This Nation’s struggle to combat the use of narcotic drugs has taken a number of forms, including education, increased expenditures for treatment and law enforcement, and the establishment of methods aimed at detecting drug abuse among workers. In this case we confront challenges to a rule promulgated by the Research and Special Programs Administration (RSPA) of the United States Department of Transportation (DOT). The rule requires extensive drug testing of employees engaged in natural gas, liquefied natural gas, and hazardous liquid pipeline operations. The International Brotherhood of Electrical Workers, Local No. 1245 (IBEW) and Oil, Chemical and Atomic Workers International Union (OCAWIU) bring this petition for review of the rule, contending *1456that it is arbitrary and capricious and unconstitutional. We affirm.
I.
FACTS AND RULEMAKING PROCEEDINGS
On July 8, 1988, RSPA issued a notice of proposed rulemaking entitled “Control of Drug Use in Natural Gas, Liquefied Natural Gas and Hazardous Liquid Pipeline Operations.” 53 Fed.Reg. 25,892 (1988). The rule called for pipeline operators to institute five different types of drug testing: (1) pre-employment; (2) post-accident; (3) randomly; (4) on the basis of reasonable cause; and (5) post-rehabilitation. Id. at 25,898-900. RSPA supported the proposed rule by citing studies regarding substance abuse and its relation to motor vehicle accidents. Id. at 25,896. RSPA also noted, however, that the number of pipeline accidents was small and that it had no evidence of a drug problem in the pipeline industry that was any greater than in the general population. Id. at 25,893.
Interested parties who commented on the rule addressed a wide range of issues, including the incidence of drug use among pipeline employees and the determination of functions for which testing should be required. Id. at 47,084-96. The random testing requirement drew special criticism. Commenters noted that there was insufficient evidence to warrant such testing. See id. at 47,086. In addition, safety standards committees1 disapproved the rule as drafted. Id. at 47,095. The Technical Hazardous Liquid Pipeline Safety Standards Committee unanimously rejected the rule as unsupported by demonstrated need. The Technical Pipeline Safety Standards Committee conditioned its approval on certain changes to the rule, including the elimination of random drug testing. 53 Fed.Reg. 47,095 (1988).
In conjunction with similar rulemaking by five other agencies of DOT,2 RSPA issued its final rule on November 21, 1988, accompanied by a report that discussed comments made by interested parties. Id. at 47,084. In this November 1988 final rule action, RSPA responded to the following issues raised during the notice and comment period: (1) the constitutionality of the rule; (2) the need for a pipeline anti-drug program; (3) the accuracy of drug test results; (4) the employees required to be tested; (5) pre-employment testing; (6) random testing; (7) post-accident testing; (8) reasonable cause testing; and (9) retesting.3 RSPA concluded generally that “[t]he majority of the commenters were opposed to one or more aspects of the proposed rule, while some commenters generally supported [it].” Id. Despite this generally adverse reaction, the agency issued the rule.
On April 13, 1989, RSPA announced that it would reevaluate the rule and delayed the dates on which the rule would take effect. 54 Fed.Reg. 14,922 (1989).4 RSPA announced on December 18, 1989, that it had completed its reevaluation; it did not alter the coverage or scope of the rule in a material way. See 54 Fed.Reg. 51,842 (1989). RSPA also reaffirmed that the rule would become effective on April 20, 1990, and August 21, 1990, depending on the size of the affected enterprise. Id.
Petitioners raise constitutional and statutory challenges to certain aspects of this regulation. They contend first that the rule generally is arbitrary and capricious because RSPA has not demonstrated a safety need in the pipeline industry that justifies imposition of drug testing. They next argue that the random drug testing *1457component of the rule is arbitrary because it is not narrowly tailored to address the alleged problem. Finally, they allege that random testing is unconstitutional because it unreasonably intrudes on those privacy interests of the employees that are protected by the Fourth Amendment.
II.
JURISDICTION
RSPA has statutory authority to issue rules pursuant to 49 U.S.C.App, §§ 1672, 1674a & 2002 (1988).5 Our court has jurisdiction under 49 U.S.C.App. §§ 1675(a) & 2005(a) (1988).
III.
DISCUSSION
In this appeal we consider three issues: (1)whether the rule requiring drug testing in the pipeline industry constitutes decision making that is arbitrary, capricious, and an abuse of discretion in violation of the Administrative Procedure Act, 5 U.S.C. § 706(2)(A) (1988); (2) whether the rule is arbitrary and capricious to the extent that it requires random drug testing without individualized suspicion; and (3) whether the provision for random drug testing violates the Fourth Amendment.6 We address each issue in turn.
