reg. 24253 (1984). "Available information from all sources, including FRA safety investigations, suggests that the problem includes 'pockets' of drinking and drug use involving multiple crew memebers (before and during work), sporadic cases of individual employees reporting to work impaired, and repeated drinking and drug use by individual employees who are chemically or psychologically dependent on those substances" (id. at 24253-24254). "Even without the benefit of regular post-accident testing," the agency observed (id. at 24254), "FRA has identified 34 fatalities, 66 injuries and over $28 million in property damage (in 1983 dollars) that resulted from the errors of alcohol and drug-impaired employees in 45 train accidents and train incidents during the period 1975 through 1983." /3/ And, the FRA noted (ibid.), "(t)he 1984 toll may set a new record." /4/ In determining to promulgate federal regulations, the FRA found wanting the existing approaches to alcohol and drug abuse among railroad employees. The agency observed that the enforcement of Rule G by the railroads, chiefly through "supervisory observations and punishment of offenders" (49 Fed. Reg. 24266 (1984)), was inadequate in many cases; "(a)t a more serious level," it added, the widespread perception of "unchecked management discretion to excuse or punish" had undermined employees' acceptance of Rule G itself (id. at 24267). /5/ The agency also found (id. at 24281) that there was a "conspiracy of silence" regarding Rule G violations, "at least partially because the conventional sanction for a violation is dismissal." Voluntary rehabilitation programs, moreover, had achieved only mixed results; the FRA determined from the data that "most problems drinkers remain unidentified and unserved after a decade of voluntary efforts" and that "(t)reatment of drug abusers presents an even more difficult topic of analysis" since very few rehabilitation programs had been providing services to drug abusers (id. at 24270). /6/ 3. a. On August 2, 1985, after reviewing extensive comments from representatives of the railroad industry, labor groups, and the general public, the FRA published the regulations that are at issue in this lawsuit (50 Fed. Reg. 31568-31579; Pet. App. 57a-78a; J.A. 8-16). In a detailed preamble to the regulations -- copies of which have been lodged with the Court for its convenience -- the agency set out its factual findings and meticulously explained its regulatory choices. The FRA found that both alcohol and drug use remained "reasonably prevalent" among railroad employees (50 Fed. Reg. 31515 (1985)), and, indeed, that "(t)he 1984 toll of alcohol and drug-related accidents may have represented a new record, at least in the train accident category" (id. at 31516). The agency added to its earlier list three 1984 train accidents in which alcohol or drug use was a causal factor. /7/ It also identified four other 1984 accidents that involved alcohol or drugs but for which the data was insufficient to assert positively that alcohol or drug use was a causal factor. /8/ Finding that "(c)onventional methods of controlling the alcohol/drug problem ha(d) proven inadequate by themselves" and that "(e)xclusive reliance on voluntary programs (was) not warranted by available information" (id. at 31527 (emphasis omitted)), the FRA concluded that "(t)he issuance of necessary and appropriate Federal regulations (was) required" (id. at 31528 (emphasis omitted)). b. The final regulations (49 C.F.R Pt. 219) consist of six subparts. /9/ Subpart C (id. Sections 219.201 to 219.213; Pet. App. 58a-68a) is entitled "Post-Accident Toxicological Testing." It provides (49 C.F. R. 219.203(a)) that railroads "shall take all practicable steps to assure that all covered employees of the railroad directly involved * * * provide blood and urine samples for toxicological testing by FRA" after any one or more of the following three circumstances: (1) a "major train accident" -- defined as one that includes either a fatality, the release of hazardous material accompanied by an evacuation or reportable injury, or damage to railroad property of $500,000 or more (id. Section 219.201(a)(1)); (2) an "impact accident" (collision) that results in a reportable injury or damage to railroad property of $50,000 or more (id. Section 219.201(a)(2)); or (3) a "train incident that involves a fatality to any on-duty railroad employee" (id. Section 219.201(a)(3)). /10/ The post-accident regulations in Subpart C, like the other portions of the final rule, apply to "covered employees" (49 C.F. 219.203(a)), defined in pertient part as persons who have "been assigned to perform service subject to the Hou(rs) of Service Act (45 U.S.C. 61-64b)" (49 C.F.R. 219.5(d); J.A. 9) -- which includes essentially all "train and engine crews, yard crews (including switchmen), hostlers, train order and block operators, dispatchers, and signalmen" (50 Fed. Reg. 31530 (1985)). The FRA recognized that other groups of employees "also play an important role in maintaining the safety of rail operations, "but it found that "(t)he available accident statistics confirm that the biggest part of the alcohol and drug problem * * * is concentrated among the() crafts that perform 'covered service.'" 49 Fed. Reg. 24286 (1984). /11/ After one of the three triggering events occurs, both blood and urine samples must be taken. The FRA noted that a blood sample "can provide a clear indication not only of the presence of alcohol and drugs but also their current impairment effects" (49 Fed. Reg. 24291 (1984)). /12/ Railroads must "make every reasonable effort to assure that samples are pRovided as soon as possible" (49 C.F.R. 219.203(b)(1)). Toward that end, employees must "be transported to an independent medical facility where the samples shall be obtained" by qualified medical personnel (id. Section 219.203(c)(1)). /13/ The regulations also establish procedures for collecting and handling the samples (id. Section 219.205). /14/ After the samples have been collected, the railroad is required to ship them by pre-paid air freight to the FRA laboratory -- presently the Center for Human Toxicology at the University of Utah (see 49 C.F.R. Pt. 219 App. B) -- for analysis (49 C.F.R. 219.205(d); Pet. App. 63a). There, the samples are analyzed using "state-of-the-art equipment and techniques to detect and quantify alcohol and drugs" (Federal Railroad Admin., U.A. Dep't of Transp., Field Manual: Control of Alcohol and Drug Use in Railroad Operation B-12 (Mar. 1986) (hereinafter Field Manual). In particular, "(e)thyl alcohol is measured by gas chromatography. Drug screens may be conducted by immunoassay and other techniques. Positive drug findings are confirmed by gas chromatography/mass spectrometry" (ibid.; see also 50 Fed. Reg. 31555-31556 (1985)). Thereafter, the FRA notifies employees of the results of the tests and affords them an opportunity to respond in writing prior to the preparation of any final investigative report (49 C.F.R. 219.211(a)(2)). /15/ Employees who refuse to provide required blood or urine samples may not perform covered service for a period of nine months, but they are entitled to a hearing concerning their refusal to take the test (id. Section 219.213)). /16/ c. Subpart D of the regulations (49 C.F.R. 219-301 to 219.309) is entitled "Authorization to Test for Cause." It authorizes (but does not require) railroads to mandate breath or urine tests (but not blood tests) for covered employees under specified circumstances. Breath tests may be ordered (1) where is supervisor has a "reasonable suspicion" that an employee is under the influence or is impaired by alcohol, based upon specific, personal observations concerning the appearance, behavior, speech, or body odors of the employee (id. Section 219.305(b)(1)); (2) in the event of a reportable accident or incident, where a supervisor has a reasonable suspicion that the employee's acts or omissions contributed to the occurrence or severity of the accident or incident (id. Section 219.301(b)(2)); or (3) in the event of certain specific rule violations, including noncompliance with a signal, excessive speeding, improper switch alignment, failure to stop short of a derail, and failure to secure a hand brake (id. Section 219.301(b)(3)). Urine tests may be ordered under circumstances (2) and (3) above (id. Section 219.301(c)(1)); they may also be ordered in cases of "reasonable suspicion," but only if two supervisors make the appropriate determination (id. Section 219.301(c)(2)(i)) and, where the supervisors suspect impairment due to a drug other than alcohol, at least one of those supervisors must have received specialized training in detecting the signs of drug intoxication (id. Section 219.3010(c)(2)(ii)). The authorizing regulations also establish procedures and safeguards for conducting breath and urine tests. Only certain designated breath-test devices may be used; the devices must be properly maintained and calibrated; and the tests must be conducted by a trained and qualified operator (49 C.F.R. 219.303(a)(1), (2), and (3)). If a breath test proves positive, the employee must be retested after 15 minutes have passed, in order to confirm the initial finding (id. Section 219.303(a)(5)). Urine, in turn, may be collected only at an independent medical facility and the collection must be supervised by personnel of that facility (id. Section 219.305(a)). Railroads must also establish with the participating medical facility procedures sufficient "to ensure positive identification of each sample and accurate reporting of laboratory results" (id. Section 219.305(b)). The testing of the urine must be performed by an independent laboratory "proficient in the testing of urine for alcohol and drugs of abuse" (id. Section 219.305(a)(1)), and each sample must "be analyzed by a method that is reliable within known tolerances" (id. Section 219.307(b)). If a screening test is positive for a substance other than alcohol, a remaining portion of the sample must be retested, using a "scientifically-recognized method capable of providing quantitative data specific to the drug (or metabolite(s)) detected" (id. Section 219.307(b)). /17/ Moreover, any laboratory that performs confirmatory testing must also participate in an external quality control program, including, where practicable, the use of blind samples (id. Section 219.307(a)(2)). Whenever the results of either breath or urine tests are intended for use in a disciplinary proceeding, the employee must be given the opportunity to provide a blood sample for analysis at an independent medical facility (49 C.F.R. 219.303(c), 219.305(d)). If an employee declines to provide a blood sample, the railroad may presume impairment from the presence of an identified controlled substance in the urine (in the absence of persuasive evidence to the contrary), but the railroads must provide detailed notice to employees of that presumption and advise employees of their rights to provide a contemporaneous blood sample (id. Section 219.309). d. The FRA identified "at least five independent, if related, bases for mandating post-accident testing." First, the agency determined that post-accident testing "is needed to guide FRA enforcement efforts under the() new rules" by "indicat(ing) the persistence, or emergence of alcohol and drug problems on individual railroads." Second, the FRA explained, "post-accident testing is necessary to the process of regulatory development" because "(e)nlightened and proportional regulation will only be possible if the true causes of major human factor accidents are know." /18/ Third, the agency reasoned, post-accident testing would provide important information to the public at large concerning the causes of major accidents. Fourth, it stated, such testing would "help to deter employees from using alcohol and drugs on the job." /19/ Finally, it said, post-accident testing is "a means of keeping the alcohol and drug problem squarely before the railroad industry" so that "progress (may be) sustained over the long term * * *." 