Determining Compensation for
Subsequent Use of Test Data Under

FIFRA: A Value-Based or
Cost-Based Standard?

INTRODUCTION

Disagreement plagues the pesticide industry over the issue of
data compensation.' At the core of this debate is the following
question: What standard, if any, does an arbitrator use in deter-
mining the compensation awarded the original submitter of test
data when it is used by the EPA at the request of a subsequent
applicant to support a "follow-on" pesticide registration under
the mandatory data-licensing scheme in FIFRA. 2 The resolution
of this question is of crucial importance to competing factions
within the pesticide industry in the wake of the Supreme Court's
recent decision in Thomas v. Union Carbide Agricultural Products Co.3

In upholding the constitutionality of FIFRA's data-licensing
scheme, the Union Carbide decision has opened the door for com-
pensation claims to be submitted to arbitration under the statute.
The debate arises from the fact that FIFRA does not provide the
arbitrator with an explicit standard or formula to be used in de-
termining the compensation to be awarded to an original data-
submitter. 4 In response, two broad alternative theories for deter-
mining compensation have been advocated by representatives of
competing factions within the pesticide industry.

Original data submitters - primarily composed of large chemi-
cal manufacturers with the resources to invent new pesticides and

1. The disagreement within the pesticide industry is reflected by the widely divergent
views presented at a recent conference on data compensation held by the National Agri-
cultural Chemical Association (NACA) in Washington, D.C. on September 11, 1985. For a
brief summary of the meeting, see NACA Data Compensation Session Hears Sharply Divergent
Opinions, Pesticide and Toxic Chemicals News, September 18, 1985, at 25-28.

2. Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) § 3(c)(l)(D), 7 U.S.C.
§ 136a(c)(1)(D) (1982), as amended by Federal Pesticide Act of 1978, Pub. L. No. 95-396, 92
Stat. 820 (1978).

3. - U.S. -, 105 S. Ct. 3325 (1985).
4. FIFRA § 3(c)(l)(D)(ii), 7 U.S.C. § 136a(c)(1)(D)(ii) (1982) ("The terms and amount

of compensation may be fixed by agreement between the original data submitter and the
applicant, or, failing such agreement, binding arbitration under this subparagraph.").



194 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

develop the costly test data required to register them - argue
that compensation should be based on the value of the test data
to a "follow-on" registrant. The essence of a value-based stan-
dard is that compensation should include more than just a share
of the testing costs; compensation should also take into account
the economic benefits enjoyed by "follow-on" registrants when
they enter the market without suffering the delay involved in de-
veloping test data and obtaining regulatory approval. In other
words, value-based compensation includes the opportunity costs
suffered by the original data submitter as a result of the regula-
tory delays involved in bringing a new pesticide to market.

In contrast, "follow-on" registrants - primarily composed of
smaller competing manufacturers and formulators - argue that
compensation should be based on an equitable sharing of the ac-
tual cost of producing the test data alone. Within this cost shar-
ing framework, there is further debate as to whether equitable
sharing should be calculated on a market share or a per capita
basis. If calculated on a market share basis, then the costs of the
test data would be divided between registrants based on a com-
parison of either their present market share or prospective market
share. If calculated on a per capita basis, then the costs of the test
data would be equally divided.

The stakes are high in this debate. In the central decision is-
sued to date, Stauffer Chemical Co. v. PPG Industries, the original
data-submitter (Stauffer) was awarded value-based compensa-
tion.5 By relying on Stauffer's data, the follow-on registrant
(PPG) entered the market five years earlier than it would have if it
had been forced to independently generate the test data and ob-
tain regulatory approval. The arbitrators measured the value of
this early market entry to PPG as approximately fourteen million
dollars, while PPG's per capita share of the actual cost of develop-
ing the data was only approximately one and one half million
dollars.

Including the Stauffer decision, only two arbitrator's decisions
have been issued under FIFRA's data-licensing scheme. 6 Both

5. Stauffer Chemical Co. v. PPG Industries, No. 16 199 077 82, Federal Mediation and
Conciliation Service (June 28, 1983) (Birch et al., Arb.) [hereinafter cited as Stauffer). For a
detailed discussion of the Stauffer decision, see text at nn. 110-38, infra.

6. Stauffer and FMC Corp. v. Tricon International, No. 16 199 0033 84G, American Ar-
bitration Association (Jan. 10, 1985) (Foy et al., Arb.) [hereinafter cited as FMC]. For a
detailed discussion of the FMC decision, see infra, text accompanying notes 139-49.



FIFRA Compensation

decisions address the issue of a standard for determining com-
pensation, but neither resolves it satisfactorily. More impor-
tantly, the arbitrators in the Stauffer decision ignored important
elements of the legislative history of the data-licensing scheme in
granting value-based compensation. Despite the absence of ex-
plicit statutory guidance, an arbitrator need not act independently
when determining "compensation." In its Union Carbide decision,
the Supreme Court is careful to point out that the legislative his-
tory of the amendments to FIFRA are "far from silent" on the
issue of FIFRA's standard for compensation. 7

This note will explore the turbulent history of the data-licens-
ing scheme in FIFRA to determine whether it provides an arbitra-
tor with guidance in choosing a standard for compensation. By
way of background, Section I will focus on the details of the 1972
and 1978 amendments to FIFRA which created the mandatory li-
cense requirement, and two recent Supreme Court decisions es-
tablishing its constitutionality. Section II will analyze the
legislative histories of both the 1972 and 1978 amendments with
an eye to understanding what they reveal about the issue of
FIFRA's standard for compensation. Section III will then analyze
the two arbitrator's decisions that have been issued thus far in
light of Congress' purposes related to compensation as revealed
in FIFRA's legislative history.

The results of these analyses support the follow-on registrant's
claims that compensation should be based on the actual costs of
developing test data. Congress enacted the compensation provi-
sions to facilitate market entry and encourage competition in the
pesticide industry without discouraging innovation. A value-
based standard that includes the opportunity costs avoided by fol-
low-on registrants defeats the competitive purposes of the com-
pensation provisions. The arbitrators in Stauffer failed to
adequately consider the legislative histories of the amendments to
FIFRA and were prejudiced by factors irrelevant to the issue
before them; as a result, they erred in awarding value-based com-
pensation. However, within a cost sharing framework, Congress
was silent on the question of how to equitably divide testing costs
and left this issue to the reasonable determination of the arbitra-
tors and parties involved.

7. 105 S. Ct. at 3340 (citations omitted).

1986]



196 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 11:193

I. FIFRA's MANDATORY DATA-LICENSING SCHEME AND

DISCUSSION OF RECENT SUPREME COURT DECISIONS

UPHOLDING ITS CONSTITUTIONALITY

A. 1972 Amendments: The Federal Environmental Pesticide Control
Act of 1972

FIFRA was first adopted in 1947; at that time it was primarily a
licensing and labeling statute designed to protect pesticide users
from misbranded and adulterated pesticides.8 License applicants
were required to submit test data and a pesticides's formula to the
Secretary of Agriculture in support of the claims made on their
labels. 9 Disclosure of "any information relative to formulas of
products" was prohibited, but the 1947 FIFRA was silent with re-
spect to the use of submitted data in support of subsequent appli-
cations and the disclosure of health and safety data.' 0

In 1972, the reach of FIFRA was expanded through the passage
of the Federal Environmental Pesticide Control Act of 1972."1
The amendments transformed FIFRA into a comprehensive regu-
latory statute. The 1972 amendments also introduced a
mandatory data-licensing scheme to FIFRA for the first time. 12

The EPA was authorized to consider data submitted by one appli-
cant for registration as support for another application, provided
the subsequent ("follow-on") applicant offered "reasonable com-
pensation" to the original data submitter.' 3 The amount of
compensation was to be negotiated between the parties; if negoti-
ations failed, the EPA was to resolve the dispute by conducting

8. FIFRA, Pub. L. No. 80-104, Law ofJune 25, 1947, ch. 125, 61 Stat. 163 (1947). For a
discussion of the early history of FIFRA, See F. GRAD, TREATISE ON ENVIRONMENTAL LAW,

§ 8.02 (1985).
9. FIFRA, Pub. L. No. 80-104, Law ofJune 25, 1947, ch. 125, §§ 4(a) and (b), 61 Stat.

167-68 (1947). See also Ruckelshaus v. Monsanto Co., - U.S. -, 104 S. Ct. 2862, 2864
(1984) (hereinafter referred to as "Monsanto"). The Department of Agriculture's responsi-
bilities under FIFRA were transferred to the newly created Environmental Protection
Agency in 1970. See Reorganization Plan No. 3 of 1970, 35 Fed Reg. 15623 (1970).

10. FIFRA, Pub. L. No. 80-104, Law ofJune 25, 1947, ch. 125, § 3(c)(4), 61 Stat. 166-67
(1947). However, in Monsanto, the Supreme Court stated that prior to 1972 "there is some
evidence that the practice of using data submitted by one company during consideration of
the application of a subsequent applicant was widespread and well known". 104 S. Ct. at
2877 (Discussion of evidence in footnote omitted). The legislative history of the 1972
amendments to FIFRA support this conclusion. See S. REP. No. 838 (Pt. II), 92nd Cong.,
2d Sess., reprinted in 1972 U.S. CODE CONG. & AD. NEWS 3993, 4040.