A. The Arbitrariness of the Drug Testing Rules Generally.
The Administrative Procedure Act requires us to “hold unlawful and set aside agency action ... found to be — (A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A) (1988). We employ two criteria to determine whether an administrative rule is “arbitrary and capricious.” First, the agency must “explain[] specifically” the reason for the rule. Bluestein v. Department of Transp., 908 F.2d 451, 457 (9th Cir.1990). See Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 2866, 77 L.Ed.2d 443 (1983). Second, the decision to establish the rule must be reasonable. “[A] reasonable, if controversial, decision ... cannot be overturned as arbitrary and capricious.” Bluestein, 908 F.2d at 456.
We assess these criteria in light of Congress’ directive that minimum federal safety standards “be practicable and designed to meet the need” for pipeline safety and the safe transportation of hazardous liquids. 49 U.S.C.App. §§ 1672(a) & 2002(a) (1988). Under these statutory provisions, the agency must consider the following factors:
(1) relevant available pipeline safety data;
(2) whether such standards are appropriate for the particular type of pipeline transportation or facility;
(3) the reasonableness of any proposed standards; and
(4) the extent to which such standards will contribute to public safety.
Id. § 2002(b)(l)-(4); see id. § 1672(á)(l)(A)-(D). The legislative history of these provisions indicates that Congress intended to guide the agency’s exercise of its rulemaking authority and to avoid the promulgation of unreasonable rules not supported by a demonstrated need in the industry. See H.R.Rep. No. 1390, 90th Cong., 2d Sess. 19-20 (1968), reprinted in 1968 U.S.Code Cong. & Admin.News 3223, 3235-36; S.Rep. No. 733, 90th Cong., 1st Sess. 7-8 (1967). We shall discuss each of these factors.
1. The Safety Need.
Here we confront the major thrust of the petitioners’ administrative law challenge. *1458They contend that RSPA can document no specific drug problem in the pipeline industry sufficient to justify testing. RSPA’s own data, they argue, show that in 1988, less than four percent of the 454 pipeline accidents and none of the twenty fatalities were attributable to operator error; and that in the three-year period from 1985 to 1988, only two fatalities were caused by human error.7 In petitioners’ view, not only is there no evidence of a drug problem, but there is no support for a safety problem generally.
RSPA accepts, as it must, the accuracy of its own data. It responded in the rule-making phase by asserting that “no commenter presented any statistically reliable data to prove there was not a drug problem [in the pipeline industry].” 53 Fed.Reg. 47,087 (1988) (emphasis added). Given the widespread societal problem of drug abuse, RSPA concluded that some drug abuse very likely exists in the pipeline industry. It is not fatal to RSPA’s argument that it cannot demonstrate that the pipeline industry has a specific drug problem. That failure alone does not establish the arbitrariness of its regulations. See, e.g., Bluestein, 908 F.2d at 457 (rejecting APA challenge of random tests in airline industry).
Even were we to assume that the drug problem in the pipeline industry is less significant than it is among the general population, the danger inherent in pipeline operations leads us to be wary of substituting our judgment for RSPA’s. The industry’s safety record is imperfect and terrible accidents have occurred. We may analogize to the nuclear power industry, for which the courts have upheld drug testing programs similar to the one at issue here. See, e.g., Rushton v. Nebraska Pub. Power Dist., 844 F.2d 562, 566 (8th Cir.1988) (upholding plan mandating pre-employmnet, pretransfer, annual, for-cause, and random drug tests); Alverado Washington Pub. Power Supply Sys., 111 Wash.2d 424, 441, 759 P.2d 427, 436 (1988) (en banc) (upholding testing in pre-employment and for-cause situations), cert. denied, — U.S. —, 109 S.Ct. 1637, 104 L.Ed.2d 153 (1989). Given the potentially catastrophic consequences of a pipeline accident, we cannot say that the rule does not relate to a perceived need for safety regulation in the industry.8
2. Appropriateness of the Standards.
In assessing the “appropriateness” of the rule, we must consider the twin concerns of practicality and the need for pipeline safety. See H.R.Rep. No. 1390, 90th Cong., 2d Sess. 20, reprinted in 1968 U.S. Code Cong. & Admin.News 3223, 3236. Petitioners argue that the rule is inappropriate because RSPA requires testing of all employees engaged in operations, maintenance, or emergency response functions, a total of fifty percent of pipeline employees (116,000). This charge of over-inclusiveness falls short because petitioners are unable to specify which employees are in safety-sensitive positions for which testing would be appropriate. When neither the petitioners nor the industry can draw dividing lines of any sort, we cannot conclude that the government’s lines are arbitrary and capricious.
