The Duke Law Journal is hosting a symposium, Intellectual Property Exceptionalism in Administrative Law, that raises a question that pops up occasionally both in criminal law and patent law: Why don’t courts apply the usual doctrines of administrative law in these areas? After all, in patent law, the Patent and Trademark Office (PTO) administers the patent laws. And in criminal law, the Department of Justice (DOJ) administers the criminal laws. But courts don’t treat these executive agencies like other agencies. The decisions of other agencies get deferential standards of review, such as Chevron deference. But courts generally reject those deferential standards when it comes to reviewing the decisions of the PTO and the DOJ.


Among scholars, one common view has been to say that this is just a mistake. The PTO and DOJ are agencies, so they should get the same deference that other agencies get. Deferring to agencies would allow smarter decision-making by expert executive agencies instead of by poorly informed courts. (For an example of this argument in patent law, see Jonathan Masur’s Supreme Court Review article, “Regulating Patents”; for an example from criminal law, see Dan Kahan’s Harvard Law Review article, “Is Chevron Relevant to Federal Criminal Law?”) From this perspective, the special treatment of criminal law and patent law is a problem in need of correction.

I have a different view. As I see it, courts generally don’t apply administrative law doctrines when reviewing the decisions of the PTO and DOJ for a good reason: Patent law and criminal law predate the regulatory state, and they work using different mechanisms. The deferential doctrines of administrative law, such as Chevron, were designed to review delegations of lawmaking authority by Congress to regulatory agencies. Deference is necessary to ensure that the power can be delegated. But patent law and criminal law do not rely on delegations of lawmaking authority. As a result, I think it would be a bad idea to apply deferential standards in patent law or criminal law.

Here’s the context, drawn largely from this article. Modern doctrines of administrative law like Chevron are a response to a specific kind of agency authority. Starting mostly in the late 19th century, Congress began to create administrative agencies that were designed to have rulemaking power. Instead of enacting statutory rules in a particular field, Congress would create an agency and tell the agency to go off and regulate, hopefully in some kind of expert way, subject to a broad grant of authority or designation to act in the public interest. This arrangement didn’t fit particularly well in the three-branch constitutional framework, as it essentially gave the power of all three branches to an agency located in the executive branch. But the courts ended up fitting this arrangement into the three-branch framework, albeit awkwardly, on the theory that Congress had delegated that power to the agency, subject to some judicial review. To give those administrative agencies the authority to act in that zone of delegated power, courts created the deferential standards of administrative law such as Chevron. Deference created a zone in which the agency would regulate. Deference was necessary not because agencies were experts, but because it was necessary step to give agencies the substantive rulemaking powers they were supposed to exercise.

Courts generally haven’t applied administrative law doctrines within patent law and criminal law because those fields do not rely on delegations of power. Both systems predate the late 19th century regulatory agency model, and both have retained their earlier mechanisms that work in a pre-regulatory-state way. In the case of criminal law, Congress doesn’t delegate to DOJ the power to say what the criminal laws mean. DOJ can execute the law by bringing a prosecution, based on the state of the law as established by the courts, but the power to say what the law is belongs solely to the courts. Similarly, Congress doesn’t delegate to the PTO the power to say what the patent laws mean. The PTO reviews patent applications, based on the state of the law as established by the courts, but the power to say what the law is belongs solely to the courts.

You might say, well, yes, this is the way it works now: But should it change? Should courts start applying administrative law doctrines in criminal law and patent law?

I think the answer is “no.” Deference doesn’t make sense without a delegation of rulemaking power. Consider the effect deference to the PTO might have in patent law, an issue I explored in my 2000 article, “Rethinking Patent Law in the Administrative State.” The idea of the patent laws is that the promise of getting a valuable patent will induce inventors to come with and disclose new inventions. The mechanism resembles that of a unilateral contract offer. If the patent applicant can satisfy the terms of the offer, as explained in the Patent Act and interpreted by the courts, the applicant is entitled to a patent. It’s kind of like trying to encourage engineers to invent a private space ship by offering a prize of $10 million to the first group that can successfully do so. The promise of the prize is what induces the desirable conduct, with the “prize” in patent law being the intellectual property right of the patent. The role of the PTO is to act as the agent of the party (Congress) that offered the prize to determine if the prize was won.

This perspective suggest to me that PTO deference would interfere with the goals of patent law. You wouldn’t want courts to defer to the PTO in patent cases for the same reason you wouldn’t be eager to invest in relying on a prize offer if you knew it was up to the offerer’s discretion to decide whether the offer was accepted. The point of making a prize offer is to induce reliance on it. To induce reliance, there has to be a way to make the offer enforceable. With a prize, that way is contract law. If you satisfy a prize offer, and the offerer decides that in light of his expertise he thinks it better that he not pay, you can go into court and get de novo review. And ex ante, if you knew that the offerer had discretion not to pay, you would be less eager to invest your time and energy trying to satisfy the prize. I think a similar dynamic explains why PTO deference would generally be a bad idea in patent law. At the margins, deference to the PTO as to whether it has to pay out by conferring a patent would just discourage reliance on the patent system’s offer.

Deference in criminal law would create analogous problems, I think. DOJ doesn’t “regulate” criminal activity. Instead, DOJ tries to achieve Congress’s goals in enacting criminal laws (such as deterrence and retribution) by bringing enforcement actions subject to whatever standard of criminality Congress adopted. The theory is that the legislature is supposed to debate and pass public legislation on what is prohibited before that law is to be enforced, so everyone can know where the line is ahead of time and the wide range of inputs and constraints of legislation is fully considered. Deferring to DOJ would frustrate this design rather than assist it. Deference would just give DOJ power to swing the hammer of federal criminal law after the act occurred, without the restrictions of enacting legislation, to conduct that the elected legislature hasn’t actually concluded was criminal. That would be contrary to all of the rule-of-law values on which the criminal law system is premised.