Are Administrative Patent Judges Unconstitutional? A Possible Dissenting View:
My friend and colleague John Duffy has drawn a lot of attention in the blogosphere and in the media for his essay concluding that the way of appointing administrative patent judges is unconstitutional. According to Duffy, the problem is that administrative patent judges of the Board of Patent Appeals and Interferences (BPAI) are appointed by the Director of the Patent and Trademark Office (PTO), whereas under the Appointments Clause they must be appointed by a cabinet officer because they exercise "significant authority pursuant to the laws of the United States . . and must, therefore, be appointed in the manner prescribed by [the Appointments Clause]." Buckley v. Valeo, 424 U.S. 1, 126 (1976). Duffy's argument has drawn a lot of attention, including possible legislative fixes and a pending cert petition.
But I'm not sure Duffy's view is right, and in this post I want to explain why.
My source of doubt is that I'm not so sure that BPAI judges exercize "significant authority" for Appointments Clause purposes. I am no expert in the Appointments Clause, but Duffy's article relies heavily on the analogy between BPAI judges and administrative decisionmakers. Because executive branch judges that rule on administrative claims have been ruled to be covered by the Appointments Clause, so should BPAI judges. This assumption seems to rely on a specific notion of how patent law fits into administrative law: Specifically, executive branch adjudications of patent claims are basically like executive branch adjudications of licenses, welfare benefits, and the like.
But does that analogy hold? I don't think it does. The reason is that patent law takes its operating principles from contract law rather than administrative law. As I argued a few years ago in Rethinking Patent Law in the Administrative State, 42 William & Mary Law Review 127 (2000), which I have just posted to SSRN, modern patent law predates the regulatory state and is based on the model of a unilateral contract offer rather than the exercise of administrative discretion.
I go into detail on this in the article, but here's the basic idea. The patent laws announce a contract offer: Anyone who can create an invention that satisfies the standards of patentability and agrees to disclose the invention to the public in a patent application becomes entitled to the quid pro quo of the patent grant. The filing of a patent application is an attempt to accept the offer. The government's ruling on the patent application reflects the judgment of the offeror as to whether the offer has in fact been been accepted and a contract exists. If the contract exists, the offeror (that is, the government) then bestows the quid pro quo of the property right of a patent. If the offeror concludes that no acceptance occurred and thus no contract exists, then the offeror will award no quid pro quo and the offeree (the applicant) must go to court to sue the offeror for breach of contract (failure to confer the patent).
With this understanding, a BPAI judge's job is conceptually very different from the job of other decisionmakers in the administrative state. The BPAI judge is a representative of the offeror, and the job is to determine if the applicant's conduct satisfied the contract. Although the position is a type of judgeship, the ultimate "decision" is essentially that of a contracting party as to whether a contract was accepted.
This is different from an administrative judge because administrative judges are working within a zone of delegated power. The basic notion of modern administrative law is that executive agencies are given substantive rulemaking power to "do the right thing" within a zone of delegated discretion; the administrative law judge has "significant authority" because he or she exercises that delegated power. But there is no similar delegated power in patent law. The modern patent law system predates the administrative state by several decades, and it is based on a contractual model instead of the later delegated power model. Given that, I'm not entirely sure that BPAI judges exercize "significant authority," and I'm not sure their method of appointment is unconstitutional.
My case is complicated by the fact that in the last few decades, the ossification of administrative law models of agency action has made it common to conceive of all decisions by agencies as the same. The fact that patent law followed different principles has been clouded. As I explain in detail in my William & Mary article, this has led to decisions both by the Supreme Court and the Federal Circuit that have tended to try to push patent law into more of a traditional administrative law model. Professor Duffy relies on these cases in support of his suggestion that the appointment of BPAI judges is unconstitutional by analogy to administrative judge cases. Given these recent cases, one response would be that whatever historical distinction separated patent law from regulatory executive action, the new cases have narrowed that difference enough that it no longer matters.