11. Federal Environmental Pesticide Control Act of 1972 (FEPCA), Pub. L. No. 92-516,
86 Stat. 973 (1972).

12. FEPCA, Pub. L. No. 92-516, § 3(c)(l)(D), 86 Stat. 973, 979-80 (1972).
13. Id.



FIFRA Compensation

administrative hearings.' 4 Compensation decisions were subject
to judicial review in a federal district court only at the instigation
of the original data-submitter, and the reviewing court was pre-
cluded from reducing the amount of compensation awarded by
the EPA. 15

The 1972 data-licensing scheme was limited by the trade secret
provisions in FEPCA.16 Original data-submitters were allowed to
designate any portions of their submitted material as "trade
secrets or commercial or financial information."' 7 The EPA was
prohibited from publicly disclosing such information, and could
not consider it in support of a subsequent (follow-on) application
without the original submitters consent.' If the EPA disagreed
with the designation of certain information as "trade secrets or
commercial or financial information," the original submitter was
authorized to institute a declaratory judgment action in federal
district court. '9

Effective operation of the 1972 mandatory data-licensing
scheme was blocked by litigation. The initial failure to establish
an effective date in the Act for mandatory licensing led to litiga-
tion over the possibility of the retroactive application of the com-
pensation requirement to data already in the EPA's files. 20 In
1975, Congress resolved this problem by amending Section three
to provide that data submitted in support of earlier applications
are subject to the compensation provisions only if submitted after
December 31, 1969.21 However, judicial decisions which ex-
panded the category of data protected as "trade secrets or com-
mercial or financial information" continued to block the effective

14. Id.

15. Id.

16. FEPCA, Pub. L. No. 92-516, §§ 10(a)-10(c), 86 Stat. 973, 989 (1972).

17. Id. at § 10(a).

18. Id. at § 10(b).
19. Id. at § 10(c).

20. See Amchen v. GAF, 391 F. Supp. 124 (N.D. Ga. 1975). Such litigation was over data
already in the EPA's files under registrations prior to the enactment of FEPCA (October
21, 1972), and prior to the effective date of its registration provisions. 38 Fed. Reg. 31862
(Nov. 19, 1973).

21. FIFRA § 3(c)(1)(D), 7 U.S.C. § 136a(c)(1)(D) (1982), as amended by the Federal
Pesticide Act of 1975, Pub. L. No. 94-140, § 12, 89 Stat. 751, 755 (Nov. 25, 1975). Under
the 1975 amendment, the obligation to pay for the use of such data applies only with
respect to applications for registrations or registrations submitted on or after the enact-
ment of FEPCA (October 21, 1972).

1986]



198 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

operation of mandatory licensing.2 2 Little room was left for the
application of the mandatory licensing scheme because such a
large percentage of the submitted test data was protected under
the rubric of trade secret. 23

B. 1978 Amendments: The Federal Pesticide Act of 1978

In response to these and other regulatory problems, FIFRA was
amended again with the passage of the Federal Pesticide Act of
1978.24 Both the mandatory licensing and trade secret provisions
of the Act were changed substantially to their present form. 2 5 In
their complexity, the current provisions reflect the difficult history
of the mandatory licensing scheme. Since the 1978 amendments,
FIFRA divides submitted test data into three categories based on
their date of submission:

[1] Data submitted in support of an application for the original
registration or new use of a pesticide which is registered after
September 30, 1978, the effective date of the 1978 amendments,
are protected from use by another applicant for ten years from
the date of registration. 26 This is known as a ten year "exclusive
use" period. The EPA may not consider such data in support of a
later follow-on application for ten years without the permission of
the person who originally submitted the data. After the expira-
tion of the ten year exclusive use period, the person who submit-
ted the original data is also entitled for the following five years to
compensation for the use of this data in the manner discussed
below.

2 7

22. FEPCA, Pub. L. No. 92-516, §§ 10(a)-10(c), 86 Stat. 973, 989 (1972). The EPA
argued for a narrow statutory definition of "trade secret or commercial or financial infor-
mation" that did not encompass data related to the safety or efficacy of a pesticide prod-
uct. However, the courts disagreed with EPA's interpretation and found that Congress
had accepted the broad definition of "trade secret" set forth in section 757 of the Restate-
ment of Torts. See Mobay Chemical Corp. v. Costle, 447 F. Supp. 811 (W.D.Mo. 1978);
Chevron Chemical Co. v. Costle, 443 F. Supp. 1024 (N.D.Cal. 1978); Dow Chemical Co. v.
Train, 423 F. Supp. 1359 (E.D. Mich. 1976).

23. The legislative history of the 1978 amendments to FIFRA indicates that these deci-
sions effectively prevented the EPA from considering the data submitted by one applicant
when reviewing a later application by another. See S. REP. No. 334, 95th Cong., 1st Sess. 7
(1977); H.R. REP. No. 663, 95th Cong., 2nd Sess. 18 (1977). See also Monsanto, 104 S. Ct.
at 2865.

24. Federal Pesticide Act of 1978, Pub. L. No. 95-396, § 2(a), 92 Stat. 820 (1978).
25. FIFRA §§ 3(c)(l)(D), 10, 7 U.S.C. §§ 136a, 136h (1982).
26. See FIFRA § 3(c)(1)(D)(i), 7 U.S.C. § 136a(c)(l)(D)(i) (1982).
27. See FIFRA § 3(c)(1)(D)(ii), 7 U.S.C. § 136a(c)(1)(D)(ii) (1982).



FIFRA Compensation

[2] Data submitted after December 31, 1969, the date set by
Congress in the 1975 amendments to provide for a starting point
for compensation of earlier data, and before September 30, 1978,
the effective date of the 1978 amendments, may be considered by
the EPA in support of an application by any other person, but
only if the new applicant has made an offer to "compensate" the
person who submitted the original data. 28 This requirement only
applies within a fifteen year period following the date of the origi-
nal submission of the data.

[3] Data submitted in support of an application for registration
prior to January 1, 1970, or data for which exclusive use or com-
pensated use periods described above have expired, may be freely
considered by the EPA in support of any other applications.29

In every case in which a follow-on applicant is required to pay
compensation, he is required to send the original data-submitter
an offer to pay for use of the data.30 If at the end of ninety days
after the date of delivery of the offer there is no agreement on the
amount and terms of compensation, nor on a procedure for
reaching such an agreement, then either party may initiate arbi-
tration proceedings by requesting the Federal Mediation and
Conciliation Service to appoint an arbitrator. The procedures
and rules of the Service apply to the proceedings and to the selec-
tion of an arbitrator.3 1

The findings and determinations of the arbitrator are "final and
conclusive, and no official or court of the United States shall have
power or jurisdiction to review any such findings and determina-
tions, except for fraud, misrepresentation, or other misconduct
by one of the parties to the arbitration or the arbitrator .... ."32

Action on the registrant's application is not to be delayed during
the fixing of compensation through arbitration. If the Adminis-
trator of the EPA determines that either the original data submit-

28. Id. Note that the "reasonable compensation" language of the 1972 data-licensing
provisions was changed to "compensate."

29. See FIFRA § 3(c)(l)(D)(iii), 7 U.S.C. § 136a(c)(l)(D)(iii) (1982).
30. FIFRA § 3(c)(l)(D)(ii), 7 U.S.C. § 136a(c)(l)(D)(ii) (1982). Under current regula-

tions, he may also file such an offer directly with the EPA to avoid the necessity of having
to make an individual offer of compensation. 40 C.F.R. § 152.86 (1985).

31. Id. The Federal Mediation and Conciliation Service has published interim final reg-
ulations which establish that the roster of commercial arbitrators maintained by the Ameri-
can Arbitration Association (AAA) will be utilized, and the FIFRA arbitration rules of the
AAA will be followed for the arbitration of pesticide compensation disputes. 45 Fed. Reg.
28105 (1980).

32. Id.

19861



200 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 11:193

ter or applicant failed to participate in the negotiation or
arbitration, or failed to comply with an agreement or arbitration
decision, then he shall order the original submitter to forfeit his
right to compensation or "deny the application or cancel the re-
gistration of the pesticide in support of which data were used
without further hearing."33 He must, however, furnish fifteen
days notice of his intent to act by certified mail.

