# Respondents Brief — American Textile Mfrs. Institute, Inc. v. Donovan

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385006_2223%3A07

## Record

- **Collection:** Supreme Court brief
- **Document type:** Respondents Brief
- **Published:** January 1, 1981
- **Citation:** 452 U.S. 490

## Text

gy ‘
Da
‘
4
B

Nos. 79-1429 and 79-1583

JAH 44 108!

In the Supreme Court of the United States PEN

OCTOBER TERM, 1980

AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC.,
ET AL., PETITIONERS

Vv.

Ray MARSHALL, SECRETARY OF LABOR,
UNITED STATES DEPARTMENT OF LABOR, ET AL.

NATIONAL COTTON COUNCIL OF AMERICA, PETITIONER
Vv.

RAY MARSHALL, SECRETARY OF LABOR,
UNITED STATES DEPARTMENT OF LABOR, ET AL.

ON WRIT OF CERTIORARI TO THE UNITED STATES COURT
OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT

BRIEF FOR THE FEDERAL RESPONDENT

WADE H. MCCREE, JR.
Solicitor General

KENNETH S. GELLER
Deputy Solicitor General

BARRY SULLIVAN

Assistant to the Solicitor General
CARIN A. CLAUSS

hes Department of Justice
Solicitor of Labor Washington, D.C. 20580
BENJAMIN W. MINTZ (202) 633-2217

Associate Solicitor

(Aen et

ALLEN H. FELDMAN
DENNIS K. KADE
DIANE E. BURKLEY
JOHN A. BRYSON
Attorneys

7 TSE aes

Department of Labor
Washington, D.C. 20210

Iu the Supreme Court of the United States

OCTOBER TERM, 1980

No. 79-1429

AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC.,
ET AL., PETITIONERS

’ Vv.

RAY MARSHALL, SECRETARY OF LABOR,
UNITED STATES DEPARTMENT OF LABOR, ET AL.

No. 79-1583
NATIONAL COTTON COUNCIL OF AMERICA, PETITIONER
v. |

RAY MARSHALL, SECRETARY OF LABOR,
UNITED STATES DEPARTMENT OF LABOR, ET AL.

ON WRIT OF CERTIORARI TO THE UNITED STATES COURT
OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT

BRIEF FOR THE FEDERAL RESPONDENT

QUESTIONS PRESENTED

1. Whether the occupational health standard for cot-
ton dust, which is based on the conclusion of the Secre-
tary of Labor that technology exists to meet the standard
and that the costs of compliance are well. within the
capability of the textile industry, is “feasible” within

(I)

af

the meaning of Section 6(b) (5) of the Occupational
Safety and Health Act of 1970, 29 U.S.C. 655(b) (5).

2. Whether Section 3(8) of the Act, 29 U.S.C. 652(8),
which provides that a health standard must be “reason-
ably necessary or appropriate to provide safe or health-
ful employment and places of employment,” requires the
Secretary to conduct a cost-benefit analysis and deter-
mine that the costs of a particular standard are “jus-
tified” by the expected benefits.

3. Whether the transfer and income protection pro-
vision of the cotton dust standard is a measure that is
“reasonably necessary or appropriate to provide safe or
healthful employment” within the meaning of Section
3(8) of the Act, 29 U.S.C. 652(8).

TABLE OF CONTENTS

OE NTL OR PG eR OE TE CET

IIT issih thinacsinssclicaanscoeiivtaubaanitioneedtanbalbeliaiblatiae ti

SIA SERPS eta RD OR RM TREN SST RS
Bis’ MORON assassin

B. Toxic effects of cotton dust -........0..0..0.20.0------.

C. The basis for the final standard ................

1. The Secretary’s health determinations and

dust control strategy -................ CALC TN

2. The feasibility of the standard ...................

a. Technological feasibility 2.0000.

b. Economic feasibility -...............0.0......

3. The medical transfer and income protec-

PN

D. The decision of the court of appeals ..............

ke REE ao ieee ena

Argument:

I. The cotton dust standard is reasonably necessary
and appropriate because it protects against a
significant risk of material health impairment...

II. The Secretary’s response to the significant health
threat posed by cotton dust is economically justi-
fied and required by the statute _......._. ‘aliases

A. A cost-benefit test is inconsistent with the
language and legislative history of Section
cL) SPRINTS cane sae Liaitibidiistalestins

B. The requirement of Section 3(8) that a
standard be “reasonably necessary or appro-
priate” does not embody a cost-benefit test..

(II)

13
14

)

37

39

52

IV
Argument—Continued Page

C. A cost-benefit test is not necessary to pre-
vent serious economic disruptions or serious
misallocation of scarce resources for com-

bating occupational disease .......................... 55
III. The Secretary correctly concluded that the cot-
ton dust standard is economically feasible ......... 63

IV. The medical transfer and income protection pro-
vision is within the Secretary’s authority to im-
pose requirements reasonably necessary and ap-
propriate to secure healthful employment .......... 65

EERE INA Ns ER Olas Stree te Me CERN SCE ETT GNI 70

TABLE OF AUTHORITIES

CITATIONS
Cases:

AFL-CIO v. Brennan, 5380 F.2d 109 ..............2..22..... 39, 51
A.L.A. Schechter Poultry Corp. v. United States,

i 62
American Frozen Food Institute v. Train, 539 F.2d

RENCE PER ERAT SLAP UN PS OEE PRO BOP SP USE OE RE 51
American Iron & Steel Institute v. EPA, 426 F.2d

BEE i adheainsiaickistie sido Dahan sachigiapictinedeipasaninins tibsbi ticiintgeeasais 51

American Iron & Steel Institute v. OSHA, 577 F.2d
825, cert. dismissed, No. 78-918 (Sept. 10,
SN, sossabbetc nab ncced quent aabicaiicencstheaieeaedsmuudoaiia 39, 51
American Petroleum Institute v. OSHA, 581 F.2d
493, aff’d on other grounds, Industrial Union De-
partment V. American Petroleum Institute, No.

TS IE GA Cher EE Tae 50
American Smelting & Refining Co. v. OSHRC, 501

| 2) | Epeaaee iii hang ciae aricieeinehniiats gecesi obsiaias noses 34

‘C.P.C. Int'l, Inc. v. Train, 540 F.2d 1829 ................ 51

Clark v. Uebersee Finanz-Korp, 332 U.S. 480........ 53
Cotton Warehouse Ass’n Vv. Marshall, No. 79-1789

SIA SUN TIE: coheed aciecaksenaiseickaccsitnn-cu cd doasibonaicbeih stic Sasi: 27

EPA V. National Crushed Stone Ass’n, No. 79-770
RMON OI MEINE” snsciccsehonesncileesstbcianebotnladtiiindesnaanacs 37, 40, 44, 51

FPC v. Hope Natural Gas Co., 320 U.S. 591 .......... 49

V

Cases—Continued Page
FTC v. Simplicity Pattern Co., 360 U.S. 55 ............ 47
Florida Peach Growers Ass’n v. Dep’t of Labor,

UU: BEE I sicipichis hice iieetsabremepe cals rs Pais a, 46
Golden State Bottling Co. v. NLRB, 414 U.S. 168.. 64
Great Atlantic & Pacific Tea Co. v. FTC, 440 U.S.

Oe igieianid akiosceieeebaliucnrenicaamonieabomiacatod 47
H & H Tire Co. v. Department of Transportation,

GED Fe Me sniiditdincictntdte whine ee 54-55
Harris v. McRae, No. 79-1268 (June 30, 1980)...... 38
Industrial Union Department v. American Petro-

leum Institute (Benzene), No. 78-911 (July 2,

UID! vastecsssaclvancten eared 27, 29, 30, 31, 34, 37, 38, 40, 45,

50, 53, 54, 56, 61, 63
Industrial Union Department v. Hodgson, 499 F.2d

er en OER aA Pee ee LL I 39, 49, 50, 63
Marshall v. Sun Petroleum Products Co., 622 F.2d

1176, cert. denied, No. 80-471 (Dec. 15, 1980)... 52

National Realty and Construction Co. v. OSHRC,
cl al msc ck. 3 6) Amn a 51
le Sale | LC) Et) Seance cen 40

National Traffic and Motor Vehicle Safety Act,
I I ne aioleschiwde 54
Be RN RID UD oocci oo cnccsccenseccceseromecnostcnsess 55

Occupational Safety and Health Act of 1970, 29
U.S.C. 651 et seq.:

See ay OP Gm, OR ec 46, 68
Section 2(b), 29 U.S.C. 651(b) ...............00....... 2
Section 2(b) (5), 29 U.S.C. 651(b) (5) .......... 68
Section 3(8), 29 U.S.C. 652(8) ...... 29, 50, 53, 66, 68
Section 5(a) (1), 29 U.S.C. 654(a) (1) .......... 46
Section 6({a), 29 U.S.C. 655 (a) .........00002...... 2,47, 52
Section 6(b), 29 U.S.C. 655 (b) ............00000..... 4,52

Section 6(b) (5), 29 U.S.C. 655 (b) (5) ....22, 26, 27,
31, 34, 39, 47, 48, 50, 53, 55, 58, 68

VII

Statutes and regulations—Continued Page
Section 6(b) (6) (A)-(B), 29 U.S.C. 655 (b)
Ue a a 44,47
Section 6(d), 29 U.S.C. 655(d) ...........00... 44,57
Section 6(f), 29 U.S.C. 655(f) _.......... 4, 24, 46, 47
Section 6(g), 29 U.S.C. 655(g) 0... 56
Secuon 10, 29 U.E.0. G68 ..............:................... 52
Walsh-Healey Act, 41 U.S.C. 35 et seq. _................ 2
MOR WRU nil hice okeGi cccesecedeccccke ccs 52
ee Ciera EN ne 3
29 C.F.R. 1910.1048 (f) (2) (v) ...-ccccccccccccceoeeoccoc---. 24, 66
Miscellaneous:
115 Cong. Rec. (1969) :
a ea Tr 48
Sy RS IEE ceric ce ete 48
Es SAN RRS EE CM NICE FRE NY Re 48

Doniger, Federal Regulation of Vinyl Chloride: A
Short Course in the Law and Policy of Toxic

Substances Control, Ecol. L. Q. 497 (1978) ...... 62
Exec. Order No. 12044, 43 Fed. Reg. 12661, 12663
OCS Sit nite ahr MDa EEO ASR) DON ER LO Nee 59
34 Fed. Reg. 7946, 7958 (1969) ............................ 2
41 Fed. Reg. (1976):
A tana ae a he el kB ee 60
SEE ihespestit aed Seatac, oe Set 14
Da OO a yeas 5
iT Silence a en ei ia ae 3, 30
De UNI seccneccn Se Ser se ie ve 9
De I Nisei alia arse ea a, 32
43 Fed. Reg. (1978) :
pp. 5987, 5944-5945, 5948 57
I TIPE Shes is a ee 20
ARETE NA RR eae OME, ae 3
Ec, Ge RPS RE TREE ERED Ces Eu De ae 7
SC CREE emg 4

Miscellaneous—Continued Page

Na SSPE ce oe SRT a 8
aI ona dhicalscibhanasaybeteniedsinipesaibeensateraens 2
ES CE a ane Re 4,7, 31, 32
Os a i Nias si cncindpenpehkatinnichinicnesepadinnmtens 6, 7, 8, 11, 31, 32
i MITTIN aciscincecdicitecndsisceasonstesticiecanaiocabenninbingedi 2
Be Nia tatdtiilnciinebdidipnibsslteniiandunanesnasiensanadoubiwosewthdsies 3, 30
EN EIR ae Se DEO 9
I iis ndaicdcecdethicteanesianasabbdswonchonciens 10
Cl aslenlinbapmatbioeie 33
i oda. colacannenennvcnondncincisnebeailovgnates 10, 30
eI. 8k scnctisnnsaiscrniputdsacdseoiteanensewoisaitte 30
PR EE tet telie li ctupniinisedecdimonsbaettaedinpdacneiovsetndoindss 10, 30
PWS ELST NS OAR RES PL 8, 10, 11, 36
p. 27360 ........ a ae Aaa oe ehh hae te 3,12
I i sani cnn inesp Rahiotbetnibbinusinenens 13
EE EERE TIRE See Sa enw acre rear a 18
a a aici es in ds cdaipeurnialeacians 3
a a 3
I TIE ih act acsiulinccecnicopnandncsentesiriniatabvar oboe 2
I PTT cicscretuitdeanvnrckensanseseavbacionsens 14, 16, 18, 65
Be el aia cad gs seaniua bnninaruasaaeen 16
Rae iia epielatatigeSiioeavacslepnapsteicenlag sheen 18
TS ital 18
EU TESTE ee ene SOA RIOR ae ee 17, 19, 64
a ldo 19
ARREST RSI ce ence PP 22, 23
sia de tas reac iipisicldaindinlen 22
aa hata cemitvinnsin dn atone 12, 36
Oe I air csguincnntsninianasnslabbonibaacsleaieele 28
Ms SEE SP AD ne SR md 23, 66, 67, 68
Ne REE RET RN CNN rs SR eT Re RE TC 23, 65
pp. 52988-52984, 58008 ...................... eee 57
i AG iid incllenacoiudinsimcieinctnhoveniiingaeessadietelbie 56
He TIE osctcasdccasocsiisinctnhsioaniecanseeanianpeonsudbs 57
Se ec snatced cae nianenensaihesananl 50
44 Fed. Reg. 58642, 58661 (1979) ............0000222--. 55
44 Fed. Reg. 58660-58661 (1979) ..........220222-22222... 55

45 Fed. Reg. (1980):

Miscellaneous—Continued Page
NRE IES REY PsA ete ee 56
I i 27

H.R. Rep. No. 92-1158, 92d Cong., 2d Sess. (1972) .. 54
H.R. Conf. Rep. No. 94-700, 94th Cong., 1st Sess.
Uc ied 40
Industrial Hygiene Field Operations Manual, U.S.
Department of Labor, Occupational Safety and
Health Administration (1980) _.......0... 57-58
Note, Cost-Benefit Analysis for Standards Regu-
lating Toxic Substances Under the Occupational
Safety and Health Act: American Petroleum
Institute v. OSHA, 60 B.U.L. Rev. 115 (1980). 61

S. 2788, 91st Cong., Ist Sess. (1969) 0. 48
S. Rep. No. 1301, 89th Cong., 2d Sess. (1966) _.... 55
S. Rep.No. 92-749, 92d Cong., 2d Sess. (1972) ...... 54
S. Rep. No. 93-132, 93d Cong., Ist Sess. (1973) ....... 44
S. Conf. Rep. No. 94-516, 94th Cong., 1st Sess.
REE ARSE IAS GS SI NEE DAN aA SIO oR OR ae OR 40

Senate Comm. on Governmental Affairs, 95th
Cong., 2d Sess., Study on Federal Regulation,
ven. ¥i.(Commm: Print 1678): 60
_Subcomm. on Labor of the Senate Comm. on Labor
and Public Welfare, 92d Cong., 1st Sess., Legis-
lative History of the Occupational Safety and
Health Act of 1970 (Comm. Print 1971) _........ 1
Webster’s Third New International Dictionary
So, Reet CRI AE eR Raw PML eS ZA ON Se a 39

OPINIONS BELOW

The opinion of the court of appeals (Pet. App. 19-
102)* is reported at 617 F.2d 636. The Secretary’s stand-
ard for occupational exposure to cotton dust, and the
statement of reasons supporting the standard, are pub-
lished at 43 Fed. Reg. 27350-27399 (1978), as amended
at 43 Fed. Reg. 28473-28474, 35032-35035, and 56893-
56894 (1978).