Perhaps that's right. But at the same time, I think that a great deal of the historical distinction remains. It includes the standards of review for legal issues, which require Chevron deference for regulatory decisions but de novo review for constructions of the requirements of patentability. And given the very high stakes involved in this issue, I think a court should at least want to look more closely at this issue before concluding that the BPAI judges are covered by the Appointments Clause.
But I'm not sure Duffy's view is right, and in this post I want to explain why.
My source of doubt is that I'm not so sure that BPAI judges exercize "significant authority" for Appointments Clause purposes. I am no expert in the Appointments Clause, but Duffy's article relies heavily on the analogy between BPAI judges and administrative decisionmakers. Because executive branch judges that rule on administrative claims have been ruled to be covered by the Appointments Clause, so should BPAI judges. This assumption seems to rely on a specific notion of how patent law fits into administrative law: Specifically, executive branch adjudications of patent claims are basically like executive branch adjudications of licenses, welfare benefits, and the like.
But does that analogy hold? I don't think it does. The reason is that patent law takes its operating principles from contract law rather than administrative law. As I argued a few years ago in Rethinking Patent Law in the Administrative State, 42 William & Mary Law Review 127 (2000), which I have just posted to SSRN, modern patent law predates the regulatory state and is based on the model of a unilateral contract offer rather than the exercise of administrative discretion.
I go into detail on this in the article, but here's the basic idea. The patent laws announce a contract offer: Anyone who can create an invention that satisfies the standards of patentability and agrees to disclose the invention to the public in a patent application becomes entitled to the quid pro quo of the patent grant. The filing of a patent application is an attempt to accept the offer. The government's ruling on the patent application reflects the judgment of the offeror as to whether the offer has in fact been been accepted and a contract exists. If the contract exists, the offeror (that is, the government) then bestows the quid pro quo of the property right of a patent. If the offeror concludes that no acceptance occurred and thus no contract exists, then the offeror will award no quid pro quo and the offeree (the applicant) must go to court to sue the offeror for breach of contract (failure to confer the patent).
With this understanding, a BPAI judge's job is conceptually very different from the job of other decisionmakers in the administrative state. The BPAI judge is a representative of the offeror, and the job is to determine if the applicant's conduct satisfied the contract. Although the position is a type of judgeship, the ultimate "decision" is essentially that of a contracting party as to whether a contract was accepted.
This is different from an administrative judge because administrative judges are working within a zone of delegated power. The basic notion of modern administrative law is that executive agencies are given substantive rulemaking power to "do the right thing" within a zone of delegated discretion; the administrative law judge has "significant authority" because he or she exercises that delegated power. But there is no similar delegated power in patent law. The modern patent law system predates the administrative state by several decades, and it is based on a contractual model instead of the later delegated power model. Given that, I'm not entirely sure that BPAI judges exercize "significant authority," and I'm not sure their method of appointment is unconstitutional.
My case is complicated by the fact that in the last few decades, the ossification of administrative law models of agency action has made it common to conceive of all decisions by agencies as the same. The fact that patent law followed different principles has been clouded. As I explain in detail in my William & Mary article, this has led to decisions both by the Supreme Court and the Federal Circuit that have tended to try to push patent law into more of a traditional administrative law model. Professor Duffy relies on these cases in support of his suggestion that the appointment of BPAI judges is unconstitutional by analogy to administrative judge cases. Given these recent cases, one response would be that whatever historical distinction separated patent law from regulatory executive action, the new cases have narrowed that difference enough that it no longer matters.
Perhaps that's right. But at the same time, I think that a great deal of the historical distinction remains. It includes the standards of review for legal issues, which require Chevron deference for regulatory decisions but de novo review for constructions of the requirements of patentability. And given the very high stakes involved in this issue, I think a court should at least want to look more closely at this issue before concluding that the BPAI judges are covered by the Appointments Clause.