Instead of delineating job titles or positions, the regulations specify job functions. An “employee” is defined as “a person who performs on a pipeline or LNG facility an operating, maintenance, or emergency-response function regulated by Part 192, 193, or 195 of this chapter.” 49 C.F.R. § 199.3 (1989).9 These “functions” appropriately *1459specify the personnel whose impairment would pose the greatest risk to public safety.10
3. The Reasonableness of the Standards.
The parties rest their arguments concerning the factor of “reasonableness” on a cost-benefit analysis of the program. Not surprisingly, they disagree about how much money will be spent and saved if the program goes forward. Although we agree with petitioners that the cost-benefit figures are uncertain, we nonetheless regard the drug testing program as a reasonable measure properly geared to safety needs. The legislative history of this provision supports our view that public safety, and not cost, shall be the predominant concern. “In determining reasonableness, safety, which is the purpose of this act, shall be the overriding consideration.” S.Rep. No. 733, 90th Cong., 1st Sess. 8 (1967).11 This is not to say that some costs, properly demonstrated, might not be too great considering their marginal contribution to safety. We can only say that petitioners have made no such demonstration here.
4. Contribution of the Rule to Public Safety.
Petitioners make no real effort to deny that the drug testing rule would make some contribution to public safety as required by 49 U.S.C.App. §§ 1672(a)(1)(D) & 2002(b)(4) (1988). We conclude that RSPA’s drug testing program contributes to public safety by enhancing pipeline safety. The program provides a means for the detection and deterrence of drug usage.
5.Conclusion.
Based on our analysis of the statutory guidelines for issuance of rules respecting pipeline safety and the transportation of hazardous liquids, we conclude that the drug testing rules formulated by RSPA are not generally arbitrary and capricious.
B. The Arbitrariness of the Random Testing Provision.
Petitioners also mount an administrative challenge to the feature of the rules requiring random drug testing. As a general matter,
an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. .
Motor Vehicle Mfrs. Ass’n, 463 U.S. at 43, 103 S.Ct. at 2867.12
*1460Petitioners argue that RSPA acted capriciously by requiring suspicionless searches without fully exploring less intrusive alternatives. In their view, RSPA could have tailored the program more narrowly by relying on drug tests administered to determine suspicion, and by limiting the class of workers to be randomly tested to those in the most safety-sensitive positions.
The RSPA did consider the alternatives to testing without individualized suspicion that were propounded by the advisory committees.13 It rejected these proposals, however, as inadequate to meet the agency’s goal. RSPA deemed one alternative to testing without individualized suspicion— pre-employment screenings' — as unsuitable by itself because such tests could be circumvented by temporary abstinence, whereas random tests would either yield discovery of drug use or serve to deter it. See 53 Fed.Reg. 25,898 (1988). Another alternative, post-accident testing, is essentially non-preventive, whereas a key aim of random testing is deterrence. See 53 Fed.Reg. 47,087 (1988) (stating RSPA’s conclusion that it had to “anticipate potential problems and act in a rational, reasonable, and practicable way to prevent ‘accidents’ from occurring”).14
The importance of deterrence is great. RSPA determined that random testing met certain needs and offered advantages unmatched by other forms of testing. In addition to providing a key “preventive” measure, random testing eliminated the opportunity for harassment, real or alleged, that can result from reasonable cause testing. 53 Fed.Reg. 25,898 (1988). Moreover, the agency’s conclusion has been borne out by the successful experience of other random testing programs.15 In our view, therefore, RSPA’s decision to adopt random testing was not an arbitrary and capricious exercise of its rulemaking authority.16
C. The Constitutionality of Random Testing in the Pipeline Industry.
The true heart of the petitioners’ case rests in their constitutional challenge. *1461Petitioners assert that random drug testing, as well as any testing without individualized suspicion, abridges the Fourth Amendment proscription against unreasonable searches and seizures. Many of the difficult threshold questions on this issue have been resolved,17 and we must therefore “balance the individual’s privacy expectations against the Government’s interests to determine whether it is impractical to require a warrant or some level of individualized suspicion in the particular context.” National Treasury Employees Union v. Von Raab, 489 U.S. 656, 109 S.Ct. 1384, 1390, 103 L.Ed.2d 685 (1989).18 The random drug testing rule raises the most difficult questions regarding the constitutionality of suspicionless testing. Thus, we focus our analysis on the propriety of the random testing.