50 Fed. Reg. 31541-31542 (1985). The FRA further explained that the authority conferred by Subpart D would assist railroads to enforce rules against drug or alcohol use by on-duty employees and enhance their ability to identify the human causes of accidents, incidents, and rule violations. 49 Fed. Reg. 24294-24295 (1984). The agency also pointed out that testing under Subpart D would measurably improve the ability of the railroads to detect alcohol and drug violations by employees (see 50 Fed. Reg. 31550 (1985)) and would thereby promote overall deterrence (see id. at 31526). The FRA emphasized, however, that Subpart D is intended to be "permissive only; no particular test or series or program of testing is required or compelled by this section, and any such test or tests actually undertaken by the railroad are done at its initiative and not directly at the instance of FRA." 49 Fed. Reg. 24295 (1984). B. The Present Controversy 1. Respondents, the Railway Labor Executives' Association and various of its member labor organizations, brought this action seeking to enjoin the FRA regulations on a variety of statutory and constitutional grounds. In an opinion issued from the bench (Pet. App. 50a-56a), the district court granted summary judgment in petitioners' favor. The court explained (id. at 52a-53a) that railroad personnel have a valid Fourth Amendment interest "in the integrity of their own bodies," but that there is also a competing "public and governmental interest in the * * * promotion of * * * railway safety, safety for employees, and safety for the general public that is involved with the transportation." In striking the appropriate balance, the court emphasized (id. at 53a) "that the railroad industry is one of the most extensively regulated industries that we have in interstate commerce; and that the regulation(s) extend() not just to the railroads themselves, but a certain amount of regulation of the employees." Applying the criteria for assessing the reasonableness of a search within a "pervasively regulated" industry (ibid.), the court noted (id. at 53a-55a) that the FRA regulations serve "a valid governmental and public interest"; that there are "objective event(s) which trigger() the testing" which are "as well defined as a set of regulations could give"; and that the regulations make a "genuine attempt * * * to limit the scope of the testing requirements to the needs and the events as they have occurred." The court accordingly held that the Fourth Amendment balance "is being struck in favor of the regulatory scheme" (id. at 53a). /20/ 2. A divided court of appeals reversed (Pet. App. 1a-49a). The court held, first (id. at 8a-13a), that drug and alcohol tests are searches within the meaning of the Fourth Amendment, and the Subpart D of the regulations involves sufficient governmental action to implicate the Fourth Amendment, even though that subpart only authorizes, but does not require, the adoption of testing procedures by otherwise private railroads. The court therefore turned to whether the regulations were constitutionally reasonable. It "agree(d) that the exigencies of testing for the presence of alcohol and durgs in blood, urine or breath require prompt action which precludes obtaining a warrant" (id. at 16a). It also acknowledged (id. at 24a) that "accommodation of railroad employees' privacy interest with the significant safety concerns of the government does not require adherence to a probable cause requirement." The court held, however (id. at 25a), that "particularized suspicion is essential to finding toxicological testing of railroad employees justified at its inception" and, because all but one provision of the regulations were wanting in that respect, /21/ the court struck them down. /22/ Elaborating on its understanding of the "particularized suspicion" standard, the court of appeals emphasized (Pet. App. 25a-26a) that "(a)ccidents, incidents or rule violations, by themselves, do not create reasonable grounds for suspecting that tests will demonstrate alcohol or drug impairment in any one railroad employee, much less an entire train crew." The court also surmised (id. at 26a) that it would "pose() no insuperable burden on the government to require individualized suspicion," noting (ibid.) that such a standard was already codified in certain Subpart D provisions (see 49 C.F.R. 219.301(b)(1) and (c)(2)), and reasoning (Pet. App. 26a-27a) that the same standard "should be incorporated into the mandatory testing provisions set out in 49 C.F.R. Section 219.201 and the other testing provisions set out in 49 C.F.R. Section 219.301(b)(2) and (3)." In addition, the court explained (Pet. App. 28a) that in the absence of a requirement of particularized suspicion, it would be troubled by a further "flaw in the reasonableness of this approach to the problem" -- the inability of the drug tests to "measure current drug intoxication or degree of impairment," rather than merely "the metabolites of various drugs, which are not evidence of current intoxication and may remain in the body for days or weeks after the ingestion of the drug" (ibid.). Reversing the judgment of the district court, the court of appeals held that "intrusive drug and alcohol testing may be required or authorized only when specific articulable facts give rise to a reasonable suspicion that a test will reveal evidence of current drug or alcohol impairment" (id. at 36a). /23/ Judge Alarcon dissented (Pet. App. 37a-49a). In his view (id. at 39a (emphasis in original)), "the activities of railway personnel are closely regulated to promote safety." He would therefore have held "that the closely regulated industry exception to the requirement of a warrant and probable cause applies to blood, breath, and urine tests of railway employees under the specific circumstances prescribed in the regulations challenged in this action" (ibid.). Moreover, Judge Alarcon would have found the FRA regulations constitutionally acceptable, even apart from the regulated nature of the industry. Noting that the court's holding was in conflict with decisions in several other circuits (id. at 42a), /24/ Judge Alarcon criticized the majority for "fail(ing) to engage in (a) balancing of interests" and for focusing instead "solely on the degree of impairment of the workers' privacy interests" (id. at 46a). A proper balance, the dissent observed (ibid.), would have recognized "that the government's compelling need to assure railroad safety by controlling drug use among railway personnel outweighs the need to protect privacy interests." "some quantum of individualized suspicion is usually a prerequisite to a INTRODUCTION AND SUMMARY OF ARGUMENT After several years of detailed study, the FRA concluded that alcohol and drug abuse by railroad employees had become a serious threat to the safe opertion of the nation's railroads. To deter that abuse, and to assist in determining the causes of significant accidents and incidents, the agency promulgated regulations that mandate blood and urine testing, and that authorize breath and urine testing, pursuant to carefully circumscribed criteria. The court of appeals invalidated those regulations, however, because they do not impose an across-the-board requirement of particularized suspicion. We believe that the court's decision is mistaken. A. "The fundamental command of the Fourth Amendment is that searches and seizures be reasonable" (New Jersey v. T.L.O., 469 U.S. 325, 340 (1985)). What is reasonable, however, "depends on the context within which a search takes place" (id. at 337). The reasonableness of a particular search practice "is judged by 'balancing its intrusion on the individuals's Fourth Amendment interest against its promotion of legitimate governmental interests.'" United States v. Montoya de Hernandez, 473 U.S. 531, 537 (1985) (citation omitted). Applying this balancing test, the Court has emphasized that a search may be reasonable under the Fourth Amendment in the absence of either a warrant or probable cause. Moreover, the Court has explained, although constitutional search or seizure(,) * * * the Fourth Amendment imposes no irreducible requirement of such suspicion." United States v. Martinez-Fuerte, 428 U.S. 543, 560-561 (1976) (footnote omitted). Thus, the Court has on several occasions identifed exceptions to the requirement of particularized suspicion "where the privacy interests implicated by a search are minimal and where 'other safeguards' are available 'to assure that the individual's reasonable expectation of privacy is not "subject to the discretion of the official in the field"'" (New Jersey v. T.L.O., 469 U.S. at 342 n.8 (citations omitted)). We believe that the balance of competing interests here strongly supports the constitutionality of the FRA regulations -- even though they generally require no showing of particularized suspicion of alcohol or drug use by railroad employees. B. The FRA regulations entail only the most minimal intrusion on employees' reasonable expectations of privacy. Three features of the regulations account for that fact. First and foremost, the testing mandated and authorized by the FRA regulations takes place as an aspect of an ongoing employment relationship. As this Court explained in O'Connor v. Ortega, No. 85-530 (Mar. 31, 1987), slip op. 7, "(t)he employee's expectation of privacy must be assessed in the context of the employment relation." This point gathers special force where employees work in an industry subject to extensive public and private regulation. Here, the heavily regulated nature of the railroad industry means that railroad employees' ordinary expecatations of privacy have been significantly diminished, wholly apart from the impact of the FRA regulations. Second, the FRA regulations tightly cabin administrative discretion. Under the testing procedures, railroad officials lack any significant discretion in deciding who to test, when to test, or how to conduct the tests. As Judge Alarcon noted in dissent below, the "'"time, place, and scope"'" of the testing are thus "brought within reasonable bounds" (Pet. App. 41a (citations omitted)). Third, the FRA testing procedures themselves are narrowly tailored to respect, to the fullest extennt practicable, the employees' reasonable expectations of privacy. Blood testing is mandated after only the most serious accidents and incidents, and blood and urine samples may be collected only at a medical facility, at the direction of medical personnel. The testing procedures are highly reliable; safeguards must be followed to ensure accurate results; and participating laboratories must meet exacting quality assurance standards. While the process of providing samples at a medical facility may impose a measure of inconvenience for some employees, there is no warrant for the court of appeals' suggestion (Pet. App. 22a) that blood, urine, and breath tests are "degrading" or "offen(sive to) human dignity." C. While the intrusion on privacy is minimal, the governmental interests at stake are substantial. The FRA has a surpassing interest in ensuring that train travel is safe for the riding public and for operating employees, and that accidents capable of releasing hazardous materials into the atmosphere do not occur. As Judge Alarcon correctly observed (Pet. App. 46a), "locomotives in the hands of drug or alcohol impaired employees are the substantial equivalents for time-bombs endangering the lives of thousands." After carefully considering the dimensions of the problem, and after analyzing but rejecting alternative solutions, the FRA determined that blood, urine, and breath testing would help deter alcohol and drug abuse and would assist in ascertaining the causes of significant accidents and incidents. The court of appeals had no basis, in law or in fact, for second-guessing that judgment. ARGUMENT THE FRA REGULATIONS ARE CONSTITUTIONAL UNDER THE FOURTH AMENDMENT A. A Warrantless Search May Be Valid Even In