The 1978 amendments also provide procedures for joint devel-
opment and the sharing of costs related to "defensive data" - i.e.
additional test data required by the EPA to maintain in effect an
existing registration of a pesticide. 34 Upon notice from the EPA
that additional test data is required, a registrant has ninety days to
either furnish evidence to the EPA that appropriate steps are be-
ing taken to secure the additional data, or agree with one or more
registrant's to either jointly develop or share in the cost of devel-
oping the additional data. 35 If the parties cannot agree on the
details of a joint development or cost sharing arrangement within
sixty days after notifying EPA of their intent to cooperate, then
either registrant may initiate arbitration proceedings by request-
ing the Federal Mediation and Conciliation Service to appoint an
arbitrator. 36 Again, the procedures and rules of the service apply
to the proceedings and to the selection of an arbitrator; and the
findings and determinations of the arbitrator are "final and con-
clusive" and not subject to judicial review except in the cases of
fraud, misrepresentation or misconduct. 37

The trade secret provisions of FIFRA were also amended by
providing for the disclosure of all health, safety and environmen-
tal data, notwithstanding the prohibition against the disclosure of
"trade secrets," provided that the use of such data is subject to
the licensing scheme outlined above.3 s A strict prohibition re-
mains on the disclosure of information that reveals "manufactur-
ing or quality control processes" or certain details related to
added inert ingredients unless the EPA Administrator determines

33. Id.
34. FIFRA § 3(c)(2)(B), 7 U.S.C. § 136a(c)(2)(B) (1982).
35. FIFRA § 3(c)(2)(B)(ii), 7 U.S.C. § 136a(c)(2)(B)(ii) (1982).
36. FIFRA § 3(c)(2)(B)(iii), 7 U.S.C. § 136a(c)(2)(B)(iii) (1982).
37. Id.
38. FIFRA § 10(d), 7 U.S.C. § 136h(d) (1982).



FIFRA Compensation

that "disclosure is necessary, to protect against an unreasonable
risk of injury to health or the environment.- 3 9

C. Judicial Action Since 1978: The Monsanto and Union Carbide
Decisions

Effective operation of the 1978 data-licensing scheme has also
been blocked by litigation. The mandatory license was repeatedly
challenged by original data-submitters as an unconstitutional
"taking" of property without just compensation in violation of the
Fifth Amendment, and the arbitration provisions were also chal-
lenged as an impermissible intrusion on the power of the judici-
ary in violation of article 111.40

After many years of litigation, both of these issues were re-
solved by the Supreme Court in the course of two decisions, ren-
dered, respectively, in 1984 and 1985. In June 1984, the
Supreme Court held in Ruckelshaus v. Monsanto Co.4 1 that the
mandatory license does not result in an uncompensated taking in
violation of the Fifth Amendment. In July, 1985, the Court held
in Thomas v. Union Carbide Agricultural Products Co. 42 that the re-
quirement for binding arbitraton does not violate the require-
ments of article III relating to proper judicial tribunals.

In resolving the Fifth Amendment issue, the Court in Monsanto
held that with respect to data submitted prior to October 22,
1972 and after September 30, 1978, the mandatory license estab-
lished by the 1978 amendments to FIFRA did not constitute a
"taking" of property. However, the Court ruled that the opera-
tion of the data-licensing scheme "may effect a taking with respect
to certain health, safety, and environmental data constituting
trade secrets under state law and designated . . . as trade secrets

39. Id.
40. Mobay Chemical Corp. v. Costle, 517 F. Supp. 252, 254 (W.D.Pa. 1981), afd sub

noma. Mobay Chemical Corp. v. Gorsuch, 682 F.2d 419 (3d Cir.), cert. denied, 459 U.S. 988
(1982) (no taking in violation of the fifth amendment); Chevron Chemical Co. v. Costle,
499 F. Supp. 732 (D.Del. 1980), afd, 641 F.2d 104 (3d Cir.), cert. denied, 452 U.S. 961
(1981) (no taking in violation of the fifth amendment); Union Carbide Agricultural Prod-
ucts Co. v. Ruckelshaus, 571 F. Supp. 117 (S.D.N.Y. 1983), vacated and remanded, 104 S. Ct.
3566, afd on rehearing, No. 76 Civ. 2913 (R) (S.D.N.Y. Sept. 4, 1984), rev'd sub. nom.
Thomas v. Union Carbide Agricultural Products Co., - U.S. -, 105 S. Ct. 3325 (1985)
(district court held arbitration provisions unconstitutionally restricted access to judicial
review in violation of Article III); Petrolite Corp. v. EPA, 519 F. Supp. 966 (D.D.C. 1981)
(no taking in violation of the fifth amendment).

41. 104 S. Ct. 2862 (1984).
42. 105 S. Ct. 3325 (1985).

1986]



202 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

upon submission to the EPA between October 22, 1972 and Sep-
tember 30, 1978. '" 43 The Court further held that if such a taking
occurs it is for a public purpose, and a Tucker Act remedy is avail-
able to original data submitters to provide just compensation for
the use of their data. 44

Challenges to the constitutionality of the arbitration scheme es-
tablished by the 1978 amendments were held to be not ripe for
review by the Court. Monsanto could not demonstrate actual in-
jury because an arbitration had not yet taken place under the stat-
ute.45 However, operation of the arbitration scheme was held to
be a precondition to a claim for compensation under the Tucker
Act.

One year later, the Court resolved the issues it had not reached
in Monsanto relating to the arbitration provisions in Thomas v.
Union Carbide Agricultural Products Co. 4 6 The challenge to the arbi-
tration provisions on article III grounds was held to be ripe for
review because, in contrast to Monsanto, actual arbitration had
taken place to fix the amount of compensation an applicant
should pay for the use of previously submitted data. 47 In addi-
tion, the EPA now had procedures in place through which appli-
cants could meet the requirements of the mandatory license, and

43. 104 S. Ct. at 2882.
44. When the United States takes property for a public purpose, the injured party is

able to seek just compensation under the Tucker Act, 28 U.S.C. § 1491 (1982).
45. 104 S. Ct. at 2882, citing Duke Power Co. v. Carolina Environmental Study Group,

Inc., 438 U.S. 59 (1978); Regional Rail Reorganization Act Cases, 419 U.S. 102 (1974).
46. 105 S. Ct. 3325 (1985). Initially, a federal district court for the Southern District of

New York held that the arbitration provisions were an unconstitutional assignment ofjudi-
cial power in violation of article III. Union Carbide Agricultural Products Co. v. Ruckel-
shaus, 571 F. Supp. 117 (S.D.N.Y. 1983). The district court, rather than strike down the
arbitration provisions alone, enjoined the entire FIFRA data use and compensation
scheme. The Supreme Court vacated this judgment, and remanded it back to the lower
court for reconsideration in light of the Monsanto decision. Ruckelshaus v. Union Carbide
Agricultural Products Co., 104 S. Ct. 3566 (1984). On remand, the district court rein-
stated its prior judgment enjoining the operation of the data-licensing scheme as violative
of article III. Union Carbide Agricultural Products Co. v. Ruckelshaus, No. 79 Civ. 2913
(RO) (S.D.N.Y. Sept. 4, 1984). In October 1984, the Supreme Court granted EPA's re-
quest for a stay of the injunction pending disposition of the appeal of the case. Ruckel-
shaus v. Union Carbide Agricultural Products Co., 105 S. Ct. 237 (1984).

47. 105 S. Ct. at 3333. One of the appellants, Stauffer Chemical Co., had engaged in an
arbitration lasting months. For a discussion of the award granted by the arbitrators, see
infra, text accompanying notes 110-38.



1986] FIFRA Compensation

numerous registrations had been issued under those
procedures .48

In upholding the constitutionality of the arbitration provisions,
the Court held that FIFRA does not confer "private rights" which
require adjudication or review by an article III court, but instead
confers "public rights." 49 The Court rejected the argument that
litigation between private parties necessarily involves private
claims to be decided by article III courts, and held that the public
nature of the remedies defines the claim as asserting essentially
public rights.50 The Court further emphasized that the arbitra-
tion provisions have their own system of internal sanctions and
rely "only tangentially, if at all, on the judicial branch for enforce-
ment."5' The danger of encroachment on the article III powers is
minimized in such a case. Although cases such as this involve the
government'.s administration of its own programs, the presence of
public rights does not depend on the federal government as a
party to the action.52

The Court left open the question of whether a private party
could initiate an action in court to enforce a FIFRA arbitration,

48. In August 1984, the EPA promulgated regulations that implemented the data-li-
censing scheme. 49 Fed. Reg. 30,884 (1984), enacting 40 C.F.R. Part 152 (1985).

Prior to August 1984, EPA's attempts to implement mandatory licensing were frustrated
by court decisions. Following the enactment of the 1978 amendments, EPA issued an
interim final rule known as their "cite-all" regulations. 44 Fed. Reg. 27,932 (1979). Re-
gardless of whether an applicant had developed his own test data sufficient to meet EPA
registration standards, the "cite-all" regulations required an applicant to rely on all infor-
mation in EPA's files relevant to an evaluation of his product, and to compensate prior
applicants accordingly. InJune 1982, the Third Circuit declared the "cite-all" regulations
invalid because the EPA had failed to follow proper notice and comment procedures re-
quired by the Administrative Procedure Act. See Mobay Chemical Corp. v. Gorsuch, 682
F.2d 419 (3d Cir. 1982).

In December 1982, EPA reproposed its 1979 "cite-all" regulations essentially un-
changed, with a proper 60 day comment period. 47 Fed. Reg. 57,624 (1982). However, in
January 1983, a district court for the District of Columbia invalidated the "cite-all" regula-
tions because they were inconsistent with options given to applicants by FIFRA. National
Agricultural Chemicals Association v. EPA, 554 F. Supp. 1209 (D.D.C. 1983). While EPA
was formulating new rules in response to this decision, the lower court in Monsanto ruled
that the mandatory data-licensing scheme in FIFRA was unconstitutional, and enjoined the
EPA from implementing it in any way. Monsanto Co. v. Acting Administrator, EPA, 564 F.
Supp. 552 (E.D.Mo. 1983). For a discussion of Monsanto, see supra, text accompanying
notes 41-45.