Petitioners make two arguments to guide our traditional balancing of government and individual interests: (1) random testing does not sufficiently serve the government’s overriding interest in safety; and (2) the coverage of testing is not sufficiently tailored to meet that end. RSPA responds generally that the first argument is grounded on an erroneous view of the law and the second is based on an incorrect view of the facts.
1. Demonstration of a Compelling Interest to Justify the Rule.
RSPA identified a safety interest as a justification for the rule: “[T]he clear public interest in assuring that certain sensi-five safety-related pipeline personnel perform their duties free of prohibited substances provides justification for testing....” 53 Fed.Reg. 47,085 (1988). RSPA, as already pointed out, also justified the rule by claiming it deterred drug use. Id. at 47,087.19
In applying the leading case in our circuit, Bluestein, we observe that RSPA need not show a particularized drug problem in the pipeline industry to merit testing. In the Bluestein court’s characterization of Von Raab, “the Court rejected the petitioners’ contention that there was insufficient evidence of a drug problem in the Customs Service to justify suspicionless testing. The Court noted that drug abuse is a pervasive social problem, and stressed that the testing program was aimed as much at deterrence as at detection.” Bluestein, 908 F.2d at 456.20
It is true that RSPA can show neither the existence of a widespread drug problem in the pipeline industry nor a direct linkage between drug impairment and the possibility of accidents. However, an estimated 65 percent of pipeline accidents “are due to such causes as operational error, the improper monitoring of corrosion, and equipment failure, all matters which can be adversely affected by the impairment of specific pipeline personnel.” 53 Fed.Reg. 25,-893 (1988). Moreover, RSPA deemed random testing necessary to insure adequate “routine construction operations, and maintenance functions, as well as emergency response activities, [which] demand skilled, *1462competent, alert and unimpaired workers to perform the functions safely.” Id.; see also RSPA, Final Regulatory Evaluation, at 2.
Bluestein and Von Raab recognize a compelling governmental interest in deterring drug use and the incident harms that may occur from such abuse in the airline industry and the Customs Service. In our view, the consequent harm that would occur from a pipeline accident is sufficient to merit the finding of a strong governmental interest in the detection and deterrence of substance abuse among pipeline workers. Indeed, the concern for public safety animates the general acceptance of drug testing by courts.21
Petitioners also contend that RSPA has not demonstrated a compelling interest in the testing of personnel whose functions are not safety-related. The difficulty with this argument is that petitioners did not cite a single job function identified in the regulations as non-safety-related. We recognize that in certain circumstances, an agency may cast too wide a net in defining the category of persons who must be subjected to random drug testing.22 Nevertheless, we cannot assume that the agency overreached solely on the basis of petitioners’ assertion.
When the government dictates safety regulations to private sector businesses, it typically focuses on general categories of duties, rather than on precise job titles. By contrast, when it imposes regulations for government jobs it has specific information that it does not possess for private sector jobs. Courts have tended, therefore, to permit regulations with more expansive job or function definitions when the rule affects the private sector, on the theory that the government has less access to information as to which positions invoke the relevant safety concerns. Compare Skinner v. Railway Labor Executives’ Ass’n, 489 U.S. 602, 109 S.Ct. 1402, 1408, 103 L.Ed.2d 639 (1989) (upholding testing for “employees assigned to perform service subject to the Hours of Service Act”)23 with Harmon v. Thornburgh, 878 F.2d 484, 486, 490-91 (D.C.Cir.1989) (holding that Justice Department testing program was too broadly drawn by its inclusion of persons who had no relation to prosecution of federal drug offenders), cert. denied sub nom., Bell v. Thornburgh, - U.S. -, 110 S.Ct. 865, 107 L.Ed.2d 949 (1990). As we have stated, the RSPA rule is defined in functional terms to cover relevant persons irrespective of their official job title.24 Were RSPA to have been more specific regarding job titles, the possibility of arbitrariness would have increased.25
We conclude by noting that seven circuit courts of appeal, including our own, have *1463upheld random drug tests for employees with safety-sensitive, security-sensitive, or public-integrity-sensitive jobs.26 The government’s interest in the safety of the pipeline industry is great. Random drug testing, as compared with other forms of testing, offers the best potential deterrent to drug use. This factor, coupled with the possibility of a catastrophic accident, is sufficient to show a strong governmental interest in random testing. Against this interest we must weigh the Fourth Amendment privacy concerns of the individuals to be tested.