JURISDICTION

The judgment of the court of appeals was entered on
October 24, 1979 (Pet. App. 26), and petitions for re-
hearing were denied on January 11, 1980 (Pet. App.
103-105). The petition for a writ of certiorari in No.
79-1429 was filed on March 14, 1980, and the petition for
a writ of certiorari in No. 79-1583 was filed on April
9, 1980. On October 6, 1980, the Court granted the
petition in No. 79-1583 in full and granted the petition
in No. 79-1429 limited to questions 1, 2 and 4 (J.A. 489).
The jurisdiction of this Court is invoked under 28 U.S.C.
1254(1).

STATEMENT

A. Introduction

Byssinosis, which is commonly known as “brown lung”
disease, is a painful and permanently disabling respira-
tory disease that is induced by the inhalation of cotton

1“Pet. App.” refers to the appendix to the petition in No. 79-
1429. “J.A.” refers to the joint appendix filed in this Court. “App.”
refers to the joint appendix filed in the court of appeals. “Leg.
Hist.” refers 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). For convenience, we shall refer to the Secretary, the
Assistant Secretary, and the Occupational Safety and Health Ad-
ministration as “the Secretary” or “the agency.”

(

(1)

2

dust.* Most victims of byssinosis are textile workers,*
among whom the disease has been known to be wide-
spread since at least the beginning of the eighteenth
century (Pet. App. 29 & n.11). The prevalence of this
disease, which Congress found to exemplify the “magni-
tude of [our occupational health] problem” (Leg. Hist.
143), played a central role in the decision to enact the
Occupational Safety and Health Act of 1970, 29 U.S.C.
651 et seg., which aims “to assure so far as possible
every working man and woman in the Nation safe and
healthful working conditions and to preserve our human
resources” (Section 2(b) of the Act, 29 U.S.C. 651(b) ).

On May 29, 1971, the Secretary, as authorized by Sec-
tion 6(a) of the Act, 29 U.S.C. 655(a), adopted a pre-
existing federal standard as the occupational health
standard for cotton dust.t* The 1971 standard identified

2 Numerous investigators have established the relationship be-
tween cotton dust and the disease. See J.A. 15-22; Pet. App. 108-
124; App. 210-217, 218-228. Despite extensive research, however,
medical science has not yet identified the particular component of
cotton dust that causes byssinosis. 43 Fed. Reg. 27352-27355 (1978).
Medical science thus broadly defines “cotton dust” to include all
dust in the working environment of employees who work with
cotton, without regard to whether the origin of the dust can be
traced directly to the cotton plant, and the Secretary has adopted
that definition for regulatory purposes. 43 Fed. eg. 27354-27355
(1978).

3 The textile industry is divided into two basic operational seg-
ments: (1) yarn manufacturing, and (2) weaving of yarn into
cloth. Between 126,000 and 200,000 workers are exposed to cotton
dust each day ir yarn manufacturing operations consisting of
opening, picking, carding, drawing, roving, spinning, winding,
spooling, twisting, and warping. See 43 Fed. Reg. 27365, 27379
(1978). Many additional workers are exposed in the weaving
process (App. 517).

* Under Section 6(a) of the Act, 29 U.S.C. 655(a), the Secretary
was authorized during an initial two-year period to adopt any
existing federal health and safety standard or any national “con-
sensus” standard. In 1969, the Secretary had set an airborne dust
concentration limit as an occupational health requirement for public
contractors under the Walsh-Healey Act, 41 U.S.C. 35(e) (34 Fed.
Reg. 7946, 7953 (1969) ). That limitation incorporated the recom-

3

a maximum permissible exposure level (PEL) for air-
borne concentrations of cotton dust and required that
employers meet that level through the use of feasible
engineering controls, with respirators to be used only
where the PEL could not be achieved by engineering con-
trols. 29 C.F.R. 1910.1000(e)

On June 19, 1978, after extensive informal rulemak-
ing proceedings, the Secretary promulgated a more strin-
gent health standard under Section 6(b) of the Act, 29

mendation of the American Conference of Governmental Industrial
Hygienists (ACGIH) that exposure be limited to 1,000 micrograms
of total dust per cubic meter of air (1000 ug/m*) averaged over
an eight-hour workday. “Total dust” includes both respirable and
nenrespirable dust fractions. In 1978, when the Secretary set the
present standard, he followed the weight of current scientific opin-
ion by keying the maximum exposure level to the respirable dust
fraction (43 Fed. Reg. 27355 (1978) ). ACGIH had likewise adopted
that measuring device in 1974, when it issued a revised recommen-
dation that exposure should be limited to 200 ug/m*, the standard
that the Secretary eventually adopted here for 2'i operations except
slashing and weaving. See 43 Fed. Reg. 27351 (1978).

In all operations except slashing and weaving, the 1971 standard
—1000 ug/m? of total dust—is roughly equivalent to 500 ug/m? of
respirable dust, as measured by the 1978 standard. See J.A. 464-465;
App. 384; 43 Fed. Reg. 27355 (1978). Slashing and weaving are
exceptional because a substantial portion of the respirable dust
comes from starch that is used in those operations and is not harm-
ful. The Secretary set a higher permissible exposure level—750
ug/m*—in those operations. 43 Fed. Reg. 27360 (1978). See page
12, infra. See also 41 Fed. Reg. 56504 (1976).

5 “Engineering controls,” for purposes of the cotton dust stand-
ard, are mainly general and local ventilation systems designed to re-
move particulate matter from the air by using suction to capture
emission source dust, which is then conveyed to a filtration system
that screens out the particulates before recirculating the air into the
working area. General ventilation systems consist of filtration sys-
tems used in conjunction with central air conditioning equipment
(43 Fed. Reg. 27363 (1978)). Local exhaust ventilation requires
enclosure of machinery and the use of suction hoods that are placed
as close as possible to the emission source; the air is then filtered
and recirculated (id. at 27364). See J.A. 79-82.

4

U.S.C. 655(b).® The new standard sets stricter PELs
governing airborne concentrations of cotton dust in var-
ious manufacturing processes and requires a program of
prophylactic work practices, education, training, and
medical surveillance. The standard allows employers four
years within which to meet the PELs through engineer-
ing controls and requires the use of respirators in the
interim. Respirators will thereafter be required where
the PEL cannot feasibly be reached through engineer-
ing controls. The standard also provides that whenever
a physician determines that an employee is unable to
wear a respirator, the employee must be given the op-
portunity to transfer to another position having a dust
level at or below the PEL, if one is available, without
loss of pay.

Petitioners, among others, sought judicial review of
the standard under Section 6(f) of the Act, 29 U.S.C.
655(f). On October 24, 1979, the court of appeals up-
held the standard as applied to petitioners (Pet. App.
19-102).

B. Toxic Effects of Cotton Dust

Reliable estimates show that at least 35,000 active and
retired American textile workers are chronically disabled
as a result of disease caused by the inhalation of cotton
dust (J.A. 347; 48 Fed. Reg. 27353 (1978)).7 In addi-

6 Section 6(b) of the Act, 29 U.S.C. 655(b), establishes procedures
to be followed in promulgating permanent occupational safety and
health standards. Among other things, the Secretary must provide
for notice and comment and the opportunity for a hearing. In the
present case, comments were received from 263 parties, and 109
witnesses appeared during 14 days of hearings held in three cities.
The administrative record exceeds 105,000 pages (Pet. App. 35).

7Such chronic disease is manifested in severe and continuous
shortness of breath, chronic cough, and severe and permanent loss
of lung function. The afflicted employee must often choose prema-
ture retirement despite efforts to continue working (43 Fed. Reg.
27352-27353 (1978)). Excess heart strain and death through heart
failure may also result from severe breathing incapacity (J.A. 64-
73).

5

tion, dose-response curves drawn by Dr. James Mer-
chant, Director of the Appalachian Laboratory for Safety
and Health, and his colleagues, based on studies of var-
ious textile mills, show that one quarter of yarn manu-
facturing workers manifest at least acute symptoms of
disease at exposure levels roughly equivalent to those set
by the 1971 standard.* The dose-response data also show
that 13% of those workers will manifest such symptoms
even under the 1978 PEL. See note 8, supra. These
figures underestimate the true extent of the disease, how-
ever, because workers with severe acute symptoms will
often leave their jobs rather than face further disability,
and will not be counted in worker surveys (J.A. 42-49).

Detection of acute byssinosis depends largely upon
worker complaints of chest tightness and breathlessness.
A standard questionnaire and classification system® are
generally used. The worker’s condition is classified as
grade 4 byssinosis if the symptoms occur only occas-
ionally on the first day of the work week. If the symp-

8 Dr. Merchant’s dose-response curve showed that, in the yarn
manufacturing processes, a 3% prevalence of all grades of byssi-
nosis occurred at a dust level of 50 wg/m*. At levels of 100 ug/m!3,
200 ug/m*, and 500 ug/m*, the percentages rose to 7%, 18%, and
26%. In the remaining processes (slashing and weaving), no sig-
nificant prevalence was found at 200 uwg/m*; percentages of 5%
and 15% were found at levels of 500 ug/m*® and 1000 ug/m? (43
Fed. Reg. 27355 (1978) ; J.A. 44, 46). Dr. Merchant’s figures gen-
erally accord with those of other investigators. See J.A. 56-57,
61-63; App. 376-385. Comparing his work with that of other re-
searchers, Dr. Merchant stated that “it is remarkable once dust
concentrations and different processes are taken into account, even
with the different methods used, different countries, [and] slight
differences in questionnaire, how similar the prevalence levels are
from country to country and process to process. So I think there
is a good degree of correlation between the investigations” (J.A.
437). Dr. Arend Bouhuys, Director of the Yale University Lung
Research Center, made essentially the same observation (J.A. 404).

® The classification system was develoned by Dr. Schilling in his
study of the Lancashire cotton mills (41 Fed. Reg. 56500-56501
(1976)).

6

toms frequently occur on the first day of every work
week, the condition is classified as grade 1. If the symp-
toms occur on the first day of every work week and
regularly recur on other workdays, the condition is clas-
sified as grade 2.’° Byssinotics classified in grades %
through 2 are clearly unhealthy. They suffer from
breathlessness and chest tightness, and they find that
their normal activities must be curtailed at work and at
home. One expert, who had studied byssinosis for 20
years, testified that in addition to discomfort at work,
“in the evening [the worker] finds it difficult to undergo
normal activities, he finds it difficult to walk rapidly
home, and tends to do very little” (J.A. 484).

Not only does acute byssinosis cause substantial dis-
comfort and interfere with normal activities, but it also
causes permanent health damage. While petitioners re-
peatedly assert (see, e.g., ATMI Br. 4-5, 49-51) that
acute byssinosis is “reversible,” that statement is incor-
rect, or at least imprecise. Symptoms may well subside
if a worker is removed from the dusty environment, but
the damage that has been done cannot be undone. If
workers are again exposed, “[their] ability to combat the
[disease] is most probably less than when they were nor-
mal to begin with” (J.A. 422-423) and repeated exposure
results in a “cumulative damaging effect” (J.A. 25).
Thus, even grade 1% byssinosis is a “pathological state’’
that carries “significant health effects.” 43 Fed. Reg.
27354 (1978) .*

10 The condition is classified as grade 3 if grade 2 symptoms are
accompanied by evidence of permanent incapacity, diminished pul-
monary effort tolerance or reduced ventilatory capacity.