Is a Tax Court judge really given discretion to "do the right thing" any more than a BPAI judge?
Is too much of a stretch to say that in this sense, the PTO is functioning more like a state government than like other administrative agencies?
State governments charter corporations, &c.
If so, you've got an interesting piece of jurisdictional trivia to support your argument. The PTO is one of the few agencies that effectively preempts state licensing authorities -- you can practice patent law before the PTO without a state bar license.
The only other example of that is for the IRS, which permits CPAs to appear in tax court. Perhaps the IRS and tax appeals could be analogized to the contract system you're talking about as well?
...what would it take to transform all administrative agencies into this form of contract referree rather than paternalistic nanny I wonder?
The theory of patents as corporations has obvious affinities with the theory of patents as contracts.
.
I believe that as a technicality, Duffy is right. "First to invent" is a statutory right, not a rules-based right, and interference practice is aimed to learn who was first to invent.
.
The unilateral contract model is okay, as far as it goes, but patent law is remarkably different from administrative contract letting (see Boeing v. Airbus).
.
As a mater of statute, a person is entitled to a patent if (list of condition snipped). It's a wide open field - not inviting an offer to perform a particular act (the traditional unilateral contract model), but encouraging full public disclosure in exchange for a time-limited right to exclude.
The origin of the first-to-invent rule is not so clearcut as you might imagine. It was adopted by Jefferson and two other cabinet members in deciding the steam boat interference shortly after the first patent act in 1790, which was silent on the question.
Whether that cuts for against Duffy I will leave for Prof. Kerr to say.
I'm curious, why should the fact that it's a wide-open field make a difference? Let's say I have an apple orchard and I make an offer: "If anyone goes to my orchard and picks apples, handing them in to me, at the end I'll weigh the apples and give you X dollars a pound that you picked." If someone goes to the orchard, picks apples, and then hands them to me, don't I still have a contractual obligation to pay them X dollars a pound?
.
My comment was origination-agnostic. Interference practice has a pure statutory basis and outcome. The BPAI judges are rulings directly on a statute, not on an administrative rule.
If you read the article, I think it provides the answer.
.
Of course you do. And I see the parallel that you are trying to draw. But let me take your example and change it ever so slightly.
.
Let's say I have an apple orchard and I make an offer: "If anyone invents a device or method that can be used to make life more pleasant, predictable or efficient, I make no guarantee as to the commercial success but I will publish the invention to the public at my own expense."
.
Do you disclose your invention? Why, or why not? The trick is to make the carrot attractive enough to obtain disclosure, but without a promise of reward - only a promise of a right to enforce exclusion of others.
.
The unilateral contract of patent law is remarkably different from your example. No promise of commercial success or reward. Just a promise of a statutory right to enforce exclusion.
.
Congress, as a matter of statute, sets the parameters for the grant of patent protection. It did not cede to the USPTO the power to decide the parameters for a patent grant.
Ah, I agree that the quid pro quo is different: my point is only that the legal mechanism is (I think) the same.
If we want to analogize to federal contracts, BPAI has a role far more akin to the Federal Court of Claims than to the initial hearings of contractual disputes by the agency one has or wants a federal contract with. Indeed these courts are equally signfiicant in the judicial pecking order -- appeals from both the BPAI and the Federal Court of Claims go to the Federal Circuit. So if we take the contracts analogy seriously, Professor Duffy isn't going far enough -- BPAI judges like Court of Claims judges ought to be appointed by the President and confirmed by the Senate! Your argument leads in the opposite direction of where you want to go.
Another point in favor of patent disputes being more significant than most regulatory disputes is that the power and duty of the federal government to issue patents is specified in the Constitution itself.