The Absence Of Particularized Suspicion 1. "The fundamental command of the Fourth Amendment is that searches and seizures be reasonable * * *." New Jersey v. T.L.O., 469 U.S. 325, 340 (1985). See United States v. Montoya de Hernandez, 473 U.S. 531, 537 (1985); United States v. Villamonte-Marquez, 462 U.S. 579, 588 (1983). Thus, the Fourth Amendment "does not denounce all searches or seizures, but only such as are unreasonable." Carroll v. United States, 267 U.S. 132, 147 (1925). The test of reasonableness, moreover, "is not capable of precise definition or mechanical application." Bell v. Wolfish, 441 U.S. 520, 559 (1979). Rather, in defining the contours of the right to be free from unreasonable searches and seizures, this Court has repeatedly said that "'the specific content and incidents of this right must be shaped by the context in which it is asserted.'" Wyman v. James, 400 U.S. 309, 318 (1971) (quoting Terry v. Ohio, 392 U.S. 1, 9 (1968)). See also New Jersey v. T.L.O., 469 U.S. at 337 ("what is reasonable depends on the context within which a search takes place"). The Court has adopted a balancing test to govern this inquiry. "The permissibility of a particular law enforcement practice is judged by 'balancing its intrusion on the individual's Fourth Amendment interests against its promotion of legitimate governmental interests.'" United States v. Montoya de Hernandez, 473 U.S. 531, 537 (1985) (quoting United States v. Villamonte-Marquez, 462 U.S. 579, 588 (1983)). See also Delaware v. Prouse, 440 U.S. 648, 654 (1979); Camara v. Municipal Court, 387 U.S. 523 (1967). This approach recognizes that not every invasion of privacy is prohibited by the Fourth Amendment, but only "arbitrary and oppressive interference by enforcement officials with the privacy and personal security of individuals." United States v. Martinez-Fuerte, 428 U.S. 543, 554 (1976). "Where a careful balancing of governmental and private interests suggests that the public interest is best served by a Fourth Amendment standard of reasonableness that stops short of probable cause, (the Court has) not hesitated to adopt such a standard" (New Jersey v. T.L.O., 469 U.S. at 341). In examining the competing interests, the Court has often found a third consideration important -- the amount of discretion left to officials in carrying out the search. See, e.g., Donovan v. Dewey, 452 U.S. 594, 601-604 (1981); Delaware v. Prouse, 440 U.S. at 654, 661; Marshall v. Barlow's Inc., 436 U.S. 307, 320-321 (1978). See also Colorado v. Bertine, 479 U.S. 367, 376-377 (1987) (Blackmun, J., concurring). Because this constraint on discretion derives from the constitutional requirement of "reasonableness" (Delaware v. Prouse, 440 U.S. at 654), the means needed to confine discretion also necessarily vary from context to context. What is required to meet the constitutional concern for controlling discretion depends on the strength of the governmental interest in a particular search practice and the impairment of privacy interests that the practice effects. 2. Applying this balancing test, the Court has held that in the context of an ordinary investigation of criminal conduct by law enforcement officers, both probable cause and a warrant are generally necessary to render a search reasonable. See United States v. Karo, 468 U.S. 705, 717 (1984); United States v. United States District Court, 407 U.S. 297, 317 (1972). But as the Court explained most recently in Griffin v. Wisconsin, No. 86-5324 (June 26, 1987), slip op. 4 (citation omitted) -- and as we show in more detail in our brief (at 17-19) in National Treasury Employees Union v. von Raab, No. 86-1879 /25/ -- exceptions from those requirements have been permitted "when 'special needs, beyond the normal need for law enforcement, make the warrant and probable-cause requirement impracticable.'" See New Jersey v. T.L.O., 469 U.S. at 340 (quoting Almeida-Sanchez v. United States, 413 U.S. 266, 277 (1973) (Powell, J., concurring)) ("although 'both the concept of probable cause and the requirement of a warrant bear on the reasonableness of a search, * * * in certain limited circumstances neither is required'"). Where, for example, "the burden of obtaining a warrant is likely to frustrate the governmental purpose behind the search" (Camara v. Municipal Court, 387 U.S. at 533), the Court has routinely held that a warrant is not required by the Fourth Amendment. See, e.g., Griffin v. Wisconsin, slip op. 7; O'Connor v. Ortega, No. 85-530 (Mar. 31, 1987), slip op. 11 (plurality opinion); New Jersey v. T.L.O., 469 U.S. at 340. The Court has likewise found that the probable cause standard is inappropriate where it would defeat the purposes that the search is designed to achieve. See, e.g., Griffin v. Wisconsin, slip op. 9; New Jersey v. T.L.O., 469 U.S. at 340-342. 3. The court below recognized (Pet. App. 16a) "that the exigencies of testing for the presence of alcohol and drugs in blood, urine or breath require prompt action which precludes obtaining a warrant." And it acknowledged (id. at 24a) "that accommodation of railroad employees' privacy interest with the significant safety concerns of the government does not require adherence to a probable cause requirement." The court nevertheless insisted (id. at 24a-25a) that a blood, urine, or breath test cannot be "justified at its inception" unless there are "reasonable grounds for suspecting that the search will turn up * * * evidence" of on-duty alcohol or drug abuse by an individual employee. To meet that standard, the court of appeals concluded, "particularized suspicion is essential" (id. at 25a). The court of appeals erred, however, in conclusively presuming that the searches at issue here cannot be reasonable under the Fourth Amendment unless there is some individualized suspicion of unlawful conduct. As this Court explained in United States v. Martinez-Fuerte, 428 U.S. at 560-561 (footnote omitted), while "some quantum of individualized suspicion is usually a prerequisite to a constitutional search or seizure(,) * * * the Fourth Amendment imposes no irreducible requirement of such suspicion." See also New Jersey v. T.L.O., 469 U.S. at 342 n.8; Camara v. Municipal Court, supra. For example, the Court has held that individualized suspicion is not required for the Coast Guard or Customs Service to board a vessel and to examine its owner's documents. United States v. Villamonte-Marquez, supra. Similarly, in Wyman v. James, 400 U.S. 309, 318 (1971), the Court upheld the right of a welfare caseworker to enter the home of a welfare recipient to ensure compliance with welfare rules, even though the statute authorizing the home visit required no showing that the recipient was out of compliance with the rules. And in United States v. Martinez-Fuerte, supra, the Court held that border officials may stop automobiles at permanent checkpoints without any individualized suspicion. This Court has also upheld warrantless searches, in the absence of particularized suspicion, in several cases involving "closely regulated" industries. In those cases, the Court has emphasized that the regulated nature of the industry reduces the individual's expectation of privacy, while at the same time affording sufficient notice and regularity to obviate the need for a warrant or particularized suspicion. In United States v. Biswell, 406 U.S. 311 (1972), for example, the Court upheld the statutorily-authorized, warrantless search of the locked storeroom of a licensed gun dealer, which resulted in the seizure of unlicensed firearms. The Court explained (id. at 316) that inspections under the statute "pose only limited threats to the dealer's justifiable expectations of privacy" since "(w)hen a dealer chooses to engage in this pervasively regulated business and to accept a federal license, he does so with the knowledge that his business records, firearms, and ammunition will be subject to effective inspection." More recently, in Donovan v. Dewey, 452 U.S. 594 (1981), the Court upheld provisions of the Federal Mine Safety and Health Act of 1977, 30 U.S.C. 813(a), which authorized warrantless inspections, without particularized suspicion, of underground and surface mines. The Court emphasized that the statute was "specifically tailored" to assure "the certainty and regularity" of the inspection program, and that "the regulation of mines * * * is sufficiently pervasive and defined that the owner of such a facility cannot help but be aware that he 'will be subject to effective inspection'" (452 U.S. at 602 (footnote and citation omitted)). See also New York v. Burger, No. 86-80 (June 19, 1987), slip op. 10 (upholding a state statute permitting warrantless searches of junkyards, in the absence of particularized suspicion). The court of appeals misread those cases when it presupposed that there must be particularized suspicion in order to test railroad employees for the presence of alcohol or drugs. As this Court made clear in T.L.O., the question whether particularized suspicion is a necessary component of reasonableness depends upon the balance of competing interests; it cannot simply be presumed. In particular, as T.L.O. instructs, exceptions may be made to the requirement of individualized suspicion may be made to the requirement of individualized suspicion "where the privacy interests implicated by a search are minimal and where 'other safeguards' are available 'to assure that the individual's reasonable expectation of privacy is not "subject to the discretion of the official in the field"'" (469 U.S. at 342 n.8 (citations omitted)). The court of appeals glossed over these factors; as Judge Alarcon observed in dissent (Pet. App. 46a), the court of appeals never "engage(d) in the balancing of interests required by the Court." By presuming a requirement of particularized suspicion, the court of appeals spared itself the more rigorous obligation to weigh all of the competing interests. As we show below, the FRA regulations should be judged by, and easily satisfy, the reasonableness standard under the Fourth Amendment: as in T.L.O., "the privacy interests implicated by (the) search(es) are minimal" and "'other safeguards' are available 'to assure that the individual's reasonable expectation of privacy is not "subject to the discretion of the official in the field."'" Moreover, the governmental interest in ensuring the safety of the general public (and of railroad employees themselves) is compelling. /26/ B. The FRA Regulations Entail A Minimal Intrusion On Employees' Expectation Of Privacy 1. The employment relationship and the regulated nature of the industry A first and quite general feature of the FRA regulations, which significantly minimizes their interference with expectations of privacy, is the fact that the blood, urine, and breath tests are conducted as an aspect of the employment relationship. As this Court explained in O'Connor v. Ortega, slip op. 6 (emphasis in original), "(t)he operational realities fo the workplace * * * may make some employees' expectations of privacy unreasonable." Thus, "(t)he employee's expectation of privacy must be assessed in the context of the employment relationship" (id. at 7). In the Ortega case, for example, the plurality found that the search of an employee's office "'involve(d) a relatively limited invasion'" of privacy, noting that "the privacy interests of governmental employees in their place of work * * *, while not insubstantial, are far less than those found at home or in some other contexts" (id. at 14). This proposition has particular force where, as here, the employee works in an industry in which his activities, and those of his employer and co-workers, are subject to extensive public and private regulation. As the Court has explained in similar settings, persons who work in such "'closely regulated'" industries have a "reduced expectation of privacy" (New York v. Burger, slip op. 10) and "in effect consent() to the restrictions placed upon (them)" (Almeida-Sanchez v. United States, 413 U.S. 266, 271 (1973)). "Railroads have been subject to comprehensive federal regulation for nearly a century." United Transportation Union v. Long Island R.R., 455 U.S. 678, 687 (1982) (footnote omitted). Beginning in 1893 with the enactment of the Safety Appliance Act, ch. 196, 27 Stat. 531, 45 U.S.C. 1 et seq., the federal government has taken an increasingly active role in regulating railroads in an effort to promote safety. In the early part of this century, Congress enacted, in short succession, the Block Signal Act, ch. 46, 34 Stat. 838, 45 U.S.C. 35, which allowed for research and later required the implementation of automatic signaling systems; the Ash Pan Act, ch. 225, 35 Stat. 476, 45 U.S.C. 17 et seq., repealed by the Federal Railroad Safety Authorization Act of 1982, Pub. L. No. 97-468, Tit. VII, Section 705, 96 Stat. 2580, which prohibited the use of locomotives equipped with ash pans that could not be dumped without employees going under the locomotive for that purpose; the Accident Reports Act, ch. 208, Section 1, 36 Stat. 350, 45 U.S.C. 38, which established a system for collecting accident data on injuires and fatalities and documenting accident causes; the Locomotive Inspection Act, ch. 103, Section 1, 36 Stat. 913, 45 U.S.C. 22, which required railroads to inspect and test locomotives to prevent boiler explosions; and the Signal Inspection Act, ch. 91, Section 441, 41 Stat. 498, 49 U.S.C. App. 26, which permitted the government to inspect railroad signal systems and, if necessary, to require modifications. More recently, Congress enacted the Federal Railroad Safety Act of 1970, 45 U.S.C. (& Supp. III) 421 et seq., a comprehensive rail safety measure that gives the Secretary of Transportation broad powers to "prescribe, as necessary, appropriate rules, regulations, orders, and standards for all areas of railroad safety" (45 U.S.C. 431(a)). Acting pursuant to these statutes and others, the FRA has promulgated a substantial volume of regulations, which prescribe the safety obligations of railroads in finer detail. Included among these are regulations prescribing procedures for enforcing the Hazardous Materials Transportation Act (49 U.S.C. App. 1809), 49 C.F.R. Pt. 209; governing railroad noise emissions, 49 C.F.R. Pt. 210; prescribing standards for track safety, 49 C.F.R. Pt. 213, freight car safety, 49 C.F.R. Pt. 215, radio usage, 49 C.F.R. Pt. 220, locomotive safety, 49 C.F.R. Pt. 229, railroad safety appliances, 49 C.F.R. Pt. 231, and railroad power brakes and drawbars, 49 C.F.R. Pt. 232; requiring the filing of railroad operating rules, 49 C.F.R. Pt. 217, reports of accidents and incidents, 49 C.F.R. Pt. 225, and reports of accidents specifically stemming from signal failure, 49 C.F.R. Pt. 233; mandating minimum markings on the tail end of passenger, commuter, and freight trains, 49 C.F.R. Pt. 221; requiring special safety glazing on locomotives, passenger cars, and cabooses, 49 C.F.R. Pt. 223; and prescribing rules governing the installation, inspection, maintenance, and repair of signal and train control systems, devices, and appliances, 49 C.F.R. Pt. 236. /27/ Given the nature of the industry, most safety regulation has focused on railroad equipment, track, and signals. But because "the safety of employees and travelers" does not depend solely on "the enactment of laws relating to mechanical appliances" (Baltimore & O. R.R. v. ICC, 221 U.S. 612, 619 (1911)), Congress, the FRA, and various state agencies have also taken steps to regulate the conduct of railroad employees. The Hours of Service Act, ch. 2939, 34 Stat. 1415, 45 U.S.C. 61 et seq., enacted in 1907 and amended several times since then, sets limits on the number of hours that certain railroad operating employees may work, recognizing that "(t)he length of hours of service has direct relation to the efficiency of the human agencies upon which protection to life and property necessarily depends" (Baltimore & O. R.R., 221 U.S. at 619). See also Atchison, T. & S.F. Ry. v. United States, 244 U.S. 336, 342-343 (1917). /28/ The comprehensive Federal Railroad Safety Act of 1970 also authorizes the regulation of employees, specifically granting the Secretary of Transportation the power to "test() * * * railroad facilities, equipment, rolling stock, operations, or persons, as he deems necessary to carry out the provisions of this subchapter" (45 U.S.C. 437(a) (emphasis added)). More recently, the Rail Safety Improvement Act of 1988, Pub. L. No. 100-342, 102 Stat. 624, has amended the existing federal rail safety statutes to extend to individuals (including employees) liability for particular safety infractions that had previously been imposed only on the railroads. See H.R. Rep. 100-637, 100th Cong., 2d Sess. 2, 20 (1988). And numerous federal regulations also bear directly on the daily activities of railroad employees. See, e.g., 49 C.F.R. 218.1 to 218.30 (requiring workmen to follow certain prescribed safety procedures when co-workers are engaged on rail tracks); id. Section 218.37 (requiring crew members to take certain precautionary measures when a train is moving at a reduced speed); id. Section 220.61 (requiring employees to follow certain procedures in transmitting train orders by radio). See also, e.g., Mass. Ann. Laws ch. 160, Sections 178-181 (Law. Co-op. 1979) (setting eyesight and experience requirements for railroad engineers and conductors). In addition to government regulation, employees' conduct is also closely regulated by the private railroads themselves. See O'Connor v. Ortega, slip op. 6. Operating Rule G has for decades prohibited the on-duty use or possession of alcohol or drugs by railroad employees. The railroads have also taken steps to enforce that rule through supervision and discipline (see pages 2 & note 1, 5, supra), and in some cases through the use of "sniffer" dogs and toxicological testing similar to the FRA procedures. See Brotherhood of Locomotive Engineers v. Burlington N. R.R., 838 F.2d 1087, 1089 (9th Cir. 1988); Railway Labor Executives Ass'n v. Norfolk & W. Ry., 833 F.2d 700, 702-703 (7th Cir. 1987). As required by federal regulation, railroads also train employees on safety rules and periodically test their proficiency in complying with those rules. See 49 C.F.R. 217.9, 217.11. Moreover, the railroads have historically imposed exacting physical requirements for employment in particular posts, including periodic physical exams that typically involve taking blood and urine samples. See Railway Labor Executives Ass'n v. Norfolk & W. Ry., 833 F.2d at 702. /29/ In short, railroad employees are an integral part of an industry that is pervasively regulated, both by the government and by the railroads themselves, all in the service of safe rail passage. As Judge Alarcon correctly concluded (Pet. App. 38a-39a), that regulatory presence significantly diminishes employees' reasonable expectation of privacy. /30/ For precisely those reasons, two courts of appeals have upheld urinalysis testing, in the absence of particularized suspicion, in industries where pervasive regulation has reduced employees' expectations of privacy. See Rushton v. Nebraska Public Power Dist., 844 F.2d 562, 566 (8th Cir. 1988) (nuclear plant engineers); Shoemaker v. Handel, 795 F.2d 1136, 1142 (3d Cir.), cert. denied, 479 U.S. 986 (1986) (race track employees). /31/ 2. The constraints on official discretion This Court has explained that "(t)he essential purpose of the proscriptions in the Fourth Amendment is to impose a standard of 'reasonableness' upon the exercise of discretion by government officials, including law enforcement agents, in order '"to safeguard the privacy and security of individuals against arbitrary invasions * * * "'" (Delaware v. Prouse, 440 U.S. at 653-654 (footnote and citations omitted)). /32/ Thus, as the Court stated in Prouse, "(i)n those situations in which the balance of interests precludes insistence upon 'some quantum of individualized suspicion,' other safequards are generally relied upon to assure that the individual's reasonable expectation of privacy is not 'subject to the discretion of the official in the field'" (id. at 645-655 (footnote omitted), quoting Camara v. Municipal Court, 387 U.S. at 532). Or, as the Court put the matter in its most recent decision concerning a "closely regulated" industry, the regulation must adivse the individual "that the search is being made pursuant to the law and has a properly defined scope, and it must limit the discretion of the inspecting officers" (New York v. Burger, slip op. 11). See also New Jersey v. T.L.O., 469 U.S. at 342 n.8. The post-accident testing procedures in Subpart C easily satisfy that standard. Under those procedures, every covered employee who is directly involved in one of the three triggering events is automatically subject to blood and urine tests. The railroads have virtually no discretion in that regard. Except in those cases of "impact accidents" or "fatal train incidents" in which the railroad "can immediately determine, on the basis of specific information, that the employee had no role in the cause(s) of the accident/incident" (49 C.F.R. 219.203(a)(3)(i)), the railroad may not carve out exceptions, nor may it select particular employees, but not others, to test (id. Section 219.203(a)(2)). Moreover, as Judge Alarcon observed (Pet. App. 41a), the "time, place, and scope" of the testing is limited. See New York v. Burger, slip op. 19; United States v. Biswell, 406 U.S. at 315. The tests must be performed "as soon as possible after the accident or incident" (49 C.F.R. 219.203(b)(1)), and they may only be conducted at "an independent medical facility" by qualified medical personnel (id. Section 219.203(c)(1)). Moreover, consistent with its overall purpose (see id. Section 219. 1), the regulations permit testing to determine only the use of alcohol or drugs; no other kinds of tests are permitted. Subpart D of the regulations, authorizing (but not requiring) breath and urine testing for cause, also tightly confines the railroads' discretion. /33/ Section 219.301(b)(3), which authorizes breath tests for employees who have been directly involved in specified operating rule violations or errors, and 49 C.F.R. 219.301(c)(1), which, in part, authorizes urine testing under identical circumstances, apply to all "covered employees" (id. Section 219.301(a)), permit tests to be conducted only within the first 8 hours following the violation or error in question (id. Section 219.301(f)), and limit the tests to the specified violations. /34/ Similarly, Section 219.301(b)(2), which authorizes breath tests for employees who have been involved in certain reportable accidents or incidents, and Section 219.301(c)(1), which, in part, authorizes urine testing under identical circumstances, likewise apply to all "covered employees" (49 C.F.R. 219. 301(a)), permit tests to be conducted only within the first 8 hours following the accident or incident (id. Section 219.301(f)), and limit the testing to specified accidents or incidents. And while there is an element of discretion under Section 219.301(b)(2) and (c)(1) -- in that a supervisor must determine that there is a reasonable basis to suspect that the employee's actions may have contributed to the occurrence or severity of the event -- that discretion is modest indeed. Unlike the authority to search conferred in Marshall v. Barlow's Inc., 436 U.S. 307 (1978), which "devolve(d) almost unbridled discretion upon executive and administrative officers" (id. at 323), 49 C.F.R. 219.301(b)(2) and (c)(1) require the employer to perform the familiar, and routine, task of assessing the nature of the occurrence and determining, in light of the objective circumstances, whether human error played a part. There is nothing "random, infrequent, or unpredictable" (Donovan v. Dewey, 452 U.S. at 599) about that judgment. 