11 As Dr. Bouhuys stated (J.A. 152-153) :

The reversible physiological changes induced by cotton dust
exposure need to be prevented and should therefore be con-
sidered in establishing a standard. The reversible effects of
cotton dust exposure on the lungs of workers constitute a re-
current insult to their lungs which is unacceptable if only for
the fact that they are associated with symptoms of cough,
chest tightness, shortness of breath, and wheezing. In addi-

oe

7

Medical science has not yet discovered the precise man-
ner in which lower grade, acute byssinosis progresses to
the level of chronic, irreversible, and disabling disease
(grade 3 byssinosis).’* Indeed, the idea of an orderly
progression may not accurately describe the route of the
disease, as Dr. Russell Harley, an ATMI expert, testified
at the hearing (J.A. 192):

[T]he assumption is often made that the disorder
progresses from % to 1 to 2 to 3 and, thus, all
grades reflect the progress of the individual’s dis-
ability. In many instances, however, there is no
progression at all. Sometimes Grade 3 seems to ap-
pear de novo or there is a jump from 1 to 3. Among
those who develop permanent disability, Grade 2
very often never occurs.

It is beyond dispute, however, that both acute and
chronic stages are attributable to inhalation of cotton
dust (43 Fed. Reg. 27352 (1978)) and that repeated
acute insults lead to chronic impairment (43 Fed. Reg.
27354 (1978) ; J.A. 153-154, 411). This progression oc-
curs because “[u]ltimately repeated bronchial construc-

tion, it is likely although not definitely established that re-
current acute effects of dust exposure contribute to the insidious
development of chronic lung disease. * * * [S]Jince the acute
effects of cotton and hemp dusts on lung function are similar,
prudence indicates the assumption that temporary hypoxemia
may also result from acute dust exposures to cotton dust. There
can be little question that any exposure which may result in
temporary disturbances of the supply of oxygen to the tissues
of the human body needs to be prevented.

12 The Secretary pointed out in the final standard that the chronic
disease caused by cotton dust is not adequately described by the
term “grade 3 byssinosis” (48 Fed. Reg. 27353 (1978)). For in-
stance, some workers demonstrate excessive rates of chronic bron-
chitis that must be deemed attributable to their dusty environment
(ibid.). Accordingly, the “suggestion that the term chronic ob-
structive pulmonary disease (COPD) [should] constitute [the]
generally accepted clinical definition of chronic lung disease * * *
has merit” (ibid.).

8

tion leads to fixed airway obstruction” (J.A. 411). The
prevention of chronic respiratory disease thus necessarily
entails the prevention of acute disease (J.A. 411; 43 Fed.
Reg. 27354 (1978) ).*

C. The basis for the final standard

On December 28, 1976, the Secretary proposed a new
occupational health standard for cotton dust. The Sec-
retary proposed, on the basis of then available health
data and considerations of technological and economic
feasibility, that the PEL should be set at 200 ug/m* for
all textile operations. The PEL was to be attained ulti-
mately through engineering controls, but respirators
would be required in the interim. Because Dr. Mer-
chant’s data showed that 13% of the relevant population
would remain unprotected even at this level, however, the
proposed rule also mandated a medical surveillance pro-
gram, specific work practices to minimize exposure, and

18 The diagnostic techniques currently available for identifying
acute respiratory disease caused by cotton dust are not wholly satis-
factory because they depend to a large extent on the accuracy of
employee responses to the standard questionnaire (see pages 5-6,
supra), and it is not uncommon for workers to conceal their symp-
toms. It is therefore “possible that one could miss true grade 2
byssinotics with a medical program that is competent, is profes-
sional, well run * * *” (J.A. 489). See also 43 Fed. Reg. 27359
(1978). Acute disease, including grade 1%, may sometimes be de-
tected through objective measurements of lung function loss (pul-
monary function decrement). A spirometer can be used to measure
lung function loss at the beginning and end of the work day, with
a decline in the volume of air expired signifying an adverse reac-
tion to dust in the working environment. The measure most com-
monly used is the Forced Expiratory Volume in the first second of
a deep breath expiration (FEV 1). See 43 Fed. Reg. 27352-27354
(1978). However, some true byssinotics complain of chest tightness
and breathlessness but show no measurable deficit in pulmonary
function, while others show pulmonary function deficit without ex-
periencing breathlessness or chest tightness (J.A. 23, 24, 198, 435;
Pet. App. 119-120). There is thus no assurance that some employ-
ees are not true byssinotics despite negative results on both the
questionnaire and the pulmonary function test.

9

employee education and training (41 Fed. Reg. 56504-
56508 (1976) ).

During the extensive informal rulemaking proceedings
that followed, the textile industry conceded that a dust
exposure limit was needed, but contended that equally
effective protection could be provided at a dust level less
stringent than the proposed standard, through reliance
on respirators and medical transfers.'* The industry also
attacked the technological feasibility of the proposed PEL
as applied to textile processing operations from spinning
through weaving,” and it claimed that the standard as a
whole was not economically feasible or achievable with-
in the proposed seven-year implementation period. On
the other hand, employee representatives attacked the
standard as too lenient and advocated a 100 wg/m*® PEL
in all operations.

1. The Secretary’s Health Determinations and Dust
Control Strategy

The Secretary found overwhelming support in the rule-
making record for the proposition that textile workers
face a severe hazard of byssinosis in most operations both
at the proposed level of 200 ug/m’® and below. Even
at 200 wg/m*, workers in all operations except slashing

14 The industry advocated that the PEL be set at 500 ug/m* for
all textile operations except slashing and weaving; for those opera-
tions, the industry recommended an exposure level of 1000 ug/m?
(43 Fed. Reg. 27355-27360 (1978)). The industry claimed that
its data showed a very low prevalence of disease at 500 ug/m? in
yarn manufacturing, thus contradicting Dr. Merchant’s results.
The industry contended, therefore, that only a small number of
medical transfers would be necessary in connection with the in-
dustry’s proposed limit and that that number of transfers was
within the capability of the industry (ibid.).

16 The industry conceded that the Secretary’s proposed limit was
technologically feasible in the initial processes of opening through
roving.

10

and weaving face a 13% risk of byssinosis. At the in-
dustry’s proposed 500 wg/m® level, that risk is doubled.’

Based on the testimony of many medical witnesses
(e.g., J.A. 401-405, 411), the Secretary concluded that
the standard could not be effective unless it was aimed
at curbing all grades of byssinosis. “[B]yssinosis repre-
sents a continuum of disease,” and “the diminished state
of health of grade % byssinotics is a stage of the pro-

16 Dr. Merchant’s dose-response curve showed the existence of
risk at exposure levels approaching zero in yarn manufacturing
operations, and the record affirmatively demonstrated that no safe
level of exposure existed (43 Fed. Reg. 27358 (1978)). The in-
dustry relied upon certain other studies in attempting to show
that the prevalence of byssinosis was not as high as Dr. Merchant
suggested. The Secretary found the industry data deficient, how-
ever, because the reported prevalence of disease was not accom-
panied by adequate dust level data in many instances. For example,
the Cone Mills survey (Pet. App. 126-136) provided no dust level
data, and the Burlington and West Point Pepperell surveys gave
only “assigned” ranges (Pet. App. 161-163, 187). A dose-response
curve similar to Dr. Merchant’s could not be constructed on the
basis of such data. Moreover, the actual exposure levels in the
mills studied by the industry were generally much lower than the
500 wg/m% level that the industry advocated and often the worker
populations in such mills had been part of an on-going worker
surveillance program. That was the difficulty with respect to the
mills studied by Morgan and Weill. 43 Fed. Reg. 27356, 27358
(1978). The Secretary noted that “[t]he Weill study in particu-
lar demonstrates the difficulties inherent in analyzing epidemiologi-
cal data collected using a mixture of either very clean or very dusty
mills.” Jd. at 27358. Furthermore, the industry’s low prevalence
claim was based on data collected through a questionnaire, and the
validity of the data could depend on the manner in which the ques-
tionnaire was administered. Indeed, other investigators had noted
a rough equivalency between rates of prevalence identified by ques-
tionnaire and by pulmonary function testing (43 Fed. Reg. 27357
(1978)), whereas the industry here found a much higher rate
through pulmonary function testing, thereby casting further doubt
on the industry data (8%-11.8% vs. 1%-3%). Because of the
marked disparities between the industry’s data and those collected
by independent public health investigators, together with the defi-
ciencies in the industry’s methodology, the Secretary regarded the
industry’s data as having questionable validity (48 Fed. Reg. 27359
(1978) ).

11

gressive development with prolonged exposure to cotton
dust to an ultimately irreversible respiratory disease”
(43 Fed. Reg. 27354 (1978)). Because acute symptoms
are often disregarded by workers and may therefore go
undetected, the Secretary stated, it is necessary to protect
such workers before the condition becomes irreversible
(43 Fed. Reg. 27354 (1978)).

The Secretary concluded that the best way to reduce
the prevalence and risk of disease is tc reduce dust levels
to the extent feasible through engineering controls. The
Secretary’s conclusion was supported by substantial medi-
cal and scientific evidence suggesting that reduction of
dust levels should be the main objective of the standard
(Pet. App. 51-52 & n. 87).

The Secretary rejected the industry’s plan for less
stringent dust control combined with reliance on transfers
and respirators. First, the Secretary doubted the indus-
try’s assumption of low disease prevalence at the 500
ug/m* level (see note 16, supra). Second, Dr. Merchant’s
study suggested that a quarter of the relevant population
would be diseased at that level, while the textile industry
claimed that it could transfer no more than 10% of its
workforce (J.A. 445-446). It was evident, therefore,
that transfers could not play a sufficiently major role in
employee protection.’* Third, respirators were not accept-
able because of their documented unreliability, their in-

7 The National Institute for Occupational Safety and Health
(NIOSH) had recommended, based on a thorough review of the
medical literature, that the new standard impose the lowest feasible
limit on dust concentrations, but in any event no higher than 200
ug/m® (J.A. 175). In addition to NIOSH’s recommendation, Dr.
Merchant testified that he favored a reduction to 100 ug/m?. (J.A.
52). Dr. Hans Weill recommended that the level be set between
200 ug/m® and 500 ug/m? (J.A. 156).

18 Under the industry plan, 60% of the diseased population could
not be transferred. Indeed, the Secretary concluded that it was
only at the 200 ug/m® level that the number of necessary transfers
would be within the capability of the industry (48 Fed. Reg. 27359
(1978) ).

12

ability to protect employees with impaired breathing, and
their associated safety hazards (48 Fed. Reg. 27384
(1978) ).

The Secretary also considered and rejected more strin-
gent exposure limits. In the yarn manufacturing proc-
esses, the Secretary concluded that the employee repre-
sentatives’ proposal of 100 wg/m* was not within the
“technological capabilities of the industry.” 48 Fed. Reg.
27360 (1978). With respect to weaving operations, the
Secretary noted that “engineering dust controls in weav-
ing may not be feasible even with massive expenditures
by the industry.” Ibid. See also J.A. 247-248 (necessary
changes would have required reconstruction of 90% of
the weave rooms). In this regard, one study had esti-
mated that nearly 60% of compliance costs would be
concentrated in the weaving sector and that the impact
on the industry would thus be severe (J.A. 74, 125-130).
Rather than require these apparently infeasible expendi-
tures, the Secretary determined that optimal worker pro-
tection would be served by concentrating “resources on
achieving 200 ug/m* in yarn manufacturing as rapidly
as feasible * * *.” 43 Fed. Reg. 27360 (1978). Ac-
cordingly, the Secretary set the roughly equivalent levels
of 200 ug/m’ in yarn manufacturing and 750 ug/m? in
slashing and weaving (see page 3, note 4, supra).

2. The Feasibility of the Standard

The Secretary also carefully assessed the technological
feasibility and the economic consequences of the final
standard. Before publishing his proposal, the Secretary
had contracted with Research Triangle Institute (RTI) to
conduct a technological assessment and economic impact
analysis. During the rulemaking, interested parties sup-
plied much additional technical and economic informa-
tion. The Secretary concluded on the basis of these data
that the final PEL’s—750 ug/m* in weaving and 200
ug/m® in the remaining operations—were both techno-
logically achievable and well within the financial capa-
bilities of the industry.

13

a. Technological Feasibility

The industry conceded that it could achieve the 200
ug/m* level in the initial processes, opening through rov-
ing,” through conventional dust control techniques, in-
cluding local exhaust ventilation. The industry main-
tained that it was technologically impossible, however,
to achieve that level in the later operations, spinning
through weaving,” and it therefore challenged the RTI
study’s premise that local exhaust ventilation was suit-
able for those operations. In setting the final standard,
the Secretary generally agreed that local exhaust ventila-
tion has greater limitations in operations beyond roving,
but he did not believe that the lower levels could not be
achieved (43 Fed. Reg. 27362-27363 (1978)). Indeed,
record evidence showed that dust levels in operations be-
yond roving were already quite low, although the indus-
try had not devoted much effort to dust control in these
operations.”"

In addition, other effective techniques were also avail-
able to reduce dust levels. Processes are not generally
isolated in the mills, but something as simple as closing
the door between the weave room and other processing
areas could reduce levels by preventing the migration of
dust (J.A. 83-84; see also J.A. 18). Enhanced filtration
of air circulated by the general ventilation system would

19 In these operations, the raw cotton is cleaned and formed into
strands suitable for spinning.

70In “spinning,” the processed cotton is spun into yarn. In
“winding,” “spooling,” “twisting,” and “warping,” the spun yarn
is packaged for weaving or other processing. In “slashing” and
“weaving,” the yarn is treated with sizing and woven into fabric.

21 In spinning, for example, raw data supplied by ATMI showed
that the dust levels of 30 out of 46 (65%) spinning operations
were already below 200 ug/m*. That sample included operations
spinning 100% cotton, which is conceded to be the dustiest variety
(App. 3490-3560). In weaving operations, the ATMI data showed
that 87% of the operations were within the 750 ug/m? limit (ibid.),
while data collected by Hovan Hocutt, a ventilation expert, showed
that 72% of the weave rooms already complied with the final
standard (J.A. 247-248, 465).