2. The Intrusion on a Reasonable Expectation of Privacy.
Privacy interests are implicated in the collection of urine samples and the analysis of their contents. See Skinner, 109 S.Ct. at 1413; Von Raab, 109 S.Ct. at 1390. In weighing the importance of this privacy interest, we must determine whether circumstances in the pipeline industry should compel us to subordinate this interest. Two considerations guide our analysis: (1) whether as a result of other circumstances employees in this industry already have a diminished expectation of privacy; and (2) whether the particular testing program minimizes the intrusion of privacy. See, e.g., Skinner, 109 S.Ct. at 1418.
Petitioners contend that, unlike the railroad workers in Skinner and the Customs Services officials in Von Raab, the pipeline workers in this case do not have a diminished expectation of privacy. They maintain that the Supreme Court upheld testing in those cases either because the workers were in an industry that “ha[s] long been a principal focus of regulatory concern,” Skinner, 109 S.Ct. at 1419, or because they were in positions with extensive background investigations and other intrusions on their privacy. Von Raab, 109 S.Ct. at 1397.
There is force in petitioners’ contention. It is true that pipeline workers have never been a focus of “regulatory concern.” Rather, the regulations affecting these workers concern performance, medical fitness for duty, or training and ability to perform their particular trades. See, e.g., 49 C.F.R. §§ 193.2705, 193.2707, 193.2711, 193.2713 (1989). Nevertheless, the regulations do suggest the potential safety hazards of the industry. These alone lead to a diminished expectation of privacy. See, e.g., Skinner, 109 S.Ct. at 1418 (“[T]he expectations of privacy of covered employees are diminished by reason of their participation in an industry that is regulated pervasively to ensure safety, a goal dependent, in substantial part, on the health and fitness of covered employees.”).27 Obviously the subjective expectations of certain affected employees do not alone determine the constitutional issue. We also *1464must consider whether pipeline workers hold objective and socially “justifiable expectations” such that they could reasonably expect to be free of additional safety regulations. See Skinner, 109 S.Ct. at 1419.
A second issue with respect to the privacy interests of the affected workers is whether the rule minimizes the nature of the intrusion.28 No matter how carefully tailored, “[a]ll urinalysis programs implicate serious privacy concerns.” Taylor v. O’Grady, 888 F.2d 1189, 1197 (7th Cir.1989). Of the types of testing imposed under RSPA’s regulations, only the random testing occurs without some prior basis of individual suspicion.
The absence of individualized suspicion increases the intrusiveness of testing on an employee’s privacy. We conclude, however, that the level of intrusiveness does not justify striking down the random drug testing rule. Moreover, pre-employment and post-accident tests are less intrusive than random testing because they are triggered by the employee’s own act or conduct, or by a definable event. Consequently, we find that such testing rules also withstand constitutional scrutiny.29 Given the different safety needs of other industries, our holding is limited to a finding that the privacy interest implicated by testing without individualized suspicion in the pipeline industry is outweighed by the government’s interest in detecting and deterring drug use.
For the foregoing reasons, we affirm the decision of the Research and Special Programs Administration.
AFFIRMED.
. The Natural Gas Pipeline Safety Act (NGPSA) and the Hazardous Liquid Pipeline Safety Act (HLPSA) require that proposed rules be submitted to established safety standards committees. 49 U.S.C.App. §§ 1673(b) & 2003(b) (1988).
. See 53 Fed.Reg. 47,023-61 (1988) (Federal Aviation Administration); id. at 47,063-82 (Coast Guard); id. at 47,101-32 (Federal Railroad Administration); id. at 47,133-54 (Federal Highway Administration); id.' at 47,155-77 (Urban Mass Transportation Administration).
. See 53 Fed.Reg. 47,085-92 (1988).
. For employers with more than 50 employees, the dates for implementation were changed from December 21, 1989, to April 20, 1990, and for employers with fewer than 50 employees, the dates were changed from April 23, 1990, to August 21, 1990. 54 Fed.Reg. 14,922 (1989).