3. The narrowly-tailored nature of the testing procedures Finally, the actual testing procedures adopted by the FRA are narrowly tailored to respect employees' reasonable expectations of privacy. Indeed, the court of appeals agreed that "(t)he manner of conducting the tests is generally reasonable," noting, in particular, that "(t)he intrusiveness of the process of urine testing has been reduced as much as is practicable" (Pet. App. 29a). Under Subpart C, the employer must transport the employee after the accident to "an independent medical facility" where the blood and urine samples will be taken by qualified medical personnel (49 C.F.R. 219.203(c)(1)). The samples, once taken, must be handled in a specified manner (id. Section 219.205(a)) and must thereafter be shipped to the FRA laboratory for analysis (id. Section 219.205(d)). The laboratory employs "state-of-the-art equipment and techniques to detect and quantify alcohol and drugs" (Field Manual B-12), including an initial screening test for drugs and a confirmatory GC/MS test in the event of a positive finding (ibid.; see also 50 Fed. Reg. 31555-31556 (1985)). Once the analysis is completed, the FRA notifies the employee of the result and offers him an opportunity to respond in writing prior to the preparation of a final investigative report (49 C.F.R. 219.211(a)(2)). And an employee who refuses to provide a sample is entitled to an elaborate hearing before any disciplinary action may be taken (id. Section 219.213). The procedures for authorized breath and urine testing under Subpart D are also carefully carfted to respect the privacy fo railroad employees. Only certain designated breath-test devices may be used; the devices must be properly maintained; and the tests must be conducted by qualified personnel (49 C.F.R. 219.303(a)(1), (2), and (3)). If an initial breath test is positive, it must be confirmed at least 15 minutes later (id. Section 219.303(a)(5)). Urine may be collected only at an independent medical facility and the collection must be supervised by personnel of that facility (id. Section 219.305(a)). The medical facility must adopt and follow procedures to ensure positive identification of each sample and accurate reporting of laboratory results (id. Section 219.305(b)). Only an independent laboratory "proficient in the testing of urine for alcohol and drugs of abuse" may perform the actual testing (id. Section 219.307(a)(1)), and the regulations require the laboratories to use rigorous testing methods sufficent to ensure accurate results (id. Section 219.307(b)). Finally, whenever the results of either breath or urine tests are intended for use in a disciplinary proceeding, the employee must be given the opportunity to provide a blood sample for analysis at an independent laboratory (id. Sections 219.303(c), 219.305(d)). The FRA regulations are thus as responsive as practicable to the privacy interests of employees. Indeed, recognizing that "(b)reath and body fluid testing techniques involve varying degrees of perceived 'intrusiveness,'" (49 Fed. Reg. 24276 (1984)), and noting its own "continuing concern over the utility and reasonableness of the post-accident testing program," the agency "substantially reduced the number of events and redefined the types of events that will trigger the requirement of testing" under Subpart C (50 Fed. Reg. 31542 (1985)). /35/ The FRA also tailored the post-accident testing to respect the medical needs of passengers and employees who may be injured in an accident or incident (49 C.F.R. 219.203(b)(2), 219.203(e)). Moreover, in deciding under Subpart D to authorize breath and urine, but not blood, testing, the agency noted that "drawing blood is invasive" in ways that the other two tests are not (50 Fed. Reg. 31556 (1985)). In these and other ways the agency carefully designed the testing procedures "to assure appropriate regulatory balance" (49 Fed. Reg. 24276 (1984)). /36/ We acknowledge that, despite these limitations, some employees may experience some inconvenience in being transported to a medical center for testing or may feel self-conscious about the actual collection process. But we reject as unwarranted the court of appeals' apparent belief (Pet. App. 22a-23a) that blood, urine, and breath testing are substantially intrusive procedures. /37/ As this Court observed in Schmerber v. California, 384 U.S. 757, 771 n.13 (1966) (citation omitted(, "'(t)he blood test procedure has become routine in our everyday life. It is a ritual for those going into the military service as well as those applying for marriage licenses. Many colleges require such tests before permitting entrance and literally millions of us have voluntarily gone through the same, though a longer, routine in becoming blood donors.'" See also Winston v. Lee, 470 U.S. 753, 761-762 & n.5 (1985). Urine and breath tests are still less intrusive, since neither requires an actual invasion of the body surface, and both involve a natural bodily function, routinely (and necessarily) conducted many times during the normal work day. Cf. Mackey v. Montrym, 443 U.S. 1, 19 (1979) (a breath test "provides the most reliable form of evidence of intoxication for use in subsequent proceedings"). Moreover, under the FRA regulations all blood and urine samples must be collected by medical personnel at medical facilities (49 C.F.R. 219.203(c)(1), 219.305(a)). Thus, as in Schmerber, the collection process is conducted "in a hospital environment according to accepted medical practices" (Schmerber, 384 U.S. at 771), and in that critical respect it is not materially different from any number of medical and physical examinations required by employers in the public and private sectors -- including the railroad industry -- as a condition of employment. The courts have approved such tests when imposed as a reasonable condition of employment. See, e.g., McDonell v. Hunter, 612 F. Supp. 1122, 1130 n.6 (S.D. Iowa 1985), aff'd, 809 F.2d 1302 (8th Cir. 1987). Cf. Bratcher v. United States, 149 F.2d 742, 745-746 (4th Cir.), cert. denied, 325 U.S. 885 (1945). C. The FRA Regulations Serve Compelling Governmental Interests 1. As respondents acknowledged below (Br. 29) and as the court of appeals recognized (Pet. App. 24a), the government has a surpassing interest in ensuring the safety of the general public and operating employees by preventing alcohol and drug impairment aboard the nation's railroads. Railroad safety and measures to secure it, as we noted above (pages 26-27, supra), have been a preoccupying concern of Congress since the turn of the century. "The first of the Safety Appliance Acts dates back to 1893, the Signal Inspection Act and the Accident Reports Act to 1910, the Hours of Service Act to 1907, and the Locomotive Inspection Act to 1911." H.R. Rep. 91-1194, 91st Cong., 2d Sess. 8 (1970). More recently, recognizing the enormous risks involved in the industry -- including the "transportation of a variety of volatile, highly toxic, and sometimes frightening hazardous materials" (id. at 9) -- Congress enacted the comprehensive Federal Railroad Safety Act of 1970, 45 U.S.C. (& Supp. III) 421 et seq., endowing the Secretary of Transportation with broad powers to "prescribe, as necessary, appropriate rules, regulations, orders, and standards for all areas of railroad safety" (45 U.S.C. 431(a)), including, without limitation, the power to "test() * * * railroad facilities, equipment, rolling stock, operations, or persons, as he deems necessary to carry out the provisions of this subchapter" (45 U.S.C. 437(a)). /38/ Acting pursuant to that broad authority (see 48 Fed. Reg. 30723 (1983)), the FRA promulgated regulations designed "to prevent accidents and casualties in railroad operations that result from impairment of employees by alcohol or drugs" (49 C.F.R. 219.1(a)). In doing so, the agency meticulously documented the safety issues that prompted its concern. "It is clear," the FRA concluded from the voluminous administrative record before it (50 Fed. Reg. 31514 (1985)), "that alcohol and drug use is sufficiently common to pose a significant safety problem." The evidence, much of which is summarized in the preamble to the regulations, revealed case after case in which alcohol and drug abuse by railroad employees had caused massive property damage, injuries and fatalities, or the release of hazardous materials. See pages 2-7, supra. And there is every reason to believe that "use of alcohol and drugs continues to play a causal role in some of the most serious rail accidents" (J.A. 188). /39/ The FRA concluded that alcohol and drug testing would reduce these tragic incidents, primarily in two ways. First, it explained (50 Fed. Reg. 31541 (1985)), testing would "help to deter employees from using alcohol and drugs on the job." While the agency recognized "that many substance-dependent employees are unlikely to be deterred from bringing their problems to work," it found that "at least half of the alcohol and drug-impaired employees involved in accidents and injuries on the railroads become impaired volitionally" (ibid.). "The disciplinary actions that will result from this program will be vivid examples to other employees who may be tempted to bring alcohol or drugs onto the railroad" (ibid.). /40/ Second, the FRA determined that a program of mandatory post-accident alcohol and drug testing would permit the agency "to determine with greater precision the causes of major accidents of interest to the public" (50 Fed. Reg. 31541 (1985)), and thereby to take further measures to safeguard the general public. "Enlightened and proportional regulation," the agency explained (ibid.), "will only be possible if the true causes of major human factor accidents are known." In the past, as the National Transportation Safety Board has reported (J.A. 131), the "(i)nvestigation of accidents has been hampered because toxicological tests for alcohol-drug use (were) not made after serious railroad accidents when the employees responsible for the operation of the train (were) not fatally injured." The FRA's post-accident testing program should help to cure that deficiency: the agency concluded (50 Fed. Reg. 31541 (1985)) that data derived from post-accident testing "may indicate the need for further alcohol and drug countermeasures" and also "provide guidance on the need for more generalized measures designed to maintain train separation, better train handling or safer yard operations." /41/ Finally, we note that the FRA adopted the alcohol and drug testing regulations only after it had first examined alternative approaches and found them wanting. See pages 5-6 & note 6, supra. While we do not believe that the FRA's consideration of less intrusive alternatives was constitutionally required, /42/ the agency's comprehensive analysis of its options reflects the essential reasonableness of the choices that it made. 2. The court of appeals did not quarrel with the importance of the governmental interests in this case (see Pet. App. 7a), but it discerned a "flaw in the reasonableness of th(e) (FRA's) approach to the problem" (id. at 28a). In particular, the court stated (ibid.), "(b)lood 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." Accordingly, it concluded (id. at 29a), "(r)equiring individualized suspicion would help to alleviate some of the harsh consequences of exclusive reliance on test results." In so reasoning, the court erred in three respects. First, we do not accept the court's premise that urine testing is irrelevant to a determination of current drug intoxication. The FRA noted that urine tests do "indicate whether the person in question is using one or more controlled substances," and it observed that "(t)he drug abuser is * * * a member of a population that presents a higher risk to safety than those who are not memebers of that population." 