14

likewise prevent recirculation of contaminated air (J.A.
11, 168-169, 414). Moreover, some witnesses recognized
that dust levels in later operations would be lowered if
the earlier processes were subject to the most efficacious
cleaning (J.A. 10-11, 287, 469). Because the industry’s
projections of achievable dust levels were based solely on
_ the difficulty of using local exhaust ventilation techniques,
the Secretary concluded that the established exposure
levels were, as a general matter, within the technological
reach of the industry. 43 Fed. Reg. 27366-27367 (1978) .

b. Economic Feasibility

In the rulemaking proceedings, the assessment of the
economic impact of the 200 uwg/m* level upon the industry
was dominated by consideration of the cost cf the engi-
neering controls necessary to achieve that level. At the
outset, the Secretary had contracted with RTI to study
the economic impact of alternative standards. That
process continued with the issuance of the proposed stand-
ard and the agency’s solicitation of views as to how the
expected costs might be financed. 41 Fed. Reg. 56498
(1976). The rulemaking proceedings produced volumi-
nous information. See page 4, note 6, supra. The
quantitative and qualitative information adduced in the
rulemaking permitted the Secretary not only to compare
the dollar estimates provided by the various parties, but
also to examine critically both the methodology and the
technological and economic assumptions upon which those
estimates were based. In this way, the Secretary was
able to make an independent feasibility determination.
In simplest terms, the Secretary concluded that RTI had
based its dollar estimate on the need for greater tech-
nology than was necessary and that the industry had
underestimated the efficacy of the technology it had costed.

The RTI study estimated that $2.4 billion would be
required to meet a uniform level of 200 ug/m® in all

*2 The issue of technological feasibility is not presented in this
case (see ATMI Br. 8 n.16).

15

textile operations; that estimate included $1.3 billion in
costs attributable to meeting the 200 ug/m* level in
weaving operations, which was not required by the final
standard (J.A. 74). The industry produced the Hocutt-
Thomas cost estimates, which were based on an ATMI
survey and methodology that the industry declined to dis-
close fully (J.A. 388-397, 465-468). The industry’s sub-
mission estimated that costs of $543 million would be
required to achieve 200 ug/m* in opening through roving,
500 wg/m*® in spinning through warping, and 1000
ug/m*® in weaving (J.A. 234-251; Pet. App. 155-156).
The industry submitted no cost estimates for lower levels
in operations beyond roving, because it claimed that fur-
ther reductions were impossible.

Other parties, including the Council on Wage and
Price Stability (J.A. 333-335), presented data and in-
formation showing that existing trends toward moderni-
zation and technology changes should be considered in
assessing the economic feasibility of a standard (J.A. 266-
286). Although the RTI and industry estimates were
both premised, for instance, on a compliance strategy
that relied exclusively on “retrofitting” or adaptation of
old equipment, the record showed that on-going mod-
ernization would necessarily result in significant cost
savings.”*

*3 For example, Dr. Parker Reist of the University of North
Carolina testified that new, high-speed carding machinery was
cleaner, more productive, and less costly per unit of production.
In addition, the elimination of picking operations through use of
high speed chute feed cards constituted a significantly less costly
alternative to retrofitting. Dr. Reist explained that high-speed
cards attaining speeds of 80-120 pounds of cotton per hour had
recently become available (J.A. 298-300). For purposes of dust
control, however, air flow requirements per card remained roughly
the same (id. at 300-301). Moreover, high speed cards can be used
to provide substantially increased capacity for the same cost of
air flow. Alternatively, the number of cards can be reduced, thereby
maintaining the same total hourly capacity at lower air costs (J.A.
293-294, 300-301, 477-480). One mill found that the expense of
retrofitting 139 low-speed cards would cover the purchase of 32

16

RTI’s estimate was overstated in three other significant
respects. First, RTI derived a unit cost per machine for
applying local exhaust ventilation to achieve the 200
ug/m® level (J.A. 88-95). The Secretary had concluded,
however, that this costly technique would not be neces-
sary for some operations. In spinning through warping,
for instance, the Secretary found that many mills had
already met the 200 ug/m* level, and that necessary re-
ductions would occur because of lower dust levels in
prior operations (without additional cost) and through
isolation of processes and enhancement of the broken
ends collection system on spinning frames.% See pages
13-14, supra. Thus, RTI’s estimate that $493 million
would be required to achieve the 200 wg/m® level in spin-
ning operations was wholly unrealistic. 48 Fed. Reg.
27366-27367 (1978). Second, RTI’s estimate assumed
that all existing machinery should be taken into account,

high-speed cards with increased capacity and lower operating costs
(J.A. 304-306).

Greater dust control will also result when a mill chooses, for
production reasons, to replace traditional ring-spinning frames
with open-end spinning frames. A traditional ring-spinning frame
is difficult to control, whereas an open-end frame has built-in
controls and requires only the filtration system (J.A. 470-471).
An open-end frame also has the dust-control advantage of eliminat-
ing both the preceding roving operation and the subsequent winding
operations. One mill found that a decreased unit cost of production
resulted from this conversion (J.A. 476-477). Although open-end
spinning is not suitable for all operations, it was expected that
this new technology will grow from 5% to 30% of the spindles in
the industry (J.A. 459).

24The Hocutt-Thomas analysis estimated that enhancement of
the broken-ends collection system would cost $17.7 million (J.A.
251, 486; Pet. App. 155-156). The Secretary recognized that local
exhaust ventilation was not as applicable in weaving operations
as in other operations, but RTI’s $1.2 billion estimate for achieving
200 ug/m* in weaving depended on that technique (43 Fed. Reg.
27367 (1978); J.A. 91-95). Similarly, the Secretary found local
exhaust ventilation largely unnecessary in spinning, winding,
spooling, twisting and warping, but RTI’s estimate included $620
million for such controls (43 Fed. Reg. 27866-27367 (1978) ).

17

regardless of whether the machinery was being used to
process pure cotton, pure synthetics, or a blend. Machines
that are isolated and process pure synthetics would not
need to be controlled, however, because the ambient dust
would not contain cotton. Moreover, blends produce sub-
stantially less dust because “synthetic fibers are virtually
trashfree and contribute little’ (J.A. 9) to the ambient
air concentration (J.A. 9, 30, 50, 90-91, 133, 136-137).
RTI admitted that this incorrect assumption could have
increased its overall compliance cost estimate by 30%
(J.A. 90-91, 133, 186-187). Third, RTI assumed that the
entire industry stood on the same footing, that no prior
dust control expenditures had been made, and that dust
levels were at uniform levels throughout the industry
(J.A. 85-90).2> However, the evidence adduced at the
hearings disclosed a range of dust levels within the in-
dustry, as well as significant pre-existing compliance with
the levels prescribed in the final standard.

In addition to the RTI estimate, the Secretary had one
other comprehensive cost estimate: the Hocutt-Thomas
calculations that were based on the ATMI data. Based
on his comparison of the two estimates, the Secretary
concluded that “[s]ince ATMI estimates are based on the
most recent data, OSHA believes them to be more realistic
than those offered by * * * RTI” (48 Fed. Reg. 27373
(1978) ). While RTI had been unable to assess the extent
of current dust controls, Hocutt had the benefit of such
information reported in ATMI’s survey (J.A. 236-238).
The Hocutt-Thomas estimates also provided a reasonable
reference in cases where the record showed that RTI’s

25 While RTI had access to some data on actual dust levels and
industry efforts to implement dust control, RTI was not able to
survey the industry to determine the extent to which its projected
costs would actually have to be incurred (J.A. 418-420). This
limitation also resulted in overstatements with respect to estimates
for nonengineering costs because the record showed, for example,
that medical surveillance programs had been established in some
mills, thereby obviating the need for incurring initial costs in that
area.

18

control cost estimates were overstated. In weaving, for
example, the Secretary noted that raising the PEL to 750
ug/m*® meant that “the need for certain engineering
controls and/or amounts of air and filtration would be
changed [and] costs [would] be reduced significantly”
(43 Fed. Reg. 27371 (1978)). Because the Secretary
believed that better dust control would result from the
expense identified by Hocutt-Thomas, he “adopted the
Hocutt-Thomas estimate for slashing and weaving which
predicted a level of 1000 ug/m* instead of the RTI esti-
mate for slashing and weaving at 200 ug/m*”’ (ibid.).

Although the Hocutt-Thomas cost estimate was prem-
ised in part upon PELs that were higher than those
mandated by the final standard, the Secretary concluded
that this estimate also was overstated in several im-
portant respects. The Hocutt-Thomas estimate, like that
of RTI, was based on retrofitting costs, and it thus over-
stated actual costs to the extent that modernization and
technological innovation would otherwise occur (43 Fed.
Reg. 27372 (1978). Similarly, the Hocutt-Thomas esti-
mate did not account for the already low levels in many
operations or the fact that more stringent controls in
earlier operations would, without additional cost, re-
sult in less dust in later operations (see pages 13-14,
supra).** Finally, the Hocutt-Thomas calculations may

26 Thus, Hocutt estimated that $41 million would be required
to meet a 500 wg/m? level in winding, spooling, twisting and warp-
ing, whereas the Secretary believed that, given the already low
levels in: these operations, further reductions could be effected
through maximum cleaning in prior operations, isolation of
processes, and enhancement of general ventilation. 43 Fed. Reg.
27362-27364, 27366-27367 (1978). Hocutt also noted that dust
capture systems on winders could handle 20,000 cubic feet of air per
minute, although he costed only the price of half that volume to
achieve 500 ug/m*. He also pointed out several spooler modifica-
tions that would improve performance beyond the 500 ug/m? that
he had calculated. For twisters, Hocutt provided no estimate be-
cause he had concluded that these machines could not be controlled
with local exhaust ventilation. The cost with respect to warpers
likewise was not estimated because Hocutt stated that dust levels

19

also have included costs for machines processing synthetic
material. These overstatements tended to offset any dif-
ferences in cost attributable to the fact that the Hocutt-
Thomas estimates were premised in part upon higher
dust levels. The Secretary was unable to adjust the
Hocutt-Thomas estimates further, however, for several
reasons, including the fact that the industry refused to
disclose much of its supporting data aad methodology
(43 Fed. Reg. 27373 (1978) ). Nonetheless, based on the
record, the Secretary believed that his adjusted estimate,
amounting to $550 million in capital costs, was suff-
ciently reliable to assess the economic impact of the
standard (ibid.).

The Secretary concluded that the industry as a whole
could well afford costs in the $550 million range, to be
incurred over a four-year period. This conclusion was
based largely on RTI’s extensive analysis of the impact
of various cost levels on the industry’s structure, its
rates of return, and its profits, in light of its competitive
position with respect to synthetics and imports (J.A. 95-
150).” RTI also reported that the industry’s capacity
was currently under-utilized and that the capital required
for compliance costs thus would not displace other capi-
tal needs (43 Fed. Reg. 27377 (1978) ).

RTI examined capital costs per dollar of sales and
found these costs to be small in the yarn manufacturing
sector for both RTI’s $211.8 million (500 ug/m*) and
$984.4 million (200 wg/m*) capital requirements esti-
mates. The increase in costs amounted, respectively, to
only 1.7 and 7.8 cents per dollar of sales (J.A. 96).
Because the capital costs of the final standard for this

would be 500 ug/m3, but he did note that these machines were
amenable to some controls (J.A. 245-250).

27 While the Secretary disagreed with RTI’s cost estimates be-
cause of certain incorrect assumptions about the current state of
dust control in the industry, as well as strategies necessary to
achieve further dust control, RTI’s analysis of the industry’s
ability to handle various levels of costs was based on unrelated
data that were not subject to the same criticism.

20

sector were estimated at $467 million, it was clear that
increased costs per dollar of sales would be small. RTI
also calculated the price increases that would be re-
quired, given various compliance costs, to maintain the
previous rate of return on investment. For the $211
million in capital costs that RTI estimated for a 500
ug/m* dust level, necessary price increases ranged from
less than 0.1 cent per dollar of sales to 1.6 cents per
dollar of sales. Given the $984 million costs RTI antici-
pated for a 200 wg/m* dust level, the necessary price
increases ranged from 0.5 cents to 6.8 cents per dollar
of sales (J.A. 97-102) .*8

In addition, RTI examined the competitive posture of
cotton and the textile industry. RTI judged competition
from synthetics not to be significant. Synthetics had pen-
etrated the market to its limits, and there remained an
irreducible demand for cotton products. Moreover, the
price differential was unlikely to remain so heavily in
favor of synthetics in view of the escalating costs of
petroleum (43 Fed. Reg. 27350, 27370 (1978) ) .28

RTI also noted that import restrictions were in force
under a number of textile trade agreements and that the
annual allowable rate of growth of imports was set in the
most recent agreement at the 6%-7% range. Moreover,
international monetary changes since 1971 had resulted
in sharply increased levels of American textile exports

28 In weaving, the final standard was calculated to increase costs
by only slightly more than 0.13 cents per dollar of sales (J.A. 124).

28a RTI noted that synthetics had gradually achieved a 70% share
of domestic fiber consumption, but that factors other than the
price differential between cotton and synthetic fiber would govern
any further shifts. It is difficult and costly for a mill to change
the ratio of cotton to synthetics in its processing machinery. More-
over, fluctuations in consumer demand for major cotton products—
men’s wear and home furnishings (sheets and towels) —is relatively
independent of price (J.A. 140-146). Arthur Figh, a leading tex-
tile financial expert and vice-president of the Chase-Manhattan
Bank, agreed that synthetics were unlikely to displace cotton any
further (J.A. 474-475).