. RSPA first issued its final rule on November 21, 1988, 53 Fed.Reg. 47,084, announced a reevaluation of the scope of the rule on April 13, 1989, 54 Fed.Reg. 14,922, and published certain minor modifications to the rule on December 18, 1989, id. at 51,842. A notice of petition for review must be filed within 90 days of the issuance of the final rule. 49 U.S.C.App. §§ 1675(a) & 2005(a) (1988). In Case No. 89-70061, the petition for review of the November 21, 1988 order was timely filed on February 10, 1989. In Case No. 89-70308, the petition for review was filed in the U.S. District Court for the Northern District of California on January 20, 1989, and was removed to this court on May 9, 1989. The two cases were consolidated by unopposed motion on January 19, 1990.
. Petitioners also challenge the post-accident and pre-employment provisions of the rule, each of which could involve testing without individualized suspicion.
. See U.S. Department of Transportation, Research and Special Programs Administration, Annual Report on Pipeline Safety, Calendar Years, 1985, 1986, 1987, and 1988 (cited in Petitioners’ Brief, at notes 7 and 17).
. We are also unpersuaded by petitioners' contention that the mechanized nature of the industry obviates the need for additional safety regulations. At oral argument, counsel for petitioners conceded that humans perform necessary safety functions related to pipeline maintenance. Failure to render proper maintenance and to monitor pipeline conditions can lead to safety hazards. See 53 Fed.Reg. 25,893 (1988).
.These functions include: responsibility for controlling gas pressure or hazardous liquid flow (49 C.F.R. §§ 192.619, 192.621, 192.623, 195.402 (1989)); inspection of testing devices that control pressure {id. §§ 192.739, 195.428); inspection for pipeline leaks and odori-*1459zation of gas (id. §§ 192.706, 192.723, 192.625; monitoring of cathodic (anti-corrosive) protection {id. §§ 192.465, 195.416); repair of pipelines {id. §§ 192.713, 192.715, 192.717); emergency response functions (id. §§ 192.615, 195.-402).
.In this context, petitioners also suggest that the agency has not demonstrated that impairment from drugs will result in accidents. Such evidence is admittedly difficult to obtain. In its Final Regulatory Evaluation, RSPA concluded that substance abuse impaired critical faculties, such as vision, perception, judgment, and motor performance. "Many routine operation and maintenance functions, as well as emergency response activities involved in pipeline operation, demand skilled, competent, alert, and unimpaired workers." RSPA, Final Regulatory Evaluation, "Drug Testing (Urinalysis), Natural Gas & Hazardous Liquid Pipeline Industry,” at 2. In our view, RSPA's reasoning was appropriate that the impairment of workers could lead to catastrophic consequences.
. The parties’ disagreement over cost estimates thus need not detain us for long. The uncertainty of estimating costs in this field also leads us to conclude that rejecting a rule solely on the basis of such uncertain data would be inappropriate.
. Petitioners contend that Motor Vehicle Mfrs. Ass'n requires an agency to consider alternatives propounded by commenters and to explain its rejection of those options. Indeed, Motor Vehicle Mfrs. Ass’m contains some language appearing to support that proposition. See, e.g., Motor Vehicle Mfrs. Ass'n, 463 U.S. at 48, 103 S.Ct. at 2869 (noting that "[a]t the very least this alternative way [requiring installation of air bags] of achieving the objectives of the Act should have been addressed and adequate reasons given for its abandonment”).
We do not read Motor Vehicle Mfrs. Ass'n so broadly. In that case, the Supreme Court took *1460the agency to task for initially adopting an alternative — the requirement of air bags- — and then abandoning that alternative without explaining its decision. See 463 U.S. at 46, 103 S.Ct. at 2868. The Court viewed this action as arbitrary because, at the very least, the agency should have explained why it abandoned the air bags option. In the instant case, the alternatives urged upon the agency by petitioners and other interested parties involved less intrusive or restrictive alternatives. In this sense, RSPA’s decision was no different from the ordinary administrative review of alternatives — some of which will be more onerous to affected persons and some less so. An agency need not explain its reason for rejecting every alternative. See id. at 51, 103 S.Ct. at 2870 ("Nor do we broadly require an agency to reconsider all policy alternatives in reaching decision.”). Certainly Motor Vehicle Mfrs. Ass’n allows us to require more explanation from an agency in certain contexts, but not perhaps as much as petitioners in this case may wish.
.The agency rejected suggestions proposed by the two advisory committees. It reasoned that both committees are advisory only; the agency need only respond to suggestions, it need not heed them, see 49 U.S.C.App. § 1673(b) (1988); id. § 2003(b). Moreover, RSPA did respond, 53 Fed.Reg. 47,094-95 (1988), and it adopted certain of the committees’ suggestions, see id.