49. 105 S. Ct. at 3335-37. In reaching this holding, the Court distinguished Northern
Pipeline Construction Co. v. Marathon Pipeline Co. 458 U.S. 50 (1982).

50. 105 S. Ct. at 3338.
51. Id.
52. 105 S. Ct. at 3336.



204 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

although it indicated that other provisions of law raise that possi-

bility. Finally, the court noted that the arbitration scheme does

not completely preclude review of the arbitration proceeding- a
provision is included for judicial review of cases involving fraud,
misrepresentation, or misconduct. 53 As a result, the Court con-

cluded that such review preserves the "appropriate exercise of
the judicial function." 5 4

In Union Carbide, the Court also left unresolved the issue of

whether FIFRA's "standard for compensation is so vague as to be
an unconstitutional delegation of legislative powers" in violation
of article I.5 5 In their briefs, appellees emphasised that the lan-
guage of FIFRA does not provide guidance on what standard to
use in determining "compensation", and further argued that
nothing appears in "FIFRA or its legislative history to define this
right or to establish a standard for its determination.- 5 6 In sup-

port of their contention, appellees cited two district court

opinions, 57 comments made by the Federal Mediation and Concil-
iation Service in the federal register,58 and statements made by
former EPA administrator Douglas M. Costle found in the legisla-
tive history of the 1978 amendments. 59

However, the Court refused to rule on this question because it

had not been adequately briefed or argued, and the issue was left
open for determination on remand. 60 In dicta, the Court cau-

tioned that "[a] term that appears vague on its face may derive

53. 105 S. Ct. at 3339. See FIFRA § 3(c)(l)(D)(iii), 7 U.S.C. § 136a(c)(1)(D)(iii) (1982).
54. 105 S. Ct. at 3339, quoting Crowell v. Benson, 285 U.S. 22, 54 (1932).
55. 105 S. Ct. at 3339 (citing A.L.A. Schecter Poultry Corp. v. United States, 295 U.S.

495 (1935)).
56. Brief for Appellees Union Carbide Agricultural Products Co. at 28-30, Thomas v.

Union Carbide Agricultural Products Co., 105 S. Ct. 3325 (1985).
57. Id. (citing Sathon, Inc. v. American Arbitration Association, No. 84 C 6019, slip op.

at 7 (N.D. II. Mar. 30, 1984), appeal dismissed, No. 84-1540 (7th Cir. July 3, 1984) ("there is
nothing in the statute (or the regulations promulgated thereunder) relating to the stan-
dard to be applied in such proceedings"); Monsanto Co. v. Acting Administrator, EPA,
564 F. Supp. 552, 561, 567 (E.D. Mo. 1983), vacated on other grounds, Ruckelshaus v. Mon-

santo Co., 104 S. Ct. 2862 (1984) ("there are no discernible guidelines outlining the fac-
tors which constitute just compensation for use of [a submitter's] data")).

58. Id., citing 45 Fed. Reg. 55,394 (1980) ("neither the House or Senate specified a
formula or other guidance on the valuation of data for compensation purposes") and 45
Fed. Reg. 28,107 (1980) ("[t]he statutory scheme of FIFRA provides that the arbitrators
will determine the standards on a case-by-case basis...").

59. Id., citing H.R. REP. No. 663, 95th Cong., 1st Sess. 59 (1977) (testimony of Douglas
M. Costle). See discussion of the context of Mr. Costle's remarks infra, text accompanying
notes 90-92.

60. 105 S. Ct. at 3339-40.



1986] FIFRA Compensation

much meaningful content from the purpose of the Act, its factual
background, and the statutory context." 6 1 The Court further
stated that "[a]lthough FIFRA's language does not impose an ex-
plicit standard, the legislative history of the 1972 and 1978
amendments is far from silent" on the question of FIFRA's stan-
dard for compensation. 62 Section two will analyze the legislative
histories of both the 1972 and 1978 amendments to FIFRA in an
attempt to discover what they say on this issue.

II. LEGISLATIVE HISTORY OF 1972 AND 1978
AMENDMENTS TO FIFRA

A. Legislative History of the 1972 Amendments: The Federal
Environmental Pesticide Control Act of 1972

The mandatory data-licensing provisions of the 1972 amend-
ments to FIFRA63 emerged as a compromise between differing
versions of the original bill (H.R. 10729) supported alternatively
by the Senate Committee on Agriculture and Forestry and the
Senate Committee on Commerce. 64 The Committee on Agricul-
ture and Forestry supported a version of the bill that prohibited
data submitted in support of an application for pesticide registra-

61. Id., quoting American Power & Light Co. v. SEC, 329 U.S. 90, 104 (1946).
62. 105 S. Ct. at 3340, citing S. CONF. REP. No. 1188, 95th Cong., 2nd Sess. 18 (1978);

S. REP. No. 334, 95th Cong., 1st Sess. 4,8,31 (1977); S. REP. No. 838, 92nd Cong., 1st
Sess. 69, 72-73 (1972); Hearings on Extending and Amending FIFRA before the Subcommittee on
Department Investigations, Oversight, and Research of the House Committee on Agriculture, 95th
Cong., 1st Sess. passim (1977).

63. FEPCA, Pub. L. No. 92-516, § 3(c)(1)(D), 86 Stat. 973 (1972). See supra, discussion
of the 1972 amendment's mandatory data-licensing scheme at text accompanying notes
12-22.

64. On November 9, 1971, the House passed a clean bill, H.R. 10729. This bill was
referred to the Senate Committee on Agriculture and Forestry on November 19, 1971.
On December 16, 1971, the Chairman of the Senate Committee on Commerce wrote to
the Chairman of the Committee on Agriculture and Forestry asking for rereferral of the
final bill to the Committee on Commerce if and when the Committee on Agriculture and
Forestry chose to report a pesticide bill. The Chairman of the Committee on Agriculture
and Forestry agreed to rereferral to the Committee on Commerce. After extensive hear-
ings, on June 7, 1972, the Committee on Agriculture and Forestry ordered that H.R.
10729, the Federal Environmental Pesticide Control Act of 1972, be both reported to the
Senate and referred to the Committee on Commerce. The Committee on Commerce held
hearings on June 15 and 19, 1972, and reported their version of the bill to the Senate on
July 19, 1972. The Committee on Commerce recommended sixty-five amendments to the
Committee on Agriculture and Forestry's version of the bill that were divided into fifteen
groups, including the amendments to data-licensing. See S. REP. No. 838 (Pt. II), 92nd
Cong., 2nd Sess. (1972), reprinted in 1972 U.S. CODE CONG. & AD. NEws 3993, 4023-28
[hereinafter referred to as "1972 S. REP. No. 838, reprinted at _-].



206 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

tion from being considered by the EPA in support of subsequent

applications without the consent of the original data-submitter. 65

This would have given original data submitters "exclusive use" of

their data without restriction. The Committee on Commerce sup-
ported an amendment that would strike the exclusive use provi-
sion from the bill, and allow the EPA to freely consider such data

in support of subsequent applications. 66 This would have given

follow-on registrants "free use" of previously submitted data in
support of their registrations.

The battle for an exclusive use provision was led by the Na-

tional Agricultural Chemical Association (NACA) - a trade asso-

ciation primarily representing large chemical companies that

develop and manufacture pesticides. The essence of NACA's ar-
gument was that "no one is going to invest substantial amounts in

research to meet. . . requirements for data if competitive compa-

nies can thereafter use that data to register and sell in competi-

tion the same product." 6 7  Proponents of exclusive use

consistently argued that its purpose was to give manufacturers
"an incentive to undertake the research necessary to develop bet-

ter and safer pesticides." 68 NACA also emphasized that exclusive
use did not directly or indirectly extend patent protections to origi-
nal data submitters because subsequent applicants always had the

option to register and market a product based on their own test
data. 6 9 While acknowledging that duplication of test data might
result, NACA argued that such replication would be beneficial to

65. See 1972 S. REP. No. 838, reprinted at 4034-35.
66. See S. REP. No. 970, 92nd Cong., 2d Sess. (1972), reprinted in 1972 U.S. CODE CONG.

& AD. NEws 3993, 4096 [hereinafter referred to as "1972 S. REP. No. 970, reprinted at _").
The legislative history suggests that "as a matter of practice but without statutory author-
ity," the EPA considered submitted test data "to support the registration of the same or
similar product by another registrant" prior to the 1972 amendments. See S. REP. No. 838,
reprinted at 4040. This view was reiterated by the Supreme Court in Monsanto, 104 S. Ct. at
2877.

67. 1972 S. REP. No. 838, reprinted at 4035.
68. Id. at 4034.
69. Id. at 4037-41. Under both the 1972 and current version of FIFRA, normal seven-

teen year patent protections run concurrently with the compensated use or exclusive use
protections provided by mandatory licensing. However in the pesticide industry, once a
potential commercial pesticide candidate is discovered it usually takes between fourteen
and twenty-two years to complete the development process. Monsanto, 104 S. Ct. at 2871.
Because of the unusually long time that it takes to bring a pesticide product to market,
patent law provides little protection for the pesticide developer.