50 Fed. Reg. 31556 (1985). While that information, standing alone, will not dispositively prove on-the-job impairment, it is nonetheless material information that a railroad is entitled to learn. The FRA accordingly explained that any results from a urine test must be "taken with specific information on the pattern of elimination for the particular drug and other information on the behavior of the employee and the circumstances of the accident" before a judgment can be reached as to how a particular accident occurred (49 Fed. Reg. 24291 (1984)). Second and relatedly, the court of appeals ignored the fact that in each case urine testing will be accompanied by blood testing or the opportunity for a blood test. See 49 C.F.R. 219.203(a), 219.305(d). /43/ After examining the literature in the field and entertaining extensive comments from the public, the FRA concluded (49 Fed. Reg. 2491 (1984)) that a blood sample "can provide a clear indication not only of the presence of alcohol and drugs but also their current impairment effects." As scientists from the National Institute on Drug Abuse testifed during the FRA regulatory proceedings, "(a)lthough urine screening technology is extremely effective in determining drug use, * * * (i)f company policies are tied to performance impairment, or use on the job, then blood would be the body-fluid to assay, since it is generally scientifically accepted that presence of drug in blood indicates very recent use" (J.A. 63 (statement of Richard A. Lindblad and J. Michael Walsh)). The sources relied upon by the court of appeals (Pet. App. 28a) do not say otherwise, since they criticized only the capability of urine testing, and did not contradict the FRA's conclusions about blood testing. /44/ Finally, the court of appeals' qualms about urine testing do not take into account the fact that the FRA regulations are designed not simply to discern on-the-job alcohol and drug use or impairment; they are also intended to deter such use or impairment, by discouraging employees from using alcohol and drugs at a time close enough to their shifts that they may remain impaired when they go on duty. See 50 Fed. Reg. 31541 (1985). Whatever its limitations as a means of discerning on-duty drug use or impairment, urine testing provides a powerful deterrent to the use of drugs by railroad employees who are about to begin a shift. /45/ 3. The court of appeals also surmised (Pet. App. 26a) that it would "pose() no insuperable burden on the government to require individualized suspicion." It noted (ibid.) that a reasonable suspicion standard is already codified as part of Subpart D (49 C.F.R. 219.301(b)(1) and (c)(2)), and it concluded that the same standard "should be incorporated into the mandatory testing provisions set out in 49 C.F.R. Section 219.201 and the other authorized testing provisions set out in 49 C.F.R. Section 219.301(b)(2) and (3)" (Pet. App. 26a-27a). This analysis is flawed at every turn. First and foremost, a requirement of individualized suspicion would effectively preclude testing in the large number of cases in which alcohol and drug abusers betray no outward signs of impairment. As the FRA determined (50 Fed. Reg. 31526 (1985)), "in a significant proportion of cases -- probably a clear majority of the cases involving violations of th(e) final rule -- there may be no external signs detectable by the lay person or, in many cases, even the physician." In particular, "research indicates that social drinkers, bartenders, and even some police officers cannot accurately judge levels of intoxication" (ibid. (citing evidence)). And the FRA found (id. at 31527) that "(i)fdetection of alcohol use is difficult, detecting the use of the wide variety of controlled substances presents a challenge that is even more formidable. Many drugs of abuse effects much more subtle or complex (and sometimes more pernicious) than alcohol" and "most drug abusers will be careful to control their demeanor when there is a threat of detection." See also J.A. 60 (NIDA testimony), 179. Second, a regulation that permits testing only upon a finding of particularized suspicion would be wholly depended on the ability and willingness of supervisors and operating employees to observe their co-workers and demand that abusers undergo the required tests. The FRA found very little reason to rely on such voluntary compliance. Rather, the agency determined that there was a "conspiracy of silence" among railroad employees concerning alcohol and drug use (49 Fed. Reg. 24281 (1984)). In addition, it found that "(t)he threat of being held liable in tort to employees and/or members of the public creates powerful disincentives for railroads either to explore the possible involvement of alcohol or drugs in the course of their own investigations or to report, as alcohol or drug-related, accidents for which other explanations are (or appear to be) sufficient" (48 Fed. Reg. 30726 (1983)). And much railroad work involves the use of small crews, operating over great distances, without the opportunity for sustained observation by peers and supervisors. Third, a particularized suspicion standard is flatly inconsistent with one of the FRA's principal objectives in requiring post-accident testing: to determine, with greater reliability and consistency, the causes of significant accidents and incidents (see 50 Fed. Reg. 31541 (1985)). As the FRA explained (ibid.), "(e)nlightened and proportional regulation will only be possible if the true causes of major human factor accidents are known." Under the court of appeals' standard, however, the agency could not require testing -- even after a major accident, involving fatalities and massive property damage -- unless, in addition to the accident itself, an employee showed individualized signs of impairment. Finally, as in United States v. Martinez-Fuerte, 428 U.S. 543, 557 (1976), imposing a requirement of particularized suspicion would reduce the deterrent effect of the alcohol and drug-testing regulations, by permitting abusers to persist in their habit until sufficient evidence accumulates against them. See J.A. 190. If the nation's railroads must await palpable evidence of on-duty impairment before they may intervene, the result could be an incalculable sacrifice of safety to employees and the public. As Judge Alarcon observed (Pet. App. 38a), "(a)n idle locomotive, sitting in the roundhouse, is harmless. It becomes lethal when operated negligently by persons who are under the influence of alcohol or drugs." In view of the minimally intrusive nature of the FRA regulations, and the surpassing importance of their objectives, there is no basis for requiring the railroad industry, and the general public, to absorb such costs. CONCLUSION The judgment of the court of appeals should be reversed. Respectfully submitted. CHARLES FRIED Solicitor General JOHN R. BOLTON Assistant Attorney General THOMAS W. MERRILL Deputy Solicitor General JAMES M. SPEARS ROBERT J. CYNKAR Deputy Assistant Attorneys General LAWRENCE S. ROBBINS Assistant to the Solicitor General LEONARD SCHAITMAN MARC RICHMAN Attorneys B. WAYNE VANCE General Counsel Department of Transportation S. MARK LINDSEY Chief Counsel GREGORY B. MCBRIDE Assistant Chief Counsel DANIEL CAREY SMITH Deputy Assistant Chief Counsel Federal Railroad Administration JULY 1988 /1/ The FRA found that "(v)irtually all railroads have adopted Rule G in one form or another" and it noted that "(t)he customary sanction for violation of Rule G is dismissal from employment." 48 Fed. Reg. 30727 (1983). /2/ The FRA noted, moreover (48 Fed. Reg. 30726 (1983)), that the available figures significantly understated the extent of the problem, since the threat of being held liable in tort and the risk of losing one's job discouraged the railroads and employees from accurately reporting the cause of accidents. /3/ Table 1 of the NPRM (49 Fed. Reg. 24256 (1984)) contains a breakdown of the 45 accidents, showing the number of fatalities and injuries and the extent of the damage to railroad property. The FRA also analyzed those accidents in finer detail (id. at 24257-24264). For example, Table 2 (id. at 24257-24260) examines the accidents specifically investigated by the FRA. Those include a 1982 derailment in Livingston, Louisiana (id. at 24259), in which, as a result of alcohol consumption by the engineer and front brakeman, hazardous material was released into the atmosphere, causing "the evacuation of an entire community of 2,700 persons, some of whom were unable to return to their homes for an extended period" (id. at 24254). Also included is a 1980 rear end collision in Pisgah, California, in which, as a result of alcohol consumption, an engineer operated the train at excessive speed and a brakemen failed to take appropriate action (id. at 24259), "result(ing) in the release of a combustible liquid from the dome of a tank car that burned for 18 hours" (id. at 24265). Table 4 of the NPRM lists employee fatalities investigated by FRA between 1975 and 1982 and sets out the laboratory results of alcohol and drug testing, as well as the nature of the accident in each case (id. at 24262-24264). /4/ The FRA further observed (49 Fed. Reg. 24254 (1984)) that "the documented data tell only a part of the story. Many alcohol and drug-tested accidents and injuries are not so recorded under the existing reporting system. From available information, it appears highly probable that because of the latitude present in that system the railroads either fail to detect or fail to report alcohol and drug involvement in a significant number of cases." The agency noted (ibid.) that "of 15 significant train accidents identified by (the National Transportation Safety Board) or FRA investigations as involving alcohol or drugs, the respective railroads reported alcohol or drug involvement in only 6." And it added (ibid.) that "(t)he under-reporting of alcohol and drug involvement is likely even more pronounced in the vast majority of accidents which do not occasion a Federal investigation." /5/ The FRA explained (49 Fed. Reg. 24267 (1984)) that "(s)uch broad discretion need not be abused to create an impression of unfairness that can erode the perceived legitimacy of the rule. It is enough if employees believe such discretion is subject to unchecked abuse." /6/ The FRA also considered several other regulatory options, including requiring railroads to report all Rule G violations; making improvements in railroad life style; using motor vehicle records to identify alcohol and drug abusers; installing devices to halt trains where the engineer is incapacitated; establishing a toll free number that employees could use to report safety violations on an anonymous basis; and requiring all railroad employees to ride at the front end of each train, on the assumption that the presence of more employees would discourage the use of drugs or alcohol. 49 Fed. Reg. 24284-24286 (1984). /7/ These included a rear end collision in Greystone, New York, in which an engineer, who later tested positive for barbiturates, failed to control the train in accordance with signal indications; a head-on collision in Wiggins, Colorado, in which an engineer was intoxicated, in which five persons were killed and two others injured, and in which nearly $4 million in damage to railroad property was sustained; and a derailment in Silver Bow, Montana, in which one person was killed, three others were injured, and over $1.3 million in damage to railroad property occurred. 50 Fed. Reg. 31520 (Table 2) (1985). See also J.A. 132-133 (letter prepared by the National Transportation Safety Board). /8/ These included a rear end collision in Newcastle, Wyoming, in which an engineer, who later tested positive for marijuana, failed to control the train in accordance with signal indications, and in which two persons died, two others were injured, and more than $1.3 million in damage to railroad property occurred; a rear end collision in Carbondale, Illinois, in which both an engineer and front brakeman failed to take appropriate action; a side collision in Camden, Arkansas, in which a fireman operating the train tested positive for marijuana and an engineer, who left the scene of the accident, was observed to have consumed alcohol; and a side collision in Alvarado, Texas, in which the front brakeman, who failed to take appropriate action, tested positive for alcohol, marijuana, and methamphetamine, the engineer, who failed to control the train, appeared to witnesses to have consumed alcohol, and in which one person died, another person was injured, and nearly $2 million in damage to railroad property took place. 