Petitioners Brief — Industrial Union Dept., AFL-CIO v. American Petroleum Institute

Supreme Court brief1980

Ask Donna

What actually matters in this document.

Text

SS eeeEEEEEEeEeeaeEeaEaEaEeEeEeEeE——————

INDUSTRIAL UNION DEPARTMENT, AFL-CIO,

PETITIONER

Vv. .

AMERICAN PETROLEUM INSTITUTE, ET AL.

RAY MARSHALL, SECRETARY OF LABOR, PETITIONER

Vv.

AMERICAN PETROLEUM INSTITUTE, ET AL.

ON WRIT OF CERTIORARI TO THE UNITED STATES

COURT OF APPEALS FOR THE FIFTH CIRCUIT

BRIEF FOR THE FEDERAL PAR’ .ES

WADE H. McCREB, JR.

Solicitor General

CARIN ANN CLAUSS FRANK H. EASTERBROOK

Solicitor of Labor Deputy Solicitor General

BENJAMIN W. MINTZ WILLIAM ALSUP

Associate Solicitor Assistant to the Solicitor General

Department of Justice

ALLEN H. FELDMAN Washington, D.C. 20530

DENNIS K, KADE

DIANE EB. BURKLEY

Attorneys

Department of Labor

Washington, D.C. 20210

Opinions below

Jurisdiction

Questions presented ......................:0cccsecceeeeeeeneeee

I TI sae tcctarcthcisinsinesdsensstentanctbiciin sens

Statement

Act ...

Previous benzene standards ................

The permanent standard ....................

The basis for the standard ................

1. Toxic effects of benzene ..............

2. The policy of the lowest feasible

level of exposure ..............-...-22-------

Re | 7. gan CRIN ALES ce Pa

4. Protective clothing -.....................

5. The percentage exclusion ..............

The decision of the court of appeals.

Summary of Argument, ......................css.csesccoseee

Argument

I. The Occupational Safety and Health

Act does not require the Secretary to

establish that the costs of particular

reductions in exposure to toxic sub-

stances are justified by quantifiable

benefits

Argument—Continued

A. The definitional phrase ‘“reason-

ably necessary or appropriate” does

not impose a cost-benefit test ........

B. Section 655(b) (5) prohibits basing

a standard on a cost-benefit test...

1. The legislative history estab-

lishes that Congress wanted the

Secretary to regulate toxic sub-

stances by considering costs

only as part of “feasibility”.

The requirement that standards

be “feasible” refers to prac-

ticability rather than to a cost-

benefit balance —.....................

3. The requirement that the Sec-

retary make decisions on the

basis of the “best available evi-

dence” precludes any role for

cost-benefit analysis if the nec-

essary data are not “avail-

pS” POURRA CR Rm ee ee EOL ea I ‘ecu’

C. Any ambiguities in the statute

should be resolved in favor of

PORT CATO av ciiiissss csosccderrsediosesercee

Il. The Secretary was not required to wait

for more evidence before restricting

dermal contact with benzene-bearing

OR: TIMING sa retessiceg ts top biasdndanecleni saamtabenitae

CAUNIININOUN 5 civscesaiccesnrceceiijatlentedadessonindcapliaediseceaies:

ORG OU BNE Poe) Wel Fea ss LU 2 8

Page

CITATIONS

Cases: Page

American lron and Steel Institute v.

OSHA, 577 F.2d 825, petitions for cert.

pending, Nos. 78-918 and 78-919 ...... . 42, 45

American Trucking Associations v. Atchi-

son, Topeka & Santa Fe Ry., 387 U.S.

397 ..... RFs Elen) Dot TA 2 oO ORR 73

American Trucking Associations y, United

States, 344 U.S. 298 - Seascale 44

Aqua Slide ‘N’ Dive C orp. v. CPSC, 569

PO: OBE. czscivcisercszes Jovnceicerisiioce ae at

Associated Industries of Ne w ‘York State,

Inc. v. United States Department of

Labor, 487 F.2d 842 ........................ Ah

Atlantic & Gulf Stevedores, Ine. v.

OSHRC, 5384 F.2d 641 2... 5A

Atlas Roofing Co. v. OSHRC, 430 U.S.

B68) tian. 1 diss atten aekdladide cs seswautoabe ; 5

Bowles vy. Willingham, 321 U.S. 508 —..... 44

Bowman Transportation, Ine. vy. Arkan-

sas-Best Freight System, Inc., 419 U.S.

SPRITES a TPR arg EUP NOY sak eh pe REN RO 77

CAB v. Hermann, 353 U.S. 329 rapteneette ot 82

Clark v. Uebersee Finanz oer 332 U.S.

ARQ) . ey ee 61

Ke, duPont ‘de ‘Nemours & Co. Vv. C ‘ollins,

432 U.S. 46 . Lath 73

Ki. 1. duPont deNen mours é Co. v. . Train,

I SE tok RO essa gated senso sateppilecasachbasebodletes 72,74

Endicott Johnson Corp. v. Perkins, 317

LAE 1G A RGN ALSSIO Ns} ae Cita Pyle ts Mosse 82

FCC vy. National Citizens Committee for

Broadcasting, 4836 U.S. 775 0000... _ 44-45

FCC vy. Schreiber, 381 U.S. 279 ......... Fase 82

Cases—Continued Page

FERC vy. Pennzoil Producing Co., No. 77-

G48 (Jan. 16, 1979) )........1.......ccsssecseness 45

Florida Peach Growers Ass’n, Inc. v.

United States Department of Labor,

ME RN ID iraccn sendasesisaleseeh teypenciettiicegeas 45

FPC vy. Florida Power & Light Co., 404

Se I lle See wkt dt lec etedngeibcleweenioisetaonans 72

FPC v. Texaco Inc., 417 U.S. 880 —........-.. 44-45

FPC y. Transcontinental Gas Pipe Line

COPD, 428 UU kSs BOO \acconerebcogevesadevarzssneser 75, 82

FTC v. Simplicity Pattern Co., 360 U.S.

RARER URE tii eet A SOS a OA 54

Great Atlantic & Pacific Tea Co. v. FTC,

No. 77-654 (Feb. 22, 1979) -................. 54

H & H Tire Co. v. United States Depart-

ment of Transportation, 471 F.2d 350.. 31

Industrial Union Department, AFL-CIO

v. Hodgson, 499 F.2d 467 ...................... 42, 45

Jarecki v. G.D. Searle & Co., 367 U.S.

GRRE RAR SIAN PO a eg ORE 58

Lilly v. Grand Trunk Western R.R., 317

RP AD MME: ha tacenteeon tas soctnsatanbietciae Del sekbowiion 72

Marshall v. Barlow’s, Inc., 486 U.S. 307.. 5

Miller v. Youakim, No. 77-742 (Feb. 22,

2 ASME SRESISTG LE ner ove Y's sets WN MAO) aa 73

Moog Industries, Inc. v. FTC, 355 U.S.

RR eee ac hes ededteenciod ccpusabibcnaeintsdanahtencedokeres 82

Mourning v. Family Publications Service,

Pea GE Meee, I iircns escitierscarictanietesnascone 37, 44

NAACP v. FPC, 425 U.S. 662 ......0000000.2... 70

Permian Basin Area Rate Cases, 390

oP Nile fe Ei attctean ten WHY Bel aoipe oe A 45, 69-70, 73

RMI Co. v. Secretary of Labor, No. 78-

3312 (6th Cir. Mar. 18, 1979) _.02022..... 42

SEC v. Chenery Corp., 332 U.S. 194 ........ 44

Cases—Continued

SEC v. Sloan, 486 U.S. 108 ............0.......2

Skidmore v. Swift & Co., 328 U.S. 134...

Society of the Plastics Industry, Inc. v.

OSHA, 509 F.2d 13801, cert. denied,

A Rs I eons eeepsenngaboscnesuntes 42,

South Terminal Corp. v. EPA, 504 F.2d

RR ate Mea acter tee aT aE RSET TR

Synthetic Organic Chemical Manufactur-

ers Association v. Brennan, 503 F.2d

pS” ESIGN BCR pr BRR ee RO Sac

Tagy Bros. & Moorhead v. United States,

BP Ss Me a ro

Thorpe v. Housing Authority of the City

of Durham, 393 U.S. 268 —.....02--002-..

Tiller v. Atlantic Coast Line R.R., 323

1a 5 GRMN RENIN Den RIALS roe BON OP Ri ee

Train v. Colorado Public Interest Re-

search Group, Inc., 426 U.S. 1-200...

Train v. Natural Resources Defense Coun-

Om, Be. GEA: Ua OO ie cn .

Turner Co. v. Secretary of Labor, 561

) pe BR REINER ON VER Eu SEAR AOR NO ee

TVA v. Hill, 487 U.S. 158.0000...

Union Electric Co. v. HPA, 427 U.S. 246..

United States v. Allegheny-Ludlum Steel

ee a Oe | RRA Rules eer SE

United States v. An Article of Drug * * *

Bacto-Unidisk, 394 U.S. 784 200000000000...

United States v. Bush & Co., 310 U.S.

1k RASA se BV Abeta an hed SRR TAT EE Lee RN

APs. IN isis ccileal blest adlairdexcacconay

United States v. Culbert, 435 U.S. 371......

United States v. Seaboard Air Line R.R.,

ee Seis We he ea

Page

73-74

45, 60

54, 74

41,57

37, 44

VI VII

Cases—Continued Page Constitution, statutes and

United States v. Storer Broadcasting Co., regulations—Continued Page

nk Re | eR SERPs an OR Ue necator 44 ; ;

Energy Policy and Conservation Act of

Universal Camera Corp. v. NLRB, 340

Ce a Re eRe ESN EE 70 1975, 42 U.S.C. 6201 et seq.:

Utah Fuel Co. v. National Bituminous ' Section 325, 42 U.S.C. 6295 .............. 57

Coal Commission, 306 U.S. 56 ..0000.002.... 82 Section 325(a), 42 U.S.C. 6295(a).... 57

Vermont Yankee Nuclear Power Corp. v.

Natural Resources Defense Council,

BG sg, MEO NI e OI icdecicisiendgocanssctsarieuibness 82

Longshoremen’s and Harbor Workers’

Compensation Act, 33 U.S.C. 901-950.. 6

‘ ‘ National Foundation on Arts and Human-

Voris v. Kikel, 346 U.S. 328 .................. 72 ities Act, 20 U.S.C. 951-960 ................ 6

Weinberger v. Hynson, Westcott & Dun- Occupational Safety and Health Act of

MAG, 412 U.S. 609 a aennnn-sonseverernnnseee 58, 61 1970, 29 U.S.C. 651 et seq.:

Wisconsin v. FPC, 373 U.S. 294 .....00000.... 82

Se SE lc 5, 52

Constitution, statutes and regulations: Gk Ee Ey EE eames Wane passim

United States Constitution: = ae reget MRIS i anand ean :

Fourth Amendment .....................--.... 5 yO RTO Tt) > | nie eee es 6

Seventh Amendment ..........0...0000000..... 5 | ae Ce COED pisces. 5, 52

Clean Air Act, Section 211(c) (2) (A) and 29 US.C. 654 (8) (2) ---.----eeneveernnreere 5

(B), to be codified at 42 U.S.C. 7545 29 U.S.C. 655 (8) ---aneenvenrevvenreee 5, 6, 7, 58

(MytSP CA) wed CB) dn 56 dhe deca Diete y 2,7

Consumer Product Safety Act, 15 U.S.C. 29 U.S.C. 655(b) (1) -------------oeeeeeooees 7, 58, 81

2051 et seq.: 29 U.S.C. 655 (b) (5) Scuensnccconspecoccccosoe passim

29 U.S.C. 655(b) (6) (A) oe. 8

15 U.S.C, 2057 .....--e-nneeeeeveeeesneeeeennees 57 | 29 U.S.C. 655(b) (6) (C) 52, 53

15 U.S.C. 2058(¢) (1) (C) -...2.......... 47,57 $0 TLEC GER IB) (BY os 7

15 U.S.C. 2058(¢) (1) (D) .......0....... 47, 57 SO UE Watley 7

15 U.S.C. 2058(¢) (2) ..............-....-00- 47 29 USS. 668(e) (8), 7

15 U.S.C. 2058(c) (2) (A) -.............. 33, 46 : 50 TAO Old) 2. 9, 52, 53

15 U.S.C. 2058(¢) (2) (C) enone. ne 29 U.S.C. 655(f) ............ 9, 30, 31, 45, 74, 75

Contract Work Hours and Safety Stand- BP ite UE) ands 7, 55

ards Act, 40 U.S.C. 327-333 .................. 6 a) Ree een meee Haven 83

By is PICS oi civcecicceccoiscccenss 8

29 U.S.C. 669(a) (6). ceconne.esccecsecsnn..... 8

VIII

Constitution, statutes and

regulations—Continued Page

Rea OO, sad hadscarcptnnncnedenbgpanns 82

RE OD > each nocccechakersecceieecamniae 8

Se Ras TOAD scsi csce terion 8

Service Contract Act of 1965, 41 U.S.C.

LE SRS EMUNREC RR! WI arr CME) Od at me eA UR 6

Walsh-Healey Act, 41 U.S.C. 35-45 ......... 6

RE Rrra IES ccqhincbeicadviiniberestabasanuescsancoomiis 9, 54

29 C.F.R. 1910.1000, Table Z-2 (1977)... 10

rE eA OF STUER I Oe i 14

29 C.F.R. 1910.1028(a) (2) (ii) -.....2......... 29, 30

29 C.F.R. 1910.1028(a) (2) (iii) 22222... 30

Be NA as: ROROMROOUED, | ectccetncecndeceycesonietene 14

29 C.F.R. 1910.1028(c) (1) 20000... 14

29 C.F.R. 1910.1028(¢) (2) 2.00020... 15, 25

BO Keel tis ROR ED | oc cksccsecsnctatwoansncecien 14

BO CLP. SRR SOOM) | usa sisecevescesscdovacei 15, 25

BO C.F Be. BORG BRS ER) oa sinsnns ceca cass csedie andes 14-15

29 C.F.R. 1910.1028 (ie) , .............,..0.ccccese ae 14

yO Li eT RS NO 15

Miscellaneous:

Comment, Implementing the Anticancer

Clauses of the Food, Drug and Cosmetic

Act, 44 U. Chi. L. Rev. 817 (1977) ...... 41

I Compact Oxford English Dictionary...... 56

Doninger,

Federal Regulation of Vinyl

Chloride: A Short Course in the Law

and Policy of Toxic Substances Control,

7 Ecol. L.Q, 497 (1978) ......,............2:-.. 63

EPA, A Preliminary Report on Population

Risk to Ambient Benzene Exposures

| RENE AS tes OES (Di we 0h Real a 70

Exec. Order No. 12044, 43 Fed. 12661,

Ee PE auicdanniealibinighnsteabibiedaihs 67

eee

Miscellaneous—Continued Page

37 Fed. Reg. 11318 (1972) ...................-..-- 21, 73

39 Fed. Reg. (1974):

IRS I 21, 73

SN OS OG ae 21, 73

(ae SG An Sr 21, 73

Se 21, 73

41 Fed. Reg. (1976) :

RUSE ES a 21, 62, 73

pp. 46750-46751... nee 21

aM TL 62, 73

42 Fed. Reg. (1977):

ON OE ee a2, 22

Oe, TN ons s--------- 13

Se ENE ee 18, 29

| 13

S350 Se 13

oc 13

OS A ES en 25

EE SESE 21

43 Fed. Reg. (1978):

OR 14

A NE 30

Mi bs BS ba 14

rm Ne a a as wcwcepapnce 21, 73

IEEE Tae A 57

SANE ER oa ea a 21, 73

oe NS tb a 30

RENTERS USE Ts OTE 21, 73

SS EIU SES ale a Se 21,73

stra AERTS A REE 21, 68, 73

SO, I TE iar setdaivelaccssnompesnn 68

I 7 stletlc dst Olcca Dagheicbinpsetneindnniocnansinnn 57

Miscellaneous—Continued

Fried, The Value of Life, 82 Harv. L.

Re hf UMRAO cn

H.R. Conf. Rep. No. 94-700, 94th Cong.,

EE ss “tReet ine ese das Mey ene

Industrial Hygiene Field Operations Man-

ual, Empl. Safety & Health Guide

(CCH) § 4483.80 (1979) .........22..........

Interagency Liaison Group’s Report on

Scientific Bases for Identifying Poten-

tial Carcinogens and Estimating Their

yo eg Dk ¢ ) Rae ae ee a

McGarity, Substantive and Procedural

Discretion in Administrative Resolution

of Science Policy Questions: Regulat-

ing Carcinogens in EPA and OSHA,

67 Geo. L.J. 729 (1979) -.......................

Rhoads and Singer, What is Life Worth?,

51 Pub. Interest 74 (Spring 1978) _....

S. 755, 96th Cong., Ist Sess. § 3 (1979)...

Senate Comm. on Governmental Affairs,

95th Cong., 2d Sess., Study on Federal

Regulation, Vol. VI (Comm. Print

gs Se TPN SOREL FTA RI LOW ROPER

S. Rep. No. 94-516, 94th Cong., 1st Sess.

TID dele ielciteg atin cia Lachdatpeetendenics ngsocsces

Subcomm. on Labor of the Senate Comm.

on Labor and Public Welfare, 92d

Cong., 1st Sess., Legislative History of

the Occupational Safety ard Health

Act of 1970 (Comm. Print 1971) _....... passim

Webster’s Third New International Dic-

UE RED wekccchciepbedstbctcewndscelnecinnes

Iu the Supreme Court of the United States

OCTOBER TERM, 1978

No. 78-911

INDUSTRIAL UNION DEPARTMENT, AFL-CIO,

PETITIONER

Vv.

AMERICAN PETROLEUM INSTITUTE, ET AL.

No. 78-1036

RAY MARSHALL, SECRETARY OF LABOR, PETITIONER

Vv.

AMERICAN PETROLEUM INSTITUTE, ET AL.

ON WRIT OF CERTIORARI TO THE UNITED STATES

COURT OF APPEALS FOR THE FIFTH CIRCUIT

BRIEF FOR THE FEDERAL PARTIES

OPINIONS BELOW

The opinion of the court of appeals (Pet. App.

la-33a)’ is reported at 581 F.2d 493. The regulation

1 We use the following abbreviations in this brief. “Pet.

App.” refers to the separate appendix to the petition in No. 78-

911. “A.” refers to the Joint Appendix filed in this Court.