BTW, the contract analogy as you stated it doesn't work. What offeror, after acceptance, is allowed to revoke the offer? That is no contract law that I know. A better mapping from patent law to a contractual metaphor is that patent law is simply a general advertisement inviting offers: it does not refer to a specific invention with specific claims. The patent applicant makes a specific offer to share information in exchange for a grant by submitting the application with its claims demarking the technological "territory" the applicant wishes to be granted. The Patent Office has a legal duty to accept all reasonable offers. Here is where the contract analogy breaks down: the BPAI is not adjudicating a breach of contract, it is adjudicating a special law requiring the federal government to accept reasonable "offers": perhaps analogous to adjudicating a dispute over who wins a federal government contract put out for bid, but even that is stretch.
Even with this fix, there is still a big problem:
BPAI judge is a representative of the offeror, and the job is to determine if the applicant's conduct satisfied the contract.
Since when does the defense lawyer also get to be the judge? Speaking of the old common law, this is a gross violation of the common law maxim forbidding one to be a judge in one's own case. A fix for this problem is the very same one Professor Duffy says the Constitution requires -- to have the BPAI judges appointed by a Cabinet official less closely associated with the Patent Office.
How does that happen in patent law? The offeror can change his mind that the contract was accepted and take back the quid pro quo, but that doesn't revoke the offer.
Since when does the defense lawyer also get to be the judge?
Again, how does that happen in patent law? If you think I owe you five bucks and I disagree and refuse to give it to you, I am not a "judge" of the dispute. I suppose I could declare my self a judge, put on a robe, and write an opinion explaining why I think I do not owe you $5. But I still don't think that makes me a judge.
.
Both the performance and the quid pro quo are uncertainties.
.
The performance is the inventor's whim, and the quid pro quo, while definite as to enforcing a right, is indefinite as to delivering an outcome.
.
Patent law is remarkably different from unilateral contract; although they share the function of enticing and encouraging all comers to attempt to obtain the promised benefit.
.
Your error is to view the offer as narrow, or as presented by the government. The offer in a patent grant is defined by the inventor. The inventor is charged with describing, with particularity, the property right that the "offeror" will protect.
You don't have much of an analogy to contract law if one party, having entered into the "contract", is allowed with no consequences to "change his mind" and breach the contract, because he decided he didn't like the terms after all.
If you think I owe you five bucks and I disagree and refuse to give it to you, I am not a "judge" of the dispute.
The BPAI has far more power of remedy than merely declining to grant a patent. Under reexamination and interference practice, for example, it can invalidate claims or entire patents already granted. The BPAI are called judges and considered by all concerned to be such. Even you are calling them judges. They render verdicts, have built up a very important body of legal precedent, the works. Indeed, just as with the decisions of of the judges of the Federal Court of Claims, the decisions of the BPAI judges are appealed to the Court of Appeals for the Federal Circuit. The Federal Circuit is a Court of Appeals, not a Court of Original Jurisdiction. Indeed the statute granting jurisdiction to the Federal Circuit speaks of "an appeal from a decision of...the Board of Patent Appeals and Interferences", not of original jurisdiction.
Under 35 USC 6(a), the members of the BPAI are three bigwigs and a group of "administrative patent judges." This rebuts two claims put forth by Professor Kerr: (1) that they are not judges, and (2) that their nature is not administrative.
Like a court, the BPAI exercises jurisdiction. It declares that "We have jurisdiction under 35 U.S.C. § 6(b)." (source)
It is quite thoroughly the case that, where the BPAI panel is chosen by the PTO itself, the PTO is being a judge in its own case and violating a core legal principle. This problem arises whether or not the disputes are "administrative" or "contractual" in nature. Given that the Court uses the Appointments Clause to address just this kind of problem with regulatory agency adjudication, it certainly should also cover the same problem with the BPAI.
I think you're missing the dynamic here. The PTO is just a contracting party, not an Article III. And if they "change their mind," they are only changing their mind as to whether there was a contract in the first place: The remedy is taking away the patent because the patent was never owed and was granted by mistake.
All this would have been much easier to understand before the Fregeau case in 1985, when the Federal Circuit erroneously took away the traditional de novo review of patent claims that used to exist in federal district court. See the dissent in Fregeau, discussed in the article.