. The heavy supervision of workers in the pipeline industry does not, as petitioners suggest, negate the need for other mechanisms to prevent accidents. As RSPA observed, extensive supervision in other transportation industries has not prevented accidents. 53 Fed.Reg. 47,-087 (1988).
. See, e.g., 54 Fed.Reg. 51,846 (1989) (discussing success of the military, Coast Guard, and private industry with unannounced random drug tests); American Fed’n of Gov't Employees v. Skinner, 885 F.2d 884, 891 (D.C.Cir.1989) (random testing may be a more effective deterrent than other forms of testing), cert. denied, - U.S. -, 110 S.Ct. 1960, 109 L.Ed.2d 321 (1990).
. See Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 865, 104 S.Ct. 2778, 2793, 81 L.Ed.2d 694 (1984) (stating that courts must not "reconcile competing political interests ... on the basis of the judges’ personal policy preferences"); National Cable Television Ass’n v. Copyright Royalty Tribunal, 724 F.2d 176, 181 (D.C.Cir.1983) (noting that the stringency of judicial review should be calibrated to the specificity of Congress' statutory directives to the agency); see generally Pierce, The Role of Constitutional and Political Theory in Administrative Law, 64 Tex.L.Rev. 469, 486-88 (1985) (discussing judicial review of agency actions).
. See Bluestein, 908 F.2d at 455, which succinctly summarized these threshold issues:
First, drug testing performed by private employers under compulsion of government regulations constitutes governmental action subject to constitutional restrictions. See Skinner, 109 S.Ct. at 1411-12. Second, because "it is clear that the collection and testing of urine intrudes upon expectations of privacy that society has long recognized as reasonable,” urinalysis “must be deemed [a] searchf ] under the Fourth Amendment.” Skinner, 109 S.Ct. at 1413; accord Von Raab, 109 S.Ct. at 1390. Third, the usual Fourth Amendment requirements of a warrant and probable cause do not necessarily apply in the drug testing context.
. See also Bluestein, 908 F.2d at 455; National Fed’n of Fed. Employees v. Cheney, 884 F.2d 603, 608 (D.C.Cir.1989), cert. denied, - U.S. -, 110 S.Ct. 864, 107 L.Ed.2d 948 (1990); Harmon v. Thornburgh, 878 F.2d 484, 488 (D.C.Cir.1989), cert. denied sub nom. Bell v. Thornburgh, - U.S. -, 110 S.Ct. 865, 107 L.Ed.2d 949 (1990).
. RSPA may also have an unspecified interest in data collection, see 53 Fed.Reg. 25,897 (1988), although it is unclear whether the government’s need for data collection is sufficient to outweigh an individual’s privacy interests, see American Postal Workers Union v. Frank, 725 F.Supp. 87, 90 (D.Mass.1989) (holding that drug testing for research purposes is an unconstitutional search in violation of the Fourth Amendment).
. The D.C. Circuit has also squarely rejected petitioners’ argument that the government must show the existence of a drug problem in the regulated industry. See, e.g., Harmon, 878 F.2d at 487 ("Nor is it necessary that a documented drug problem exist within the particular workplace at issue.”) (citation omitted).
. See, e.g., Skinner v. Railway Labor Executives’ Ass’n, 489 U.S. 602, 109 S.Ct. 1402, 1419, 103 L.Ed.2d 639 (1989); Von Raab, 109 S.Ct. at 1393; American Fed’n of Gov’t Employees v. Skinner, 885 F.2d 884, 890-91 (D.C.Cir.1989), cert. denied, - U.S. -, 110 S.Ct. 1960, 109 L.Ed.2d 321 (1990); National Fed'n of Fed. Employees, 884 F.2d at 610; Thomson v. Marsh, 884 F.2d 113, 115 (4th Cir.1989) (per curiam); Rushton v. Nebraska Pub. Power Dist., 844 F.2d 562, 565 (8th Cir.1988); Transport Workers' Union v. Southeastern Pa. Transp. Auth., 884 F.2d 709, 712 (3d Cir.1988).
. See, e.g., Von Raab, 109 S.Ct. at 1397 (remanding to determine whether Customs Service defined category of employees with access to sensitive information too broadly to meet its reasonable goal); Taylor v. O'Grady, 888 F.2d 1189, 1197 (7th Cir.1989) (holding invalid a policy that required all corrections officers, rather than just those with contact with inmates, to submit to drug testing); Harmon, 878 F.2d at 490-91 (holding invalid certain Justice Department regulations that required testing for persons not engaged in the prosecution of federal drug offenders).