“Tr.” refers to the transcript of the rulemaking hearings

(1)

in question and the Secretary’s accompanying expla-

nation (Pet. App. 34a-252a) appear at 43 Fed. Reg.

5918-5970.

JURISDICTION

The judgment of the court of appeals (Pet. App.

254a-255a) was entered on October 5, 1978 (A. 37).

The petition for a writ of certiorari in No. 78-911

was filed on December 7, 1978, and the petition in

No. 78-1036 on December 26, 1978. The petitions

were granted and the cases consolidated on February

21, 1979 (A. 38-39). The jurisdiction of this Court

rests on 28 U.S.C. 1254(1).

QUESTIONS PRESENTED

1. Whether a health and safety standard for a

toxic substance promulgated by the Secretary of .

Labor pursuant to 29 U.S.C. 655(b) must satisfy a

cost-benefit test.

conducted in 1977 prior to the final standard. “Exh.” refers

to exhibits received during the rulemaking proceeding.

“Amend. Tr.” refers to the transcript of the hearings in

May 1977 on the amendment to the standard. (For the

Court’s convenience, an index to the voluminous record has

been prepared by the parties and will be lodged with the

Clerk.) “Leg. Hist.” references are to the Subcomm. on Labor

of the Senate Comm. on Labor and Public Welfare, 92d Cong.,

1st Sess., Legislative History of the Occupational Safety and

Health Act of 1970 (Comm. Print 1971). The Act confers the

rulemaking power at issue on the Secretary of Labor, who, in

turn, has delegated it to the Assistant Secretary of Labor for

Occupational Safety and Health. For convenience, we shall

refer to the Secretary, the Assistant Secretary, and the Occu-

pational Safety and Health Administration as “the Secretary.”

2. Whether a health and safety standard that is

supported by substantial evidence based on the best

available scientific knowledge may be set aside be-

cause of the possibility that future scientific studies

will provide better evidence concerning the need for

the standard.

STATUTE INVOLVED

Section 3(8) of the Occupational Safety and

Health Act of 1970, 29 U.S.C. 652(8), provides:

The term “occupational safety and health

standard” means a standard which requires con-

ditions, or the adoption or use of one or more

practices, means, methods, operations, or proc-

esses, reasonably necessary or appropriate to

provide safe or healthful employment and places

of employment.

Section 6(b) (5) of the Act, 29 U.S.C. 655(b) (5),

provides:

The Secretary, in promulgating standards

dealing with toxic materials or harmful physical

agents under this subsection, shall set the stand-

ard which most adequately assures, to the extent

feasible, on the basis of the best available evi-

dence, that no employee will suffer material im-

pairment of health or functional capacity even

if such employee has regular exposure to the haz-

ard dealt with by such standard for the period of

his working life. Development of standards

under this subsection shall be based upon re-

search, demonstrations, experiments, and such

other information as may be appropriate. In ad-

dition to the attainment of the highest degree of

health and safety protection for the employee,

other considerations shall be the latest available

scientific data in the field, the feasibility of the

standards, and experience gained under this and

other health and safety laws.

STATEMENT

On February 10, 1978, the Secretary of Labor

promulgated an occupational health standard restrict-

ing employee exposure to benzene. The standard (1)

sets a maximum permissible exposure level (PEL)

in the workplace of 1 part per million (ppm) of air-

borne benzene, averaged over an eight-hour period

with a ceiling exposure for a 15-minute period of

5 ppm; (2) requires engineering and work prac-

tice controls as the primary means to achieve this

level; and (8) where employees are exposed to liquids

containing benzene, requires the use of protective

equipment to prevent skin and eye contact with liquid

benzene. The court of appeals set aside the first two

aspects of the standard because, in the court’s view,

the requirements are not “reasonably necessary” to

provide healthful employment within the meaning of

Section 652(8). The court set aside the protective

clothing requirement because it concluded that the

requirement is not based on the “latest available

scientific data in the field” within the meaning of

Section 655 (b) (5).

ieee assis ite tS

A. The Occupational Safety and Health Act ?

Congress enacted the Occupational Safety and

Health Act of 1970 (OSHA) “to assure so far as

possible every working man and woman in the Na-

tion safe and healthful working conditions.” 29

U.S.C. 651(b). See generally Atlas Roofing Co. v.

OSHRC, 430 U.S. 442, 444-445 (1977).

The Act imposes two duties on employers. One is

a “general duty” to furnish employment and a place

of employment “free from recognized hazards that

are causing or are likely to cause death or serious

physical harm * * *.” 29 U.S.C. 654(a)(1). Em-

ployers also must obey “occupational safety and

health standards” set by the Secretary of Labor.

29 U.S.C. 654(a)(2). These are standards “which

require[] conditions, or the adoption or use of one or

more practices, means, methods, operations, or proc-

esses, reasonably necessary or appropriate to provide

safe or healthful employment and places of employ-

ment.” 29 U.S.C. 652(8).

The Act authorizes the Secretary to issue three

types of standards. During the first two years of

the Act, 29 U.S.C. 655(a) permitted the Secretary

to adopt as'an occupational health or safety standard

any preexisting “national consensus standard” or

2 This Court has reviewed the OSHA enforcement scheme

in the light of constitutional challenges raised under the

Fourth and Seventh Amendments. Marshall v. Barlow’s, Inc.,

486 U.S. 307 (1978); Atlas Roofing Co. v. OSHRC, 430 U.S.

442 (1977). This case is the first occasion for the Court to

consider the scope of the Secretary’s statutory power to set

occupational health and safety standards.

“established Federal standard.” 29 U.S.C. 655(a).’

In the event that there was more than one such stand-

ard, the Secretary was required to select the standard

that assured “the greatest protection of the safety or

health of the affected employees.” 29 U.S.C. 655(a).

Several thousand such standards are in force; they

are the great bulk of OSHA standards.

The second kind of standard is the “emergency

temporary standard,” which the Secretary may issue

to take immediate effect when he determines that

employees are exposed to grave danger from toxic

materials or harmful physical agents and concludes

’ A “national consensus standard” is any occupational] safety

or health standard that ‘‘(1) has been adopted and promul-

gated by a nationally recognized standards-producing organi-

zation under procedures whereby it can be determined by the

Secretary that persons interested and affected by the scope

or provisions of the standard have reached substantial agree-

ment on its adoption, (2) was formulated in a manner which

afforded an opportunity for diverse views to be considered

and (3) has been designated as such a standard by the Sec-

retary, after consultation with other appropriate Federal

agencies.” 29 U.S.C. 652(9). The term “established Federal

standard” means any “operative occupational health and

safety standard established by any agency of the United

States * * *.” 29 U.S.C. 652(10). As a result of this provi-

sion and 29 U.S.C. 653(b) (2), a number of standards issued

under other statutes—including the Walsh-Healey Act, 41

U.S.C. 35-45; the Service Contract Act of 1965, 41 U.S.C.

351-357; the Contract Work Hours and Safety Standards Act,

40 U.S.C. 327-333; the Longshoremen’s and Harbor Workers’

Compensation Act, 33 U.S.C. 901-950; and the National

Foundation on Arts and Humanities Act, 20 U.S.C. 951-960—

became occupational health and safety standards under OSHA.

The standards issued under other statutes were adopted with-

out the need for repromulgation.

that such emergency standards are necessary to pro-

tect employees. 29 U.S.C. 655(c).*

The third type of standard is the permanent stand-

ard. 29 U.S.C. 655(b) prescribes a rulemaking pro-

cedure for fashioning permanent standards to super-

sede standards previously adopted under Section 655

(a) or to provide new standards where none previ-

ously existed. Section 655(b) generally permits the

Secretary to set whatever standard he determines is

appropriate for health or safety protection.® Where,

however, the Secretary sets a permanent standard for

toxic materials or harmful physical agents, Section

655(b) (5) requires the Secretary to

set the standard which most adequately assures,

to the extent feasible, on the basis of the best

available evidence, that no employee will suffer

material impairment of health or functional

capacity even if such employee has regular ex-

posure to the hazard dealt with by such standard

for the period of his working life. Development

4 Hmergency temporary standards may remain effective no

more than six months. 29 U.S.C. 655 (c) (3).

‘The Secretary must determine “that a rule should be

promulgated in order to serve the objectives of the [Act],”

29 U.S.C. 655(b) (1), and must give “due regard to’ the

urgency of the need for mandatory safety and health stand-

ards for particular industries” and to the “recommendations

of the Secretary of Health, Education, and Welfare * * *.”

29 U.S.C. 655(g%). Whenever a permanent standard “differs

substantially from” a national consensus standard, the Secre-

tary must explain “why the rule as adopted will better effect-

uate the purposes of” the Act. 29 U.S.C. 655 (b) (8).

of standards under this subsection shall be based

upon research, demonstrations, experiments, and

such other information as may be appropriate.

In addition to the attainment of the highest de-

gree of health and safety protection for the em-

ployee, other considerations shall be the latest

available scientific data in the field, the feasibility

of the standards, and experience gained under

this and other health and safety laws.

The Act establishes a National Institute for Occu-

pational Safety and Health (NIOSH) to assist the

Secretary in his tasks. The standards adopted by

the Secretary, including toxic-substances standards,

may permit variances in certain circumstances, but

economic hardship is not a ground for a variance.’

6 NIOSH is required, on the basis of research, demonstra-

tions and experiments, to “develop criteria dealing with toxic

materials and harmful physical agents and substances which

will describe exposure levels that are safe for various periods

of employment, including but not limited to the exposure

levels at which no employee will suffer impaired health or

functional capacities or diminished life expectancy as a result

of his work experience.” 29 U.S.C. 669(a) (3). (NIOSH per-

forms this duty as the Secretary’s delegate. 29 U.S.C. 671

(a}.) NIOSH is to publish annually a list of all known toxic

substances by generic family or other useful grouping, and

the concentrations at which such toxicity is known to occur.

If a potentially toxic substance is not covered by any standard,

NIOSH must so advise the Secretary. 29 U.S.C. 669(a) (6)

and (e).

7 Any employer may apply to the Secretary for a “tempo-

rary” variance where, among other things, he is unable to

comply because the necessary construction or alteration of

facilities cannot be completed by the effective date of the

standard. 29 U.S.C. 655(b) (6) (A). An employer is entitled

to a “permanent” variance if he establishes that the condi-

Section 655(f) authorizes pre-enforcement review

of a new standard by the court of appeals at the

request of any person adversely affected by the

standard. Petitions for review must be filed, if at all,

within 60 days after promulgation of the standard.

In such proceedings ‘“[t]he determinations of the

Secretary shall be conclusive if supported by substan-

tial evidence in the record considered as a whole.”

29 U.S.C. 655(f).

B. Previous Benzene Standards

Benzene * has long been known to cause a number

of nonmalignant blood diseases, some fatal. See pages

tions, practices, means, methods, operations, or processes used

or proposed to be used by it will provide employment and

places of employment as safe and healthful as those that would

prevail if it complied with the standard. 29 U.S.C. 655(d).

Thus, the statute is “cost-effective” in the sense that it permits

employers to meet the performance standards in any way that

provides the designated level of protection. Although eco-

nomic hardship may not be a ground for a temporary vari-

ance (see Leg. Hist. 1188, 1201), the Act provides for small

business loans to finance the cost of compliance. 15 U.S.C. 636.

8 Benzene is a clear, colorless, noncorrosive, highly flamma-

ble hydrocarbon (C,H,) with a strong, rather pleasant odor.

Benzene’s low boiling point and high vapor pressure cause, jt

to evaporate rapidly under ordinary atmospheric conditions

(Pet. App. 69a). Benzene can be present naturally in small

quantities (a few parts per billion). It is manufactured in

substantial quantities for a wide variety of industrial uses

(id. at 69a-70a). It is used as a solvent and in the manufac-

ture of other organic chemicals, pesticides, detergents, sol-

vents and paint removers. Industries currently using benzene

include the chemical, petroleum, rubber fabricating, adhesive,

varnish and paint industries (id. at 70a). The petrochemical

and petroleum refining industries produce 94% of the total

domestic production of benzene (id. at 69a). Recovery of

15-20, infra. More recently, it has been determined

that benzene causes leukemia.

For many decades before benzene’s carcinogenic

effects were established, groups responsible for rec-

ommending levels of exposure to benzene had reduced

the recommended levels again and again.’ In 1971,

shortly after the Act was passed, the Secretary

adopted a preexisting national consensus standard

for benzene under the authority of Section 655(a).

See 29 C.F.R. 1910.1000, Table Z-2 (1977). That.

standard allowed an eight-hour time-weighted average

(TWA) exposure of 10 ppm, with a maximum con-

centration of 25 ppm averaged over a 10-minute

period. Concentration was not exceed 50 ppm at any

time. This standard was based on benzene’s known

nonmalignant effects and not on any possible leukemia

hazard (Pet. App. 7la; Exh. 2A-159 at 6-7).

coal-derived benzene, primarily as a by-product of the coking

process in steel mills, account for 6% of domestic production

(ibid.). Only 11 other chemicals and one other hydrocarbon

(ethylene) are produced in greater tonnage than benzene in

this country (ibid.).

®See Pet. App. 7la. In 1946 the American Conference of

Government Industrial Hygienists (ACGIH) recommended a

threshold limit value (TLV) of 100 ppm. Its recommendation

was reduced to 50 ppm in 1947. In 1948 it followed the

Massachusetts Department of Occupational Hygiene in adopt-

ing a TLV of 35 ppm. In 1963 ACGIH once against reduced

the recommended level, this time to 2u ppm. And in 1969

the American National Standards Institute, a voluntary stand-

ards-setting organization, adopted a TLV of 10 ppm. This

reduction was based on the known hazards of benzene other

than cancer. ACGIH reduced its recommended TLV to 10

ppm in 1974.

—

In 1974 NIOSH advised the Secretary that “the

possibility that benzene can induce leukemia cannot

be dismissed” (Exh. 2-3 at 75).'° NIOSH observed,

however, that the literature dealing with the effects

of benzene on exposed workers consisted primarily

of case history reports rather than statistically valid

studies, and it concluded that there was no “conclu-

sive” proof that exposure to benzene causes leukemia

(Exh. 2-3 at 22-23, 40, 74-75). NIOSH therefore rec-

ommended no changes in the current standard’s TWA

of 10 ppm. On the basis of benzene’s nonmalignant

effects, however, NIOSH recommended that the ceiling

of exposure to benzene for a 15-minute period be

reduced to 25 ppm, and that increased monitoring,

medical surveillance and labeling requirements be

imposed (Pet. App. 72a). |

Between 1974 and 1976 several epidemiological

studies *’ demonstrated a statistically significant risk

that workers occupationally exposed to benzene would

develop leukemia. In 1976 the National Research

Council reviewed the available data and concluded

that benzene must be considered a carcinogen (Exh.

10 NIOSH’s report can be found at 42 Fed. Reg. 22516

(1977).

11 [In an epidemiological investigation, scientists select a

group (“cohort’) of workers who have received exposure

to a chemical for a number of years. They then compare the

number of cancers in that group with the number of cancers

in a similar group that has not been exposed to the chemical,

in order to determine if exposure to that chemical is associated

with the formation of cancers.

2-4 at iii). In 1976 NIOSH reported to the Sec-

retary that the accumulated clinical and epidemio-

logical studies conclusively established that benzene

causes leukemia (Exh. 2-5 at 10). NIOSH stated

that it could not ascertain a “safe’’ level for benzene

exposure (ibid.), and it recommended that permis-

sible exposure be lowered to a two-hour TWA of 1

ppm in the air (Pet. App. 72a). Shortly thereafter,

NIOSH “strongly recommendfed]” that the Secre-

tary immediately promulgate an emergency temporary

standard (Pet. App. 72a; Exh. 2-6 at 1).

7 January 1977 the\Secretary issued voluntary

“Guidelines for Control of Occupational Exposure to

Benzene.” These guidelines recommended that ex-

posure not exceed 1 ppm (Pet. App. 73a; Exh. 2-

44). In April 1977 NIOSH~released to the Secretary

the preliminary results of another study showing a

risk of leukemia among workers exposed to benzene

and again urged him to issue an emergency tempor-

ary standard. On May 3, 1977, the Secretary issued

an emergency standard, which set a maximum ex-

posure limit of 1 ppm. 42 Fed. Reg. 22516 (1977).

12 The National Research Council is the principal operating

agency of the National Academy of Sciences, a congressionally

chartered, quasi-official organization of distinguished scien-

tists and technologists which investigates and reports upon

broad scientific problems relevant to the public welfare.

13 Several petitions for review were filed. See Industrial

Union, AFL-CIO v. Bingham, No. 77-1895 (D.C. Cir.) ;

American Petroleum Institute v. OSHA, No. 77-1516 (5th

Cir.). The Fifth Circuit stayed the emergency temporary

standard. No decision on the merits was issued by either

court. The emergency temporary standard was never enforced.

C. The Permanent Standard

On May 27, 1977, the Secretary published a pro-

posed permanent standard for benzene. 42 Fed. Reg.

27452 (1977). The proposed standard established

an exposure limit of 1 ppm averaged over an eight-

hour period, with a ceiling of 5 ppm during any 15-

minute period. The proposal also required the use of

gloves, masks, or other apparel to prevent eye contact

with benzene and repeated or prolonged skin exposure

to liquid benzene. In addition, monitoring of airborne

concentrations of benzene, medical surveillance, label-

ing, and recordkeeping were proposed (id. at 22521-

22522, 27454). The agency proposed to exempt work

operations during the first year of the standard

where the only exposure to benzene was from liquid

mixtures containing 1% or less benzene. After the

first year, the exception was to apply only to liquids

containing 0.1% or less benzene (id. at 27460-27461)."

Public comment on the proposal was invited (id. at

27452).

The agency received extensive comments from nu-

merous parties, and a hearing was conducted from

July 19 through August 10, 1977. Ninety-five per-

sons appeared as witnesses to represent industries

including refining, petrochemical, oil and gas pro-

duction, aviation fueling, and coke ovens and coke

14 The proposal also excluded activities related to storage,

distribution, and sale of gasoline as a fuel subsequent to its

discharge from bulk terminals. The agency explained that this

area was the subject of an ongoing study and that it intended

to assess the situation after conclusion of that study (42 Fed.