21

(J.A. 146-150). RTI concluded that small changes in the
relative prices of domestic and foreign textile products
would have “very little impact” on the domestic textile
industry, “and, in general, the trade agreements dampen
the effects on domestic markets of small changes in rela-
tive prices” (J.A. 150).

In charting the maximum potential inflationary im-
pact, as well as the long-run economic consequences on
employment levels and cotton products demand, RTI as-
sumed the occurrence of the necessary price increases that
it had calculated. RTI found that these long-term effects
would be minimal even for costs as high as $984 million.”
Thus, as ATMI contends (ATMI Br. 27), RTI did per-
form analyses that assumed a full pass-through of costs,
but, significantly, RTI concluded that the long-term detri-
mental effects of such price increases were not significant.
In view of this conclusion and of RTI’s determination that
interfiber and international competition were not con-
stricting factors, RTI’s judgment that prices would rise
over time to generate revenues that would cover costs
(J.A. 172-173) was fully warranted.

By comparing ratios of costs to annual profits, RTI also
analyzed the short term effect of compliance costs with
respect to financing. These ratios indicated both the in-
ternal availability of funds and the industry’s ability to

?° Using a price elasticity analysis, RTI calculated that price in-
creases due to $984.4 million in compliance costs would decrease
demand by only 3%-4% and that compliance costs of $211 million
would lead to a demand decrease less than 1% (J.A. 113-124). Thus,
only a small decrease in demand would result from $467 million in
compliance costs, and the impact on cotton growers would be mini-
mal. National Cotton Council’s contrary assertion (Br. 20-22) is
based on the grossly inaccurate assumption that compliance costs
would exceed $2 billion. Moreover, National Cotton Council’s con-
tention (Br. 16) that RTI erroneously used short-term elasticities is
insubstantial. RTI explained that long-term elasticities were un-
available (J.A. 114-115). In addition, RTI’s judgment that com-
petition from synthetics and imports was not significant meant that
the long-term effects of price increases would not differ substantially
from those calculated from short-term elasticities.

22

borrow from external sources. Although RTI assumed
for these calculations that prices would rise to maintain
pre-standard profits, this assumption was supported by
studies showing relatively small short-run demand price
elasticities, indicating a capacity for passing on cost in-
creases through price increases (J.A. 113-120).* These
ratios were substantial for the nearly $2.4 billion in capi-
_tal costs that RTI had estimated for a 200 uwg/m! level,
especially in the weaving sector, but they were sig-
nificantly smaller for the $221 million estimated for a 500
ug/m* level. The cost effects were an element in the Sec-
retary’s feasibility determination that led to the decision
to set a higher limit in weaving operations (J.A. 125-
130). As the Secretary explained, the estimated costs of
the final standard—$550 million—fell between those fig-
ures, and the unfavorable ratios identified by RTI ac-
cordingly would decline dramatically (48 Fed. Reg. 27378
(1978) ).

On the basis of these calculations, the Secretary con-
cluded that the $550 million compliance costs required
by the final standard were within the industry’s financial
ability, given a four-year period for implementation of
engineering controls, and would have an adverse impact
on only a few marginal firms (48 Fed. Reg. 27378
(1978) ). The Secretary also concluded that the standard
meets the statutory criteria set forth in Section 6(b) (5)
of the Act, 29 U.S.C. 655(b) (5), because the standard’s
requirements are necessary to provide the required pro-
tection and are economically and technologically feasible
(43 Fed. Reg. 27378-27379 (1978)). The Secretary re-
jected the contention that he was required to perform a
cost-benefit analysis after he had determined that the

8° RTI also determined that, given the extreme assumption that
no price increases could occur, all sectors of the yarn manufac-
turing industry would remain profitable, on the average, for the
low costs of the 500 ug/m? level, and only one sector would ex-
perience negative return for the large costs of the 200 ug/m? level
(J.A. 97-99).

23

new standard would yield appreciable benefits ** and that
it was feasible. The Secretary noted that Congress had
performed a categorical “cost-benefit” analysis in this
area, striking the balance to provide for the maximum
feasible protection against significant risks of material
health impairment. Because the Secretary concluded that
the standard fulfilled Congress’ mandate, he had no au-
thority to weigh necessarily speculative costs and benefits
to promulgate a different standard (43 Fed. Reg. 27378
(1978) ).

3. The Medical Transfer and Income Protection
Provision

Respirators will be widely used for a substantial in-
terim period under the final standard. Consequently, the
Secretary addressed the difficulties involved in the use of
respirators and established protections for employees
who are unable to use them (43 Fed. Reg. 27386-27387
(1978) ; J.A. 164-166, 217-218, 312-315, 316-319, 320-
321).*° Whether an employee can wear a respirator is
a determination that must be made by a physician, but
experience with past medical screening programs has
shown that many workers deny or minimize work-related
health problems because they fear being discharged or
transferred to lower paying jobs (J.A. 210-212, 218-220,
323-324, 435-436). Thus, the Secretary recognized that
accurate information could be obtained and health ade-
quately protected only if employees understood that they
would not be adversely affected because of their candor
(43 Fed. Reg. 27387 (1978)). Moreover, the Secretary

31 A qualified calculation based on Dr. Merchant’s dose-response
data showed that the new standard would prevent nearly 5,000 new
cases of byssinosis each year (43 Fed. Reg. 27379 (1978) ).

32 While it was generally agreed (J.A. 314-315, 323-324, 383-384)
that such persons should be transferred, the industry contended that
transfers should be limited to workers who manifest grade 3 symp-
toms, that transfers should be voluntary, and that the workers’
pre-transfer pay and benefits should not be guaranteed (J.A. 383-
384, 446-447).

24

recognized that income protection is necessary to fulfill
Congress’ intent to place safety costs upon employers,
rather than employees (ibid.) Finally, the Secretary
determined that transfers only of grade 3 byssinotics—
that is, those workers who had already contracted chronic
disease—would not satisfy the prophylactic purposes of
the Act. Accordingly, the Secretary provided in the
standard that when a physician determines that an em-
ployee cannot wear a respirator, the employee must be
allowed to transfer to another position having a dust
level at or below the PEL, if one is available, with no
redttion in wages or employment rights. 29 C.F.R.
1910.1043 (f) (2) (v) (Pet. App. 8-9).

D. The Decision of the Court of Appeals

On petition for review under 29 U.S.C. 655(f), the
court of appeals upheld the cotton dust standard as ap-
plied to petitioners. The court observed that the ap-
propriate standard of review is whether substantial evi-
dence in the record as a whole supports the Secretary’s
determinations and that its role is “to ensure that the
agency has (1) acted within the scope of its authority;
(2) followed the procedures required by statute and by
its own regulations; (3) explicated the bases for its de-
cision; [and] (4) adduced substantial evidence in the
record to support its determinations” (Pet. App. 43;
footnotes omitted).

The court of appeals extensively reviewed the record
in light of petitioners’ contentions. The court concluded
that the Secretary had justifiably relied upon the Mer-
chant dose-response data and that the Secretary’s deter-
mination that the standard should seek to eliminate acute
byssinosis, as a means of eliminating the chronic dis-
ease, was based on the law and the evidence (Pet. App.
46-53).

Although the industry claimed that the Secretary had
ignored its proposed alternative, the court of appeals
found that the Secretary had carefully considered that
plan and had correctly concluded, on the basis of sub-

25

stantial evidence, that the industry’s studies were deficient
(Pet. App. 47-49). The court noted in addition that the
Secretary, in disfavoring respirators, had acted on the
basis of uncontradicted evidence of health risk and pur-
suant to the express statutory policy of placing the bur-
den of compliance on employers (id. at 49-50). Likewise,
the Secretary had properly rejected the unsupported as-
sumption that sufficient jobs existed to permit large
seale reliance on a transfer program (id. at 49). The
court also found extensive evidence showing the neces-
sity for principal reliance upon a stringent dust control
strategy (id. at 51-52).

The court of appeals found ample proof of technological
feasibility, based on technical data and dust readings
submitted by the industry, RTI, and NIOSH, all of which
showed that certain firms within the textile industry had
already substantially complied with the requirements of
the standard (Pet. App. 55-61). With respect to economic
feasibility, she court concluded that the Secretary had
“marshal [led] sufficient support and analysis for [his] de-
cision to use the industry’s capital cost estimate” (id. at
65). In the court’s view, the evidence supported the con-
clusion that RTI’s cost estimates were overstated to the
extent that they were based on the retrofitting of old
machines, without taking account of new machinery that
would be installed in the ordinary course (id. at 63).
Moreover, the Secretary’s assessment that the costs es-
timated by the industry would result in greater dust re-
duction than the industry had assumed also justified the
conclusion that “even the industry capital estimates were
exaggerated * * *” (id. at 65). Given the unavailability
of more precise data and estimates, due in part to the
industry’s refusal to cooperate, the court found that the
Secretary had reasonably relied on the industry’s cost
estimate, as adjusted (id. at 65-66).

The court of appeals further concluded that the Secre-
tary had properly relied on evidence showing that in-
dustry revenues would rise to cover compliance costs

26

(Pet. App. 67-69). While recognizing that the standard
might result in the closing of some marginal firms, the
court noted that the feasibility of a standard must be de-
termined by reference to the industry as a whole. Thus,
the court found substantial evidence to support the Secre-
tary’s determination that the standard would not affect
the viability of the textile industry.

The court of appeals also squarely rejected the conten-
tion that the Secretary was required to perform a cost-
benefit analysis to show either that the standard was
feasible or that it was “reasonably necessary or appro-
priate” (Pet. App. 69-77). The court noted that Congress
had already struck the balance between cosis and bene-
fits in Section 6(b) (5) of the Act, 29 U.S.C. 655 (b) (5),
and that, in contrast to other statutes, neither the lan-
guage nor the legislative history of the Act contained any
suggestion of a cost-benefit test. Congress had clearly
rejected attacks premised on the ground that the Act was
too expensive and had directed “that any severe risk to
employee health must be eliminated or reduced if feasible
means to do so exist” (Pet. App. 74) (emphasis sup-
plied). Thus, the court was not free to impose the ad-
ditional requirement of a cost-benefit test.

Finally, the court of appeals upheld the Secretary’s
authority to assure the successful implementation of the
respirator and medical surveillance program by requiring
income protection for transferred employees. The court
found that this provision was “reasonably necessary or
appropriate to provide safe and healthful employment”
because, absent the provision, employees who are unable
to wear respirators “would be forced to suffer continuing
exposure to impermissibly high dust levels or else risk
disadvantageous transfers, or even risk losing their jobs”
(Pet. App. 97). This cost, like other costs of health pro-
tection, was appropriately assigned to the employer (id.
at 97-98) .*

83 The court of appeals remanded the standard, as applied to the
cottonseed processing industry, for a reassessment of its economic
impact on those employers (Pet. App. 86-93). Although the court

27

SUMMARY OF ARGUMENT

The Secretary set the present cotton dust standard,
which was upheld in relevant part by the court of ap-
peals, based on his determination that the prior stand-
ard presented a significant risk of material health
impairment that the new standard would substantially
reduce. The cotton dust standard therefore meets the
threshold requirement identified in Industrial Union De-
partment v. American Petroleum Institute (“Benzene”),
No. 78-911 (July 2, 1980). In accordance with the
statutory mandate that “the standard [be set] 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 [he] has regular exposure to the hazard
dealt with by such standard for the period of his working
life” (Section 6(b) (5) of the Act, 29 U.S.C. 655 (b) (5)),
the Secretary considered voluminous technological and
economic evidence and concluded that the requirements of
the standard are feasible. The Secretary also considered
alternatives submitted by the industry and employee
representatives. The industry alternative was rejected
because it did not provide the degree of protection re-
quired by the statute, and the employee representatives’
alternative was rejected because it was not feasible. The
Secretary’s determinations were based on substantial evi-
dence, as the court of appeals correctly held, and the judg-
ment below should be affirmed.

upheld the standard as applied to cotton classers and warehouses,
the Secretary has since suspended application of the standard to
these employers pending administrative reconsideration of the
health risk in light of this Court’s decision in Industrial Union
Department v. American Petroleum Institute, No. 78-911 (July
2, 1980). See 45 Fed. Reg. 50328-50329 (July 29, 1980). Accord-
ingly, the Secretary acquiesced in a writ of certiorari to vacate
the judgment as to these employers and to remand the standard
to the agency. See Cotton Warehouse Ass'n Vv. Marshall, No. 79-
1789 (Oct. 6, 1980).

28

Petitioners challenge the standard on three grounds,
all of which lack merit. First, petitioners contend that
the Secretary must not only meet the Benzene test and
then determine that the standard is economically and tech-
nologically capable of achievement, but that he must also
conduct a cost-benefit analysis and determine, with re-
spect to each standard, that the anticipated costs bear a
“reasonable relationship” to the anticipated benefits. The
language and scheme of the Act provide no basis for
that contention, and the extensive legislative history of the
Act contains not a single suggestion that the Secretary
must weigh costs and benefits in setting standards. More-
over, because Congress did not contemplate that the
Secretary would weigh costs and benefits, there is noth-
ing in the Act or legislative history to guide the Secretary
in ascertaining whether a “reasonable relationship” exists
or in determining what value he should place on human
life and freedom from suffering. At bottom, petitioners’
argument speaks only to public policy considerations.
Whatever the merit of that argument, it is one that must
be addressed to Congress.