FIFRA Compensation

the goal of establishing the safety of a pesticide and its effect on
the environment. 70

The Committee on Commerce was opposed to the exclusive
use provisions for two main reasons. First, it emphasized that ex-
clusive use would create "barriers to entry in the pesticide indus-
try" far greater than those provided by patent law which would
have an anticompetitive effect on the industry. 7' Second, the
Committee on Commerce argued that by requiring manufactur-
ers to invest resources in costly duplicate testing, the exclusive
use provisions stifled the very incentive to invest in new research
and development which it ostensibly was designed to en-
courage. 72 The Committee on Commerce was supported in its
views on the anticompetitive effect of exclusive use by the Attor-
ney General, who advised Congress that instead of furthering
"the remedial purposes of the bill," the provision "serves rather
to insulate the first applicant for registration from the competi-
tion of later applicants." 73 The EPA also stated that "the effect of
this provision is to afford additional economic protection, foster
monopoly, and it may tend to restrict pesticide business to large
manufacturers. "74

The data-licensing scheme in the 1972 amendments emerged
as a compromise position between the exclusive use and free use
poles of this debate. Congress demonstrated its desire to en-
courage market entry and competition in the pesticide industry by
denying exclusive use protection to original data-submitters. In
addition, by granting original data-submitters the right to "rea-
sonable compensation," Congress recognised that allowing the
free use of test data might dampen the incentive to undertake new
research and development.

It is inconsistent with the market entry and competitive goals of
this compromise to argue that reasonable compensation should
be measured by the value of the test data to a follow-on regis-
trant. If the opportunity costs that are avoided by follow-on reg-

70. Id. at 4041.
71. 1972 S. REP. No. 970, reprinted at 4096-4103.
72. Id. at 4096.
73. Id. at 4096-98 (April 28, 1972 letter from Acting Attorney General Richard G.

Kliendienst). The Attorney General's views were shared by two law professors who were
invited to comment on the exclusive use provision. Id. at 4098-4103 (Feb. 21, 1972 letter
from John G. Stedman, Professor of Law at the University of Wisconsin, and Feb. 9, 1972
letter from John J. Flynn, Professor of Law at the University of Utah).

74. 1972 S. REP. No. 838, reprinted at 4043.

1986]



208 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

istrants are included when calculating reasonable compensation,
then, effectively, the original data-submitter is granted the
equivalent of exclusive use protection for the use of his data. Fol-
low-on registrants would be required to pay compensation costs
equivalent to the costs they would incur if they were prevented
from relying on previously submitted data by an exclusive use
provision and forced to develop the test data themselves. In re-
jecting NACA's exclusive use proposals, Congress intended to fa-
cilitate market entry for smaller chemical manufacturers and
avoid duplicate testing; while a value-based standard might avoid
duplicate testing to a degree, it would unquestionably defeat the
competitive considerations that led to the creation of the
mandatory data-licensing scheme. Therefore, Congress intended
that reasonable compensation be based on an equitable sharing
of the actual costs of developing test data.

This conclusion, drawn from an analysis of the general pur-
poses of the data-licensing scheme in the 1972 amendments, is
supported by specific language in the legislative history. The
Committee on Agriculture and Forestry, in responding to the ob-
jection that an exclusive use provision would unduly extend pat-
ent protections, stated:

If the product is not patentable or if the patent protection has
expired, there is nothing to prevent a competitor from register-
ing a similar product. Under such circumstances, the first appli-
cant has no opportunity to recover his research costs and little
incentive for undertaking that research. 75

Further on, in response to objections that exclusive use would
lead to duplicate testing, the Committee stated:

The Committee on Agriculture and Forestry does not believe
that there will be any great diversion of funds to duplicate test-
ing. Rather this provision is likely to result in equitable sharing of
research costs. There would be little reason for duplicate testing if
a second registrant could share in the test data of the first by
paying part of the CoSt.

7 6

Finally, in a section entitled "Legislative intent with respect to
section 3 (c)(1)(D)," Senate Report No. 838 states:

Thus it was decided that fairness and equity require a sharing
of the governmentally required cost of producing the test data used
in support of an application other than the originator of such

75. Id. at 4034 (emphasis supplied).
76. Id. (emphasis supplied).



FIFRA Compensation

data. If no agreement can be reached, the Administrator is
vested with authority to determine the reasonable share of the
cost of the test data used, including subsequent reallocations upon
requests for use of such data by additional applicants. 77

As the Court in Union Carbide suggests, the legislative history of
the 1972 amendments is not silent on the issue of FIFRA's stan-
dard for compensation. It indicates that Congress intended "rea-
sonable compensation" to be calculated as an equitable share of
the actual cost of developing test data. 78 A value-based standard
that includes the opportunity costs incurred by the original data-
submitter would conflict with both Congress's general purpose in
creating a mandatory data-licensing scheme in FIFRA, and spe-
cific language in the amendment's legislative history.

B. Legislative History of the 1978 Amendments: The Federal Pesticide
Act of 1978

The legislative history of the 1978 amendments is more com-
plex than that of the 1972 amendments; however, many of the
same parties pitched battle over many of the same policy consid-
erations and proposals debated in 1972. The current data-licens-
ing scheme finally emerged as a compromise crafted in a
Conference Committee convened to resolve differences between
a House amendment and the final Senate bill. 79 Before examin-
ing this final compromise, it is important to look at the contours
of the debate which led to this impasse between the two houses of
Congress.

Mirroring its stance in 1972, NACA initially proposed that
mandatory licensing be eliminated, and that original data-submit-

77. Id. at 4092 (emphasis supplied).

78. This conclusion is shared by EPA administrative law judge Harwood who ruled in
the only two decisions rendered under the 1972 data-licensing scheme that Congress in-
tended follow-on registrants to pay only a share of the actual cost of producing the test
data. Judge Harwood rejected proposals that original data-submitters be awarded royalty-
like compensation that would include the opportunity costs avoided by follow-on regis-
trants. See Ciba-Geigy Corp. v. Farmland Industries, Inc., Nos. 33, 34 & 41 (Aug. 19,
1980), Final Order issued (April 30, 1981), affirmed by the Administrator (July 28, 1981);
See also Union Carbide Agricultural Products Co. v. Thompson-Hayward Chemical Co.,
Dkt No. 27 (July 13, 1982).

79. See H. CONF. REP. No. 1560, 95th Cong., 2nd Sess. (1978), reprinted in 1978 U.S.
CoDE CONG. & AD. NEws 2043; S. CONF. REP. No. 1188, 95th Cong., 2nd Sess. (1978); See
also 124 CONG. REC. 29756 (1978) (statement of Sen. Leahy).

1986]



210 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 11: 193

ters be granted twelve year exclusive use protection.8" NACA
again argued that "it is illogical to expect that a company will risk
the investment in research and testing if the fruits of that re-
search, if successful, are to be immediately available to competi-
tors who have no obligations to support the failures." 8' In
addition, NACA argued for eliminating mandatory licensing be-
cause the "determination of 'reasonable compensation' would be
an administrative nightmare." 8 2

At first, NACA's proposals for exclusive use protection were
opposed by the Pesticide Formulator's Association (PFA) 8s - a
trade association representing small-to-medium sized pesticide
manufacturers and formulators. In essence, the PFA argued that
the 1972 data-licensing scheme could work, provided that Con-
gress removed the confusion over the definition of "trade secret"
and established guidelines for determining "reasonable compen-
sation." The PFA suggested guidelines that would establish a
"cost basis" for compensation based on one and one-half (1 '/2)

times the original cost or "reproduction costs," whichever was
less.8 4 Compensation would take place over five years and be ap-
portioned according to a market share comparison.

In a last minute compromise, NACA and the PFAjoined in sup-
port of a proposal to create a ten year exclusive use period for
submitted test data. The PFA agreed to this compromise in part
because it included a provision that allowed for the equitable
sharing of the cost of developing "defensive data" - i.e. data de-
manded by the EPA to maintain ("defend") the registration of a
previously registered pesticide.8 5 NACA's willingness to equita-
bly share the costs of developing defensive data was considered to
be an important concession by the PFA because the research and
development costs of the industry as a whole had significantly in-
creased in prior years as a result of more stringent EPA regula-

80. Extending and Amending FIFRA: Hearings Before the Subcommittee on Department Investiga-

tions, Oversight and Research of the House Committee on Agriculture, 95th Cong., 1st Sess. 187,
198 (1977) [hereinafter cited as 1977 House Hearings].

81. Id. at 199.

82. Id. at 198.
83. Id. at 231-47.
84. Id. at 236-37.
85. Id. at 514-23. See also Extension of the Federal Insecticide, Fungicide, and Rodenticide Act:

Hearings Before the Subcommittee on Agricultural Research and General Legislation of the Senate Com-

mittee on Agriculture, Nutrition and Forestry, 95th Cong., 1st Sess. 3-6, 69-73 (1977) [hereinaf-
ter cited as 1977 Senate Hearings].



FIFRA Compensation

tions pertaining to previously registered pesticides. The
compromise proposed that disputes over the sharing of defensive
data costs be submitted to binding arbitration outside of the
EPA.