50 Fed. Reg. 31521 (Table 2a) (1985). The FRA "call(ed) attention" to these accidents "because they indicate the possible involvement of alcohol and drugs in some of the more serious events of the past year and the difficulty faced by the investigating agencies, particularly in the absence of clear procedures to ensure that toxicological samples will be promptly obtained" (id. at 31516). See also J.A. 131 (National Transportation Safety Board). /9/ The present litigation principally involves Subparts C and D of the regulations. In addition to those sections, Subpart A (J.A. 8-16) sets out some general provisions (49 C.F.R. 219.1 to 219.21), Subpart B (Pet. App. 57a-58a) states a general prohibition of alcohol and drug use (49 C.F.R. 219.101, 219.103), Subpart E establishes policies for identifying alcohol and drug abusers and referring them for treatment (id. Sections 219.401 to 219.407), and Subpart F creates a system for pre-employment screening (id. Sections 219.501 to 219.505). The constitutionality of Subparts A, B, E, or F is not presently at issue. /10/ The FRA concluded that the three triggering events were "of substantial public interest" and were "accidents for which, based on FRA's experience, causal determination is often extremely difficult." 50 Fed. Reg. 31542 (1985). The agency recognized, however, that the "burdens on employees and the railroads should be subject to reasonable limitations" (id. at 31543). It therefore adopted a "scaled-down" version of the triggering events relative to what had been initially proposed (see id. at 31542-31453). See pages 34-35 note 35, infra. /11/ The regulations also provide that railroads must test "each and every operating employee assigned as a crew memeber of any train involved in the accident or incident" (49 C.F.R. 219.203(a)(2); Pet. App. 60a), unless, in the case of an "impact accident" or "fatal train incident," the "railroad representative can immediately determine, on the basis of specific information, that the employee had no role in the cause(s) of the accident/incident" (49 C.F.R. 219.203(a)(3)(i)). FRA's decision to cover all operating employees in a particular train crew "reflects the fact that operating employees are most often at fault in alcohol and drug-related accidents and that some alcohol and drug-caused accidents in the past have involved apparent sequential or simultaneous failures of performance by two or more crew members. It also reflects the extreme difficulty of distinguishing fault and degrees of fault immediately after the more substantial accidents subject to this section." 50 Fed. Reg. 31544 (1985). /12/ See also J.A. 63 (testimony from the National Institute on Drug Abuse). /13/ The regulations make clear, however, that, in the aftermath of an accident, a covered employee must be permitted to discharge whatever duties "may be necessary for the preservation of life or property" (49 C.F.R. 219.203(b)(2); Pet. App. 61a). Moreover, "(n)othing in this subpart shall be construed to limit the discretion of a physician to determine whether drawing a blood sample is consistent with the health of an injured employee or an employee afflicted by any other condition that may preclude drawing the specified quantity of blood" (49 C.F.R. 219.203(e); Pet. App. 62a). /14/ Those regulations, in turn, make reference (49 C.F.R. 219.205(a) and (c); Pet. App. 62a-63a) to the FRA Field Manual -- a booklet prepared by the agency that provides guidance to the industry in complying with the regulations. /15/ Unlike the Customs drug-testing program at issue in National Treasury Employees Union v. von Raab, No. 86-1879, there is no prohibition on the release of FRA testing results to prosecutors. Compare Pub. L. No. 100-71, Section 503(e), 101 Stat. 471. /16/ At a disqualification hearing -- whose format was derived, in part, from comments submitted by the American Civil Liberties Union (50 Fed. Reg. 31546 (1985)) -- an employee "can contest whether there was a refusal, whether the accident or incident required testing (or whether, in a marginal case, the railroad representative made a good faith determination that it did), and whether, in a case where blood was refused based on health concerns, the refusal was made in good faith and based on medical advice" (ibid.). /17/ The FRA made clear that an immunoassay would not constitute an acceptable confirmatory test (49 C.F.R. 219.307(b)) because immunoassay tests "generally carry the potential for a known percentage of 'false positive' results." 50 Fed. Reg. 31555 (1985). Instead, although the FRA did not specifically prescribe the use of the gas chromatography/mass spectrometry test -- in order not to "usurp the role of the scientific community in establishing the reliability of existing assays and development of new assays" (id. at 31556) -- the agency noted that GC/MS is generally "the current state of the art with respect to confirmation of drugs of abuse" (id. at 31555) and observed that its own laboratory would be using GC/MS in conducting the confirmatory testing of post-accident samples under Subpart C (id. at 31556). /18/ As the National Transportation Safety Board (NTSB) (an independent federal agency responsible for investigating major transportation accidents) explained in support of the FRA regulations, "one of the major difficulties in addressing the alcohol/drug abuse problem continues to be inadequate statistics." Moreover, the Board reported, "(i)nvestigation of accidents has been hampered because toxicological tests for alcohol-drug use are not made after serious railroad accidents when the employees responsible for the operation of the train are not fatally injured." J.A. 130-131. /19/ The NTSB agreed that post-accident blood and urine testing, as proposed by the FRA, "would be a powerful incentive to comply with the existing rules prohibiting alcohol and drug use" (J.A. 135). /20/ The district court rejected (Pet. App. 51a-52a) as meritless repondents' other constitutional and statutory challenges to the regulations. /21/ The court of appeals did not invalidate 49 C.F.R. 219.301(b)(1) and (c)(2), which authorize breath and urine testing on a "reasonable suspicion" of alcohol or drug impairment or intoxication. /22/ In insisting on a requirement of particularized suspicion, the court of appeals rejected (Pet. App. 16a-17a) the district court's finding "that the() regulations should be evaluated under the standards applicable to administrative inspections of pervasively regulated industries." The court explained (id. at 18a) that "(a)ll of the decisions in this line of cases have upheld warrantless searches of property, not of persons," and it "decline(d) to make such an extension in this case." It also stated (id. at 19a) that "(a)lthough some railroad safety regulations are directed at employees, * * * the vast bulk of safety legislation is directed at owners and managers of railroads, not their employees." /23/ The court agreed, however (Pet. App. 29a), that "(t)he manner of conducting the tests is generally reasonable in that they are performed in medical facilities. 49 C.F.R. Section 219.203(c)." And the court acknowledged (ibid.) that "(t)he intrusiveness of the process of urine testing has been reduced as much as is practicable in that only personnel of the medical facility may supervise the sample collection. 49 C.F.R. Section 219.305(a)." The court also rejected respondents' other challenges to the FRA regulations predicated on statutory grounds (Pet. App. 32a-34a), the right to privacy (id. at 34a-35a), and the equal protection component of the Due Process Clause (id. at 35a-36a). /24/ The court itself acknowledged (Pet. App. 30a) that its decision "may be seen as conflicting with decisions of other circuits" and it criticized or distinguished (id. at 31a-32a) decisions of the Fifth Circuit in National Treasury Employees Union v. von Raab, 816 F.2d 170 (1987), cert. granted, No. 86-1879 (Feb. 29, 1988); the Seventh Circuit in Division 241 Amalgamated Transit Union v. Suscy, 538 F.2d 1264, cert. denied, 429 U.S. 1029 (1976); the Eighth Circuit in McDonell v. Hunter, 809 F.2d 1302 (1987); and the Third Circuit in Shoemaker v. Handel, 795 F.2d 1136, cert. denied, 479 U.S. 986 (1986). /25/ A copy of the government's brief in von Raab has been served on respondents. /26/ Because post-accident testing under Subpart C requires the taking of a blood sample, it must be regarded as a search within the meaning of the Fourth Amendment. See Schmerber v. California, 384 U.S. 757, 767 (1966). The breath and urine testing procedures under Subpart D, however, present more complicated questions. Because those procedures merely authorize, but do not require, testing by private railroads (see 50 Fed. Reg. 31551 (1985)), we submit that they do not implicate the Fourth Amendment at all. The "exercise of the choice allowed by * * * law(,) where the initiative comes from (a private party) and not from the State, does not make its action in doing so 'state action' * * *." Jackson v. Metropolitan Edison Co., 419 U.S. 345, 357 (1974) (footnote omitted). See also Flagg Bros. v. Brooks, 436 U.S. 149, 160 n.10, 164-166 (1978); Rendell-Baker v. Kohn, 457 U.S. 830, 841-842 (1982); Blum v. Yaretsky, 457 U.S. 991, 1004 (1982). Rather, "a State normally can be held responsible for a private decision only when it has exercised coercive power or has provided such significant encouragement, either overt or covert, that the choice must in law be deemed to be that of the State" (ibid.). The court of appeals adduced no such "coercive power" or "significant encouragement" in its analysis of the Subpart D regulations. We recognize that this Court, in Railway Employees Dep't v. Hanson, 351 U.S. 225, 232 (1956), held that certain otherwise private conduct was state action because federal law (there, the Railway Labor Act, 45 U.S.C. 152) had preempted state statutes forbidding the private conduct. Here, the FRA promulgated Subpart D in part to preempt state and local rules that might otherwise have prohibited breath and urine testing by private employers (see 49 Fed. Reg. 24294 (1984)). See also 49 C.F.R. 219.13(a). However, for the reasons set out in our brief (at 24-30) in Communications Workers v. Beck, No. 86-637, a copy of which has been supplied to respondents, we do not believe that the state-action analysis in Hanson should be applied outside its particular context; this Court did not say otherwise in its decision in Beck, as it had no occasion to decide the question (see Communications Workers v. Beck, No. 86-637 (June 29, 1988), slip op. 24). If the Court concludes that the Subpart D regulations do constitute state action, however, then we believe that those regulations should be regarded as authorizing a search within the meaning of the Fourth Amendment. In our brief in von Raab (at 24-25 n.18), we suggested that a urine test required as part of pre-employment fitness examination, where the employee knows of the test in advance, entails such a minimal interference with any reasonable expectation of privacy that it may not be a search under the Fourth Amendment. The post-accident and post-incident testing authorized by Subpart D, however, does not share those or substantially equivalent features, and thus, assuming it can be regarded as state action, should be considered a Fourth Amendment search. /27/ In addition to the expansive scope of federal regulation, the railroad industry has historically been subject to wide-ranging safety regulation by the States, in the exercise of their traditional police powers. As the Court declared more than 100 years ago, "(c)ommon carriers exercise a sort of public office, and have duties to perform in which the public is interested," Munn v. Illinois, 