Second, petitioners contend that the Secretary’s econ-
omic feasibility determination was not based on “substan-
tial evidence.” While the Secretary had voluminous
evidence concerning that question, petitioners apparently
challenge his determination on the ground that he relied
on parts of various cost estimates, rather than on the
totality of a single submission. The record shows that the
Secretary carefully considered the evidence and tested the
assumptions underlying the various estimates, and that
his conclusions were fully reasoned and fully explained.
The Act clearly requires that the Secretary base his de-
cision on “the best available” evidence, and there is no
reason to believe that Congress intended for him to ab-
dicate his judgment by accepting or rejecting one or
another submission in its entirety. Rather, the Secretary
is required to analyze critically the evidence that he re
ceives, as he did here.

29

Finally, petitioners assert that the Secretary lacks au-
thority to require the transfer to other available jobs,
without loss of pay, of workers who cannot wear res-
pirators. The Secretary found that the respirator pro-
gram would not be effective in protecting against the risk
of disease without a transfer and income protection pro-
vision, because workers in unhealthful workplaces who
are unable to use respirators are unlikely to come forward
with that information if they would thereby face unem-
ployment or, at the very least, transfer to a lower-paying
job. Moreover, transfers without income protection would
place the costs of compliance on workers, contrary to
Congress’ intention that that burden should rest on em-
ployers. The Secretary thus properly concluded that the
transfer and income protection provision is a measure
“reasonably necessary or appropriate to provide safe or
healthful employment.” Section 3(8) of the Act, 29
U.S.C. 652 (8).

ARGUMENT

I. THE COTTON DUST STANDARD IS REASON-
ABLY NECESSARY AND APPROPRIATE BE-
CAUSE IT PROTECTS AGAINST A SIGNIFICANT
RISK OF MATERIAL HEALTH IMPAIRMENT

In Industrial Union Department v. American Petro-
leum Institute (“Benzene”), No. 78-911 (July 2, 1980),
a plurality of the Court held that the Secretary may
promulgate a new OSHA standard only when he has met
the threshold burden of showing that the existing stand-
ard exposes employees to a significant risk of material
health impairment that the new regulation will reduce
(slip op. 31-32; slip op. 2 (Powell, J., concurring) ). As
the plurality noted, the Secretary must be afforded sub-
stantial discretion in determining whether risks are sig-
nificant: “[W]hile the Agency must support its finding
that a certain level of risk exists by substantial evidence,
we recognize that its determination that a particular
level of risk is ‘significant’ will be based largely on policy

30

considerations” (slip op. 45 n.62).*% The cotton dust
standard clearly meets the threshold requirement identi-
fied in Benzene,

The previous standard for cotton dust allowed an ex-
posure level of 1000 ug/m* of total dust, or approxi-
mately 500 uwg/m* of respirable dust, in most textile
operations. See page 3, note 4, supra. The dose-
response curve based on the studies of Dr. Merchant and
his colleagues, which the Secretary determined to be “the
most thorough and objective evaluation of byssinosis
prevalence in the United States textile industry” (43 Fed.
Reg. 27355 (1978) ),* indicated that workers exposed at
that level face a one in four risk of contracting some
grade of byssinosis and a one in 15 risk of grade 2
byssinosis (ibid.; J.A. 40; 41 Fed. Reg. 56504 (1976) ).
The risk of grade 3 byssinosis (chronic disease) is far

34 The plurality also remarked that “OSHA is not required to
support its finding that a significant risk exists with anything ap-
proaching scientific certainty” (slip op. 45). Rather, the Secretary
must be afforded “some leeway where [his] findings must be made
on the frontiers of scientific knowledge” (ibid.). See also slip op. 2
(Burger, C. J., concurring) ; slip op. 3-4 (Powell, J., concurring).

35 While industry representatives criticized various aspects of
the Merchant studies (see pages 9-10, supra), the Secretary
concluded that Dr. Merchant’s findings provide ‘“‘a reliable assess-
ment of [the] health risk to cotton textile workers from cotton
dust” (43 Fed. Reg. 27357 (1978)). The Secretary noted that “the
Merchant data has been published and subject to peer review for
several years without any major reported challenge to its basic
soundness” (ibid.). Indeed, only Dr. Moon Suh of Burlington In-
dustries presented any colorable criticism of the Merchant study
and, as the Secretary pointed out, Dr. Suh’s criticisms were con-
vincingly rebutted at the hearing by Dr. Merchant and others.
See 43 Fed. Reg. 27357-27358 (1978). As the Secretary stated,
“the dose-response data of Merchant, supported by numerous other
studies * * *, is the best available evidence and * * * it provides
the Agency with sufficient evidence upon which to take action * * *.”
43 Fed. Reg. 27358 (1978). Contrary to ATMI’s contention (ATMI
Br. 14-15), the Secretary’s standard was based not merely on “pol-
icy,” but on documentation of serious disease.

31

greater than one in 1000.** That risk is manifestly sig-
nificant. See Benzene, supra, slip op. 44-45. Likewise, as
demonstrated at some length above (see pages 4-12,
supra), it is beyond question that byssinosis constitutes
a “material impairment of health or functional capacity”
in both its acute and chronic forms. Thus, the Secretary
reasonably determined that the previous cotton dust
standard presented an unacce;table risk of material im-
pairment.**

36 Dr. Bouhuys estimated that out of a total population of 235,000
retired and 200,000 active textile workers, 35,000 are permanently
disabled (J.A. 154-155; 43 Fed. Reg. 27353 (1978) ).

37In assessing the risks posed by cotton dust, the Secretary
properly considered all grades of byssinosis for several reasons.
First, the Secretary found, based on the evidence (J.A. 152-153,
421-423, 434-435), that “grade 14 byssinosis and associated pul-
monary function decrements are significant health effects in them-
selves and should be prevented so far as possible” (43 Fed. Reg.
27354 (1978)). Second, the Secretary also found, based on sub-
stantial scientific testimony (see pages 6-8, supra; J.A. 151-152),
that the prevention of chronic disease necessarily requires the pre-
vention of grade 1% byssinosis and other acute responses to cotton
dust because byssinosis is a “continuum of disease,” and persons
manifesting acute symptoms are more susceptible to chronic disease.
In many cases, therefore, further exposure “progresses inevitably to
the chronic irreversible respiratory diseases associated with cotton
dust exposure” (43 Fed. Reg. 27354 (1978)). See note ll, supra.
The Secretary’s determination to set a standard designed to elim-
inate these acute symptoms therefore constitutes a sound and justi-
fiable policy of prevention. See Benzene, supra, slip op. 45. More-
over, as the court below correctly concluded (Pet. App. 52-53),
Section 6(b)(5) of the Act, 29 U.S.C. 655(b) (5), not only permits
but requires the Secretary to act to prevent chronic disease by
protecting workers from the milder forms of byssinosis. That sec-
tion directs the Secretary to set permanent standards for occup-
tional exposure to “toxic materials or harmful physical agents” at
the level “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 exposure to the hazard dealt with by
such standard for the period of his working life” (emphasis added).
For many workers, a lifetime of exposure at current levels will
result in chronic debilitating, often fatal disease.

32

Although the new standard will not wholly eliminate the
risk of byssinosis, the Secretary found that it.would sub-
stantially reduce that risk. According to the Merchant
studies, the new PEL would halve—to 13% —the preva-
lence of all grades of byssinosis for workers in all op-
erations except slashing and weaving. Given the estimated
125,000 to 200,000 persons at work in those portions of
the industry (J.A. 154; 43 Fed. Reg. 27353 (1978) ),
this reduction would represent a substantial increase in
health benefits. Moreover, the record evidence also in-
dicated that the new standard would eliminate many cases
of chronic disease (41 Fed. Reg. 56505 (1976); 43 Fed.
Reg. 27354 (1978)). For example, Dr. Arend Bouhuys,
Director of the Yale University Lung Research Center,
stated that “it is also probable that implementation of the
[200 wg/m*] standard would greatly decrease the preva-
lence of disabling chronic lung disease among cotton tex-
tile workers in the future” (J.A. 152; see also J.A. 411).

Indeed, petitioners do not contest the validity of the
Secretary’s judgments that the previous standard pre-
sented a significant risk of material health impairment or
that various provisions of the new standard will reduce
that risk. Apart from their assertion that the Act re-
quires a cost-benefit analysis (see pages 37-63, infra),
the thrust of petitioners’ attack on the new standard
rests on their assertion (ATMI Br. 11-12, 39-40) that
the Secretary’s standard is no more effective than the in-
dustry alternative, which combines a 500 ug/m* PEL
with reliance on medical surveillance, employee transfers,
and widespread respirator use. Both the Secretary and
the court of appeals correctly rejected petitioners’ factual
assertion that their alternative would be equally or ap-
proximately as effective in curbing byssinosis.**

38 Petitioners’ contention that the industry alternative would be
equally effective is based on certain dubious factual assertions. In
the main, petitioners rejected the Merchant data upon which the
Secretary relied, focusing instead on industry studies that purported
to show a much lower prevalence of byssinosis among textile work-
ers. The assessment of those conflicting data obviously required ap-

33

The Secretary’s rejection of the industry alternative
was based principally on the recognition that a 26% risk
of »yssinosis exists at the 500 ug/m® dust level advocated
by the industry.*® While the Secretary’s standard would

plication of the agency’s expertise, and the Secretary’s judgment
Should not lightly be set aside. As the court of appeals remarked,
“OSHA gave serious attention to the medical studies * * * proffered
to support the industry’s alternative * * * and concluded that their
validity had largely been discredited” (Pet. App. 47-48). The court
below carefully reviewed the Secretary’s extensive discussion of the
industry’s studies (see 43 Fed. Reg. 27356-27359 (1978) ), sum-
marized the Secretary’s reasons for rejecting those studies (Pet.
App. 48 n.76), and concluded that the Secretary’s “judgment on this
issue is sufficiently supported on this basis” (id. at 48-49).

8° The validity of Dr. Merchant’s risk estimate was challenged
at the hearing by Dr. Moon Suh of Burlington Industries (see note
35, supra). Dr. Suh’s contentions were thoroughly ventilated at the
hearing and decided adversely to petitioners (ibid.). Petitioners’
present reliance (ATMI Br. 9 n.19) on Dr. Suh’s assertions is
therefore unjustified.

ATMI also seeks to discredit (ATMI Br. 8-9 & nn. 18, 19) the
Merchant data on the ground that they measure only prevalence
rather than dose-response, and that the prevalence of the disease
observed by Dr. Merchant may be related to “presumably higher”
dust exposures in the past. Dr. Merchant testified, however, that
he carefully designed his study to minimize that problem by choos-
ing mills in which’ “the process had been stable, where the machin-
ery in place had been stable and had not been changed recently,
[and where] the population was relatively stable. So we felt we
could minimize the major differences in dust levels in the past”
(J.A. 437-438). Although Dr. Merchant and the agency both
acknowledged “some weaknesses” in data, such as the small size
of some of the populations studied, Dr. Merchant concluded that the
data could be used to draw valid dose-response curves with ac-
cepted statistical techniques. That conclusion was thoroughly tested
at the hearing, and the Secretary concluded that it was supported
by the weight of scientific opinion (see pages 5, 10, supra). While it
is true, as ATMI observes (ATMI Br. 9 n.19), that Dr. Merchant
stated that his data could not be used to make a conclusive de-
termination as to the prevalence of byssinosis in the entire textile
industry, that is irrelevant. The objective of the study was to de-
velop dose-response data and, for this purpose, Dr. Merchant sought
to “get a good distribution of the dust over a large variety of dust

34

eliminate the risk for 13% of the affected population, the
industry alternative would simply attempt to identify,
through medical surveillance, those employees who have
already exhibited symptoms of acute byssinosis and would
permit those employees to transfer before their disease be-
came demonstrably chronic. See generally J.A. 383-384.
This industry alternative proposal is unsatisfactory for
' several reasons.

First, the plan is based on the invalid assumption (see
page 8, note 13, supra), that a bright line separates
chronic and acute forms of the disease. Second, the plan
ignores the fact that persons suffering from acute by-
ssinosis are materially impaired (see page 30, note 35,
supra), and it does nothing to prevent the onset of that
condition. Third, the plan assumes, contrary to the opin-
ions expressed by the industry’s own experts, that medi-
cal surveillance can be a useful primary method of disease
control at a 500 ug/m* exposure level, rather than simply
a useful secondary control measure.** Fourth, the plan’s

measurements” (J.A. 485). These conditions were available at
the mills that Dr. Merchant studied and he was thus able to
“find striking dose-response relationships” (J.A. 485). In sum, the
Secretary acted here on the “best available evidence,” as required
by the statute. 29 U.S.C. 655(b) (5). See Benzene, supra, slip op. 45.

*° Medical surveillance as a primary means of combating disease
was repeatedly disfavored in the rulemaking proceeding, even by
Dr. Imbus, the industry’s principal medical expert. Dr. Imbus
stated that “[mJedical surveillance programs are not a substitute
for diligent efforts to control dust” (J.A. 60), and that, “in order
for a medical surveillance program to provide any real protection
you must first of all have a manageable situation” (ibid.). Dr.
Imbus recounted one area in an all-cotton plant with a 400 ug/ms
dust level, where there was significant incidence of symptoms and
decrement of FEV; he noted that “[e]ven though dust levels here
do not look inordinately high, it is obvious that further action needs
to be taken to lower them, and medical surveillance is strictly an
interim measure” (id. at 59). See J.A. 439-441, 445 (Dr. Imbus;
Dr. Merchant). See also American Smelting & Refining Co. Vv.
OSHRC, 501 F.2d 504, 515 (8th Cir. 1974) (“[Bliological monitor-
ing did not eliminate or even reduce the hazard; it merely disclosed
it. * * * Workers should not be subjected to hazardous concentra-

35

reliance on transfers and respirators is simply unreal-
istic even on its own terms. The number of transfers the
industry can easily accommodate is necessarily low and
would be far below the number necessary at the PEL
that the industry proposed. Dr. Harold Imbus testified
that “in my experience I don’t think [the number of
workers who could transfer] could be a very high per-
centage. Certainly somewhere less than 10 percent, at
any rate” (J.A. 446). Respirators are also unsatis-

tions of airborne lead; biological monitoring should complement
* * * safe air, but is not a substitute for a healthful working envir-
onment.”’).