86

The EPA supported a proposal that would keep the 1972 data-
licensing scheme intact, but exclude health and safety data from
protection as "trade secrets." Both the EPA and the Department
of Justice opposed the industry compromise for the same market
entry and anti-monopoly considerations which led them to fight
"exclusive use" in 1972.87 The EPA emphasized that:

Exclusive use and data compensation are alternative means of
protecting the economic interests of the developer of registra-
tion data but each has a different impact on competition within
the industry and ease of entry into the market. . . .[e]xclusive
use of data will reduce competition and the right to entry in the
market as contrasted by the Administration's bill [that provides
for data compensation]. 88

In a series of letters to Congress, the Department of Justice
agreed with this conclusion. 89

Both the EPA and the Department of Justice commented on
FIFRA's standard for compensation during these early hearings.
EPA Administrator Costle testified that his agency "lacked exper-
tise" in this area and was "uncomfortable in the role of judge as
to the economic value of the data in question." Consequently, he
urged Congress to "make more explicit what factors it feels are
pertinent in determining reasonable compensation.- 90 Taken out
of context, this testimony has been relied on in support of the
contention that Congress failed to provide arbitrators with a stan-
dard for determining compensation under the 1978 amendments
to FIFRA.9' This argument conveniently ignores the fact that Mr.
Costle made his comments before Congress had begun serious

86. See 1977 House Hearings at 518-19, 522-23; 1977 Senate Hearings at 72-73.
87. See 1977 Senate Hearings at 148-50, 271-74. See also Economic Impacts of ProposedAmend-

ments to FIFRA, EPA Ofc. of Pesticide Programs (June 20, 1977), reprinted in S. REP. No.
334, 95th Cong., 1st Sess. 58-68 (1977).

88. 1977 Senate Hearings at 148-49.
89. S. REP. No. 334, 95th Cong., 1st Sess., 89-103 (1977) (June 8, 1977 letter from Asst.

Attorney General Patricia M. Wald opposing initial NACA proposals, and June 15, 1977
letter from Asst. Attorney General Wald opposing the "industry compromise.").

90. H. REP. No. 663, 95th Cong., 1st Sess. 59 (1977), reprinted in 1978 U.S. CODE CONG.

& AD. NEWS 1998, 2032. [hereinafter cited as 1977 H. REP. No. 663].
91. Brief for Appellees Union Carbide Agricultural Products Co. at 28-29, Thomas v.

Union Carbide Agricultural Products Co., 105 S. Ct. 3325 (1985).

1986]



212 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

consideration of the 1978 amendments, and refers only to EPA's
limited experience under the 1972 data licensing scheme. How-
ever, his testimony does help explain why Congress later chose to
delegate the responsibility of determining compensation to an ar-
bitrator with experience in the pesticide industry, rather than the
EPA.

9 2

More importantly, Assistant Attorney General Patricia M. Wald,
in a letter to Congress on behalf of the Department of Justice,
emphasized the connection between the standard used in deter-
mining compensation and the competitive purposes of the Act's
mandatory data-licensing scheme. She warned that:

[t]he compensation provisions could be used to effect anti-com-
petitive ends in dealing with both competing manufacturers of
technical pesticides and formulators of end-use products from
the technical materials. There are inherent problems in deter-
mining what constitutes 'reasonable' compensation for data.
Thus, the provision could provide a tool for extending patent
monopoly by pricing data out of reach of small competing man-
ufacturers and for exercising control of the end-use and/or
end-users after initial sale of the patented material. We con-
sider both types of action inconsistent with competitive
policy.

93

Her comments highlight the danger involved in adopting a stan-
dard for compensation that exceeds an equitable sharing of the
actual cost of developing test data. The application of a value-
based standard that includes the opportunity costs avoided by fol-
low-on registrants when they gain immediate market entry runs
the risk of pricing small manufacturers out of competition.

Following the recommendation of its subcommittee, the Senate
Committee on Agriculture adopted a bill (S. 1678) that did not
fully accept either the industry compromise or the EPA's propos-
als, but clearly favored the Administration's position. 9 4 The
Committee eliminated the "trade secret" protection accorded
health and safety data under the 1972 amendments by giving
"trade secrets" a narrower definition, and rejected industry's ar-
guments for "exclusive use" protection. 95 However, original data
submitters were given the right to compensation for seven

92. 1977 H. REP. No. 663, reprinted at 2016.
93. S. REP. No. 334, 95th Cong., 1st Sess. 91 (1977) (June 8, 1977 letter from Asst.

Attorney General Patricia M. Wald).
94. Id. at 7-8, 17-19. See also 123 CONG. REC. 25706-710 (1977).
95. Id. at 8.



FIFRA Compensation

years.9 6 If the parties could not agree on compensation, then
they were to submit to arbitration under procedures almost iden-
tical to those provided by present law. 97 With respect to "defen-
sive data," joint development or cost sharing arrangements were
permitted and an arbitration provision was included to resolve
any disputes.98 On the question of a standard for compensation,
the Committee simply stated that "[t]he subcommittee agreed
that the amount and terms of reasonable compensation would be
determined by the parties involved- the original submitter of
data and the later applicant who wishes to rely on it. . . ."99 In
keeping with the market entry and competitive purposes of
mandatory licensing, the Bill's floor manager, Senator Leahy,
stated that the compensation "mechanism protects the data de-
veloper's right to recover his data generation costs while guarantee-
ing small companies entry to the market and protecting them
against an unfair competitive situation."' 100

The House passed a bill (H.R. 8681) that was closer to the in-
dustry compromise than the Administration's proposals.' 0 ' In a
crucial markup session of the Subcommittee on Department In-
vestigations, Oversight, and Research, a proposal to provide for a
ten year period of compensated use was rejected in favor of in-
dustry's ten year "exclusive use" proposals by a roll call vote of 6
to 5.102 However, concessions on the trade secrets issue led to a
compromise proposal that provided for five years of "exclusive
use" protection to be followed by five years of compensation for
the use of data by follow-on registrants. 0 3 "Defensive data" was
accorded the same protections as original data.' 0 4 If the parties
could not agree on compensation, then they were to submit to
arbitration under procedures similar to present law. 10 5 There
were minor differences between the House and Senate bill's lan-
guage with respect to the procedures to be used in the case of
arbitration.

96. Id.
97. Id. at 8, 17-18.
98. Id. at 8, 18-19.
99. Id. at 8.
100. 123 CONG. REC. 25706 (emphasis supplied).

101. 1977 H. REP. No. 663, reprinted at 1996-98.
102. Id. reprinted at 2013-16.
103. Id. reprinted at 2015.
104. Id.
105. Id. reprinted at 2016.

1986]



214 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

The final version of the 1978 amendment's data-licensing
scheme emerged as a compromise drafted by the Conference
Committee formed to resolve the differences between the House
and Senate. The Conference Committee granted ten year "exclu-
sive use" protection to data submitted after the enactment of the
1978 amendments. 0 6 In introducing the compromise, Senator
Leahy, a member of the Conference Committee, emphasised that:

[bly limiting exclusive use to data pertaining to pesticides in-
gredients not previously registered, and only for a 10 year pe-
riod, the conferees have largely confined exclusive use
coverage to chemicals that are patentable and for a term gener-
ally shorter than the patent life enjoyed by pesticides. The an-
ticompetitive aspects of exclusive use have therefore been
essentially neutralized. 10

7

The Conference Committee granted 15 years of compensation
for the use of data submitted between the enactment of the 1972
and 1978 amendments, and the use of "defensive data."' 0 8 Since
patent protections had already expired on most of the data sub-
mitted between 1972 and 1978, exclusive use protection would
have had a chilling effect on market entry and competition, and
effectively extended patent protections afforded to original data
submitters. The extension of exclusive use protection to the sub-
mitters of defensive data would also have resulted in wasteful du-
plication and discouraged joint development ventures. In
addition, providing defensive data with exclusive use protection
would have taken away the concession won by the PFA in return

for their support of NACA's exclusive use proposals.
Therefore, the Conference Committee enacted a data-licensing

scheme that would encourage innovation and competition, dis-
courage duplication, and assist the small pesticide manufacturer
and formulator. In the words of Senator Leahy:

The Conference substitute on the data issue is an appropriate
compromise between the need to encourage innovation and
the need to provide an equitable and less expensive process for
producing pesticide safety data. I believe it recognises the
needs of both large and small firms. 10 9

106. FIFRA § 3(c)(1)(D)(i), 7 U.S.C. § 136a(c)(1)(D)(i) (1982). See text accompanying
note 26, supra.

107. 124 CONG. REC. 29756 (1978).

108. FIFRA § 3(c)(I)(D)(ii), 7 U.S.C. § 136a(c)(l)(D)(ii) (1982) and FIFRA § 3(c)(2)(B),
7 U.S.C. § 136a(c)(2)(B) (1982). See supra. text accompanying notes 28, 35-37.

109. Id.



FIFRA Compensation

Ten years of exclusive use protection is accorded to data submit-
ted after the enactment of the 1978 amendments in order to en-
courage future research and development. As a balance to the
exclusive use protections won by industry for newly submitted
data, the compensation provisions were enacted to facilitate mar-
ket entry and encourage competition.