. The Hours of Service Act defines "employee” to be "an individual actually engaged in or connected with the movement of any train,” 45 U.S.C. § 61(b)(2) (1988); see also id. §§ 63a(a)(l), 63a(d) (regulating hours of service for persons "engaged in installing, repairing or maintaining signal systems”).
. See supra note 9 and accompanying text.
. Regulations issued on the basis of job titles within the pipeline industry would have presented at least two potential difficulties. First, the possibility exists that two different companies use different titles for persons who perform similar functions. The regulations might affect one and not the other. Second, a pipeline company could purposefully or inadvertently subvert the regulations if it were to change the title of the positions affected by the testing provisions. Petitioners have not identified job func*1463tions among those specified in the rules that are irrelevant to the ultimate safety of the transportation of hazardous liquids and natural gas by pipelines. We therefore decline their invitation to second-guess the agency. See, e.g., Skinner, 109 S.Ct. at 1419 n. 9.
. See, e.g., Bluestein v. Department of Transp., 908 F.2d 451, (9th Cir.1990) (airline industry personnel); Taylor v. O’Grady, 888 F.2d 1189, 1199 (7th Cir.1989) (correctional officers in regular contact with inmates); American Fed'n of Gov’t Employees v. Skinner, 885 F.2d 884, 893 (D.C.Cir.1989) (various transportation workers), cert. denied, - U.S. -, 110 S.Ct. 1960, 109 L.Ed.2d 321 (1990); National Fed’n of Fed. Employees v. Cheney, 884 F.2d 603, 615 (D.C.Cir.1989) (Army civilian guards), cert. denied, - U.S. -, 110 S.Ct. 864, 107 L.Ed.2d 948 (1990); Thomson v. Marsh, 884 F.2d 113, 115 (4th Cir.1989) (per curiam) (civilian workers in Army chemical weapons plant); Harmon v. Thornburgh, 878 F.2d 484, 496 (D.C.Cir.1989) (Justice Department employees with clearance for top-secret information), cert. denied sub nom. Bell v. Thornburgh, - U.S. -, 110 S.Ct. 865, 107 L.Ed.2d 949 (1990); Guiney v. Roache, 873 F.2d 1557, 1558 (1st Cir.) (per curiam) (police officers carrying firearms or engaged in drug interdiction efforts), cert. denied, - U.S. -, 110 S.Ct. 404, 107 L.Ed.2d 370 (1989); Rushton v. Nebraska Pub. Power Dist., 844 F.2d 562, 567 (8th Cir.1988) (nuclear power plant engineers); Transport Workers’ Union v. Southeastern Pa. Transp. Auth., 884 F.2d 709, 713 (3d Cir.1988) (mass transit workers).
. In one of the industries affected by RSPA's regulations, for example, a requirement is already in place that "personnel assigned operating, maintenance, security, or fire protection duties [at a liquefied natural gas facility] not have any physical condition that would impair performance of their assigned duties.” 49 C.F.R. § 193.2711 (1989). The drug testing provision for such workers complements that preexisting regulation.
. Petitioners do not contend that the testing procedures themselves are intrusive. These procedures call for use of stalls or partitioned areas for privacy. See 49 C.F.R. § 40.25(a)(2) (1989).
. In addition to challenging the constitutionality of testing without individualized suspicion, petitioners also challenge a post-accident testing provision which directs that "all reasonable steps” be taken for collection of a urine sample if the employee is "injured, unconscious, or otherwise unable to evidence consent to the drug test.” 54 Fed.Reg. 51,850 (1989). These regulations appear to be unprecedented. Other transportation agencies that require post-accident testing do not mandate such procedures. See, e.g., 46 C.F.R. §§ 4.06-10, 4.06-20 (1989) (Coast Guard); 49 C.F.R. § 391.113 (1989) (Federal Highway Administration); 49 C.F.R. § 653.15 (1989) (Urban Mass Transit Administration). Although we have some doubts about these procedures, we do not interpret the regulation to be unconstitutional. We have no reason to assume *1465that the regulations anticipate procedures that will result in actual intrusion into employees’ bodies and the RSPA's response to comments supports that no intrusion is intended. 54 Fed.Reg. 51,843 (1989).