Reg. 27453 (1977)).

by-products. Comments were received from repre-

sentatives of other industries, such as analytical and

research laboratories, paint manufacturing, construc-

tion, maritime, and rubber manufacturing. Repre-

sentatives of the affected workforce, including a num-

ber of employees who had been exposed to benzene,

also testified (Pet. App. 74a-75a). The record of

this proceeding includes a hearing transcript of 3,500

pages and more than 1,100 comments and exhibits.

On February 10, 1978, the Secretary issued the

permanent standard (Pet. App. 34a-66a).° The

standard, codified at 29 C.F.R. 1910.1028, requires

that occupational exposure to airborne benzene be

reduced so that no employee is exposed to more than

1 ppm of benzene in the air, averaged over eight

hours, or to more than 5 ppm of airborne ben-

zene averaged over 15 minutes. Section 1910.1028

(c)(1). The standard requires monitoring of ex-

posure levels and requires employers to advise em-

ployees of their exposure levels. Section 1910.1028

(c). Employers must implement engineering controls

and work practices to bring exposures within permis-

sible limits or as low as possible, and until they

do so must require employees to wear respirators.

Section 1910.1028(f). The standard also requires

labeling of products containing benzene (Section

1910.1028(k)), medical surveillance (Section 1910.-

543 Fed. Reg. 5917-5970 (1978). Errata appear at 43

Fed. Reg. 13561-13563 (1978). — ce printed appendix to the

petition in No. 78-911 does not contain the corrections, which

we will note as necessary.

1028(i)), and employee training (Section 1910.1028

(j)). Direct skin and eye contact with liquids con-

taining benzene is prohibited, and protective clothing

must be provided to employees who are likely to be

exposed to such liquids (Section 1910.1028(c) (2)

and (h)).

D. The Basis for the Standard

The Secretary issued a lengthy statement of find-

ings and conclusions based on the administrative

record, explaining why he adopted the standard (Pet.

App. 67a-25la).

1. Toxic effects of benzene

The Secretary concluded that the evidence in the

record conclusively establishes that benzene causes

leukemia, a cancer of the white blood cells that causes

rapid death (id. at 98a-99a, 122a)."° The Secretary

summarized the voluminous evidence and studies

linking benzene with leukemia (id. at 99a-132a). One

study, for example, showed that workers exposed to

benzene had a risk of leukemia 20 times greater than

the normal population (7d. at 100a); another study

showed a risk five times greater (7d. at 105a)."* In-

16 Despite the fact that half of the adults diagnosed with the

subtype of leukemia (acute myelogenous leukemia or AML)

most commonly associated with benzene have remissions aver-

aging from six to eight months, the prognosis of the disease re-

mains poor. Those who enter into remission have life expec-

tancies of 12 to 18 months; those who do not generally die

within three to six months (Pet. App. 99a).

‘7 In 1977 Dr. Peter F. Infante of NIOSH reported the re-

sults of his retrospective epidemiological study of leukemia

dustry participants in the hearing did not, for the

most part, challenge the conclusion that benzene

causes leukemia (Pet. App. 123a). They argued,

however, that a threshold existed below which ex-

posure to benzene was safe and that 10 ppm, the

among benzene-exposed workers employed at Goodyear’s

Akron and St. Mary’s plants between 1940 and 1949 (Exhs.

2-51, 2A-271). Causes of death among these workers were

determined from death certificates. Dr. Infante and his col-

leagues reported that this study showed a statistically signifi-

cant five-fold increased risk of dying of leukemia, and a 10-

fold increased risk of dying of the type of leukemia (AML)

most commonly associated with benzene, when compared to

two control groups.

Dr. Infante made several assumptions that tend to under-

estimate the cancer risk. First, the status of a number of

employees could not be determined. They were assumed to be

alive. Second, Dr. Infante relied on death certificates to deter-

mine the cause of death; these tend to understate the incidence

of leukemia. Third, Dr. Infante excluded deaths before 1950.

Some victims may have died in the period 1940-1949 (see

Exh. 2-51).

Dr. Marvin Sakol, a hematologist, testified that he had ob-

served nine leukemia cases among the workers studied by Dr.

Infante, almost twice the number reported in the Infante

study (Tr. 286-289, 307, 308-315; Exh. 61). He also observed

12 new leukemia cases between 1960 and 1973; at least five of

these had worked at the plant studied by Dr. Infante (Tr.

315-316). He testified that during the 1950s, when workers’

compensation programs did not recognize leukemia as a com-

pensable employment-related disease, physicians would sign

death certificates stating that the cause of death was aplastic

anemia, rather than a diagnosed leukemia, so the family would

receive compensation (Tr. 311). Dr. Sakol also testified that

the use of death certificates to find the causes of death leads

to underestimation of the number of cases because the im-

mediate cause of death of leukemia patients was often a re-

sulting complication of the disease, such as a heart attack

(Tr. 307-308).

existing standard, was within the safe range (id. at

123a-124a). This argument was based chiefly on

“negative” epidemiological studies of certain groups

of employees, exposed to low levels of benzene, who

did not contract leukemia. The thresholds were said

to lie between the exposure levels of these studies and

the higher exposure levels at which cancers have been

shown to occur.

Although the issue whether thresholds exist is

argued “in a manner similar to those arguing relig-

ion” (Exh. 217.32.H.16 at 1), the majority of the

cancer specialists at the hearing testified that there

are no data reliably establishing that a safe level

exists for benzene.’* The Secretary concluded that no

threshold for benzene could be found because (i)

given difficulties in methods, negative studies are

generally less informative than positive ones (see Pet.

App. 123a-124a), and the specific negative studies

presented by the industrial parties, each of which the

18 See the testimony of Dr. Robert N. Ligo of NIOSH (Tr.

1225; Dr. Herman F. Kraybill of the National Cancer Institute

(Tr. 775-778, 1178-1180, 1210) ; Dr. Charles Shaw of NIOSH

(Tr. 401-402) ; Dr. Peter Infante of NIOSH (Tr. 1066) ; Dr.

John A. Zapp, Jr., testifying for the Manufacturing Chemists

Association (MCA) (Exh. 149-D at 15); Dr. Bernard Gold-

stein of the New York University Medical Center (Tr. 352) ;

Dr. Leonard J. Goldwater, testifying for the MCA (Tr. 2500,

2520-2521); Dr. William R. Gaffey, testifying for respondent

American Petroleum Institute (Tr. 1872). This testimony was

consistent with the 1970 report of the National Cancer Insti-

tute’s (NCI’s) Ad Hoc Committee on the Evaluation of Low

Levels of Environmental Carcinogens, which states: “for

carcinogenic agents a ‘safe level for man’ cannot be established

by application of our present knowledge” (Exh. 2.B.272 at

1). NCI reaffirmed this view at the rulemaking hearing (Tr.

1178-1179). ;

Secretary discussed, are defective;'® (ii) the available

data concerning workers exposed to benzene do not

show the levels of exposure, and it is therefore im-

possible to construct a risk-assessment (or dose-

response) curve that would plot the risk of cancer as

a function of the degree of exposure; *’ and (iii) the

biological mechanism by which benzene causes leu-

kemia is not known (Tr. 363; Exh. 217.32.H.26 at

25).

The existence of a risk of cancer even at levels of

exposure lower than 10 ppm also was supported by

the three affirmative considerations. First, many phy-

sicians subscribe to the theory that “all it takes to

start the carcinogenic process is to have one molecule

of the carcinogen hit the proper site that they now

have identified as DNA in the nucleus of the suscept-

ible organ site” (xh. 97 at 55). If this theory is cor-

rect, there is an ever-present risk of cancer at any ex-

posure, although the risk decreases as the exposure

level decreases and increases with the duration of

the exposure. Second, there is known to be great vari-

' For example, some of the studies did not separate work-

ers who were exposed to benzene from those who were not in

computing the study population’s risk of leukemia. Others fail-

ed to perform follow-up investigations of workers once they

had left the plant, thereby excluding two groups who regu-

larly have a statistically higher incidence of disease than

active workers—retirees and people who terminate employ-

ment early. Still others failed to indicate information necessary

to permit any evaluation of their validity, such as how the

study population was chosen and how long workers had been

exposed to benzene (see Pet. App. 107a-112a).

*° See pages 62-69, infra.

ation in the degree of susceptibility of persons to ben-

zene’s leukemogenie effects, suggesting that the bio-

logical mechanism by which benzene causes cancer

is influenced by age, prior medical problems, genetic

factors, exposure to other carcinogens, and other vari-

ables (Pet. App. 117a-118a). Thus, even if certain

employees would be safe at certain levels of exposure,

other employees might not be safe. The more sus-

ceptible employees cannot be identified in advance of

exposure (id. at 184a). Third, even if a high level

of exposure is necessary to induce leukemia where

the duration of exposure is short, repeated or con-

tinual exposures to low concentrations may be equally

sufficient to cause leukemia (Pet. App. 116a-117a;

Tr. 420, 3209) .”"

The Secretary also concluded that benzene causes

nonmalignant blood diseases, some fatal, for which

no level of safety was known (Pet. App. 96a-98a) ,”

21 The mathematical techniques that have been suggested for

computing thresholds give extraordinarily low figures. For

example, Dr. Leonard J. Goldwater calculated that the thres-

hold for vinyl chloride is .03 parts per billion (Tr. 2524-2526).

Similarly, Dr. Herman F. Kraybill testified that scientists are

generally suggesting that “10° molecules, 10° or 10°” could in-

duce cancer (Tr. 1179). An employee exposed at 1 ppm would

inhale far more than 10° molecules of benzene in an hour.

(The Secretary, however, believes that these techniques do not

yield reliable estimates of the risk of cancer.)

“2 Benzene reduces the number of red blood cells, white blood

cells and platelets (jd. at 82a). A decline of all three is known

as aplastic anemia, which may be fatal (/d. at 82a-84a). De-

cline in the number of red blood cells leads to anemia which,

if chronic, may lead to exacerbation of coronary insufficiencies

or other heart conditions (ibid.). A decline in the number

of white blood cells reduces the natural immunity of the body

and caused chromosomal damage.” These hazards,

he concluded, independently called for reductions in

exposure (id. at 97a-98a).

2. The policy of the lowest feasible level of exposure

Had the Secretary found a safe and achievable

level of exposure to benzene, he would have set the

standard at that level (Pet. App. 183a-190a). Hav-

ing found no safe level of exposure to benzene, the

Secretary was required to decide how stringent the

standard should be. He considered two kinds of ap- -

proach to this task.

One body of public health opinion holds that, where

a safe level of exposure to a carcinogen is not known,

exposure should be kept at zero or, if that is not

from disease (ibid.). Loss of platelets impairs clotting and

may lead to hemorrhages (id. at 82a-83a). Some nonmalig-

nant blood disorders of employees exposed to benzene have

been fatal (id. at 89a). The record shows instances of blood

abnormalities from exposure in the 25-40 ppm range and

aplastic anemia from exposures as low as 5-10 ppm (id. at

89a; Exhs. 154 (Ott Study), 159A, 2B-304). Dermal contact

with liquid benzene may cause erythema, blistering, dermati-

tis and secondary dermal infections (Pet. App. 80a).

2° See Pet. App. 126a-132a. Several studies show chromo-

somal damages below 25 ppm, perhaps as low as 10 ppm (id.

at 128a; Exh. 5B at 8-3). It has been suggested that the

chromosomal damage is associated with leukemia or may

result in genetic mutations (Exhs. 2-94, 2-102, 132). It is not

now possible to determine conclusively what these chromo-

somal aberrations mean in terms of demonstrable health

effects. The Secretary noted, however, that “chromosomal

damage représents an adverse biological event of serious

concern which may pose or reflect a potential health risk and

as such, must be considered in the larger purview of adverse

health effects associated with benzene” (Pet. App. 132a).

= Cteseae

a )360 j

possible, at the lowest feasible level (Pet. App. 183a-

190a). Considerable medical and public health opin-

ion supports this policy (Tr. 148-149, 328, 351, 1229,

1999-2000; Exhs. 2.B.272 at 1-2, 2-5 at 10). In

view of the command of 29 U.S.C. 655(b) (5) that

the Secretary adopt “the” standard that ‘“‘most ade-

quately assures, to the extent feasible” that “no”

employee will suffer material health impairment even

if he is exposed for his entire working life, the Sec-

retary adopted this approach, as he has in promul-

gating numerous other OSHA standards for carcino-

gens.** The Secretary found that the lowest feasible

level is 1 ppm (Pet. App. 187a).”°

*% In setting standards for carcinogens and other toxic ma-

terials for which no known safe level of exposure can be

found, the Secretary has consistently adopted the lowest feasi-

ble level and expressly declined to implement a cost-benefit

approach. See 37 Fed. Reg. 11318 (1972) (asbestos dust) ; 39

Fed. Reg. 3756, 3758 (1974) (14 substances) ; 39 Fed. Reg.

85890, 35892 (1974) (vinyl chloride); 41 Fed. Reg. 46742,

46750-46751 (1976) (coke oven emissions); 43 Fed. Reg.

9GRk, 27378-27379 (1978) (cotton dust) ; 43 Fed. Reg. 52952,

52963 (1978) (lead dust) ; 43 Fed. Reg. 45762, 45790 (1978)

(acrylonitrile).

The Secretary also has published a proposed regulation that

would permit him to establish categories of toxic substances.

See 42 Fed. Reg. 54148 (1977) (Identification, Classification

and Regulation of Toxic Substances Posing a Potential Car-

cinogenic Risk).

2>The Secretary also found that prudent public health

policy would lead to the 1 ppm level without regard to ben-

zene’s causation of leukemia. After observing that benzene

was known to cause blood diseases at exposures less than

100 ppm, the Secretary observed (id. at 97a-98a) :

It is customary to use a safety factor of 10-100 or greater

depending on the seriousness of the toxic effects and the

Industry witnesses suggested a different approach

(see A. 40-58, 77-94, 135-148, 158-179). They con-

tended that the degree of health protection should be

determined through a cost-benefit technique, which

would measure as closely as possible the number of

lives that would be saved at progressively lower levels

of exposure. Under this approach the Secretary

would set the level of exposure at the point where the

additional lives saved by further reduction are not

worth the additional cost to save them (Pet. App.

160a-161a). Whether the cost is worthwhile would be

determined by comparing the estimate of the number

of lives that would be saved by the new standard with

the number of lives (and other health or environmen-

tal benefits) that might be achieved by requiring in-

dustry to apply the same funds to meet other environ-

mental or health standards. At a minimum, industry

argued, workers ought to be willing to tolerate health

risks comparable to those government has, in other

contexts, implicitly found to be “acceptable.”

The Council on Wage and Price Stability, in a

somewhat similar fashion, urged OSHA to investi-

gate the incremental costs and benefits of alternative

standards ‘“‘to determine whether a proposed standard

should be made even more stringent or perhaps re-

nature of the data being relied upon. That is, the lowest

levels at which effects had been observed would be re-

duced by the safety factor chosen in establishing the

exposure limit. Taking this approach in the case of ben-

zene would lead to a permissible exposure substantially

less than 10 ppm without regard to the issue of leukemia

and the view that no safe Jevel can be established for the

carcinogenic risk.

laxed in some circumstances” (A.-131). The Council

concluded that, although the procedures it suggested

to estimate the health impact of alternative benzene

standards would not “yield an accurate calculation of

incremental benefits, they would provide a range of

health benefit estimates that should be included in an

economic analysis in order to aid OSHA in setting

the final standard” (A. 130).

The Secretary explained his decision not to adopt

these approaches on two grounds. First, the Secretary

concluded that the lack of data made it impossible to

quantify the health risk at various levels of expo-

sure, and prudent public health policy required that

exposure to a known carcinogen. be reduced as much

as possible until such time as it can be shown that

a safe threshold of exposure exists (id. at 160a).

Second, even if data were available to quantify the

risk, Section 655(b) (5) requires the adoption of the

standard which assures maximum benefit (i.e., pre-

vention of sc.ious illness or death), constrained only

by the limits of feasibility (Pet. App. 162a) :

In making judgments about specific hazards,

OSHA is given discretion which is essentially

legislative in nature. However, in setting an

exposure limit for a substance like benzene,

OSHA may not substitute cost benefit criteria

for the legislatively determined directive of pro-

tecting all exposed employees against material

impairment of health or bodily function. Where

the health effectiveness of the alternative ap-

proaches are extremely uncertain and likely to

vary from situation to situation, OSHA must

adopt the compliance strategy which provides

the greatest certainty of worker protection even

if that approach carries with it greater economic

burdens for the affected employers. In the case

of the benzene standard, the evidence in the rec-

ord indicates that the costs of compliance are

not overly burdensome to industry.

26 Although the Secretary disclaimed any obligation to

quantify the health benefits and standards, in response to the

contention that standards should be set by weighing health

benefits against costs the Secretary stated (Pet. App. 160a) :

There is general agreement that benzene exposure

causes leukemia as well as other fatal diseases of the

bloodforming organs. In spite of the certainty of this

conclusion, there does not exist an adequate scientific

basis for establishing the quantitative dose response re-

lationship between exposure to benzene and the induction

of leukemia and other blood diseases. The uncertainty

in both the actual magnitude of expected deaths and

in the theory of extrapolation from existing data to the

OSHA exposure levels places the estimation of bene-

fits on “the frontiers of scientific knowledge.” While the

actual estimation of the number of cancers to be pre-

vented is highly uncertain, the evidence indicates that

the number may be appreciable. There is general agree-

ment that even in the absence of the ability to establish a

“threshold” or “safe” level for benzene and other car-

cinogens, a dose response relationship is likely to exist;

that is, exposure to higher doses carries with it a higher

risk of cancer, and conversely, exposure to lower levels is

accompanied by a reduced risk, even though a precise

quantitative relationship cannot be established. In light of

the uncertainties in this area of scientific knowledge,

OSHA believes that it is required by prudence and by

the statutory mandate to adopt a highly protective posture

) in considering the evidence for health benefits.

The Secretary found that 629,000 employees are affected by

the benzene standard and that at least 30,000 employees are

exposed now to levels between 1 ppm and 10 ppm (see Pet.

App. 141la-155a).

3. Feasibility

The Secretary recognized that standards set under

Section 655(b)(5) must be “feasible” (Pet. App.