The shortcomings of any medical surveillance program are obvi-
ous. Workers may be less than totally candid in reporting symptoms
of chest tightness and breathlessness (see J.A. 436, 485), and in-
creasing familiarity with questionnaires may alter responses (J.A.
410-412). See also Pet. App. 30 n.13. Pulmonary testing may also
fail to disclose symptoms of acute byssinosis (J.A. 192-194, 435).
Moreover, the success of a medical surveillance program depends
upon the abilities and training of the persons administering the pro-
gram and “(t]he ability of all manufacturers to implement a high
quality medical surveillance program * * * is a problem” (J.A. 436).
For example, many workers testified that test results had been with-
held from them (J.A. 209-216, 218-220), which would signify a mis-
understanding of the gravity of acute byssinosis by persons ad-
ministering such tests. In addition to the usual problems associ-
ated with medical surveillance programs, the industry proposal pre-
sented further problems by suggesting that principal reliance should
be placed on pulmonary function testing, with questionnaires to be
used only when such tests showed a deficit (J.A. 383-384). The in-
dustry’s “active and effective medical surveillance program” (ATMI
Br. 11) would therefore fail to detect many persons suffering from
the disease. See note 13, supra.

*! Edward Baier, Deputy Director of NIOSH, and Dr. Merchant
expressed similar views. Baier remarked that “only a small pro-
portion of the workers can be effectively reassigned * * *” (J.A.
175) and Dr. Merchant stated that “[i]f you have a significant pro-
portion of your people affected there may not be adequate jobs at
low dust levels * * *” (J.A. 440-441). The Secretary took such
factors into account. The Secretary acknowledged that, under his
standard, 13% of textile employees would still be affected by the
disease, but he concluded that it “is not beyond the capacity of a

36

factory as a primary means of combating this disease, as
the record showed (J.A. 165-166, 316-322; App. 2374),
because they are both unreliable and, like voluntary trans-
fers, place responsibility for compliance on employees,
rather than on employers. As the Secretary noted (438
Fed. Reg. 27384 (1978) ):

Some hearing participants recommended a control
strategy involving principal reliance on respiratory
protection, to reduce the cost of compliance. How-
ever, respirators have many disadvantages which pre-
clude primary reliance or co-reliance on respiratory
protection on an equal basis with engincering and
work practice controls. The many difficulties with
respiratory protection were enumerated at the hear-
ing by respirator expert Bruce Held (Tr. 483) [J.A.
412-413]: because of difficulties in face fit, it is dif-
ficult to know whether the respirator actually pro-
vides adequate protection; respirators, by interfering
with vision, hearing, and mobility, can cause safety
problems; some employees cannot wear respirators
because of breathing difficulties. Finally, it is not
appropriate to place the burden of compliance prin-
cipally on the employee, as would be the case if
respiratory protection were the principal means of
reducing employee exposure.

In sum, the Secretary demonstrated that the previous
standard exposed employees to a significant risk of ma-
terial health impairment, which the new standard will
substantially reduce. Although the industry contends that
it suggested a far less costly alternative, the record
shows that the Secretary justifiably found that that al-
ternative was not nearly as effective.

typical plant to move those [still] affected to a low exposure area.”
43 Fed. Reg. 27359 (1978). The industry alternative could not
possibly achieve these beneficial results. Thus, the court of appeals
correctly concluded that “[t]he industry was unable to show that
job transfers would be available in sufficient number to respond to
the likely prevalence of byssinosis in plants with dust levels at the
500 ug/m® proposed by the industry” (Pet. App. 49).

37

II. THE SECRETARY’S RESPONSE TO THE SIGNIFI-
CANT HEALTH THREAT POSED BY COTTON
DUST IS ECONOMICALLY JUSTIFIED AND RE-
QUIRED BY THE STATUTE

As we have demonstrated above, the Secretary satisfied
the Benzene threshold requirement in promulgating the
cotton dust standard. The question thus presented is one
that the plurality expressly reserved in that case: what
further criteria must the Secretary use in determining
the degree of protection that the statute requires in
setting a particular standard. Petitioners contend (ATMI
Br. 36-47; NCC Br. 22-23) that the statute requires the
Secretary to weigh costs and benefits and to determine
that there is a “reasonable relationship” between the two
factors. Otherwise, they assert, a standard could drive an
industry to the brink of ruin or result in a serious mis-
allocation of limited resources under the Secretary’s defi-
nition of “economically feasible’ (AMTI Br. 31).

Petitioners’ submission should be rejected for two rea-
sons. As a factual matter, there is no evidence to indi-
cate that the Secretary’s administration of the Act, which
has been carried out in a responsible and practical man-
ner, will lead to such results. See pages 55-58, infra.
More important, neither the language of the statute
nor its legislative history contains any suggestion that
Congress intended to impose upon the Secretary the duty
to conduct a cost-benefit analysis in setting health stand-
ards for toxic substances.*? Rather, the Secretary’s statu-

42 In some statutes, Congress has expressly determined that ad-
ministrative agencies should perform a cost-benefit analysis in carry-
ing out the legislative policy. On other occasions, Congress has not
imposed such a requirement. See, e.g., EPA v. National Crushed
Stone Ass’n, No. 79-770 (Dec. 2, 1980). The decision whether to
require that regulations be cost-justified is clearly within the prov-
ince of the legislative branch, and it is not the role of the courts
to make that policy judgment anew. As the Court remarked in
TVA V. Hill, 347 U.S. 153, 194 (1978): “Our individual appraisal
of the wisdom or unwisdom of a particular course consciously se-
lected by the Congress is to be put aside in the process of inter-

38

tory duty is clear: he must set standards that protect
against significant risks of material harm to employee
health, to the extent such protection is technologically and
economically feasible.

To be sure, in determining the feasibility of a stand-
ard, the Secretary is obliged to consider the financial
health and viability of an industry. But that is far dif-
ferent from requiring an individualized and heavily sub-
jective balancing of costs and benefits in every case. A
decision to engraft such a requirement upon this statu-
tory scheme would not be consistent with either the
legislative design or the proper function of the Court in
construing this remedial legislation. See Whirlpool Corp.
v. Marshall, 445 U.S. 1, 18 (1980).

In enacting the Occupational Safety and Health Act,
Congress was not unaware of the need to give appropriate
attention to practical considerations in enforcement of the
statutory scheme. Congress therefore chose the term
“feasible” to moderate the stringency of standards prom-
ulgated under the Act. Congress was likewise aware that
costs and benefits would have to be balanced, and it struck
the balance itself; it determined that a “reasonable re-
lationship” exists when costs are feasible and their
expenditure will reduce a significant risk of material
health impairment. If the Secretary were required to
make his own determination of a “reasonable relation-
ship” between the cost and benefits of a particular stand-
ard, he would necessarily be empowered to strike the
balance at a different and often less protective point than

preting a statute. Once the meaning of an enactment is discerned
and its constitutionality determined, the judicial process comes to
an end.” See also Benzene, supra, slip op. 2 (Burger, C.J., con-
curring); Harris v. McRae, No. 79-1268 (June 30, 1980), slip op.
26. Whatever the intrinsic value of cost-benefit analysis, and we
think petitioners grossly overstate both its practicability and its
utility in this context, Congress neither explicitly nor implicitly re-
quired the Secretary to employ that technique in prescribing stand-
ards governing exposure to toxic substances in the workplace.

39

Congress intended.* Neither the language nor the
legislative history of the Act countenances such a scheme.

A. A cost-benefit test is inconsistent with the language
and legislative history of Section 6(b)(5)

1. The principal provision of the Act governing health
standards for toxic substances, Section 6(b) (5) of the
Act, 29 U.S.C. 655(b) (5), is emphatic in its mandate:
“The Secretary * * * shall set 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 [regulated toxic substance] for the pe-
riod of his working life’ (emphasis added). The lan-
guage of Section 6(b) (5) contains no hint of a cost-
benefit requirement. The role that the costs of a standard
are intended to play is expressed in the phrase “to the
extent feasible,” and the ordinary meaning of “feasible”
is “capable of being done.” See Webster’s Third New
International Dictionary 831 (1976). The word does not
refer to the desirability or benefit of doing something;
it refers, instead, to the ability to do something.“ Cost-

*8 A cost-benefit analysis would not always yield a less protective
standard. If an industry’s production process were extremely haz-
ardous and its product readily replaceable, it might be that closing
most or all of the industry would be cost-beneficial. But Congress
clearly did not grant the Secretary the authority to regulate in
that fashion. The courts have uniformly construed the feasibility
limitation as prohibiting the Secretary from protecting workers by
putting most of their employers out of business. American Iron &
Steel Institute v. OSHA, 577 F.2d 825, 835 (3d Cir. 1978), cert.
dismissed, No. 78-918 (Sept. 10, 1980); AFL-CIO v. Brennan, 530
F.2d 109, 123 (3d Cir. 1975); Industrial Union Department v.
Hodgson, 499 F.2d 467, 478 (D.C. Cir. 1974). The scheme that
Congress directed the Secretary to follow—reduction of significant
risks at feasible costs—is something quite different from that im-
plied by conventional cost-benefit balancing tests.

*# Congress has consistently employed the word “feasible” in
health and safety legislation with this difference in mind, including
in one statute enacted the same week as OSHA. Section 211(c) (2)

40

benefit analysis, on the other hand, involves a specific
method for assessing the desirability of proposed actions.

In setting OSHA standards, the Secretary has first
determined the level of protection at which employees
will no longer face a significant risk of material harm;
he has then determined whether that level is technologi-
cally and economically capable of achievement. If he
finds that the level of protection is not technologically
and economically feasible, the provisions of the standard

(A) and (B) of the Clean Air Act authorizes EPA to regulate,
control or prohibit automobile fuel additives. Control or prohibition
of additives that endanger the public health is permitted after the
agency has considered “other technologically or economically feasi-
ble means of achieving emission standards * * *.” In contrast,
control or prohibition of additives that may impair effective opera-
tion of emission control devices is permitted only after consideration
of “a cost benefit analysis” comparing device performance with and
without the control or prohibition. 42 U.S.C. (Supp. III) 7545
(c)(2)(A) and (B) (emphasis added). See Benzene, supra, slip
op. 23 n.27 (Marshall, J., dissenting) (“This approach demonstrates
a legislative awareness of the difference between a feasibility con-
straint and a constraint based on weighing costs and benefits’).

Later Congresses have also recognized this difference. For pur-
poses of the 1977 effluent limitations established by the Federal
Water Pollution Control Act Amendments of 1972, Congress re-
quired the “best practicable control technology currently available,”
which was to be determined by consideration of, inter alia, ‘the
total cost of application of technology in relation to the effluent
reduction benefits to be achieved” (33 U.S.C. 1311(b)(1)(A), 1314
(b)(1)(B)). With respect to the 1987 effluent limitations, how-
ever, Congress required the “best available technology economically
whievable”; while “costs” were to be considered, no direct compari-
son of costs and benefits was required. See EPA v. National
Crushed Stone Ass'n, supra, slip op. 5-6. And Section 325(a) of
the Energy Policy and Conservation Act of 1975, 42 U.S.C. 6295,
which requires the promulgation of energy conservation standards,
specifies that a standard must be “economically justified.” A stand-
ard was economically justifiable only if the Secretary found that
the benefits of the standard outweighed its costs and that the stand-
ard was “economically feasible.” “Feasibility” was defined as “ca-
pable of being carried out.” S. Conf. Rep. No. 94-516, 94th Cong.,
Ist Sess. 172 (1975) ; H.R. Conf. Rep. No. 94-700, 94th Cong., 1st
Sess. 172 (1975). But see 42 U.S.C. (Supp III) 6296.

41

must be adjusted. Requiring the Secretary to adjust the
standard further through the cost-benefit technique that
petitioners advocate would necessarily violate the “no-
material-impairment” mandate of the statute by depriv-
ing employees of protection that is “feasible.”

2. The cost-benefit gloss advocated by petitioners can-
not be justified by resort to legislative history. That
history unambiguously reflects Congress’ determination
that, in the case of toxic substances, affordable costs of
reducing significant risks of material harm are cate-
gorically outweighed by the benefits derived from the
prevention of premature death, disability, and disease.
Congress’ concern with the pernicious effects of exposure
to toxic substances in the workplace was a pri
impetus for passage of the Act. Leg. Hist. 142-143, 159-
160, 411-415, 503, 510, 845-846, 1061. Congress con-
demned the continued unregulated presence in the work-
place of the so-called “silent killers” such as cotton
dust, toxic substances whose harmful effects had been
fairly documented; Congress was also concerned that
those effects would be replicated geometrically due to the
rapid introduction of new and yet untested chemicals and
processes, *®

45 Leg. Hist. 1044-1045 (remarks of Rep. Udall).

46 The Senate Committee on Labor and Public Welfare stated
(Leg. Hist. 142-143) :

Occupational diseases which first commanded attention at
the beginning of the Industrial Revolution are still undermin-
ing the health of workers. Substantial numbers, even today,
fall victim to ancient industrial poisons such as lead and mer-
cury. Workers in the dusty trades still contract various res-
piratory diseases. Other materials long in industrial use are
only now being discovered to have toxic effects. In addition,
technological advances and new processes in American indus-
try have brought numerous new hazards to the workplace.
* * * [N]Jew materials and processes are being introduced into
industry at a much faster rate than the present meager re-
sources of occupational health can keep up with. It is esti-
mated that every 20 minutes a new and potentially toxic chemi-
cal is introduced into industry. New processes and new

42

Congress recognized that the boldness of its legislation
would have to match the seriousness of the problems pre-
sented. The Act would thus have to fill the “unconscion-
able gap between the minimal protection being given to
[employees] and the protection they need” (Leg. Hist.
1045 (remarks of Rep. Udall)). Generally, the level of
protection would have to be more stringent than that
provided by the previously voluntary consensus stand-
ards.*7 With respect to toxic substances, Congress made
clear that workers should remain healthy even if they
spent their entire working lives in the industry. It was
not acceptable to expect workers to be harmed, perhaps
irreparably, and then move on to safer employment. As
Senator Dominick remarked (Leg. Hist. 502-503; em-
phasis added) :

[W]hat we are doing now is to say that the Secre-
tary has got to use his best efforts to promulgate
the best available standards, and in so doing, that

sources of energy present occupational health problems of
unprecedented complexity.