TVK,
That is only true as to the facts, and traditionally that was not even true as to the facts: the patent applicant could file a de novo action in the Federal District Court and the BPAI would get no deference at all. Exactly like a breach of contract case.
The Federal Circuit messed this up in 1985 in the Fregeau case, which I discuss at length in the article. (That's the Fdeeral Circuit case I was referring to in the post.)
To be clear, I am not claiming that BPAI judges are not called "administrative judges" in the Patent Act. But I am not sure why you think the label is determinative of anything.
In other words, the BPAI are judges wielding most of the coercive powers of an Article III court. Certainly far more powers than your man-on-the-street simply refusing to pay the $5 he doesn't think he owes.
.
Attempts to shoehorn patent law into the contract model styled as a unilateral contract are flawed. To the extent your conclusion depends on a contract analogy, it too is flawed.
.
The PTO administers statutory law that prescribes subject matter and conditions for the granting of a right to exclude. The "contracting party" is the public, as against the inventor.
But self-help is not available to a contracting party simply because it changed its mind as to the validity of a contract. Self-help is only available in limited circumstances where the other party has breached a contract that included a security interest (e.g. "repo men"). Under your analogy, the patentee has not breached any contract. To overturn a contract's validity and thereby recover property that one has already granted under the contract, one must go before a real court with a real judge. If the judge is one of the contracting parties he must recuse himself -- to not do so would be a gross breach of legal ethics.
Similarly, when federal land is sold to a private citizen, the federal government cannnot simply change its mind and show up one day to take the land back from the grantee. If it thinks the grant was illegal, it must AFAIK prove its case to an Article III court to get the land back.
.
Contract law is premised on definition of an offer and a reward. Patent law defines each of "offer" and "reward" on an abstract level, one layer removed from the business action.
But isn't the substantive right at issue more important than the format of the trial? If you think I owe you $5 because you think we had a contract, and I agree to let you put on witnesses pursuant to the Federal Rules of Evidence to prove your case, and I then decide, after the trial, that I don't owe you the $5, that is still just a question of contract law. Or so it seems to me.
cbolt,
I disagree, for all the reasons I set out in my article.
.
I assume that this is an expression of disagreement with my overarching opinion that patent law doesn't fit in the unilateral contract model.
.
I've made my counterpoints in a casual fashion. I also said, in opening, that I predicted the outcome would favor "your side" (even if it's wrong in principle and logic). And so, we disagree, but "so what."
Orin queries: Under what body of law?
UCC sections 2 and 9.
No. If the judge is also my opponent's lawyer, I'm going to focus my efforts on getting him recused and save most of my worrying about the substantive issues for later.
Of the powers being regulated by the Appointments Clause, procedural powers are more important than the kind of substantive jurisdiction. The significance of an official's authority is far better estimated by looking at the procedural powers (of discovery, remedy, etc.) wielded by the official than by looking at the area of substantive law over which the official has authority. Conflicts of interest can arise in any area of substantive law where a federal agency uses its own agents, selected by itself, to judge its own cases, and that is certainly true of any area analogous to contract law.
I think we'll have to agree to disagree. FWIW, though, I'm not sure why the details of the UCC are relevant to these questions.
Right now, I can be reasonably comfortable with my 101, 102, and 112 arguments to the BPAI. But last year's KSR decision has introduced so much variability into the determination of obviousness that it often seems a coin flip whether you will win or lose based on the same set of facts and arguments.
And that is another part of why I am troubled by the contract analogy. In patent law, the rules of the game (or more accurately the interpretation of the laws) is a moving target. You draft claims based on the laws of today, prosecute them under the laws of three, four, or five years from now, and then enforce them under the laws of a couple years after that. Contrast that with contracts, which are primarily enforced based on the understanding of the parties at the time the contract was entered into.