132a-133a). To assess the economic feasibility of

the proposed standard, the agency commissioned

Arthur D. Little, Inc., to study the economic effect

of the proposed standard on the affected industries.

The Secretary thoroughly reviewed the Little study

(id. at 134a-158a).

The Little study considered compliance costs with

respect to each of 20 affected industries (Pet. App.

142a-153a). It concluded that the aggregate first-

year operating costs would be in the range of $187

million to $205 million. Recurring annual costs there-

after were estimated at approximately $34 million.

One-time investment in engineering controls was esti-

mated at approximately $266 million (id. at 134a).

After comparing these costs with information on the

profitability and capitalization of the affected indus-

tries, the Secretary concluded that there is no “seri-

ous question concerning the economic feasibility of

the standard or the ability of the regulated indus-

tries to bear the additional costs” and that the com-

pliance costs are “well within the financial capability

of the covered industries” (id. at 135a, 163a).

4. Protective clothing

The proposed and final standard prohibited all eye

and skin contact with certain liquids containing ben-

zene and required protective clothing where employ-

ees may have repeated eye or skin contact with such

liquids (Section 1910.1028(h), 42 Fed. Reg. 27461

(1977) ; Section 1910.1028(c) (2), and (h), Pet. App.

36a, 42a). This requirement is based on the Secre-

tary’s policy, consistent with recognized industrial

hygiene practice, that “in dealing with a carcinogen,

all potential routes of exposure (i.e. inhalation, inges-

tion, and skin absorption) be limited to the extent

feasible” (id. at 192a).

During the hearings, Dr. Howard Maibach testified

on behalf o” respondent American Petroleum Insti-

tute. (Tr. 2459; A. 58) that, although he had no

personal information about whether benzene may be

absorbed through the skin, the extent of absorption

(including the effect on absorption of temperature,

solution strength, repeated application, etc.) could

be resolved conclusively by conducting tests on ani-

mals using Carbon-14 (Tr. 2459-2490; A. 58-77, 149-

157). “[W]ith modern analytic and biological tech-

nology,” he stated, “it should be very simple, very

quickly, and inexpensively to answer this” (Tr.

2467; A. 68). He stated that the tests could be

conducted in six to twelve weeks (Tr. 2487; A. 75)

and at one point testified that “in the six weeks nec-

essary to do these experiments, you will have your

answer” (Tr. 2488; A. 76). Dr. Maibach did not say

that he was conducting such tests or that he intended

to do so. Although the hearing record was held open

until September 2, 1977, neither Dr. Maibach nor

anyone else submitted the results of any such experi-

ment.

27 The test involves applying radioactive benzene solution

to animal skin and later measuring the amount of radioactive

benzene in the animai’s feces and urine to determine how

much of the benzene had penetrated the skin.

The Secretary concluded that five considerations

supported the view that dermal contact should be

prohibited. First, “the view is generally accepted

that [benzene] has the ability to be absorbed through

the skin * * *” (Exh. 2-69 at 1; A. 107, 119-120; Tr.

860; A. 40; see Pet. App. 198a).** Physicians believe

that benzene is absorbed because benzene is a power-

ful lipid solvent—a characteristic which indicates it

can break through the protective cutaneous barrier

(Exh. 2: A-187 at 11). Experiments show skin ab-

sorption of benzene in animals (Exhs. 2-31 at 12, 2-

46 at 1-2; A. 95-96; Exh. 2-69 at 1-2; A. 107-108)

and in humans (Exh. 2-69; A. 119-120).

Second, even if benzene were not absorbed through

intact skin, it may be absorbed through damaged

skin (Pet. App. 1938a). As Dr. Maibach testified,

scientific experts “ordinarily assumed that penetra-

tion of any molecule will be greater through damaged

skin than undamaged skin” (Tr. 2469-2470; A. 64-

65). Benzene itself can damage the outermost pro-

tective cutaneous layer (see Exhs. 2-3 at 5, 53; 2-48;

28 See also Exhs. 2-31 at 12; 2-48; 2:A.187 at 11.

2° The Secretary did not consider these studies to be de-

finitive, either because of their lack of control data or because

of their failure to consider factors that the scientific com-

munity has found to be relevant to the permeability of human

skin (Pet. App. 193a; see Tr. 2463-2464; A. 60-61). On the

other hand, studies indicating that benzene is not absorbed

in humans (Exhs. 2-46, 2-47; A. 95-106; Tr. 2462-24638; A. 59-

60) were not considered sufficient because they are based on

analysis of the sulfate ratio in the subjects’ urine, a factor

too insensitive and non-specific for useful application (Exhs.

2-69 at 3; A. 109; 2:A-112 at 3; Tr. 2462-2463; A. 59-60).

2:A-111 at 9; 2:A-112 at 2; 2:A-187 at 6, 11), thus

increasing the likelihood of absorption.

Third, benzene often is added to or present as a

contaminant in industrial solvents containing other

chemically-related substances. See Exh. 5A at 4-19,

4-27 to 4-28, 4-31 to 4-32. Some of these chemically-

related substances, such as xylene and toluene, are

readily absorbed through the skin (see Exh. 2-3 at

25). Such substances can act as “vehicles” or car-

riers, which result in or increase the extent of skin

absorption of benzene (Pet. App. 194a).

Fourth, it is difficult to imagine a situation where

dermal contact with benzene will not also result in

airborne exposure to the vapors from that liquid.

Thus, whatever exposure occurs via skin absorption

simply adds to that from inhalation. It is therefore

advisable to guard against dermal contact in order

to prevent exposure in excess of the permissible safe

exposure level set for ambient concentrations (Pet.

App. 194a).

Finally, apart from the question of absorption, di-

rect contact with liquids containing benzene can

cause erythema and blistering (Pet. App. 80a). Pro-

longed or repeated skin contact with liquids contain-

ing benzene has been associated with the development

of a dry, scaly dermatitis, and with secondary dermal

infections (ibid.).

5. The percentage exclusion

The proposed standard exempted “work operations

where the only exposure to benzene is from liquid

mixtures containing 1% (0.1% after one year from

the effective date of this standard) or less of ben-

zene by volume, or the vapors released from these

liquids.” Section 1910.1028(a) (2) (ii), 42 Fed. Reg.

27460 (1977). The exemption would have applied

both to inhalation and to personal contact with liq-

uids. The original announcement called for public

comment on this proposed exclusion and suggested

that the final standard might not contain any such

exclusion.*

The standard promulgated on February 3, 1978,

had ‘no percentage exclusion. In response to con-

tentions by industry that the standard was not tech-

nologically feasible without a percentage exclusion,

and well after the petitions for review of the un-

amended standard were filed, the Secretary announced

on March 28, 1978, proposed rulemaking to amend

30 The announcement stated (42 Fed. Reg. 27452 (1977)):

One of the major issues involved in this proceeding is

whether the final standard should exempt certain speci-

fied operations from its coverage. For the purpose of

raising this issue, the proposal, as drafted, does not

cover * * * operations in which the only exposure to

benzene is from liquid mixtures which contain no more

than 0.1% benzene (1% benzene during the first year

after the effective date of the standard). The question

of whether these types of operations should be exempted

is raised herein, and some of the reasons for the sug-

gested exclusion of these operations are set out below.

If these operations are excluded from the final standard

issued as a result of this rulemaking, they would still

be covered by the existing benzene standard in § 1910.-

1000. OSHA seeks public comment and evidence on these

issues.

the standard to exclude certain liquids containing

benzene. 43 Fed. Reg. 12890 (1978).

On May 27 and 28, 1978, the Secretary conducted

hearings on the proposed percentage-exclusion amend-

ment. During the hearing, counsel for respondent

Rubber Manufacturers Association requested a 180-

day stay of the standard, which had already been

issued, while Dr. Maibach conducted tests on skin

absorption (Amend. Tr. 15-23; A. 181-187). The

Secretary replied that he would consider such a study

but that he would not delay the standard while the

study was being conducted (Amend. Tr. 20-21; A.

185-186).

On June 21, 1978, the Secretary announced a per-

centage exclusion almost identical to the one origin-

ally proposed. 48 Fed. Reg. 27971 (1978); A. 239-

275."

E. The Decision of the Court of Appeals

On petitions for review under 29 U.S.C. 655(f),

the court of appeals set aside the standard.” The

31 The original proposal would have exempted work opera-

tions where the only exposure is from liquids containing 1%

or less benzene during the first year and 0.1% thereafter.

Section 1910.1028 (a) (2) (ii), 42 Fed. Reg. 27460 (1977). The

final amended standard exempted work operations where the

only exposure is from liquids containing 0.5% or less benzene

by volume during the first three years of the standard and

0.1% thereafter. Section 1910.1028(a) (2) (iii) ; A. 275.

82 The initial petition for review was filed in the Fifth Cir-

cuit by respondent American Petroleum Institute. Petitions

subsequently filed in other circuits were transferred to the

Fifth Circuit and consolidated for review (see A. 1-11). Other

parties intervened in the Fifth Circuit.

The first petition for review was filed on February 2, 1978,

well before the amendments to the standard had been promul-

court began by observing that Section 655(f) explic-

itly prescribes the standard of review: ‘The deter-

minations of the Secretary shall be conclusive if sup-

ported by substantial evidence in the record considered

as.a whole.” The court observed that there are both

factual and policy determinations in rulemaking. Sec-

tion 655(f) subjects factual determinations to the

substantial evidence test (Pet. App. 5a). With re-

spect to policy determinations by the Secretary, the

court concluded, it need only determine whether ‘the

Secretary exercised his decisionmaking power within

the limits imposed by Congress” (ibid.).

Applying this standard, the court of appeals held

that substantial evidence supports the Secretary’s

factual findings that (a) benzene causes cancer (Pet.

App. 17a-18a); (b) benzene causes diseases, some

fatal, other than cancer (id. at 7a); (c) exposure at

the 10 ppm level poses a risk of leukemia (id. at 18a)

and reduction of the level to 1 ppm would result

in health benefits (ibid.). The court did not dis-

gated (A. 12-18). The amendment was issued the day before

oral argument. The court of appeals called for supplemental

briefing on the effect of the amendment. It concluded that it

did not need to review the amendment and reserved the ques-

tion of its feasibility. Petitions for review of the amendment

are still pending in the court of appeals. Because the altera-

tions of the standard ultimately did not affect the court of

appeals’ disposition of the case, it is unnecessary to review

the propriety of this proceedure. Compare Tagg Bros. & Moor-

head v. United States, 280 U.S. 420, 444-445 (1930), with

H & H Tire Co. v. United States Department of Transporta-

tion, 471 F.2d 350, 353 n.10 (7th Cir. 1972).

pute the Secretary’s finding that the standard is eco-

nomically and technologically feasible. The court

acknowledged that it is not now possible to determine

the relationship between exposure and risk (id. at

19a n.28).

The court nevertheless set aside the standard. It

concluded that the Secretary must ascertain the num-

ber of lives (and other health benefits) that the

standard will save and determine whether their value

bears a “reasonable relationship” to the cost of im-

plementing the standard (Pet. App. 19a-20a).** The

court found this requirement in the Act’s definition

of an occupational health and safety standard, 29

U.S.C. 652(8), which defines such a standard to be

one “which requires conditions, or the adoption or

use of one or more practices, means, methods, oper-

ations, or processes, reasonably necessary or appro-

ate to provide safe or healthful employment” (em-

phasis added). The court stated that this provision

is analogous to a statutory requirement in the Con-

sumer Product Safety Act that prohibits any con-

sumer product safety rule unless the Consumer

Product Safety Commission finds that the rule is

“reasonably necessary to eliminate or reduce an

unreasonable risk of injury associated with such

83 The court of appeals stated that it did not require an

“elaborate” cost-benefit analysis (Pet. App. 17a). Nonetheless,

it did require the Secretary (1) to approximate the number

of lives that the standard will save, and (2) to balance these

benefits against the cost of the standard in determining the

level of protection to accord employees (id. at 19a-22a). For

convenience, we refer to these two requirements as a cost-

benefit test.

product.” 15 U.S.C. 2058(c)(2)(A). Having held

in Aqua Slide ‘N’ Dive Corp. v. CPSC, 569 F.2d 831,

842 (5th Cir. 1978), that the latter provision re-

quires the agency to find that the benefits of its rule

bear a reasonable relationship to the costs, the court

“decline[d] to construe the precisely similar require-

ments of these two acts differently or to read words

out of the OSHA legislation” (Pet. App. 14a-15a).

To the Secretary’s argument that 29 U.S.C. 655

(b) (5) calls for the “highest degree of health and

safety protection” and requires the selection of the

standard that “most adequately assures, to the extent

feasible” that no employee will suffer material im-

pairment of health even if he is exposed for his

entire working career, the court responded that the

Secretary is not given “unbridled discretion” to cre-

ate “absolutely risk-free workplaces regardless of

cost” (Pet. App. 15a). The statute limits the Sec-

retary to “feasible” standards based on “the best

available evidence.” These provisions suggested to the

court “pragmatic limitations” on the Secretary’s de-

cisions (id. at 16a). In any event, the court thought,

in view of the parallel purposes of OSHA and the

Consumer Product Safety Act it must be “guided by

Aqua Slide’ in determining whether the Secretary

had chosen properly (ibid.).

The court held that the Secretary passed the

“costs” part of its test by relying on the Little study

to show that the cost of compliance was feasible

(Pet. App. 17a). But the court stated that the Sec-

retary failed the “benefits” part of the test because

he could not show that “measurable” benefits would

result (id. at 18a). The court conceded that there is

“general agreement in the scientific community that

exposure to carcinogens at iow levels is safer than ex-

posure at higher levels” and that it is rational to con-

clude that a reduction in the limit from 10 to 1 ppm

would result in “some benefit” (zbid.). As rational

as this deduction might be, however, “mere rational-

ity” was not sufficient. The Secretary must show

“measurable benefits” (ibid.).*

The court acknowledged that the Secretary is un-

able to satisfy this requirement because the lack of

data concerning the effects of exposure at low levels

makes an estimate of benefits “impossible” (Pet.

App. 19a n.23). Nonetheless, it concluded that the

standard could not be upheld because “Congress pro-

vided that OSHA regulate on the basis of knowledge

rather than on the unknown” (id. at 20a). The court

speculated that the necessary data could be obtained

in the future (7d. at 19a n.23, 20a). (This specula-

tion was based on three items of evidence that we

discuss at note 56, infra.)

84 Although disclaiming any obligation to satisfy a cost-

benefit test, the Secretary argued that the standard would

satisfy such a test because the standard is within the financial

capability of the industry, because exposure to benzene is

unsafe at any level, and because exposures to lower levels

would be safer than exposures at higher levels. Conceding

that “some” health benefits would result from the lower stand-

ard, the court nonetheless rejected this argument, holding that

“some” benefits did not satisfy its test, which requires benefits

to be quantified (Pet. App. 18a-19a).

The court also set aside the prohibition of dermal

and eye contact with liquids containing benzene.

The court held that the Secretary’s finding that

dermal contact poses risk of cancer would ‘‘pass mus-

ter” when viewed under the traditional substantial-

evidence test (Pet. App. 25a). The court held, none-

theless, that the Secretary was obliged to “acquire”

the radioactive benzene test recommended by Dr.

Maibach before promulgating any standard based on

skin absorption (id. at 26a-27a). No such experi-

ment had been conducted before the Secretary prom-

ulgated his standard, but Dr. Maibach’s testimony

led the court to believe that definitive results could

be made “readily available.” The court held that the

Secretary had therefore violated Section 655(b) (5)’s

requirement to set standards based on “research, dem-

onstrations, experiments, and such other information

as may be appropriate,” and on the “latest available

scientific data in the field.”

Because all of the standard’s other requiy2ments

were “tied to” the permissible exposure limit and the

dermal contact prohibition, the court set aside the

entire standard (Pet. App. 33a).

SUMMARY OF ARGUMENT

The court of appeals has required the Secretary of

Labor to show that the costs of regulating toxic sub-

stances under the Occupational Safety and Health

Act are commensurate with the benefits of the regu-

lation. Although cost-benefit analysis may be a use-

ful tool in ordinary cases of regulation, OSHA is no

ordinary statute, and 29 U.S.C. 655(b) (5), the toxic

substances provision, is no ordinary command. Sec-

tion 655(b) (5) instructs the Secretary to ensure that

“no employee will suffer impairment of health or

functional capacity even if such employee has regu-

lar exposure to the hazard dealt with by such stan-

dard for the period of his working life.” The Sec-

retary “shall” set “the” standard necessary to ac-

hieve the level of exposure that “most adequately as-

sures, to the extent feasible,” that the no-material-

impairment criterion is met. It recapitulates its pur-

pose as “the attainment of the highest degree of

health and safety protection for the employee * * *.”

The statute simply leaves no room for the conclusion

that the Secretary must balance costs against bene-

fits in setting an exposure level.

The court of appeals found its cost-benefit require-

ment in 29 U.S.C. 652(8), which defines a health and

safety standard as a rule regulating practices and

conditions of employment “reasonably necessary or

appropriate to provide safe or healthful employ-

ment.” But a cost-benefit standard can be placed in

this definitional clause only by main force. The leg-

islative history of the entire Act contains not one

mention of a cost-benefit study requirement, and

there is no hint that Section 652(8) was intended

to modify the substantive requirements of 29 U.S.C.

655(b) (5). Under the statutory scheme, an occu-

pational health and safety standard is “reasonably

necessary or appropriate” if it carries out the goals

identified in the substantive parts of the statute. The

benzene standard does so, and it therefore meets the

statutory requirements. Here, as elsewhere in the

law, the phrase “reasonably necessary” is a descrip-

tion rather than a limit. See, e.g., United States v.

An Article of Drug * * * Bacto-Unidisk, 394 U.S.

784, 786 n.2, 791-792 (1969); Mourning v. Family

Publications Service, Inc., 411 U.S. 356, 369 (19738).

The toxic substances provision itself requires the

Secretary to achieve the lowest possible risk within

the range of feasibility. The language, which is

quoted above, shows that a more particular cost-bene-

fit balance is unnecessary. The legislative history

shows that Congress meant what it said. The prin-

cipal concern of the draftsmen of the Act was toxic

substances. The statutory language was attacked re-

peatedly as unyielding; the supporters of the statute

did not reject this analysis but embraced it as a de-

scription of their purpose. When a compromise finally

was struck, Congress decided to restrict the scope of

Section 655(b) (5) to toxic or hazardous substances

—rather than, as originally proposed, all industrial

hazards—but to leave the force of the section un-

changed. The nature of the compromise is readily

understood: proponents of regulation obtained a

sweeping statute concerning toxic substances, while

opponents confined the strongest section of the stat-

ute to this particular hazard.