Recent scientific knowledge points to hitherto unsuspected
cause-and-effect relationships between occupational exposures
and many of the so-called chronic diseases—cancer, respira-
tory ailments, allergies, heart disease, and others.

* * * * *

Studies of particular industries provide specific emphasis
regarding the magnitude of the problem. For example, despite
repeated warnings over the years from other countries that
their cotton workers suffered from lung disease, it is only
within the past decade that we have recognized byssinosis as
a distinct occupational disease among workers in American
cotton mills.

See also id. at 844-846 (House Report).

47 The House Report emphasized the minimalist character of
national consensus standards (Leg. Hist. 848) :

Voluntary standards set by the consensus method represent
the lowest common denominator of change in the status quo
that is necessary to achieve acceptance. Now that we have
chosen the force of law to command acceptance of safety and
health regulations, this method should not apply.

43

he should take into account that anyone working in
toxic agents or physical agents which might be harm-
ful may be subjected to such conditions for the rest
of his working life, so that we can get at something
which might not be toxic now, if he works in it a
short time, but if he works in it the rest of his life
it might be very dangerous; and we want to make
sure that such things are taken into consideration in
establishing standards * * *.

Congress was not unmindful that these requirements
would entail substantial compliance costs. Senator Eagle-
ton characterized these costs as “reasonable and neces-
sary costs of doing business” (Leg. Hist. 1150) because
“we can no longer permit profits to be dependent upon
an unsafe or unhealthy worksite” (id. at 1151). Senator
Yarborough, a principal sponsor of the legislation, echoed
this understanding, stating that the cost of compliance
was “the price we should pay for the 80 million workers
of America” (id. at 444). Senator Yarborough also re-
sponded to the assertion by some critics that the costs
of protection would be too great. He noted that the in-
creased costs “would be put into consumer goods” (id.
at 444), and further stated (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 allow-
ance under workmen’s compensation and _ social
security? * * *

on * a *

We are talking about people’s lives, not the in-
difference 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.

Congress anticipated that new standards could pur-
chase greater protection only through greater costs, but
it was not prepared to allow feasible standards to be

44

relaxed to accommodate particular employers. For in-
stance, Congress recognized that small firms might face
relatively greater economic burdens, but it refused to ease
those burdens by lowering standards. Instead, Congress
chose to assist such firms through government loans.
Leg. Hist. 525-526; 15 U.S.C. 686(b) (5). See S. Rep.
No. 93-132, 93d Cong., 1st Sess. 3 (1973). Moreover,
Congress explicitly provided that temporary variances
from a standard could not be based on economic hardship.
29 U.S.C. 655(b) (6) (A)-(B); Leg. Hist. 1188, 1201.
Compare 29 U.S.C. 655(d). To guarantee the continued
existence of every employer would permit employers who
have lagged behind in safety and health protection to
perpetuate their competitive advantage and frustrate the
achievement of the legislative purpose.** See EPA v.
National Crushed Stone Ass’n, supra, slip op. 14-18.
Congress recognized that the standards to be imposed
might significantly increase the cost of doing business,
but it adjudged that the costs nonetheless should be
imposed so long as they were feasible for the industry
as a whole.

No Member of Congress so much as suggested that,
because of cost-benefit considerations, the Secretary must
tolerate risks of material health impairment, caused by
toxic substances, when those risks could be eliminated
feasibly. Given the length and thoroughness of the legis-
lative inquiry prior to passage of the Act, the omission

48 Congress found little natural economic incentive for employers
to invest capital in safety and health protection. Because employees
often move from job to job, “an employer investing in current occu-
pational health precautions would receive no future benefits, not
even in the reduction of workmen’s compensation costs. He might
be paying for the past neglect of another employer. Uniform en-
forcement will also reduce or eliminate the disadvantage that a
conscientious employer might experience where inter-industry or
intra-industry competition is present.” Leg. Hist. 854 (House Re-
port). Moreover, “many employers—particularly smaller ones—
simply cannot make the necessary investment in health and safety,
and survive competitively, unless all are compelled to do so.” Leg.
Hist. 144 (Senate Report).

45

of any reference to a matter as significant as a cost-
benefit analysis “cannot be deemed inadvertent.” Ben-
zene, supra, slip op. 32 (Marshall, J., dissenting).

Petitioners support their contention that the statute
requires a cost-benefit justification principally by ref-
erence to qualifying language contained in the Act:
“most adequately” assures, “to the extent feasible,” that
no “material” impairment will occur (ATMI Br. 31-34).
The Secretary does not dispute that this language man-
dates consideration of practical concerns, including costs,
in the setting of standards. Neither Congress nor the
Secretary has ever contended that this legislation does
or should license the agency to attempt to achieve ab-
solute occupational safety and health. But neither do
these qualifying phrases insist upon a balancing of costs
and benefits as the assurance that standards will be
realistic.

A standard can “most adequately” assure protection
when the benefits to be produced are significant and the
costs are feasible. Furthermore, “material” impairment
plainly refers to the severity of the health or safety con-
sequences and to the degree of risk of suffering those
consequences. Benzene, supra, slip op. 31-33 n.48. The
concept of “material impairment” unquestionably limits
the Secretary’s standard-setting authority. Assessing the
gravity and likelihood of adverse health consequences,
however, does not logically entail a comparison of those
consequences with the costs of avoiding them.** Hence, |
Congress’ decision that “the” standard must protect
against “material” impairment, rather than “any” im-
pairment, does not suggest that avoidance of a “ma-

49 For example, two substances may be equally likely to cause
cancer. It may cost very little to minimize exposure to one, while
it may be difficult and very costly to prevent exposure to the other.
Contrary to petitioners’ contention (ATMI Br. 38-40), it makes
little sense to regard the potential impairment caused by unregu-
lated exposure to one substance as more “material,” or the risk of
one as less “significant,” than the other.

46

terial” impairment must be justified on a cost-benefit
basis.

3. The pattern and structure of the Act as a whole
also demonstrate that Congress did not intend health
protection in this area to depend on an individualized
balancing of costs and benefits with respect to every
standard. Congress’ express purpose was to create safe
and healthful workplaces “so far as possible” (Section
2 of the Act, 29 U.S.C. 651). The Act imposes on em-
ployers the general duty to furnish workplaces “which
are free from recognized hazards that are causing or are
likely to cause death or serious physical harm * * *”
(Section 5(a) (1) of the Act, 29 U.S.C. 654(a) (1)). It

5° The legislative history indicates that Congress substituted the
term “material impairment” for “any impairment” to ensure that
the Secretary would not be required to regulate trifling or remote
hazards such as mosquito or black-fly bites. Leg. Hist. 345 (re-
marks of Sen. Dominick). Petitioners also rely (ATMI Br. 34) on
the compromise reached by the conference committee, which re-
sulted in the use of the “substantial evidence” scope of judicial
review, rather than the “arbitrary and capricious” standard. See
Section 6(f) of the Act, 29 U.S.C. 655(f). The substantial evidence
standard reflects a legislative judgment that the Secretary should
develop standards on a relatively formalized basis by providing a
hearing and a record capable of review. Leg. Hist. 298-299, 884,
1053. But these procedural concerns do not show that Congress
intended to protect industries from all costs that were not, in their
view, “reasonably related” to benefits. Florida Peach Growers
Ass’n V. Department of Labor, 489 F.2d 120 (5th Cir. 1974), cited
by petitioners (ATMI Br. 34), suggests only that Congress was
concerned about avoiding the imposition of “arbitrary burdens” on
employers (489 F.2d at 128). The Fifth Circuit cites no legislative
history to elucidate the manner in which a burden might be “arbi-
trary,” and the sole reference relied on by ATMI here is totally in-
apposite. ATMI relies (ATMI Br. 34 n.70) on the remarks of Sena-
tor Cook with respect to the “arbitrary and capricious” test for
judicial review. Leg. Hist. 342-344. But these remarks related to
judicial review of the Secretary’s veto of a state plan for regulating
occupational health, not to judicial review of an occupational health
standard or to the prospect that costs might be out of proportion
to benefits. Jbid. Accordingly, this legislative compromise sheds no
light on the issue whether a standard must satisfy a cost-benefit test.

47

is not enough that a workplace be “reasonably free” of
hazards; it must be as free of serious hazards as pos-
sible. See National Realty and Construction Co. v.
OSHRC, 489 F.2d 1257, 1265-1266 (D.C. Cir. 1973).
Likewise, in choosing among alternative existing stand-
ards during the first two years after the Act became
effective, the Secretary was required to “promulgate the
standard which assures the greatest protection of the
safety or health of the affected employees” (Section 6 (a)
of the Act, 29 U.S.C. 655(a)). Where a new standard
deals with toxic substances, Section 6(b) (5) requires that
the Secretary adopt “the” standard that “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 exposure to the hazard dealt with
by such standard for the period of his working life”
(29 U.S.C. 655(b) (5)). Finally, variances may be ob-
tained, but only on limited grounds that do not include
financial hardship. 29 U.S.C. 655(b) (6) (A)-(B), 655
(f). Leg. Hist. 1188, 1201.

The statute thus was designed to ensure the highest
possible degree of employee protection against significant
risks of material harm, particularly in the case of toxic
substances. The urgent call for highly protective stand-
ards to prevent significant risks of material impairment
derived from a lifetime of exposure, the refusal to allow
variances for economic hardship, and the provision of
economic relief only through loans for small businesses,
all show that Congress carefully considered and limited
the role that cost considerations should play here, a role
that does not include the lowering of feasible standards.
There is no warrant to read into this statute “what
Congress has studiously omitted.” FTC v. Simplicity
Pattern Co., 360 U.S. 55, 67 (1959). See Transamerica
Mortgage Advisors, Inc. v. Lewis, 444 U.S. 11, 20 (1979) :
Great Atlantic & Pacific Tea Co. v. FTC, 440 U.S. 69,
78-79 (1979).

48

Petitioners contend, however, that the Secretary’s view
of his mandate would push the industry to the brink of
ruin (ATMI Br. 31). In our view, Congress’ insertion
of the feasibility requirement was specifically designed to
preclude that draconian result and to assure that stand-
ards would be “reasonable” and “practical.” We agree
that the economic feasibility requirement limits the Sec-
retary’s authority to set standards for toxic substances.
Moreover, were an economic feasibility test as ineffectual
as the industry suggests, Section 6(b) (5) would be a
harsh provision indeed. Consistent with the intent of
Congress, however, the Secretary has construed the
“capable of achievement” criteridn in a reasonable and
practical fashion that provides a meaningful brake on
regulation.™

51 Accordingly, petitioners’ argument is not advanced even if, as
they assert (ATMI Br. 32-33), Congress regarded the terms “rea-
sonable” and “practical” as a gloss on “feasible,” because the Secre-
tary’s application of the feasibility limitation results in reasonable
and practical standards. Petitioners rely in the main on the re-
marks of Senator Javits, the original author of the feasibility
amendment, to suggest that the term was meant to include some-
thing in the nature of “reasonableness” or “practicality.” But
petitioners’ reading of the legislative history is deficient because
Senator Javits did not use those words to describe his amendment.
Senator Javits introduced the Nixon Administration’s originai bill
(S. 2788, 91st Cong., 1st Sess. (1969) ), which would have required
the board responsible for issuing health and safety standards to
submit its proposals to certain organizations for a report on “the
technical feasibility, reasonableness and practicality of such stand-
ard.” Leg. Hist. 38-39; 115 Cong. Rec. 22511 (1969). Contrary
to ATMI’s assertion, Senator Javits did not “describe” the report
in terms of “reasonableness” and “practicality.” The “description”
of the report was made by President Nixon in his message concern-
ing the Administration bill. See 115 Cong. Rec. 22511, 22517
(1969). Senator Javits specifically disassociated himself from the
Administration’s bill on a variety of issues, stating that “it may be
desirable to go somewhat further than this bill does in strengthen-
ing the hand of the Federal Government in this area” (id. at
22512). Significantly, when Senator Javits actually described his
feasibility amendment, he consistently used only the word “feasi-
ble”; he never equated “feasible” with “reasonable” or “practical”

49

As the court stated in Industrial Union Department v.
Hodgson, 499 F.2d 467, 478 (D.C. Cir. 1974), Congress
did not “intend[] to protect employees by putting their
employers out of business.” The Secretary’s overall pur-
pose, therefore, is to assure that the industry is able
to finance the costs of compliance, as well as other “rea-
sonable and necessary costs of doing business” (Leg.
Hist

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385006_2223%3A07. Public record. Not legal advice.
