The proposed Hart-Nelson amendment No. 3 would eliminate the provision of section 3 (c) (1) (D) which would prohibit the use of one applicant's data to support the application of another. The Senate Committee on Agriculture and Forestry did not adopt this portion of the amendment, although it adopted the proviso clause of the amendment that the Administrator may refer to any applicant's data in making a determination of the adequacy of the test data under consideration.

We consider this provision as so amended to be fully justified. The reasonableness of the provision does not depend upon an appraisal of the particular data under consideration.

If it has no competitive advantage to the second applicant who wishes to piggyback, he is not injured by his inability to rely upon it. On the other hand, if it would have a competitive advantage to the second applicant, it, by the same token, has a proprietary value to the registrant who originated the data. He should not be deprived of this proprietary value in the absence of compelling policy considerations. We see no such considerations here.

The second applicant is free to develop such data by making the required investment, or by making contractual arrangements with the original applicant.

The protection under section 10 from disclosure of trade secrets does not obviate the necessity for the protection of the piggyback preclusion of section 3 (c) (1) (D), because even in the absence of disclosure the administrator, in the absence of the preclusion, could give the second applicant the benefit of another's data by merely taking official notice of the data in his files.

I have noticed in the statement that Senator Dominick has prepared for presentation to the committee the reference to this provision. In it he refers to a letter from the Attorney General which in substance opposes this provision and suggests its deletion.

I have reviewed this letter. The letter is signed by Mr. Kleindienst and, I believe, is based upon a misunderstanding or a lack of understanding of the bill as it is written.

In the letter signed by Mr. Kleindienst, he relies upon or refers to two decisions decided by the U.S. Supreme Court in 1964, the Sears, Roebuck Co. v. the Stiffel Co., and Compco v. Day-Brite, and he says in his letter if there should, however, be a genuine trade secret problem, the administrator is otherwise able, of course, to protect trade secrets under section 10 of the bill.

Section 3 (c) (1) (D) in H.R. 10729 would in effect give trade secret or packing-type protection to test data. We believe that that statement does not reflect the circumstances under which the bill will be operating. It is not necessary that the trade secret data be revealed to a second applicant in order for him to benefit by it. The administrator, in the absence of this provision, could merely take official notice of the fact that he has this other data in his file and transpose it then to the file of the second application. By doing so, he would grant registration to the second applicant, upon the basis of trade secrets compiled by the first applicant at a cost of possibly several million dollars in money and several years in time.

We believe the justification for protection from disclosure of trade secrets will not be questioned. However, in the administration of this section, we do foresee substantial areas of disagreement concerning the types of data which are required in support of an application for registration and are entitled to protection as a trade secret.

Trade secrets are protected from disclosure not only by section 10 of H.R. 10729, but also by the provisions of the United States Code, title 5, section 552, the Freedom of Information Act, and title 18, section 1905, which makes it a criminal offense for an employee of the U.S. Government to divulge a trade secret and other specified categories of information without specific statutory authority. Neither of these statutes nor H.R. 10729, however, define a trade secret or provide any meaningful guidelines.

A special committee of the National Agriculture Chemicals Association, consisting of individuals who have the responsibility for research development for their companies, has recently made a study of the types of information which must be submitted in support of a registration and attempted to appraise the categories of information which they would consider under usual industry practices to constitute a trade secret and the types which would not.

The committee, through this study, sought to determine general guidelines based on customary industry practice, but recognized that what constitutes a trade secret in a particular case will rest on the facts of that case.

As a guide in its appraisals, the committee used the definition of a trade secret as set forth in the Restatement of Torts 1939, and in the proposed Uniform Trade Secret Protection Act which currently is being prepared by a committee of the American Bar Association.

These two definitions are set forth as appendixes B and C. The committee and I, as counsel, working jointly from these definitions, developed the following primary criteria as a basis for the appraisal of what constitutes a trade secret:

(1) Did the acquisition of data represent a substantial investment in: (a) time? (b) money?

(2) To what extent have members of the industry sought to impose conditions of secrecy or nondisclosure on such information?

(3) To what extent could the information be obtained by competitors? What would be the time and cost requirements?

(4) Would the knowledge of the information be of advantage to a competitor? If so, how would competitors be able to use the information to advantage?

The committee then appraised against these criteria the different types of data which must be submitted to support an application.

Some of the types of data, when considered on the basis of customary industry practices, were considered to be valid trade secrets. Other types were considered not to constitute trade secrets because of failure to meet one or more criteria when judged on customary industry practices.

In some classifications, it considered that the detailed data normally would constitute a trade secret because of its value to a competitor and the customary practice of the industry in making a special effort to keep such data confidential. At the same time it concluded that a summary of such detailed data would not normally constitute a trade

secret.

Under such an interpretation summaries of data could be made available for public inspection while the detailed data could be protected from disclosure.

This would accord with the procedure which the Food and Drug Administration has proposed for similarly required data. Under the proposed procedure for compliance with the Freedom of Information Act as published in the Federal Register of May 5, 1972, at page 9130, the Food and Drug Administration proposed to treat safety and effectiveness data as trade secrets and to comply at the same time with the requirements of the Freedom of Information Act by releasing summaries of the data. This interpretation is based on the definition of a trade secret in the Restatement of Torts referred to above.

We believe that a similar practice can be followed under H.R. 10729, as it was reported by the Senate Committee on Agriculture and Forestry, and in doing so protect valid trade secrets, and at the same time meet the need of the public to be informed on the nature of the research work which has been conducted, and a summary of the results. Mr. Chairman, in view of the pressure of time and in view of the fact that you have permitted by statement to go into the record as presented, I think that I will just not summarize or read the remainder of it, and leave whatever time remains for questioning by the committee. Senator INOUYE. I thank you very much, sir.