The word “feasible,” although introducing cost

considerations into the Secretary’s decisions, does not

permit a cost-benefit test to be employed in setting

toxic substances standards. No one suggested in the

legislative debates that “feasible” has any meaning

other than its usual one: “capable of being achieved.”

No Member of Congress suggested that a standard is

not “feasible” if, although achievable, it is too costly

in relation to its benefits. The absence of such a sug-

gestion is telling. Any construction of the statute

that introduces a cost-benefit balance through the

“feasibility” provision would render meaningless the

non-material-impairment criterion of Section 655(b)

(5). There is simply no reason to conclude that Con-

gress drafted an internally contradictory statute.

The Fifth Circuit’s approach is at war with Sec-

tion 655(b) (5) in yet another way. Section 655(b)

(5) requires the Secretary, in issuing standards for

toxic materials, to act on “the bes! available evi-

dence.” This provision was inserted because Congress

was aware of the “frightening gap” between “our

knowledge about [toxic] substances, and what we

need to know” (Leg. Hist. 1044-1045). Congress

therefore authorized the Secretary to act on the

basis of the “best available evidence” because “it

[was] not intended that the Secretary be paralyzed

by debate surrounding diverse medical opinions.” /d.

at 848. The Fifth Circuit’s cost-benefit rule creates

precisely the type of paralysis that Congress intended

to avoid. In the present case, the best available evi-

dence for benzene is simply inadequate to perform

the type of risk assessment and cost-benefit analysis

ee ee

desired by the Fifth Circuit. In the Secretary’s view,

he will be unable to issue any benzene standard that

satisfies the court’s test. Congress could not have in-

tended to establish a cost-benefit requirement that

could not be met on the basis of the best available

evidence.

The court of appeals also erred in vacating the

protective clothing requirement. The court held that

the Secretary should not have established this provi-

sion without first “acquiring” the results of a series

of tests using radioactive benzene suggested by Dr.

Howard Maibach. But the Secretary’s decision to

include this provision without first acquiring the

radioactive-benzene studies did not violate the re-

quirement of 29 U.S.C. 655(b) (5) that the standard

be based on the “best available evidence” or the re-

quirement that it be based on “research, demonstra-

tions, experiments, and such other information as

may be appropriate.” The protective clothing re-

quirement was based on “research, experiments, and

demonstrations.” It also was based on the best evi-

dence “available.” Dr. Maibach’s study was then not

“available.” Nor is it available today, except for

“preliminary results.” The predictions by him and

industry counsel that studies with conclusive results

could be performed in only six to twelve weeks proved

unfounded. By requiring the Secretary to postpone

the standard until allegedly “readily” available

studies are conducted, the court of appeals again has

introduced the type of paralysis that Congress sought

to avoid.

ARGUMENT

THE OCCUPATIONAL SAFETY AND HEALTH ACT

DOES NOT REQUIRE THE SECRETARY TO ESTAB-

LISH THAT THE COSTS OF PARTICULAR REDUC-

TIONS IN EXPOSURE TO TOXIC SUBSTANCES ARE

JUSTIFIED BY QUANTIFIABLE BENEFITS

The court of appeals held that the Occupational

Safety and Health Act conditions the Secretary’s

ability to order reductions in exposure to toxic sub-

stances on his ability to show that the benefit to the

workers’ health bears some quantifiable relationship

to the costs of the regulation. The court conceded

that no such requirement can be found in 29 U.S.C.

655(b) (5), which authorizes and sets the criteria

for toxic substances regulations. The court found its

requirement, instead, in a definitional provision of

the Act that speaks of standards “reasonably neces-

sary or appropriate” for public health. 29 U.S.C.

652(8). In the court’s view, a cost-benefit rule is

reasonable public health policy, and thus to be read

into the Act through the definitional section. The

court of appeals thought that it makes simple good

sense to require the agency to determine that the

benefits of its regulations are worthwhile when meas-

ured against the costs.

But the court of appeals was not at liberty to insist

that the Secretary of Labor follow its approach in

dealing with toxic substances. Although many stat-

utes permit or require the kind of cost-benefit assess-

ment that the court of appeals thought to be desirable,

other statutes do not. See, e.g., Union Electric Co.

v. EPA, 427 U.S. 246 (1976) (the Clean Air Act

contains no requirement that state implementation

plans for air pollution be economically or technologi-

cally feasible, much less cost-beneficial) ; Comment,

Implementing the Anticancer Clauses of the Food,

Drug and Cosmetic Act, 44 U. Chi. L. Rev. 817

(1977) (discussing several statutes that regulate use

of carcinogens without regard to cost). When a stat-

ute requires regulation in the name of safety without

regard to the balance between costs and benefits,

or even feasibility, neither the Executive Branch nor

a court may require such a balance to be struck. TVA

v. Hill, 437 U.S. 158, 198-195 (1978).

The toxic-substances provision of the Act does not

go quite that far. 29 U.S.C. 655(b) (5) states that

the Secretary “shall” set “the” standard that “most

adequately assures, to the extent feasible” that ‘no

employee will suffer material impairment of health

or functional capacity even if such employee has

regular exposure to the hazard dealt with by such

standard for the period of his working life.” The

section then recapitulates its goal as “the attainment

of the highest degree of health and safety protection

for the employee * * *.” A statute written in such

terms is an unmistakable instruction from Congress

that employee health is to be achieved even at great

cost.> The policy of Congress is that cost may be

considered only to determine what is “feasible.” In

the pages that follow, we show that neither the defi-

nitional provision of 29 U.S.C. 652(8) nor the “feasi-

bility” language of Section 655(b)(5) requires the

Secretary to engage in the sort of cost-benefit balanc-

ing proposed by the court of appeals.

A. The Definitional Phrase “Reasonably Necessary Or

Appropriate” Does Not Impose A Cost-Benefit Test

The Fifth Cireuit concluded that 29 U.S.C. 652

(8), which defines an “occupational safety and health

standard” as a rule that requires conditions, means,

methods or processes “reasonably necessary or ap-

propriate to provide safe or healthful employment,”

imposes a cost-benefit requirement on toxic substances

35 Every court of appeals (other than the Fifth Circuit in

this case) that has considered the role of costs in standard

setting has held (either explicitly or implicitly) that the

Secretary need not undertake a cost-benefit study before

establishing a toxic substances regulation. American Iron and

Steel Institute v. OSHA, 577 F.2d 825 (3d Cir. 1978), peti-

tions for cert. pending, Nos. 78-918 and 78-919 (coke oven

emissions) ; Society of the Plastics Industry, Inc. v. OSHA,

509 F.2d 1301 (2d Cir.), cert. denied, 421 U.S. 992 (1975)

(vinyl chloride) ; Industrial Union Department, AFL-CIO Vv.

Hodgson, 499 F.2d 467 (D.C. Cir. 1974) (asbestos dust). Cf.

Turner Co. v. Secretary of Labor, 561 F.2d 82, 86 (7th Cir.

1977) (ban on “life-threatening hazards” such as industrial

carcinogens would not be subject to any cost-benefit test, even

though regulation of other hazards under a section other than

Section 655(b) (5) is so subject); RMI Co. v. Secretary of

Labor, No. 78-3312 (6th Cir. Mar. 13, 1979), slip op. 10-12

(same).

—

standards issued under 29 U.S.C. 655(b) (5). The

court of appeals is wrong. The phrase “reasonably

necessary or appropriate” is not a limitation on the

Secretary’s powers or a substantive standard of any

sort. It is, rather, a description of the rules produced

by adherence to the criteria laid down elsewhere in

the statute.”

Even if the phrase is given substantive meaning,

however, it is not a brake on the exercise of powers

86 Nothing in the lengthy legislative history of the Act sug-

gests that the “reasonably necessary or appropriate” lan-

guage is a description of the Secretary’s duty, rather than (as

we suggest) a description of rules that result if the Secretary

conscientiously carries out the duties established by the sub-

stantive sections of the statute.

All but one of the original bills (see Leg. Hist. 659)

contained a definition of “occupational safety and health

standard” similar to the definition enacted as 29 U.S.C.

652(8). Some used only the phrase “reasonably necessary”

rather than “reasonably necessary or appropriate,” but other-

wise they were identical. Leg. Hist. 3, 35, 78, 235, 532, 601,

631, 683, 725, 768, 937. Both the Senate and the House ver-

sions of the Act used the phrase “reasonably necessary or

appropriate.” Id. at 235, 937. No suggestion was made any-

where in the committee reports that the definition was in-

tended to limit the scope of permissible standards. See id. at

166, 866. No such claim was made in the floor debates. The

absence of any such assertion is telling, inasmuch as critics of

the Act in both chambers assailed the Act for its requirement

in Section 655(b) (5) that “absolute” health protection be

provided to employees regardless of cost (see pages 59, 4a, 6a,

infra). Indeed, so far as we can find, no mention was ever

made of the phrase, aside from repeating it in the section-by-

section explanations in the committee reports. In short, from

beginning to end, the legislative record is devoid of evidence

that the definition was meant to modify or even to describe

the Secretary’s duties under 29 U.S.C. 655 (b) (5).

granted by other sections of the same Act. Where

“the empowering provision of a statute states simply

that the agency may ‘make .. . such rules and regu-

lations as may be necessary to carry out the provi-

sions of [an] Act,’ * * * the validity of a regulation

promulgated thereunder will be sustained so long as

it is ‘reasonably related to the purposes of the enabling

legislation.’ Mourning v. Family Publications Serv-

ice, Inc., 411 U.S. 356, 369 (1973) (footnote omit-

ted). For example, where a statute authorized the

FDA to promulgate drug regulations “necessary”’ to

achieve the goal of a public health statute, this Court,

in sustaining the regulations, observed that “[i]t is

enough for us that the expert agency charged with

the enforcement of remedial legislation has determined

that such regulation is desirable for the public health

* * *” United States v. An Article of Drug * * *

Bacto-Unidisk, 394 U.S. 784, 786 n.2, 791-792 (1969).

Statutes that authorize regulations “necessary or ap-

propriate” to carry out the provisions of the statute

have been similarly construed.” FPC v. Texaco Inc.,

87 Where a federal statute authorizes an agency to issue

regulations to further the goals of a statute, this Court regu-

larly has held that the courts may not set aside the agency’s

regulations unless they are inconsistent with the goal of the

statute or violate a requirement of the statute. Bowles V.

Willingham, 321 U.S. 508, 519 (1944) ; SEC v. Chenery Corp.,

332 U.S. 194, 207 (1947) ; American Trucking Associations V.

United States, 344 U.S. 298 (1958) ; United States v. Storer

Broadcasting Co., 351 U.S. 192, 201-203 (1956); Thorpe v.

Housing Authority of the City of Durham, 393 U.S. 268, 280-

281 (1969) ; FCC v. National Citizens Committee for Broad-

417 U.S. 380, 394 (1974) ; Permian Basin Area Rate

Cases, 390 U.S. 747, 776 n.40 (1968) ; United States

v. Bush & Co., 310 U.S. 371, 380 (1940). And,

where a statute permitted the ICC to establish “rea-

sonable” regulations concerning railroad car avail-

ability and service, this Court concluded that the ICC

could “rationally” conclude that its regulations served

the statutory purpose despite the contention of the

railroads that the short-term injury of the regula-

tions far outweighed the long-term benefits. United

States v. Allegheny-Ludlum Steel Corp., 406 U.S. 742,

752-753 (1972).

-casting, 436 U.S. 775, 796-797 (1978); FERC v. Pennzoil

Producing Co., No. 77-648 (Jan. 16, 1979), slip op. 8.

The Act provides that ‘“[t]he determinations of the Secre-

tary shall be conclusive if supported by substantial evidence in

the record considered as a whole.” 29 U.S.C. 655(f). The

courts of appeals agreed that this substantial evidence test

requires that factual determinations made by the Secretary

be supported by substantial evidence. Policy determinations

made by the Secretary need only be consistent with the statu-

tory goal and any statutory requirements. See Associated

Industries of New York State, Inc. v. United States Depart-

ment of Labor, 487 F.2d 342, 347-350 (2d Cir. 1973) ; Florida

Peach Growers Ass’n v. United States Department of

Labor, 489 F.2d 120, 127-129 (5th Cir. 1974): Industrial

Union Department, AFL-CIO vy. Hodgson, supra, 499 F.2d

at 472-476; Synthetic Organic Chemical Manufacturers As-

sociation Vv. Brennan, 508 F.2d 1155, 1158-1160 (3d Cir.

1974) ; Society of the Plastics Industry, Inc. v. OSHA, supra,

509 F.2d at 1303-13804; American Iron and Steel Institute

v. OSHA, supra. The Fifth Circuit accepted this construction

of Section 655(f) (Pet. App. 4a-5a). We believe that this con-

clusion, about which the parties are not in dispute, is correct.

It shows that the usual principles of review of regulations,

discussed in the preceding paragraph and in the text, must be

applied here.

The phrase “reasonably necessary or appropriate

to provide safe or healthful employment” is in this

tradition. So long as the standard meets the other

requirements of the Act, Section 652(8) permits the

Secretary to issue any standard that he rationally

concludes is appropriate for the protection of the

health and safety of employees under the substantive

provisions of the Act.

The Fifth Circuit’s reliance on Aqua Slide ‘N’ Dive

Corp. v. CPSC, 569 F.2d 831 (5th Cir. 1978), which

construes part of the Consumer Product Safety Act,

was unwarranted. Although both statutes refer to

standards “reasonably necessary” to achieve certain

goals, the structures of the statutes are otherwise

quite dissimilar. The “reasonably necessary” phrase

in the Cortsumer Product Safety Act, 15 U.S.C. 2058

sumer Product Safety Commission to set consumer

product safety standards that are “reasonably neces-

sary to eliminate or reduce an unreasonable risk of

injury * * *” (emphasis added). The phrase thus

appears in the clause granting the power to regulate,

rather than in a collateral definition. The Consumer

Product Safety Commission’s responsibility, moreover,

is to find that its rules are “reasonably necessary to

eliminate or reduce an unreasonable risk of injury”;

the Commission must consider “the probable effect of

such rule upon the utility, cost, or availability of [the

affected] products” and “any means of achieving the

objective of the order while minimizing adverse effects

on competition or disruption or dislocation of manu-

facturing and other commercial practices.” 15 U.S.C.

2058(¢c)(1)(C), (D) and (2). Before the Commis-

sion may ban a product altogether as unsafe it must

find that “no feasible consumer product safety stand-

ard * * * would adequately protect the public from the

unreasonable risk of injury associated with such prod-

ucts.” 15 U.S.C. 2058(c)(2)(C). The Consumer

Product Safety Act thus recognizes that some prod-

ucts (e.g., knives) pose a risk of harm that should be

tolerated in light of their benefits; the Act permits

regulation only of risks that are ‘unreasonable.’ *

In contrast, OSHA does not limit the Secretary to

regulating only “unreasonable” health risks; it au-

thorizes the Secretary to eliminate, if feasible, all

risks of material impairment that toxic substances

pose to employees’ health.

*S The Fifth Circuit observed in Aqua Slide that the CPSC

cannot determine whether an “unreasonable risk” is pre-

sented without balancing the severity of the injury that may

result from the product, discounted by the unlikelihood of

the injury, against the harm the regulation itself will im-

pose on manufacturers and consumers. 569 F.2d at 839.

Consumer product safety risks may be reasonable because

the very feature that presents a risk (e.g., the sharp edge of a

knife) also offers a benefit. Consumers ought to be able to

choose whether the risk to them is worth the benefit to them,

and it is understandable that only hazards which are “unrea-

sonable” are regulated. In contrast, occupational health haz-

ards pose risks only to employees. The benefits (lower produc-

tion costs) accrue in the main to employers and consumers.

It is therefore understandable that 29 U.S.C. 655(b) (5) auth-

orizes the Secretary to regulate all risks of material impair-

ment to health caused by toxic substances rather than merely

“unreasonable” health risks.

B. Section 655(b)(5) Prohibits Basing A Standard On

A Cost-Benefit Test

If the Court agrees with the argument that 29

U.S.C. 652(8) does not require the application of

cost-benefit criteria in making decisions about the

level of exposure to toxic substances, it then must

decide whether 29 U.S.C. 655(b) (5) requires cost-

benefit analysis. We have pointed out above (see

pages 41-42, supra) that the language of the statute

contains no hint of such a requirement. To the con-

trary, the Secretary “shall” select “the” standard

that “most adequately assures, to the extent feasible”

that “no” employee shall suffer, even if exposed to

the toxic substance “for the period of his working

life.’ The Secretary has followed this mandate by

ascertaining the level of exposure to a substance at

which employees will net risk incurring the disease

and then determining whether this exposure level is

“feasible.” If it is not feasible to attain the safe level,

the Secretary adjusts the standard accordingly. But

a court may not override this process with a cost-

benefit balancing that could well require the Secre-

tary to tolerate additional occupational illness in order

to achieve lower costs when a stricter standard would

be feasible; that squarely violates the no-material-

impairment rule of the statute.

But no statute is so plain that resort to the legis-

lative history is forbidden. See, e.g., United States

v. Culbert, 485 U.S. 371, 374 n.4 (1978); Train v.

Colorado Public Interest Research Group, Inc., 426

U.S. 1, 10 (1976). We therefore examine the legis-

lative history of the Act and show that nothing in

a ae)

the history supports the court of appeals’ creation

of a cost-benefit test. We then turn to a number of

logical problems in the interpretation of the “feasi-

bility” language of Section 655(b) (5) on which re-

spondents rely.

1. The legislative history establishes that Congress

wanted the Secretary to regulate toxic substances

by considering costs only as part of “feasibility”

Foremost among the concerns of Congress in enact-

ing OSHA was the ever-increasing number of toxic

materials and harmful physical agents in the work-

place. Leg. Hist. 142-148, 159-160, 411-415, 503, 510,

845-846, 1061. The Senate Committee on Labor and

Public Welfare, for example, reported that toxic

agents were a pernicious threat to employee health:

Workers in the dusty trades still contract various

respiratory diseases. * * * Carcinogenic chemi-

cals, lasers, ultasonic [sic] energy, beryllium

metal, epoxy resins, pesticides, among others, all

present incipient threats to the health of work-

ers. * * * It is estimated that every 20 minutes

a new and potentially toxic chemical is intro-

duced into industry. New processes and new

sources of energy present occupational health

problems of unprecedented complexity.