And that doesn't take into account the effect on issued patents. Sure, issued patents have the benefit of a presumption of validity - but how strong would that be if the allowance was based on the findings of an illegally constituted BPAI panel? But then, a rehearing by the BPAI, or a de novo hearing by a trial or appeals court would be under different laws, notably the KSR decision on obviousness mentioned above.
Even in the normal cases of original patent issuance, how are BPAI judges, imposing the patent law determined by Congress and the Federal Courts any different than Tax Court judges, which impose the tax law determined by Congress and the Federal Courts? The PTO has very little regulatory authority, the Tax Court has none.
A problem with basing a characterization of BPAI judges based on a unilateral contract model of Patent law is that courts have increasingly disfavored unilateral contracts. Unilateral contracts nowadays are limited to those offers where acceptance is explicitly limited by performance or a reward/ad with very definite terms of how to be accepted.
The Congress's general "offer" for patents fails to be a valid offer because it is indefinite and illusory. You only have to look to KSR and it's multi-factor "common sense" test to see that the offer of a patent is indefinite (or at least not sufficiently definite to constitute a unilateral offer). It's illusory in the sense that the US Govt is the one extending the offer, and will only promise to perform if in it's own discretion (albeit a different branch), you have satisfied the contract. It's like saying "I'll promise to pay you for a painting if I decide, in my own absolute judgment, I like it, according to several quasi-objective criteria."
Also the legal defenses available for Patent cases don't really jive with the unilateral contract model. Inequitable conduct, which is essentially is fraud before the PTO, will prevent enforcement of the patent and can be raised by the party being sued for infringing the patent. This is a defense to both an injunction and money damages. In a trespass case, you don't have a defense by alleging that the landowner defrauded the party that conveyed the property to him (except possibly an unclean hands defense for only equitable remedies). The seller of land would have a remedy, but not the trespasser. The same is true of the defense of invalidity. It would match with your model if the defense could only be raised in the PTO in a re-exam hearing, but rather they can be raised by the alleged infringer in the infringement case. Although a defect in a deed may prevent a successful trespass action, a breach of the land contract won't create a defense in a 3rd party trespasser if the deed has actually been conveyed. The deed supersedes the contract, but the patent does not supersede your theoretical unilateral contract. Or what about the defense of prosecution history estoppel? It's not like parol evidence from the land contract estopps the landowner from arguing something different in interpreting the deed as to a trespass action.
I don't even want to try to imagine an analogy to patent misuse.
(1) an accused infringer must prove the patent invalid by clear and convincing evidence (e.g., why not a preponderance standard? 35 USC 282 does not require a clear and convincing proof standard; it is judge-made based on the presumption that the PTO has done a competent job substantively analyzing and issuing the patent);
(2) the PTO is entitled to deference as to how they analyzed prior art in conjunction with issuing the claims (theoretically, a litigant has a higher burden to demonstrate invalidity once the PTO has had a reference or patentability argument before it);
(3) the "good faith and fair dealing" presumed for contracts would be something litigants could assert to invalidate a patent regardless of the specific inequitable conduct standard in 37 CFR 1.56;
(4) the Federal Circuit is wrong to limit the scope of what the PTO can look at in the context of reexamination proceedings (see In re Portola, I think); and
(5) it would be consistent with standing issues. If I am an accused infringer having the patent asserted against me, why should I have "standing" to challenge the patent's validity if the PTO and the patentee are happy parties to a contract? Put another way, if I live next door to someone who is charging me with trespassing on his property, do I really have standing to challenge whether the contract he used to buy the property 15 years ago was valid, in connection with a defense to a trespass charge? Wouldn't the court there say he's listed as owner on a properly recorded deed, just as a patentee has to show that he is the owner/assignee, and that ends the inquiry? Could I avoid the standing problem by arguing that the patent "contract" was contrary to the public's interest and/or is unconscionable as an enforcement against me?
Just curious...
Cheers.
If you read the article, I think it will provide answers to your questions.