94 U.S. 113, 130 (1877). Under that broad conception of the police powers, the Court has upheld a vast array of state railroad regulation. See, e.g., Brotherhood of Locomotive Firemen & Enginemen v. Chicago, R. I. & P. R.R., 393 U.S. 129 (1968) (state "full crew" law); Lehigh Valley R.R. v. Board of Public Utility Comm'rs, 278 U.S. 24 (1928) (state plan requiring railroad to bear large costs for eliminating grade crossing); Pacific Gas & Elec. Co. v. Police Court, 251 U.S. 22 (1919) (city ordinance requiring railroad to sprinkle water on surface of street to prevent accumulation of dust); Great N. Ry. v. Minnesota, 246 U.S. 434 (1918) (state decision requiring railroad to build sidewalk where its right-of-way crosses the street); Seaboard Air Line Ry. v. Railroad Comm'n, 240 U.S. 324 (1916) (state order requiring railroad to make and maintain a particular track crossing); St. Louis & S.F. Ry. v. Mathews, 165 U.S. 1 (1897) (state statute imposing insurer's liability on railroad for damages to property injured by fire from locomotive engines); Missouri P. Ry. v. Mackey, 127 U.S. 205 (1888) (state statute imposing liability on railroads for all damage to employees due to negligence of fellow servants). /28/ As noted above (page 8, supra), the FRA regulations apply to those employees who are subject to the Hours of Service Act (49 C.F.R. 219.5(d) and (e)). Thus, the testing at issue in this case is limited to those employees whose fitness for duty has been a matter of congressional concern for more than 80 years. /29/ The court of appeals acknowledged (Pet. App. 21a) that the railroad industry has a practice of imposing health and fitness requirements, but it discounted that consideration because, as the court put it, fitness requirements "have been matters of private negotiation between railroads and employees not matters of federal regulation." But in assessing an employee's expectation of privacy, it should not matter that it is the employer, rather than the government, that constricts the realm of otherwise private conduct. The "expectations of (privacy of) employees in the private sector * * * may be reduced by virtue of actual office practices and procedures, or by legitimate regulation." O'Connor v. Ortega, slip op. 6. /30/ The court of appeals recognized that "the railroad industry has experienced a long history of close regulation" (Pet. App. 19a). It nevertheless found that factor unpersuasive, because "the vast bulk of safety legislation is directed at owners and managers of railroads, not their employees" and thus the court did not believe that federal regulation "has diminished the individual railroad employee's expectation of privacy in his person or his body fluids" (ibid. (emphasis in original)). It is not surprising, however, that in a capital intensive industry like railroading, the "bulk" of safety regulations would relate to the fixed facilities and equipment, not to personnel. But as this Court explained in New York v. Burger, slip op. 13 n.16, "the sheer quantity of pages of statutory material is not dispositive of this question." Rather, the issue is whether a railroad employee, engaged in a business already subject to far-reaching regulation, could reasonably find it intrusive when his employer -- or his employer acting at the government's behest -- requires modest toxicological test in the further service of railroad safety. /31/ The court of appeals distinguished the Shoemaker case, in part because the court concluded (Pet. App. 21a) that the Secretary of Transportation, unlike the Racing Commission in Shoemaker, lacked the authority to prescribe qualification standards for railroad employees. In fact, however, the very provision on which the court relied in drawing that conclusions expressly excepts from the stated limitation "such qualifications as are specifically related to safety." 45 U.S.C. 432 (a). /32/ See also United States v. Ortiz, 422 U.S. 891, 895-896 (1975); United States v. Brignoni-Ponce, 422 U.S. 873, 882 (1975). /33/ The court of appeals upheld (Pet. App. 26a-27a) two provisions in Subpart D -- 49 C.F.R. 219.301(b)(1) and (c)(2) -- which authorize, respectively, breath and urine tests upon a finding of reasonable suspicion of drug or alcohol impairment. We therefore do not discuss those provisions in this section. /34/ Moreover, in selecting the particular safety rule violations that would trigger authorized testing, the FRA selected only those violations that "involve() the potential for a serious train accident or grave personal injury, or both." 50 Fed. Reg. 31553 (1985). /35/ In response to critical comments, the agency, among other things, raised the threshold of property damage from $150,000 to $500,000, concluding that a lower threshold would have required testing in many cases that do not involve alcohol or drug abuse (50 Fed. Reg. 31542 (1985)). It also adopted a "scaled-down" requirement of testing after an on-duty fatality, electing not to require testing after any loss of eye or limb because to do so would encompass many cases in which the injury happened "somewhat unpredictably" (id. at 31543). In addition, the FRA limited testing after "impact accidents" to instances in which there is either a reportable injury or damage of at least $50,000. The agency acknowledged that impact accidents usually result from "human performance failures," but it nevertheless restricted the scope of this triggering event because it "recognize(d) that burdens on employees and the railroads should be subject to reasonable limitations" (ibid.). /36/ What is more, Subpart E of the regulations, which is not at issue in this case, affords employees that have a substance abuse problem an opportunity to obtain counseling and rehabilitation before a positive test result will lead to disciplinary action. /37/ Indeed, the court of appeals suggested (Pet. App. 22a) that urinalysis is as "degrading" and "offen(sive to) human dignity" as body cavity searches. /38/ More generally, this Court's cases confirm the government's compelling interest in assuring the safety of the nation's transportation systems. See, e.g., Brock v. Roadway Express, Inc., No. 85-1530 (Apr. 22, 1987) (plurality opinion), slip op. 8 ("accepting as substantial the Government's interest in promoting highway safety"); South Dakota v. Dole, No. 86-260 (June 23, 1987), slip op. 5; New York v. Class, 475 U.S. 106, 118 (1986) ("governmental interest in highway safety * * * is of the first order"); Mackey v. Montrym, 443 U.S. 1, 17 (1979) (describing "the paramount interest * * * in preserving the safety of (the) public highways"); New York Transit Authority v. Beazer, 400 U.S. 568, 591 (1979) (footnote omitted) ("the 'no drugs' policy now enforced by (the Transit Authority) is supported by the legitimate inference that as long as a treatment program (or other drug use) continues, a degree of uncertainty persists"); Atchison, T. & S. F. Ry. v. United States, 244 U.S. 336, 342-343 (1917) (describing the Hours of Service Act, which placed a ceiling on the number of consecutive hours that railroad employees could work, as "essential to public and private welfare"); Missouri P. Ry. v. Mackey, 127 U.S. 205, 210 (1888) ("the hazardous character of the business of operating a railway would seem to call for special legislation with respect to railroad corporations, having for its object the protection of their employees, as well as the safety of the public"). /39/ Perhaps the most notable of these recent accidents is the January 4, 1987 collision between an Amtrak passenger train and three Conrail locomotives at Chase, Maryland. In that accident, sixteen persons died and another 174 were injured. "Post-accident test results showed that the engineer and brakeman of the Conrail locomotives -- who failed to observe signal indications and ran through a switch into the path of the high-speed Amtrak train -- had marijuana metabolites in their blood and urine after the accident. (The brakeman also tested positive for PCP in the urine.)" J.A. 188. The NTSB concluded that "the leading contributing cause of this accident was the Conrail engineer's impairment by marijuana" (ibid). /40/ There is good reason to believe that the FRA post-accident testing program has begun to achieve some of its desired deterrent effects. From February 10, 1986, through December 31, 1987, 349 events occurred that qualified for post-accident testing. Of the employees tested pursuant to Subpart C, a total of 88 (or 5.8%) tested positive for either alcohol or drugs. Although the comparison is somewhat inexact, that figure compares favorably to the 16% positive rate found in the seven-year period prior to the issuance of the regulations for employees killed in railroad accidents and incidents. J.A. 187. /41/ Joseph W. Walsh, Associate Administrator for Safety for the FRA, has stated that in the absence of post-accident testing, government officials would not have been able to determine the cause of the Chase, Maryland collision. J.A. 189. /42/ As we explain at greater length in our brief in the von Raab case (86-1879 Br. 39), "(t)he reasonableness of any particular governmental activity does not necessarily or invariably turn on the existence of alternative 'less intrusive' means." Illinois v. Lafayette, 462 U.S. 640, 647 (1983). /43/ There may be instances in which, due to the nature of the particular accident, "the interval between the accident and time blood is drawn may be greater than the time required for certain drugs to be eliminated from the blood." 49 Fed. Reg. 24291 (1984). In such cases, a urine test may provide more useful data. /44/ For example, in Jones v. McKenzie, 833 F.2d 335 (D.C. Cir. 1987), one of the sources relied upon by the court of appeals, the D.C. Circuit considered the constitutionality of a urinalysis program for District of Columbia bus drivers. At the page cited by the court below, the D.C. Circuit explained that the immunoassay screening test employed by the District was inadequate because it could not show on-duty impairment. The D.C. Circuit did not, however, consider the sufficiency of any more precise urinalysis methodology, nor did it have before it a testing program, like the FRA program, that relies on blood as well as urine samples. The other sources cited by the court of appeals are equally unhelpful. While Dr. Dubowski did state, at the pages cited by the court below, that urine testing cannot discern present drug impairment, he also explained elsewhere in his article that "(b)ecause the blood stream is the primary pathway for drug distribution throughout the body, presence of a drug in the blood or its components usually signifies recent intake into the body, particularly when it is found in relatively high concentration and in the form of the parent drug rather than as its metabolites or biotransformation products." Dubowski, Drug-Use Testing: Scientific Perspectives, 11 Nova L. Rev. 415, 427 (1987). The article written by Karen Hudner, as its title suggestd, likewise dealt only with urine testing, and not with blood testing. Hudner, Urine Testing for Drugs, 11 Nova L. Rev. 553 (1987). The same is true of the remaining article, authored by Paul Joseph. Joseph, Fourth Amendment Implications of Public Sector Work Place Drug Testing, 11 Nova L. Rev. 605, 632 & nn. 91-93 (1987). /45/ We note that blood testing is appropriate where, as here, one central purpose of testing employees is to determine whether drug or alcohol use was the cause of a particular accident. On the other hand, where, as in cases such as the program adopted by the Customs Service in National Treasury Employees Union v. von Raab, No. 86-1879, employee testing is designed to determine whether an employee is a drug user, and thus whether he presents a significant risk to safety or security, blood testing would be less useful than urine testing, since drug residues generally pass out of the blood only a few hours after use, while they remain in the urine for longer periods. See 49 Fed. Reg. 24291 (1984).