Recent scientific knowledge points to hitherto

unsuspected cause-and-effect relationships —be-

tween occupational exposures and many of the

then so-called chronic diseases—cai cer, respira-

tory ailments, allergies, heart disease, and others.

In some instances, the relationship appears to be

direct: asbestos, ionizing radiation, chromates,

and certain dye intermediaries, among others,

are directly involved in the genesis of cancer. In

other cases, occupation exposures are implicated

as contributory factors. The distinction between

occupational and nonoccupational illnesses is

growing increasingly difficult to define.

Id. at 142.”

Death from such “silent killers” *° was occurring

because many employers had little economic incentive

to provide healthful workplaces. “[M]any employers

—particularly smaller ones—simply cannot make the

necessary investment in health and safety, and sur-

vive competitively, unless all are compelled to do so.”

Id. at 144 (Senate Report). Because employees often

moved from job to job, “an employer investing in

current occupational health precautions would re-

ceive no future benefits, not even in the reduction of

workimen’s compensation costs. He might be paying

for the past neglect of another employer.” Id. at 854

(House Report).

In these circumstances, Congress concluded, the

best approach was to impose a duty on all employers

to provide workplaces as free from health hazards

as possible. Congress recognized that achievement of

this goal would require substantial investment in

health protection (Leg. Hist. 444, 510, 525-526, 1004,

1030, 1040, 1150-1151), and that these costs “would

be put into consumer goods * * *.” Id. at 444 (Sen.

Yarborough).** But Congress concluded that the ex-

39 The House Committee on Education and Labor likewise

identified exposure to toxic substances as a principal concern.

Id. at 844-846.

40 Leg. Hist. 1044-1045 (Rep. Udall).

41 Congress anticipated that to some extent the cost of com-

pliance would be recovered in savings in health care, work-

Pains oe

pense of providing a safe and healthful place to work

is a necessary and reasonable cost of doing business.

Senator Eagleton observed that “[t]he costs that will

be incurred by employers in meeting the standards

of health and safety to be established under this bill

are, in my view, reasonable and necessary costs of

. doing business. Whether we, as individuals, are moti-

vated by simple humanity or by simple economics, we

can no longer permit profits to be dependent upon an

unsafe or unhealthy worksite.” Jd. at 1150-1151.

Senator Williams, perhaps the principal author of

the Act, stated that the cost of compliance was “the

price we should pay for the 80 million workers in

America.” Id. at 444.”

Congress recognized that health standards would

significantly increase the cost of doing business. It

required these costs to be borne so long as it was

feasible to do so. This fixes the balance between costs

men’s compensation, lost wages and lost productivity. Jd. at

142, 844-845, 1061.

42 Senator Yarborough, a co-sponsor of the bill that was

ultimately enacted, replied to the claim that the bill would be

expensive (id. at 510):

One may well ask too expensive for whom? * * * Is it

too expensive for the employee who loses his hand or

leg or eyesight? Is it too expensive for the widow trying

to raise her children on meager allowance under the

workmen’s compensation and social security? * * *

* * * * *

We are talking about people’s lives, not the indifference

of some cost accountants. We are talking about assuring

the men and women who work in our plants and factories

that they will go home after a day’s work with their

bodies intact.

and benefits in a way that favors health. Conse-

quently, Congress decided, economic hardship may not

be a ground for obtaining a variance from a standard.

29 U.S.C. 655 (b) (6) (C), 655(d) ; Leg. Hist. 160. For

toxic substances, it decided that “the” standard which

“most adequately assures, to the extent feasible” that

“no” employee would suffer material health impair-

ment during his entire working life must be adopted.

But no Member of Congress so much as suggested

that, because of cost-benefit considerations, the Secre-

tary must tolerate risks, caused by toxic substances,

that could be eliminated feasibly. Indeed, we can find

no reference in the legislative history to cost-benefit

techniques._in any context whatsoever. The silence

carries a persuasive message.**

The pattern and structure of the Act also demon-

strate Congress’ preference for health protection with- —

out regard to a particular balance against costs. Con-

gress declared its purpose to be “to assure so far as

possible” that every worker in the nation has a safe

and healthful working environment. 29 U.S.C. 651(b).

The Act imposes on all employers a duty to “furnish

to each of his employees employment and a place of

employment which are free from recognizable hazards

that are causing or are likely to cause death or seri-

ous physical harm to his employees.” 29 U.S.C. 654

(a)(1). In adopting national consensus standards

and established federal standards during the first two

48 The Appendix to this brief contains a further discussion

of the genesis of the toxic substances provision.

years of the Act, the Secretary was required to choose

the standards that would result in improved safety

or health. In the event of a conflict.among possible

standards, the Secretary was required to “promul-

gate the standard which assures the greatest protec-

tion of the safety or health of the affected employ-

ees.” 29 U.S.C. 655(a). If the Secretary adopts a

permanent standard that differs from an existing

national consensus or established federal standard, the

Secretary must explain why the change will better

effectuate the purposes of the Act. 29 U.S.C. 655

(b) (1). Where the standard deals with toxic sub-

stances, 29 U.S.C. 655(b) (5) requires that the Sec-

retary adopt “the” standard which “most adequately

assures, to the extent feasible, that no employee will

suffer material impairment of health or functional

capacity even if such employee has regular exposure

to the hazard dealt with by such standard for the

period of his working life.” Variances from stand-

ards may be obtained, but ‘only on carefully limited

grounds—financial hardship is not a ground for a

varianee.“* 29 U.S.C. 655(b)(6)(C) and 655(d).

44 Although financial hardship is not an excuse for noncom-

pliance, the Secretary’s enforcement program takes hardship

into account in working out abatement programs consistent

with the purposes of the Act. The Secretary’s Industrial

Hygiene Field Operations Manual, which sets forth proced-

ures for enforcement of health standards, states:

The employer’s economic cost of abatement will not

be considered to be a factor in the issuance of a citation.

However, if the cost of utilization of effective engineering

or administrative controls, or combination, which would

bring the employer into compliance with permissible

See note 7, supra. Financial hardship is to be re-

lieved directly only through small business loans to

assist employers in complying with standards. 15

U.S.C. 636.

In short, at every turn in the statute Congress

expressed a preference for maximum feasible health

protection.** That Congress failed to qualify this goal

in any way with a requirement that the benefits be

measurable, refused to allow variances on the ground

of financial hardship, and limited its only financial

relief to small business loans demonstrates that Con-

gress carefully did not require any cost-benefit test.

The court of appeals erred in reading into the statute

“what Congress has studiously omitted.” FTC v.

Simplicity Pattern Co., 360 U.S. 55, 67 (1959);

exposure limits would so seriously jeopardize his financial

condition as to result in the probable shutdown of his

establishment or a substantial part, then only a personal

protective equipment program should be required in the

interim, and an extended abatement date should be

granted to permit the extended implementation of feasible

engineering or administrative controls taking into account

the employer's financial condition. The burden of proof

of economic infeasibility rests upon the employer. Any

abatement date based on economic grounds shall be ap-

proved by the Regional Administrator.

Empl. Safety & Health Guide (CCH) { 4483.80, § 3.8(b) (1)

(1979). See also Atlantic & Gulf Stevedores, Inc. v. OSHRC,

534 F.2d 541 (3d Cir. 1976).

45 See SEC v. Sloan, 486 U.S. 103, 121 (1978); Train v.

Natural Resources Defense Council, Inc., 421 U.S. 60, 86

(1975) (scope of agency’s rulemaking powers determined in

part by examination of the pattern and structure of the

statute in question).

Susaeriins.

Great Atlantic & Pacific Tea Co. v. FTC, No. 77-654

(Feb. 22, 1979), slip op. 8-9.

2. The requirement that standards be “feasible” re-

fers to practicability rather than to a cost-benefit

balance

29 U.S.C. 655(b) (5), which authorizes toxic sub-

stance standards, has both a procedural directive and

a substantive criterion.*® Procedurally, the Secretary

must consider the best available evidence, the latest

experience in the field; he must base standards on

experiments, research, and such other information as

may be appropriate. Substantively, the Secretary

must set “the” standard that “most adequately as-

sures, to the extent feasible” the “highest degree of

health and safety protection.” There is no room in

this scheme for the Secretary to set standards by

weighing the value of life and against the cost

of their preservation. Congress has done that; it

struck the balance in favor of maximum health pro-

tection, subject only to the requirement of feasibility.

The “lowest feasible level” policy of the Secretary (see

pages 20-25, swpra) satisfies the imperative. The

Fifth Circuit’s cost-benefit test violates it.’

46 The language is quoted in full at page 3, supra.

4? This does not mean that the Secretary may never weigh

costs and benefits where sufficient data exist to de so. To the

contrary, the Secretary is required to set priorities in the

use of his own time and resources according to the severity of

the need for achieving better health and safety protection

against various hazards. 29 U.S.C. 655(¢). In determining

where the need for prompt rulemaking is the greatest, the

Secretary may take the relative costs and benefits of the

Respondents may contend that the phrase “to the

extent feasible” in Section 655(b) (5) itself imposes

a cost-benefit requirement. The court of appeals did

not accept this argument, and for good reasons.

The word “feasible” does not require any cost-

benefit test. The ordinary meaning of “feasible” is

“capable of being done.” See Webster’s Third New

International Dictionary 831 (1971).*° The word

does not refer to the desirability or benefit of doing

something; it refers, instead, to the ability of doing

something. That is how Congress has consistently

employed “feasible” in health or safety legislation.

Congress’ understanding is clearly conveyed by an-

other statute enacted the same week as OSHA.” This

existing standards and the proposed standards into account.

But once the Secretary decides to issue a standard for toxic

substances, he must satisfy the requirements of Section

655 (b) (5).

48 See also I Compact Oxford English Dictionary 9738:

“Capable of being done, accomplished or carried out; possible,

practicable.”

49 Section 211(c) (2) (A) and (B) of the Clean Air Act

authorizes the EPA to regulate, control or prohibit automobile

fuel additives. Control or prohibition of additives that en-

danger the public health was permitted only after “consider-

ation of other technologically or economically feasible means

of achieving emission standards * * *.” In contrast, control

or prohibition of additives which may impair effective opera-

tion of emission control devices was permitted only after

consideration of ‘a cost benefit analysis” comparing device

performance with and without the control or prohibition. 42

U.S.C. 7545 (c) (2) (A) and (B) (emphasis added).

Later Congresses likewise recognized the difference be-

tween a feasibility requirement and a requirement that evo-

oe a2

Court, too, has understood that “feasible” means

capable of achievement. See Union Electric Co. v.

EPA, supra. Respondents do not dispute that the

benzene standard is feasible in this sense—it can be

achieved at bearable cost with available technology.”

Moreover, if “to the extent feasible” means “‘to the

extent justified by the costs of doing so,” then most

of Section 655(b)(5) is sapped of meaning. The

words requiring the Secretary to set “the” standard

nomic effects be balanced or considered. The Consumer Prod-

uct Safety Act of 1972 permits the Consumer Product Safety

Commission to regulate product hazards after it considers

the risks presented by a product and the economic effects of

its proposed safety rule. 15 U.S.C. 2058 (c) (1) (C) and (D).

The Commission may take the more drastic step of banning

the product altogether only after finding that no “feasible”

safety rule would provide sufficient protection. 15 U.S.C. 2057.

Section 325 of the Energy Policy and Conservation Act of

1975, 42 U.S.C. 6295, requires promulgation of energy con-

servation standards. Section 325(a) specifies that a standard

must be “economically justified.” A standard is economically

justifiable only if the Secretary of Commerce finds both that

the benefits of the standard outweigh its costs and that the

standard is “economically feasible.” “Feasibility” is defined

as “capable of being carried out.” S. Rep. No. 94-516, 94th

Cong., Ist Sess. 172 (1975) ; H.R. Conf. Rep. No. 94-700, 94th

Cong., lst Sess. 172 (1975).

°° Economic considerations necessarily play a substantial

role in the Secretary’s determination that a particular level of

exposure is feasible. It may be technologically possible to make

benzene on the moon, but if doing so were the only way to

reduce exposure, the suggestion would not be feasible. The

Secretary takes economic considerations into account in de-

termining what is feasible. For example, considerations of

economic feasibility were important in setting the level of

exposure to cotton dust in certain textile operations (43 Fed.

Reg. 27360 (1978) ), and the Secretary allowed the inorganic

lead industry additional time (as much as 10 years) to imple-

that “most” adequately assures that “no employee

will suffer material impairment of health * * * even

if [exposed to the hazard] for the period of his work-

ing life’ would become nugatory. A construction

that renders state language meaningless should be

avoided if possible. Weinberger v. Hynson, Westcott

& Dunning, 412 U.S. 609, 683 (1973); Jarecki v.

G.D. Searle & Co., 367 U.S. 308, 307 (1961). Here

there is no reason why these words should be disre-

garded—they are the heart of the statute. Indeed,

giving the “feasibility” language a cost-benefit mean-

ing would turn the statute on its head. Standards

for hazards other than toxic substances and harmful

physical agents are governed by other sections that

lack the “feasibility” phrase. If the feasibility lan-

guage is the source of a cost-benefit requirement, then

only toxie substances rules need to meet such a test.

That is topsy-turvy, because the entire structure of

the Act establishes that Congress wanted toxic sub-

stances placed under controls far more strict than

those applicable to simple hazards (such as slippery

floors, which also are subject to regulation). See

pages 6a-7a in the Appendix, infra.

The submission that “feasible” means “capable of

being achieved” within the range of acceptable costs

ment the exposure levels for similar reasons (43 Fed. Reg.

52978, 53008 (1978) ). In this case the Secretary examined the

capitalization of firms, the profitability of businesses, and other

factors in determining that the cost of the standard was bear-

able. See Pet. App. 134a-163a. Because respondents do not dis-

pute the economic feasibility of the benzene standard, it is un-

necessary to discuss here what criteria the Secretary should

use in making these judgments.

5 atte eee CA

ee ol Pee «

(see note 50, supra) is supported by the legislative

history of Section 655(b)(5). Section 655(b) (5)

was inserted to assure “so far as possible” that

standards would prevent material health impairment:

Such standards should be directed at assuring,

so far as possible, that no employee will suffer

impaired health or functional capacity, or di-

minished life expectancy, by reason of exposure

to the hazard involved, even though such expo-

sure [to the hazard] may be over the period of

his entire working life.

Leg. Hist. 147 (Senate Report) (emphasis added).

In the Senate members objected to the original

language of Section 655(b)(5), which would have

required the Secretary to establish the standard that

“most adequately and feasibly assures” that no em-

ployee will suffer material health impairment. It

_ was argued that this language would mandate “abso-

lute safety,” which is an “impossibility.” Leg. Hist.

480-481. Other criticism, in the same vein, is dis-

cussed in the Appendix, infra, 4a-8a. In response to

this criticism, Congress restricted Section 655(b) (5)

to standards for “toxic materials or haimful physical

agents,” rather than all standards, as was originally

proposed, and modified the feasibility language to its

present version. Leg. Hist. 502-508. It is difficult

to believe that this debate would have occurred with-

out mention that the word feasible was intended to

impose a cost-benefit requirement, if that were Con-

gress’ purpose. Not once, however, throughout the

entire legislative history of 29 U.S.C. 655(b) (5) is

there any suggestion of such a meaning. In sum, the

plain language of Section 655(b) (5) and its legisla-

tive history are irreconcilable with the Fifth Circuit’s

cost-benefit test.

3. The requirement that the Secretary make decisions

on the basis of the “best available evidence” pre-

cludes any role for cost-benefit analysis if the

necessary data are not “available”

Section 655(b) (5) instructs the Secretary to act

on the basis of the “best available evidence.” Con-

gress was aware of the “frightening gap” between

“our knowledge about [toxic] substances, and what

we need to know” (Leg. Hist. 1044-1045). The best-

available-evidence provision was inserted because “it

[was] not intended that the Secretary be paralyzed

by debate surrounding diverse medical opinions.” Jd.

at 848. This provision fortifies the conclusion that a

cost-benefit analysis is unnecessary under the stat-

ute; even if such an analysis sometimes can be con-

ducted, the data are not always “available.” Where,

as here, “the factual finger points [but] does not

conclude” (Society of the Plastics Industry, Inc. v.

OSHA, 509 F.2d 1301, 1808 (2d Cir.), cert. denied,

421 U.S. 992 (1975) ), the Secretary can promulgate

standards now whether or not more complete evi-

dence could be made available in the future.

The significance of the best-available-evidence pro-

vision is two-fold. First, it supports our construc-

tion that the Act never requires the Secretary to en-

gage in cost-benefit balancing. Because Congress was

aware of the gaps in medical research concerning

toxic substances,’ Congress could not have intended

51 As of 1969 more than 6,000 chemicals were in use in oecu-

pational settings, and it was estimated that every twenty

me ——

to impose a cost-benefit requirement that in numer-

ous instances would frustrate all regulation because

even the best available evidence would not be good

enough. To construe the statute to require quantifi-

cation not possible even with the “best available evi-

dence” would “impute to Congress a purpose to

paralyze with one hand what it sought to promote

with the other.” Weinberger v. Hynson, Westcott &

Dunning, supra, 412 U.S. at 631, quoting Clark v.

Uebersee Finanz-Korp., 382 U.S. 480, 489 (1947).

Second, even if the Act imposes a cost-benefit re-

quirement where evidence is available to assess costs

and benefits, that requirement must yield where the

best available evidence is insufficient. Either con-

struction of the best-available-evidence clause re-

quires reversal in this case, because the court of ap-

peals conceded that the best available evidence is not

good enough to enable the Secretary to determine the

relationship between exposure to benzene and risk of

cancer (Pet. App. 19a n.23).