A value-based standard for determining compensation that in-
cludes the opportunity costs avoided by follow-on registrants
defeats the competitive purposes of the compensation provisions
in the 1978 amendment's data-licensing scheme. The Depart-
ment of Justice's prescient concerns over the potential anticom-
petitive effects of the compensation provisions will become a
reality if arbitrators choose to adopt a value-based standard. The
legislative histories of both the 1972 and 1978 amendments to
FIFRA indicate that Congress intended "compensation" to be
calculated as an equitable share of the actual cost of developing
test data.

Within this cost sharing framework, the question remains as to
what constitutes an equitable sharing of the cost of developing
test data. Should testing costs be divided on a market share basis,
per capita basis, or some other method?

Arguably, the competitive goals of the compensation provisions
favor the choice of a market share division of costs over a per
capita division. However, this reasoning ignores the fact that
these goals are substantially realized through the benefits which
follow-on registrants receive from avoiding the opportunity costs
suffered by the original data submitters. Once small manufactur-
ers and formulators are given genuine access to the market, the
division of the costs of test data on a per capita basis rather than a
market share basis would have a minimal impact on the competi-
tive structure of the pesticide industry.

In conclusion, the legislative histories of both the 1972 and
1978 amendments to FIFRA indicate that Congress intended that
compensation be based on an equitable sharing of the costs of
developing test data rather than on the value of the data to a fol-
low-on registrant. However, the legislative histories are silent on
the issue of how to equitably divide testing costs.

1986]



216 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 11:193

III. ARBITRATION DECISIONS UNDER THE 1978 AMENDMENTS:

THE STAUFFER AND FMC DECISIONS

At present, only two arbitrator's decisions have been issued
under the mandatory data-licensing scheme enacted by the 1978
amendments to FIFRA: Stauffer Chemical Co. v. PPG Industries,
Inc. I10 and FMC Corp. v. Ticon International.II While both cases
address the issue of a standard for determining compensation,
each is concerned with a different aspect of this question because
they were brought under different provisions of FIFRA.

This section will analyze the Stauffer decision and demonstrate
that its choice of a value-based standard for determining compen-
sation is inconsistent with the intent of Congress as revealed in
the legislative histories of the amendments to FIFRA. In addi-
tion, it will briefly discuss the FMC decision and outline some of
the useful suggestions it makes with respect to the equitable divi-
sion of the costs of test data.

A. The Stauffer Decision

The Stauffer decision was rendered in an arbitration brought
under section 3 (c) (1) (D) (ii) of FIFRA and relates to compensation
for the use of submitted test data by a follow-on registrant.1 12

The underlying issue in Stauffer is whether compensation should
include the value of the test data to the follow-on registrant or be

110. Stauffer Chemical Co. v. PPG Industries, Inc., No. 16 199 077 82, Fed. Mediation
and Conciliation Service (June 28, 1983) (Birch et al., Arb.) [hereinafter cited as Stauffer].

Shortly after the arbitrator's decision was issued, PPG filed an action against both Stauf-
fer and the EPA in federal district court for the District of Columbia to set aside the award.
PPG Industries, Inc. v. Stauffer Chemical Co., No. 83-1941 (D.D.C. filed July 7, 1983). In
its complaint, PPG challenged the FIFRA data-licensing scheme as an unconstitutional vio-
lation of article III, and in the alternative, argued that the arbitrator's were guilty of "mis-
conduct" under the statute (7 U.S.C. § 136a(c)(l)(D)(ii)) because their grant of value-
based compensation was ultra vires, i.e. outside the scope of their authority as vested by
Congress. Stauffer cross-claimed against the EPA also seeking to have the data-licensing
scheme invalidated as violative of article III, and counterclaimed against PPG seeking
damages in the amount of the award should the statute be struck down, or, enforcement of
the award. Both parties article III challenges were resolved by the Supreme Court in its
Union Carbide decision. See supra note 40. However, PPG's claim that the arbitrator's deci-
sion constituted "misconduct" is still being litigated before the federal district court for
the District of Columbia. The specific issues raised by that claim are outside the scope of
this note.

11. FMC Corp. v. Tricon International, No. 16 199 0033 84G, American Arbitration
Association (Jan. 10, 1985) (Foy et al., Arb.) [hereinafter cited as FMC].

112. FIFRA § 3(c)(l)(D)(ii), 7 U.S.C. § 136a(c)(l)(D)(ii) (1982). Seesupra text accompa-
nying note 28.



FIFRA Compensation

based on the actual cost of developing the test data. An analysis
of the Stauffer decision is particularly important because, contrary
to the intent of Congress, the arbitrators used a value-based stan-
dard in awarding compensation.

The Stauffer decision, issued on June 28, 1983, was the first to
be issued in an arbitration brought under the data-licensing
scheme enacted by the 1978 amendments. In part, the existence
of the decision eliminated barriers to standing in Union Carbide,
and enabled the Supreme Court to uphold the constitutionality of
FIFRA's arbitration provisions.' 13

The Stauffer Chemical Co. ("Stauffer") is the originator of a
corn herbicide known generically as "butylate."' 1 4 Stauffer ob-
tained a basic United States patent for butylate in 1959; conse-
quently, its patent protection expired in 1976.115 Stauffer first
registered butylate with the U.S. Department of Agriculture in
1968,116 and has maintained sales of the product since that time.
Following the enactment of the 1978 amendments, PPG Indus-
tries, Inc. ("PPG") applied to the EPA for the registration of bu-
tylate of its own manufacture, and relied on test data originally
submitted by Stauffer after December 31, 1969 in support of its
registration.' 17 In compliance with FIFRA, PPG made an offer to
compensate Stauffer for the use of the data; when the parties
could not agree on an amount of compensation, Stauffer initiated
an arbitration under the Act." 18

After months of testimony consuming 2700 pages of transcript,
an arbitration tribunal appointed by the Federal Mediation and
Conciliation Service issued an unanimous decision. The arbitra-
tors granted Stauffer a compensation award reflecting "a fair
share of both the cost of testing required by current regulations
and of the 'opportunity costs' that PPG avoided not only by avoid-
ing the normal regulatory delay but also by its comprehensive
copying of Stauffer's approved labels." ' 19

In calculating the award to Stauffer, the arbitrators first deter-
mined that $2,930,000 (in 1983 dollars) was the actual cost Stauf-

113. Thomas v. Union Carbide Agricultural Products Co., 105 S. Ct. at 3333.
114. Stauffer, supra note 105, at par. 11.
115. Id. at par. 11, n.6.
116. Id. at par. 11.

117. Id. at par. 12.
118. Id. at par. 13.
119. Id. at par. 18.

19861



218 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

fer incurred in developing the test data which PPG used in
support of its own registration. 20 This figure excludes the cost
of data submitted by Stauffer before 1970, as well as data from
numerous tests "found to be substantially duplicative of other
submitted data and therefore not required in support of PPG's
registration."'12 ' However, the figure does include the cost of
pre-1970 data resubmitted after 1969 that were not substantially
duplicative of other data submitted. Using $2,930,000 as the ba-
sis, the arbitrators awarded Stauffer one half of these costs-
$1,465,000 - to represent a "fair share" of the testing costs.' 2 2

This per capita division of testing costs was made without discuss-
ing alternative methods of division.

In addition, the arbitrators granted Stauffer compensation
based on the value of the data to PPG. They held that:

PPG has. . . obtained an important economic benefit in being
able to start marketing its butylate some five years earlier than
it could have without reliance on Stauffer's data. By contrast,
Stauffer incurred a corrresponding opportunity cost or loss of
profits when it had to delay its marketing of butylate . . . For
this cost avoidance or benefit of early market entry PPG like-
wise owes Stauffer compensation. 12 3

However, because of the speculative nature of this value-based
compensation, the arbitrators determined that it "can best and
most reasonably be expressed as a fraction of the sales actually
made and to be made by PPG over the next several years."' 2 4

As a result, by extrapolating from assumptions concerning
PPG's prospective market share and profits related to sales of bu-
tylate, the arbitrators established a formula for determining run-
ning compensation to be paid to Stauffer for a period of ten years
based on PPG's actual future sales. This formula is described in
the decision as follows:

Running Compensation (R) for every pound of butylate manu-
factured or otherwise acquired by PPG in the United States in
the ten calendar years from 1983 through 1992, both inclusive,
and not discarded but sold or otherwise transferred by PPG for
use by third parties, which R shall be computed according to
the formula

120. Id. at par. 20.

121. Id.

122. Id. at par. 21.
123. Id. at par. 14.
124. Id.



FIFRA Compensation

R= $0.15/lb. X Y/I
Wherein Y is the Producer Price Index for Finished Goods
published by the U.S. Bureau of Labor Statistics for the month
of November preceding the year in which the respective sale or
other transfer is made, and wherein I is such Producer Price
Index published for November 1982, with the proviso, however,
that the total number of pounds of butylate on which this com-
pensation R is due in any year after 1987 shall in no event ex-
ceed the highest total number of pounds of butylate on which
such compensation was due in any of the calendar years 1983
through 1987, both inclusive. 12 5

Based on the arbitrator's calculation that PPG's butylate sales in
the first five years should total approximately 47,250,000
pounds, 12 6 and assuming steady growth in sales for the following
five years, PPG can expect to pay Stauffer approximately 14 mil-
lion dollars plus inflation costs over ten years in addition to the
lump sum of $1,465,000 discussed earlier.