In your prepared statement, Mr. Conner, you have cited two very informative case histories that seek to demonstrate why data submitted in support of the registration should not be used in support of another registration.

Mr. Conner, am I correct in assuming that existing law contains no prohibition against using another registrant's data in support of a registration application, and that this has been done in the past?

Mr. CONNER. Mr. Chairman, that is correct. The present law enacted in 1947 has no provision in it relative to this issue one way or another.

I participated in the drafting of this act in 1949. I also participated in drafting the model State bill which later became the pattern for this Federal act. There was no thought at the time of the enactment of this act in 1947 that the data submitted by one applicant would ever be released for public inspection.

This did not appear to be an issue at that time.

It is since the enactment of the Freedom of Information Act a few years ago requiring the release of data in the files of the Government agencies, except that which constitutes trade secrets, or other specific exemptions, that this issue has become more important.

It is true, however, that even in the absence of the release of the data for publication that as a matter of administrative procedure, the administrators of the act have used data prepared by one applicant in support of another application.

It is the feeling of our industry that this is unfortunate, and that it has had the effect of keeping off of the market a number of pesticides that otherwise would have been placed on the market, because of the fact that they were older pesticides that did not have packing protection, and that under the current standards would have required considerable testing as to their safety.

No one company felt that it was feasible for it to undertake that investment with the knowledge that if it did and subsequently obtained a registration of a nonpatentable product, that any others in the industry then could obtain a similar registration without the necessity of doing independent research.

So as a matter of practice, we feel that this administrative practice which has existed, had the effect of keeping off of the market a number of products that otherwise would have come on the market, either because they were older pesticides that did not have patent protection, or because of the nature of the pesticide, say, a new one, that was not patentable, and under those circumstances the companies did not feel they were warranted in putting it on the market.

Senator INOUYE. But the release of such information has not caused any lack of incentive to develop new pesticides, has it?

Mr. CONNER. Mr. Chairman, there has been no release of information yet under the act, at least to any substance. That is because it is only after the enactment of the Freedom of Information Act that this information became available. EPA published for the first time, in December of last year, its proposal to release some of the information in its files in accordance with requests that had been made for it. It made it clear, however, that it would not release data that had been submitted and was considered to be trade secrets.

There have been no data released, to my knowledge, to date except that which clearly did not constitute trade secrets.

The problem arises here in defining what part of this extensive registration data that must be prepared constitutes a valid trade secret, and what does not.

Some of it clearly does not. But the large part of it clearly does, or at least we believe it clearly does, and the problem to date in the release of the information has been segregating that which does constitute a trade secret from that which is not.

Again, much of the information that had been presented previously was not marked either as trade secret or confidential, because the registrants did not consider that it would ever become an issue, the Freedom of Information Act not having been enacted at that time.

Senator INOUYE. It has been suggested that the prohibition against the use of data by other applicants is a waste of resources, because you would require the second applicant to duplicate work already done, and would thereby reduce his competitive position in the pesticide industry. It has been argued that the money and resources that they would have expended in duplicating these tests could be used for other purposes. Now, the Justice Department in its letter suggested that such an entry fee moreover acts regressively for it is more of a burden for the small manufacturer. What is your response to that?

Mr. CONNER. Mr. Chairman, I believe Dr. Phillipson might either answer that better, or call on some of the personnel from the companies who have the active responsibility in their companies in developing such data, and I believe that they could more appropriately answer that than could I.

Mr. PHILLIPSON. I do not feel that this would be a duplication of effort. Any company that feels that it is worthwhile to make the investment has the option of perhaps buying this information from

85-424-72-10

a company that is already in, and I feel that if you allow free access to this type of data, that you, in essence, are going to discourage the research of new and innovative compounds.

You are going to discourage innovation in already existing compounds, because a company simply cannot afford the risk of doing a significant amount of development and then having anybody have free access to this data.

We can't commit companies' resources to the use of hormones, viruses, steroids, bacteria, or other biological control agents, since these are not patentable, and we cannot explore chemical compounds and reduce them to practice since we would be afforded no protection in the area.

I think from the standpoint of innovation, of new uses of already existing compounds, of perhaps researching in areas of compounds or other types of products with less ecological impact, that this free access would tend to destroy and disturb and certainly make a company reluctant to make the investment in something that anybody would have access to.

I don't feel, also, Mr. Chairman, that it is the type of regulation that is injurious to the small company.

By the very nature of the work that is necessary to take a compound from initial screening to final marketplace and production is a tremendous investment, and these are the types of things that have discouraged small companies rather than regulation that pertains to the exclusive use of data.

Mr. CONNER. Mr. Chairman, could I add this, and again, referring to Mr. Kleindienst's letter that contained the provision that you just quoted.

To me this shows the difference between what he, I believe, was discussing there and the situation with which the industry is faced in this provision.

The two cases to which he referred, the Sears case and the Compo case, were cases in which the companies had developed lighting fixtures of a novel design, and they had obtained a design patent.

Sears, the defendant in those cases, copied those designs. They were sued for infringement of their patent and for unfair competition. The Court held that the patent was invalid. The Supreme Court held that the defendants could not be held liable for merely copying a design of a lighting fixture which was not subject to patent under our patent laws because it was a nonpatentable idea.

Now, that is all the Court held there. Here the equivalent would be if the company, say, the Congress would have said to the two companies that developed the designs, you must make data, you must make available for public inspection any data in your files relating to the safety of these lighting devices, or how they can best or most cheaply be manufactured.

We would be talking about the same thing.

Here all they were talking about was an unpatentable design. Here we are talking about something that, I believe, up until now, has always been recognized as being subject to a valid trade secret that is protected by nondisclosure to competitors, that is, information which a company develops that is of commercial value to it and would be of commercial value to its competitors.