This is enough, we submit, to show that the Fifth

Circuit’s approach should be rejected. But because

the court of appeals apparently thought that its cost-

benefit test would not be excessively burdensome, we

minutes a new and potentially toxic substance was introduced

into industry. Health authorities could demonstrate safe levels

of exposure for only 500 of these substances. Leg. Hist. 412,

1044. Emphasizing the “frightening gap between our knowl-

edge about [toxic] substances, and what we need to know,”

Rep. Udall stated that it was not possible to calculate “how

many workers have been struck down by silent killers in the

form of corrosive chemicals, noxious fumes, and debilitating

dust particles that have become commonplace ingredients of

the manufacturing process.” Leg. Hist. 1044-1045.

try to show in the next few pages that the court is

wrong. Its two-step test imposes the very paralysis

that Congress sought to avert.

Under the court of appeals’ approach the Secretary

must quantify the cancer risk, that is, construct a

dose-response curve showing how much cancer is

caused by each of the low levels of exposure in ques-

tion (1 to 10 ppm). Once the cancer risk is meas-

ured, with whatever degree of accuracy the court

requires, the Secretary must demonstrate that the

health benefits bear a “reasonable relationship” to the

cost of achieving them. Although serious methodo-

logical obstacles attend the second step,” the Secretary

will be unable to satisfy even the first step in many

cases.

52 Even when available data are sufficiently reliable to per-

mit the probable health benefits of a standard to be predicted,

accurately, there are problems in using cost-benefit criteria to

set an exposure standard where, as here, it is possible to set a

“feasible” exposure limit and reduce the risk of employees

more directly. The most important difficulty in setting a

higher-than-feasible exposure level using cost-benefit tech-

niques is placing an economic value (implicit or explicit) on

life, health, bereavement, pain and suffering, and other con-

sequences of occupational disease and death. There is no gen-

erally accepted method for valuing such considerations, even

implicitly. See Senate Comm. on Governmental Affairs, 95th

Cong., 2d Sess., Study on Federal Regulation, Vol. VI, at

XXIV (Comm. Print 1978). For a discussion of the various

methods occasionally proposed to quantify the value of life and

other intangibles, see 41 Fed. Reg. 46742, 46751 (1976) (coke-

oven-emissions preamble). See generally Fried, The Value of

Life, 82 Harv. L. Rev. 1415 (1969) ; Rhoads and Singer, What

is Life Worth?, 51 Pub. Interest 74 (Spring 1978). Second,

there is no general agreement concerning whether a stream

of health benefits and saved lives extending into the future

should be discounted to present value in the same way that

wees RL. oad gchean it! 4 2 T te 1 een

a eee a ES

~ . $ ‘ ‘ + ‘ | =. (25

The court of appeals’ approach seems to require

the Secretary simply to identify the benefits of reduc-

ing exposure from 10 ppm to 1 ppm, an entire order

of magnitude. Relatively accurate measurements of

such benefits eventually may be within the reach of

medical science, as the court of appeals thought. But

the court of appeals’ analysis cannot logically be con-

fined to changes in this magnitude. It is entirely for-

tuitous that the national consensus standard of ben-

zene exposure was 10 ppm; the court of appeals’

approach would require the Secretary to identify the

benefits of moving from 5 ppm to 1 ppm, if that had

been the proposed change. Or if the Secretary had

proposed a decrease from 10 ppm to 5 ppm, the court

of appeals would have demanded to know the benefits

of that course. The smaller the proposed decrease in

question, the more difficult it is to learn its benefits,

and thus the more difficult it becomes to satisfy the

court oi appeals’ requirements.

In fact, despite the superficial reasonableness of

the court’s approach, its test could not be satisfied

until the Secretary has learned and can quantify with

future economic values are generally discounted in awarding

tort damages. Third, the Secretary has found that industry

tends to overestimate the cost of implementing proposed

standards, which makes it difficult accurately to weigh benefits

against costs. For example, industry asserted that the Secre-

tary’s vinyl chloride standard would seriously disrupt the

industry. In fact, however, the industry promptly complied,

at a cost much lower than it had projected, by devising new

technology to meet the standard. Vinyl chloride production

has continued to increase. See Doniger, Federal Regulation of

Vinyl Chloride: A Short Course in the Law and Policy of

Toxic Substances Control, 7 Ecol. L.Q. 497 (1978).

useful precision the benefits to be obtained by reducing

exposure through each part of the critical range. Un-

less the Secretary can learn these savings, he will be

confronted with the demand in tomorrow’s case that he

show what part of a reduction in exposure contributes

to what part of the saving of life. For example, if 10

years from now the Secretary learns with satisfactory

certainty that a reduction in exposure to benzene from

10 ppm to 1 ppm would save 100 lives every year,

industry will respond: ‘How do you know that a

reduction to 1 ppm is necessary? What if all of the

Savings are achieved by a drop to 5 ppm, making the

further reduction superfluous?” And that response

will have substantial force if, as is often the case,

the first step of a reduction (say, from 10 ppm to 5

ppm) can be achieved quite easily, while the further

five-fold reduction to 1 ppm is costly. If 95% of the

cost of reducing exposure from 10 ppm to 1 ppm lies

in the reduction from 5 ppm to 1 ppm, then—if the

court of appeals is correct—the Secretary surely must

show what benefits are achieved by the reduction

from 5 ppm to 1 ppm. It will not be enough to show

the benefits of the entire drop; they must be shown

segment-by-segment.”

58 Jn other words, the court’s approach would call for sub-

stantial evidence that risk drops as, say, a linear function of

exposure rather than as a quadratic function. If the dose-

response curve looks more like a parabola than like a straight

line, the gains from reducing exposure below a certain level

may become vanishingly small. But because of difficulties in

the data (discussed in the text infra), the method by which

results are extrapolated from high levels to low ones must be

based on a non-verifiable choice of shape for the curve. The

It should be clear, however, that medical science

cannot supply sufficiently useful approximations in

_such an exercise. The very obstacles that prevented the

“Sucitiy from attempting to measure the benefits of a

reduction from 10 ppm to 1 ppm demonstrate this.

Given the deficiencies in studies of the consequences

of human exposure to benzene, it will be years be-

fore science accumulates enough information on even

so substantial a decrease. This is so principally be-

cause there is very little information available on the

actual concentrations of benzene at which health ef-

fects have been observed. Unless concentrations can

be identified with some accuracy, scientists cannot ex-

trapolate from their observations. Moreover, because

leukemia often does not become manifest until long

after the exposure to its cause (see Pet. App. 117a),

it will be decades before all of the cancers caused by

the exposure at the 10 ppm level are discovered (see

Exh. 2-58 at 50-56). The exposure records of many

workers exposed to benzene at other levels in other

years may be lost—if any were kept—and this hinders

reconstruction of a group of exposed workers for pur-

decision whether to try to fit a linear rather than a parabolic

curve to the known observations, given the lack of adequate

knowledge about how cancer is caused, is a choice of policy

rather than of fact. In these circumstances the choice of one

approach over another could not be justified by substantial

evidence. Extrapolations based on a model so chosen thus

could not answer the court of appeals’ apparent demand for

“harder” data. See McGarity, Substantive and Procedural Dis-

cretion in Administrative Resolution of Science Policy Ques-

tions: Regulating Carcinogens in EPA and OSHA, 67 Geo.

L.J. 729, 735 n.27, 806 (1979).

poses of comparison.™ Finally, because workers move

from job to job, and may be exposed to different car-

cinogens in different jobs, it is quite difficult to deter-

mine which of the observed cancers in any group are

attributable to exposure (or to how much exposure)

to benzene.

These formidable difficulties pertain to the meas-

urement of the incidence of cancer at the 10 ppm

level. The difficulties in estimating the amount of

cancer at the 1 ppm level—or any other level that is

not now employed—are even more serious. If anal-

ysis of the data concerning cancers occurring at high

levels of exposure leads to uncertainties, extrapola-

tions to lower levels from these data are even mvure

uncertain (Pet. App. 119a-120a).

The court of appeals did not give any reason for

concluding that these difficulties and uncertainties

can be overcome. The Secretary, who has a substan-

tial scientific staff and a decade’s experience in the

** For example, the Infante study of Goodyear employees

exposed between 1940 and 1949 (see note 17, supra) was

frustrated by the unavailability of sufficiently accurate data

about exposure levels (Exh. 2-5 at 3; Tr. 806, 813-814, 1055-

1056, 1156, 2506-2507). The data showed exposure peaks in

some cases exceeding 200 ppm, but they did not reveal how

much of the total exposure was attributable to these peaks and

how much to the approved levels (19 ppm to 100 ppm). It

cannot be determined whether the workers wore respirators

in some situations, thus reducing exposure. Dr. Infante con-

cluded that no reliable estimates of exposure could be estab-

lished, and thus that it was not possible to establish a cor-

respondence between the level of exposure and risk (see note

56, infra). The other studies in the record suffer from the

same deficiencies (Pet. App. 99a-104a, 120a-122a, 160a).

once

aaa

BEET Sade al inn re nth cs CA oh A EELS i ret

administration of toxic substances regulation, has con-

cluded that they cannot, at least within the foresee-

able future. The Secretary’s view that there is much

to be learned, and much progress to be made, before

risk assessments could be devised to meet the court

of appeals’ demands is consistent with the Interagency

Regulatory Liaison Group’s Report on Scientific Bases

for Identifying Potential Carcinogens and Estimat-

ing Their Risks (Feb. 6, 1979). This Report, pre-

pared by representatives of OSHA, the Consumer

Product Safety Commission, the Food and Drug Ad-

ministration and the Environmental Protection

Agency, concluded that the state of the art permits

only a “rough estimate * * * which may be useful

in setting priorities among carcinogens and in ob-

taining a very rough idea of the magnitude of the

public health problem posed by a given carcinogen.”

Id. at 76, 77. And even that much—which does not

include determination of measurable benefits of the

sort envisioned by the Fifth Circuit—can be achieved

only if the data are available.”

55 We do not mean by this that analyses of the costs of

achieving a given level of protection are useless or should be

avoided. Far from it; they provide information that is in some

cases very valuable. The President has encouraged their use.

Exec. Order No. 12044, 43 Fed. Reg. 12661, 12663 (1978).

In recognition of their value, the President has proposed leg-

islation that would, if adopted, require any agency in promul-

gating a major rule to prepare “an analysis of the projected

benefits and the adverse economic and other effects of the

rule” and either an explanation “why the rule attains its ob-

jectives, in a manner consistent with applicable statutes, with

less adverse economic effects than the other alternatives con-

The Secretary, under a statutory obligation to use

the best “available” evidence, resolved the uncertain-

ties presented here in favor of safety. We have dis-

cussed in detail at pages 15-25, supra, how the Secre-

tary weighed the competing considerations and how

he resolved the numerous uncertainties. The court

of appeals did not fault the Secretary’s performance

in this respect. Because the Secretary did the best

he could with the best “available” evidence, the court

of appeals’ insistence that the Secretary create more

sidered” or “why an approach entailing greater adverse

economic effects was selected.” S. 755, 96th Cong., 1st Sess.

§ 3 (1979). Our point here, rather, is that the cost analysis of

this sort is quite different from a cost-benefit balancing for

setting the level of protection in the first instance.

The Secretary also uses risk-assessment techniques where

that is possible to evaluate whether a reduction in exposure

would produce some benefit, and thus to show some health

need for the standard. See note 47, supra. For example, in the

coke-oven-emissions case, the Secretary conducted such a

study. The results are discussed at note 15 of cur memo-

randum in response to the petitions in Nos. 78-918 and 78-919.

But as that discussion shows, even those studies have obvious

limitations.

The Secretary makes even greater uses of risk assessments

when dealing with non-carcinogens. For example, in estab-

lishing the permissible exposure level for lead, the Secretary

was able to assess the extent of risk, identify different popu-

lations that faced risk thresholds at different levels of expos-

ure, and consider these factors in setting the permissible ex-

posure level. See 43 Fed. Reg. 52952, 52960-52972 (1978).

In some instances “safe” levels of exposure to non-carcinogens

can be determined. The determination of safe ‘levels, once

made, can be the basis for satisfying the statutory criteria

without reducing exposure to the lowest feasible levels. Here,

too, however, the technique was used to identify the nature

and extent of health needs rather than to balance costs against

benefits.

vis AN de call

FLERE LN tenets he Oe A\emn Dey Mh ha Fae asa

evidence was unjustified. Cf. Permian Basin Area

56 The court of appeals suggested that the Secretary await

the results of animal studies before ordering reductions in

exposure to benzene (Pet. App. 20a n.25). But, as the court

recognized, it has not been shown that benzene causes cancer

in animals. It has, however, been shown that benzene causes

cancer in humans; no one disputes this. Under these circum-

stances, it is difficult to understand what increase in pre-

cision might be achieved by waiting for the accumulation of

evidence that, by any standard, is not “available.” See Pet.

App. 119a-120a (discussing animal studies).

The court of appeals also concluded that some evidence now

available showed that more “valid extrapolations” would be-

come available in the future (Pet. App. 19a n.23). Aside from

the obvious fact that what might become available in the

future is not available now, the court of appeals’ excursion

into the risky business of scientific prediction is open to seri-

ous dispute; it is at odds with the record and with the Secre-

tary’s conclusions.

First, the court relied on Dr. Richard Wilson’s “preliminary

risk assessment” showing that only two deaths every six years

would be averted by the reduction in exposure to 1 ppm.

(A. 176). The Secretary, however, rejected Dr. Wilson’s con-

clusion because: (1) Dr. Wilson conceded that the reliability

of the data on which he based his “guess” (Tr. 2772; A.92)

was “quite poor” (Tr. 2720) ; (2) Dr. Wilson gave no basis for

his conclusions at the hearing (Tr. 2743, 2790-2798), and,

although he agreed to supply this information to the Secretary

(Tr. 2752-2753), he never did so; (3) Dr. Wilson improperly

dismissed the Infante study as unreliable (Pet. App. 107a) ;

and (4) Dr. Wilson underestimated the number of workers

who are regularly exposed below 10 ppm benzene concentra-

tions. See Pet. App. 161a-162a.

Second, the court asserted that Dr. Lawrence Partridge, a

member of the staff of the Secretary’s economic consultant

(Arthur D. Little, Inc.), stated that his firm could “perform a

cost-effectiveness analysis for the benzene standard.” In fact,

the Secretary’s consultant stated only that his firm was gener-

ally qualified to perform cost-effectiveness studies. He did not

testify that sufficient data to perform such a study existed

Rate Cases, supra, 390 U.S. at 776 1.40; NAACP vy,

FPC, 425 U.S. 662, 672 (1976) (Powell, J., concur-

ring).

* * * * +

In sum, the Secretary’s position is that the toxic

substances provision of OSHA neither requires nor

permits him to establish exposure levels on the basis

of tradeoffs, identified by cost-benefit techniques, be-

tween health benefits and costs. Even if the statute

permits such a tradeoff, it does so only when the evi-

dence to make such a decision is “available,” which

is not the case for benzene. This is not to say that

costs and benefits play no role in the statutory

scheme. They clearly do. There must be substantial

evidence for concluding that a lower exposure level

for benzene (Tr. 512). To the contrary, his economic report

specifically stated the data did not exist with regard to ben-

zene (Iixh. 5A at 3-5).

Third, the court of appeals observed that, after the close

of the record in this case, the EPA had released a preliminary

report on the relationship between cancer and benzene. See

Preliminary Report on Population Risk to Ambient Ben-

zene Haeposureas (1978) (A. 14-29). The EPA, however,

emphasized that its estimates were devised by “very crude

methods” (A. 25). This report, moreover, was based on

the data collected by Dr. Infante, who stated that, in his

view, the exposure data were too imprecise to establish a

dose-response relationship (Tr. 874, 927, 928, 1050, 1056, 1129-

1130, 1155-1156; see also Dr. Goldstein, Exh. 75 at 2-8). The

Secretary agreed with Dr. Infante. That EPA may have

concluded that its risk estimate is usable for its own purposes

(such as confirming that the levels of benzene now found in

the ambient air pose a public health hazard) is no basis for

disturbing the Secretary’s judgment that the same evidence

is not useful for a quite different purpose. Cf, Universal Cam-

era Corp. V. NLRB, 340 U.S. 474, 476 (1951).

smn. tet Ren semi oll

will be beneficial; here there was such evidence, as

the court of appeals acknowledged (Pet. App. 18a).

Risk assessment can be employed to identify safe

levels of exposure for some substances (see note 55,

supra), And the cost of compliance can be considered

in at least five ways: (1) in setting regulatory pri-

orities (see note 47, supra); (2) in defining what

level of exposure can be achieved feasibly (see pages

56-59 and note 50, supra); (8) in. selecting the

method by which employers must achieve the lowest

feasible level of exposure; (4) in granting vari-

ances under which less expensive, but equally effec-

tive, methods of control can be substituted for those

specified by the Secretary (see note 7, supra); and

(5) in the course of enforcing the standards (see

note 44, supra). Our point is not that the statute

has deprived the Secretary of the flexibility necessary

for the establishment of intelligent) public health

policy. But it has forbidden the kind of tradeoff re-

quired by the court of appeals, and the court’s re-

quirement that the Secretary produce an “estimate

of expected benefits” (Pet. App. 20a), and determine

by a weighing of those benefits against the cost that

the regulation bears a “reasonable relationship” to

quantifiable benefits, is improper.

*? See South Terminal Corp, v. EPA, 504 F.2d 646, 655-656,

657, 676 n.33 (lat Cir. 1974): where Conyress has forbidden

the EPA to balance costs against benefits under a particular

statute, “[t]urning down necessary controls because they are

economically burdensome is impermissible; selecting the least

burdensome effective alternative simply makes good sense.”

C. Any Ambiguities In The Statute Should Be Resolved

In Favor Of Greater Safety

The very most that can be said for the court of

appeals’ approach is that the phrases “reasonably

necessary or appropriate to provide safe or healthful

employment” and “to the extent feasible” are argu-

ably ambiguous. The court of appeals resolved the

ambiguity by selecting a cost-benefit requirement .

that, in the court’s view, best served the public in-

terest. The court overlooked two principles that

should have dissuaded it from resolving the perceived

ambiguity as it did.

First, OSHA is remedial legislation dealing with

the public health and safety. Many cases hold that

ambiguities in such legislation should be resolved in

favor of safety rather than, as the court of appeals

held, in favor of lower costs for industry.”