Inexplicably, the arbitrators granted Stauffer this windfall with-
out even mentioning the legislative histories of the amendments
to FIFRA, and after a cursory and incomplete discussion of the
purpose of the compensation provisions.' 2 7 In their decision, the
arbitrators suggest that apart from a desire to encourage the de-
velopment of new pesticides, the purpose of data-licensing is the
"avoidance of wasteful use of scientific manpower in duplicative
testing and governmental review thereof."' 28 Further on, they
tacitly acknowledge the competitive purposes of data-licensing by
claiming that the prevention of the manpower waste inherent in
duplicative testing "directly encourages competition and lower
pesticide costs to the public .... 129

While one purpose of data-licensing is to avoid duplicative test-
ing, the Act, as well as its legislative history, does not indicate that
it is only concerned with the waste of manpower; to the contrary,
the presence of the joint development and cost sharing elements
of the defensive data provisions demonstrates Congress' desire to
avoid the economic waste inherent in duplicative testing as
well. 130 In addition, one reason Congress rejected exclusive use

125. Id. at "Compensation Award," par. B.
126. Id. at par. 22.
127. Id. at pars. 4 & 7.
128. Id. at par. 4.
129. Id. at par. 7.
130. FIFRA § 3(c)(2)(B), 7 U.S.C. § 136a(c)(2)(B) (1982).

1986]



220 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 11:193

protection in favor of compensation in 1972 was their concern
that the economic costs of duplicate testing would stifle the ability
of smaller companies to invest in the new research that the data-
licensing scheme was designed to encourage. 13 1

More importantly, the arbitrator's claim that competition and
lower pesticide costs results solely from preventing the waste of
manpower inherent in duplicative testing is without substance. A
look at the legislative histories of the amendments reveals that
Congress recognized that the key to compensation and lower
costs was the removal of barriers to entry into the pesticide mar-
ket. 132 As a result, in 1978 they rejected exclusive use protection
for data submitted between 1970 and 1978 in favor of the com-
pensation provisions in an attempt to facilitate market entry for
follow-on registrants.133 If these registrants are now required to
pay a value-based compensation that reflects the opportunity
costs which they avoid as a result of immediate registration, then
the barriers to market entry have been rebuilt. Under such cir-
cumstances, the availability of scientific manpower is of little
consequence.

Why did the arbitrators grant Stauffer a windfall? They may
have been unduly prejudiced by two factors specific to this case.
First, PPG is one of the largest chemical companies in the U.S.;134

they are not representative of the small manufacturer or formula-
tor whose competitive position Congress sought to enhance by
creating the compensation provisions. However, the size of a
specific follow-on registrant should be irrelevant to an award of
compensation. An administrative nightmare would result if arbi-
trators are asked to link the amount of compensation awarded to
the specific competitive position of a follow-on registrant.

Second, when PPG registered butylate they not only relied on
Stauffer's submitted test data but also copied Stauffer's approved
labels.' 35 The arbitrators acknowledge that Stauffer failed to
copyright their labels and, consequently, they were "in the public
domain for anyone to copy."' 3 6 However, the arbitrators cite

131. 1972 S. REP. No. 970, supra n. 66 reprinted at 4096.
132. See supra text accompanying notes 93-109.
133. See supra text accompanying notes 106-09.
134. In 1985, PPG Industries was ranked by Fortune Magazine as the 89th largest U.S.

industrial corporation with sales of approximately 4.2 billion dollars. Vol. 111, No. 9 For-
tune 268 (April 29, 1985).

135. Stauffer, supra note 105, at pars. 15-16.
136. Id. at par. 15.



FIFRA Compensation

what they term as PPG's "slavish" copying as support for their
decision to award Stauffer value-based compensation. 3 7 Even if
the arbitrators believe that PPG's decision to copy Stauffer's un-
protected labels was unethical, it was avowedly legal and not a
basis for awarding any amount of compensation.

The arbitrator's decision to grant Stauffer compensation based
on the value of the data to PPG is contrary to the competitive
goals of the data-licensing scheme. As section II demonstrates,13 8

the intent of Congress to facilitate market entry and encourage
competition will be defeated if follow-on registrants are required
to pay compensation that approximates the costs they would in-
cur if they developed the test data themselves.

B. The FMC Decision

In contrast to Stauffer, the FMC decision was rendered in an ar-
bitration brought under section 3(C)(2)(B) of FIFRA and is con-
cerned with allocating the costs of developing defensive data
between two companies holding registrations of the same pesti-
cide. 13 9 The issue of whether a value-based standard or cost-
based standard should be used in relation to defensive data is
moot because this provision is explicitly concerned with
"shar[ing] in the cost of developing such data."' 40 As a result,
the underlying issue in FMC is whether a market share basis, per
capita basis, or some other method should be used in allocating
the costs of developing the data. While the FMC decision does
not resolve this issue, it does suggest some useful guidelines to be
used in dividing the costs of test data.

The FMC decision is limited in scope. It is an initial decision
requested by the parties to resolve the specific issue of allocating
the costs of developing additional data in support of their regis-
trations of "ethion."' 4 ' FMC argued that the costs should be di-
vided "equally or per capita."' 42 Tricon argued that the costs
should be allocated "on the basis of market share."' 43 In their
decision, the arbitrators did not choose one method of allocating

137. Id. at pars. 15-16.
138. See supra text accompanying notes 63-109.
139. FIFRA § 3(c)(2)(B), 7 U.S.C. § 136a(c)(2)(D) (1982). See supra text accompanying

notes 34-37.
140. FIFRA § 3(c)(2)(B), 7 U.S.C. § 136a(c)(2)(D) (1982).
141. FMC, supra note 111, at 1-2.
142. Id. at 1.
143. Id.

1986]



222 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 11:193

costs over another; instead they suggested certain guidelines con-
cerning the equitable division of the costs of test data. Following
the issuance of the initial decision, the parties reached a
settlement.

The arbtrators began with two basic principles concerning cost
sharing. First, they emphasized that the statute itself does not re-
quire any specific method of allocating costs. 14 4 Second, they
pointed to the legislative history of both the defensive data provi-
sions and section 3(c)(1)(D) in holding that Congress intended
that the costs be shared in some way. 14 5 From this foundation,
they made the following findings.

First, both interest costs and inflation should be included in de-
termining data development costs. 146 Second, if market share is
chosen as a factor in determining cost allocation, then it should
be based on a party's prospective market share "over a period of
time that is at least as great as the period of time over which the
costs to be allocated will be amortized under prevailing industry
standards."' 14 7 A company's past market share might be decep-
tive because companies are capable of transfering their registra-
tions. Third, if a market share allocation is adopted, the
arbitrators determined that consideration should be given to the
volume of a company's sales both in the U.S. and overseas in
countries where a U.S. registration is sufficient to meet their regu-
latory requirements.

148

The FMC decision refuses to make a choice between a per cap-
ita and market share division of the costs of test data. To the
contrary, the arbitrators conclude by stating:

The per capita and market shares formulae strike us as only the
initial negotiating positions of two parties who are seeking an
agreement on sharing any type of cost. We don't think either
formulae is likely to be the formula finally agreed on by two
willing parties having very different shares of the market. More-
over, we are not satisfied that per capita and market share are
the only bases for sharing costs and would like to explore other
factors that might be considered by two willing parties who
were seeking an agreement on cost sharing. 149

144. Id. at 4.
145. Id. at 4-6.
146. Id. at 2, 6.
147. Id.
148. Id. at 2, 7.
149. Id. at 9-10.



FIFRA Compensation

The ambiguity of the FMC initial decision is reflective of the
legislative histories of the amendments. Section II pointed out
that within a cost sharing framework, the legislative histories are
silent on the issue of how to equitably divide testing costs. 150

Both Congress and the FMC decision correctly recognize that this
issue is best left to the reasonable determination of the parties
involved.

CONCLUSION

In the current debate, the follow-on registrants are correct in
asserting that compensation should be based on an equitable
sharing of the actual costs of developing test data. Congress en-
acted the compensation provisions in an attempt to encourage
competition without discouraging innovation in the pesticide in-
dustry. Value-based compensation rebuilds the barriers to mar-
ket entry and competition which Congress tried to tear down
when it chose compensation over exclusive use protection for
data submitted between 1970 and 1978. In Stauffer, the arbitra-
tors failed to adequately consider the purposes of the compensa-
tion provisions and were prejudiced by factors irrelevant to the
issue before them; consequently, they erred in awarding value-
based compensation. The arbitrators in FMC correctly recog-
nized that within a cost sharing framework, the task of equitably
dividing testing costs is left to the discretion of the arbitrators and
parties involved, while also suggesting some useful guidelines for
dividing testing costs. Most importantly, arbitrators should heed
the Supreme Court's warning that Congress was "far from si-
lent"'15 1 on the question of a standard for compensation. In re-
sponse they should award cost-based compensation consistent
with the purposes of FIFRA's data-licensing scheme.

Geoffrey H. Coil

150. See supra text accompanying notes 112-13.
151. Thomas v. Union Carbide Agricultural Products Co., 105 S. Ct. at 3340.

1986]