Second, since OSHA was enacted, the Secretary

has construed the toxic substances provision of 29

U.S.C. 655(b) (5) as requiring the achievement of

the lowest achievable level of exposure where a safe

level cannot be determined. The Secretary has pro-

mulgated eight standards regulating exposure to

carcinogens and other toxic substances (see note

24, supra). In almost all of the eight, the Sec-

retary has explained that no safe level of exposure

58 See, e.g., E.l. duPont deNemours & Co. Vv. Train, 430

U.S. 112, 183 (1977) ; FPC v. Florida Power & Light Co., 404

U.S. 453, 469 (1972); United States v. Seaboard Air Line

R.R., 361 U.S. 78, 83 (1959) ; Voris v. Eikel, 346 U.S. 328, 333

(1953) ; Tiller v. Atlantic Coast Line R.R., 323 U.S. 574, 577

n.2 (1945) ; Lilly v. Grand Trunk Western R.R., 317 U.S. 481,

486 (1948).

te sate le nd Cea Re AD we

ie i eS AE it Ws nce hel

to such substances can be determined, concluded that in

such circumstances the standard must be set at the

lowest achievable level, and rejected arguments that

the statute requires the level of exposure to be set by

reference to cost-benefit criteria.” The agency’s expe-

rience over the course of a decade, the Secretary re-

cently said, “has demonstrated to us the futility of

applying traditional cost-benefit analysis and method-

ology to the workplace health context.” 43 Fed. Reg.

45762, 45790 (1978) (explanation of the acryloni-

trile standard).

The Secretary’s consistent and well-articulated con-

struction of the Act is entitled to substantial defer-

ence. EL’. I. duPont deNemours & Co. v. Collins, 432

U.S. 46, 53-54 (1977); Miller v. Youakim, No. 77-

742 (Feb. 22, 1979), slip op. 18 (and cases cited

therein). An agency enjoys substantial leeway when

it must interpret a new statute and fashion a solu-

tion on the basis of imperfect data. Permian Basin

Area Rate Cases, 390 U.S. 747, 777 (1968). Defer-

ence to the agency’s approach is all the more appro-

priate where, as here, the agency’s interpretation is

supported by the agency’s practical experience in

implementing the statute, see American Trucking

Associations v. Atchison, Topeka & Santa Fe Ry., 387

U.S. 397, 416 (1967) ; Skidmore v. Swift & Co., 323

® See especially 39 Fed. Reg. 35890, 35892 (1974) (vinyl

chloride) ; 41 Fed. Reg. 46742, 46751 (1976) (coke oven

emissions) ; 43 Fed. Reg. 27350, 27378-27389 (1978) (cotton

dust) ; 43 Fed. Reg. 52952, 52963 (1978) (lead dust) ; 43 Fed.

Reg. 45762, 45790 (1978) (acrylonitrile). See also 37 Fed.

Reg. 11318 (1972) (asbestos dust) ; 39 Fed. Reg. 3756, 3758

(1974) (14 selected carcinogens).

U.S. 184, 140 (1944), and the statute in question is

a complex one requiring scientific expertise, see

Train v. Natural Resources Defense Council, Inc.,

421 U.S. 60, 87 (1975), HM. 1. duPont deNemours &

Co. v. Train, 430 U.S, 112, 135 n.25 (1977). Cer-

tainly the Secretary’s construction is “sufficiently rea-

sonable to preclude the court of appeals from substi-

tuting its judgment for that of the Agency.” Train

v. Natural Resources Defense Council, Inc., supra,

421 U.S. at 87.

THE SECRETARY WAS NOT REQUIRED TO WAIT

FOR MORE EVIDENCE BEFORE RESTRICTING

DERMAL CONTACT WITH BENZENE-BEARING

LIQUIDS

The Secretary determined that dermal contact with

liquids containing benzene creates a risk of cancer,

and he accordingly ordered workers to wear protec-

tive clothing to avert that risk (see pages 25-28,

supra). The court of appeals held that the Secre-

tary’s decision that benzene is absorbed through the

skin is supported by substantial evidence (Pet. App.

25a). Despite the statutory command in 29 U.S.C.

655(f) that “[t]he determinations of the Secretary

shall be conclusive if supported by substantial evi-

dence” and despite the command in 29 U.S.C. 655 (b)

(5) that texie substances regulations shall be based

on the best “available” evidence, the court set aside

the dermal contact rules on the ground that other

evidence that might become available in the future

would throw more light on the subject.

—

tall

Dee wlhe a ide in’. Litt oe -

The court of appeals was wrong for several rea-

sons. First, in setting aside a standard despite the

presence of substantial evidence, the court disre-

garded 29 U.S.C. 655(f). Second, in considering the

prospect of a study not in the record, the court over-

looked the rule that “review of administrative deci-

sions is to be confined to ‘consideration of the deci-

sion of the agency * * * and of the evidence on which

it was based’.” PC vy, Transcontinental Gas Pipe

Line Corp., 423 U.S. 326, 831 (1976), quoting from

United States vy. Carlo Bianchi & Co., 3873 U.S. 709,

714-715 (1963). Third, the court transgressed 29

U.S.C. 655(b) (5), which limits decisions to the

“available” evidence.

The court of appeals’ instruction to the Secretary

to await the production of additional evidence could

he justified only if the phrase “best available evi-

dence” in Section 655(b)(5) means “the best evi-

dence capable of being produced.” But the legislative

history of the “best available evidence” rule, which

we discuss at pages 60-61, supra, shows that the

phrase was intended by Congress to mean “the best

evidence in existence,” rather than “the best evidence

‘capable of being produced.” Congress was aware of

the potential for paralysis that medical disputes could

introduce into the administrative process. There is

always more evidence just around the corner. Con-

gress intended the “best available evidence” approach

to free the Secretary from the grip of claims that he

should wait for just one more study, sure to be eon-

clusive. As the House Committee put the matter

(Leg. Hist. 848): “The Committee feels it is vital

that when the Secretary sets an occupational health

standard he do so on the basis of the best available

professional evidence; it is not intended that the Sec-

retary be paralyzed by debate surrounding diverse

medical opinions.” By insisting that the Secretary

wait for just one more study—which the court of

appeals thought would be definitive—the court of

appeals introduced the very paralysis that Congress

wanted to avoid.

The circumstances of this case well show the im-

portance of the “best available evidence” rule and

the difficulties caused by insisting on the accumula-

tion of more evidence. The court of appeals held

that the agency should have obtained and considered

the results of tests using radioactive benzene. The

court said that these tests would have provided con-

clusive answers in six to twelve weeks to the ques-

tion whether benzene is absorbed through the skin,

as well as several other allied questions (Pet. App.

25a-26a). The agency thus violated, in the court’s

view, the directive in Section 655(b) (5) that stand-

ards be based on “research, demonstrations, experi-

ments” and “the latest available scientific data in

the field.” The court did not deny that the record is

replete with “research,” “demonstrations,” and “ex-

periments” on the skin-absorption issue and included

the best scientific evidence available at the time of

the rulemaking proceeding.” The court held, how-

causes dermatitis, which, aside from its own health effects, in-

—

ee |

anes i

a ee ee

ORI AOD DLE ish St Os

‘

;

ever, that Section 655(b) (5) requires the Secretary

“to acquire” (Pet. App. 27a) a radioactive benzene

study before issuing any dermal prohibition and pro-

tective clothing requirement."' ,

creases the risk that benzene will be absorbed through the

damaged skin. The court dismissed the agency's theory that

dermal contact causes dermatitis as a “post hoc rationaliza-

tion” (Pet. App. 22a). This was hardly so; this risk was

specifically relied on by the agency (id. at 198a).

"' Respondent Rubber Manufacturers Association has con-

tended that the Secretary erred in refusing to reopen the

record and stay the standard for 180 days while Dr. Maibach

conducted the radioactive benzene tests. Although the court

below did not discuss the question, the Secretary did not err

in refusing the request. This Court has never reversed a re-

fusal by an agency to reopen an administrative record except

on a showing of abuse of discretion. See Bowman Trans-

portation, Inc. Vv. Arkansas-Best Freight System, Ine., 419

U.S. 281, 294-295 (1974). Here there was no abuse of dis-

cretion. The statute itself put industry on notice that toxic

substances are subject to regulation, and it was common

knowledge that benzene is toxic. The central question for in-

dustry, then, was not whether the Secretary would begin to

investigate exposure to benzene, but how soon he would do

so (given the substantial numbers of other toxic substances

subject to regulation). Industry could have begun the process

of research as soon as the statute was enacted; it did not need

to wait until the initial notice of rulemaking. And questions

concerning the extent of absorption of benzene into the body

through exposed surfaces were obvious candidates for. re-

search, The Secretary’s notice of rulemaking, in any event,

clearly advised industry that the proposed standard would

include a prohibition against dermal contact with liquid ben-

zene. All interested parties were invited to present their views

at the rulemaking hearing. Dr. Maibach appeared on behalf

of the American Petroleum Institute and presented his views.

Industrial users of liquid benzene also presented their views

(Exhs. 6-48, 6-59). Eventually the administrative record

The requirement that the standard be based on the

“latest,” “best,” ‘available’ evidence was clearly

satisfied in this case. There was no radioactive ben-

zene study in existence, although the industry had

ample time to prepare one. And, as things turned

out, there was no reason to wait.

Dr. Maibach’s preliminary report was supplied to

the agency for the first time in January 1979. It

supports the Seeretary’s decision because it shows

that benzene is absorbed througn the skin and thus

confirms the Secretary’s conclusion that dermal con-

tact with benzene poses a cancer risk. The attempt

to perform a radioactive benzene test also illustrates

the pitfalls in setting aside a standard based on the

speculation that evidence can be created readily. Dr.

was closed, and the Secretary concluded that skin absorption

of benzene posed a cancer risk and issued a standard that pro-

hibited dermal and eye contact with liquid benzene. The re-

quest to reopen came several months thereafter during a later

proceeding directed solely to the issue whether the standard

should be amended to exclude workplaces where the only

exposure resulted from certain liquids with trace amounts

of benzene. The request was not accompanied by the actual

evidence sought to be added but was, in reality, merely a

request that the standard be stayed for 180 days while Dr.

Maibach conducted the very test that the parties were free to

conduct and submit a year earlier during the rulemaking

- proceeding. The agency’s representative refused the request

on the ground that it would be inconsistent with the narrow

scope of the amendment proceeding (Amend. Tr. 20-21; A.

185-186). He explained that the agency wouid consider such

study once it was available. But the agency would not defer the

standard for 180 days to await a study whose “results could

not possibly be known at this time” (ibid.). This decision

was not an abuse of discretion.

Maibach testified that the radioactive benzene tests

not only would determine whether benzene is absorbed

through the skin but also would answer a number of

other issues, such as the effect of repeated applica-

tions on the skin, and the effect of temperature vari-

ations. The court of appeals was influenced by the

prospect that “conclusive” answers to so many ques-

tions were “readily” available in a short time through

a simple study (Pet. App. 26a). But the time proved

to be not-so-short. Dr. Maibach’s “preliminary” re-

sults were received by the Secretary one-and-a-half

years after Dr. Maibach testified that his study could

be completed in six to twelve weeks.” The results

proved to be not-too-conclusive.” Moreover, Dr. Mai-

*2 Dr. Maibach testified on August 3, 1977. In May 1978

the Secretary was advised by respondent Rubber Manufac-

turers Association that Dr. Maibach had begun his tests;

the Association similarly advised the court of appeals at

oral argument in June 1978 (Pet. App. 27a n.27). The pre-

liminary report of Dr. Maibach was submitted in January

1979 in the form of an appendix to the brief of the Rui»er

Manufacturers Association in opposition to the petitions for

certiorari. The preliminary report states that additional

studies will have to be conducted. Perhaps the additional

studies will be furnished as an appendix to another brief

during the coming months.

"* As to the rate of absorption, the Maibach study is subject

to serious question. The study was conducted in a manner that

underestimates the rate of absorption. First, the subjects were

exposed to only a single dose of benzene (Rubber Mem. in

Opp. App. 7a), whereas in real life some employees are ex-

posed repeatedly to liquids containing benzene. Dr. Maibach

testified in 1977 that the number of doses might well make an

important difference (Tr. 2467; A. 63). Second, much of the

dose applied in the Maibach study would have evaporated

bach’s study did not address any issue except to con-

firm the Secretary’s conclusion that benzene may be

absorbed through the skin and to suggest that the

rate of absorption depends on the amount applied.”

from the skin, thus reducing the amount of benzene available

for penetration. Dr. Maibach testified in 1977 that benzene

is very volatile and evaporates rapidly (Tr. 2468; A. 63).

The Maibach subjects were kept outdoors, which would have

accelerated evaporation. Third, in the Maibach study liquid

benzene was applied to the forearm of a rhesus monkey; the

forearms of rhesus monkeys and humans may not be compara-

ble for absorption purposes. Fourth, the rate of abscerption

through human skin varies greatly with the part of the body,

as Dr. Maibach testified (Tr. 2466, 2480; A. 62, 71). Studies

on the rate of absorption through the forearm thus may not

indicate the usual rate of absorption. Fifth, the rate of ab-

sorption usually increases with the temperature (Tr. 2479;

A. 70). The study does not state the temperature at which the

subjects were maintained. Sixth, the Maibach study, which

relies on a single application of benzene to the test subjects,

dees not take into account the fact that repeated contact with

benzene can blis‘*~ ‘r. skin and reduce its ability to prevent

penetration. In 1{ ~ Dr. Maibach testified that penetration

‘normally is greate. trough damaged skin than undamaged

skin (Tr. 2469; A. €.). Seventh, Dr. Maibach based his “pre-

liminary judgment” about the comparative amount of ben-

zene absorbed through the skin and inhaled through the lungs

on a computation provided by respondent Rubber Manufac-

turers Association (Rubber Mem. in Opp. App. 10a-lla).

Dr. Maibach stated that “[f]urther experiments are necessary

in order to draw conclusions as to the comparative amount

of benzene absorbed” (id. at 10a). The Maibach study itself

admits several of these limitations.

** Dr. Maibach concluded that not very much benzene is

absorbed through the skin. Even if the conclusion about the

rate of absorption were reliable (but see note 63, supra), it

would not cast doubt on the dermal contact prohibition. Be-

cause there is some absorption, because this adds to whatever

rs a

i hal are taba ae a

Phar bot when Riad” <

errant Mlb eudiisn 1 3cine

eM Pine nant a tchindon®

tN NN he iN eel ate sn Prt.

oP at Deo et ORY 6le te AE NSE? PT Boer Tt

Putten <x

Dr. Maibach proposed further experiments to tackle

the additional issues that the court of appeals thought

would be resolved in Dr. Maibach’s initial effort.

This train of events illustrates our point that in-

dustry often will be able to point to toxicological re-

search that allegedly will provide conclusive answers

to scientific uncertainties. It illustrates the mischief

that can be caused when health standards are post-

poned in the hope that one more study will settle the

matter. These considerations underscore Congress’

wisdom in requiring the Secretary to consider only

available evidence in discharging his responsibility

to protect the health of employees.” If the Secre-

tary were required to wait until the evidence had

been supplied—however long that might take—

the power to control the timing of new regulations

would be effectively transferred from the Secretary

to the affected parties, who would have every incen-

tive to delay.

Finally, the Secretary did not violate the re-

quirement that the standard be based on “research,

demonstrations, experiments, and such other informa-

is inhaled, and because there is no known safe level of ex-

posure, the Secretary was required to adopt the lowest feasible

level of exposure. It is unquestioned that the use of protective

clothing to prevent dermal or eye contact with benzene-

bearing liquids is feasible.

** Of course, if, after the Secretary has promulgated a

standard, new evidence comes to light casting doubt on the

wisdom of any standard, then any interested person may ask

the Secretary to make appropriate revisions in the standard.

29 U.S.C. 655(b) (1).

tion as may be appropriate.” The dermal prohibition

was based on research, demonstrations and experi-

ments. Moreover, this language permits the Secre-

tary to detefmine the extent to which it is “appro-

priate’ to acquire research, demonstrations, and

experiments or other information in setting stand-

ards.” This “discretion, entrusted as it is to the

agency Congress believed best capable of making the

expert judgment required, should be set aside by a

court only on the clearest abuse of discretion. No

such abuse exists here.”

“6 The Senate committee, which added this language, made it

clear that the phrase “as may be appropriate’”’ modifies not

only “other information” but also “research, experiments

fand] demonstrations’: “The committee intends that [the]

standards promulgated under section 6(b) shall represent

feasible requirements, which, where appropriate, shall be —

based on research, experiments, demonstrations, past experi-

ence, and the latest available scientific data.’”’ Leg. Hist. 147

(emphasis added).

*T Although the Secretary has no independent research cap-

ability, 29 U.S.C. 669(c) permits him to contract with private

organizations for “studies” to carry out his responsibilities

under the Act. The extent to which the Secretary contracts

for independent studies before issuing a standard is a de-

cision concerning “how best to allocate its resources,”’ which,

at most, is reviewable only for an abuse of discretion. Moog

Industries, Inc. v. FTC, 355 U.S. 411, 414 (1958) ; Wisconsin

Vv. FPC, 373 U.S. 294, 318-314 (1963). Cf. FPC v. Transcon-

tinental Gas Pipe Line Corp., supra, 423 U.S. at 333; Utah

Fuel Co. v. National Bituminous Coal Commission, 306 U.S.

56, 61 (1939) ; Endicott Johnson Corp. v. Perkins, 317 U.S.,

501, 506-507 (1943); CAB v. Hermann, 353 U.S. 322, 323

(1957) ; FCC v. Schreiber, 381 U.S. 279, 290 & n.17 (1965) ;

Vermont Yankee Nuclear Power Corp. v. Natural Resources

Defense Council, Inc., 435 U.S. 519, 524-525 (1978). NIOSH,

CONCLUSION

The judgment of the court of appeals should be

reversed, and the case should be remanded with in-

structions to dismiss the petitions for review.

Respectfully submitted.

WADE H. McCCREE, JR.

Solicitor General

FRANK H. EASTERBROOK

Deputy Solicitor General

WILLIAM ALSUP

Assistant to the Solicitor General

CARIN ANN CLAUSS

Solicitor of Labor

BENJAMIN W. MINTZ

Associate Solicitor

ALLEN H. FELDMAN

DENNIS K. KADE

DIANE E. BURKLEY

Attorneys

Department of Labor

May 1979

the only agency with independent research capability under

the Act (29 U.S.C. 669), concluded that benzene is readily

absorbed thro

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.