Further Questions About the Scope of the Dep't of Education's Authority Under Title IX
by Michael C. Dorf
In recent weeks, I wrote two blog posts (here and here) as well as a Verdict column critical of some aspects of the Dep't of Education (ED)'s notice of proposed rulemaking with regard to Title IX. I had thought I was done with that topic, but some reactions to the column (in the comments section, via an email exchange with Prof. Josh Blackman, and on criminal defense attorney Scott Greenfield's blog) raised interesting questions that I think merit further discussion. Hence, this fourth entry in the "trilogy." By way of preview, I will suggest that ED's theory on a key point relies on a very broad view of discrimination that Republican administrations and the Supreme Court have typically rejected.
My Verdict column acknowledges that the Trump/DeVos ED has the legal authority to rescind the Obama-era guidance under which colleges and universities were required to adopt certain procedures to ensure that they responded adequately to sexual assault complainants; insofar as the Republican ED opts for lighter regulation of covered institutions than a prior Democratic administration adopted, that is well within the scope of the authority delegated to the agency by Title IX. However, I went on to note that in various respects the new rules do not simply deregulate; rather, they impose procedural obligations on federally funded educational institutions that will make it more difficult for those institutions to find students responsible for sexual assault.
In a comment on the Verdict column, my colleague Prof. Kevin Clermont suggested that my view of agency authority was too narrow and would, if applied more broadly, hamstring agency action across a wide range of areas where many people (including both of us) would think agencies ought to have rulemaking authority.
My argument in the column was that rules that make it harder to obtain a finding of sexual assault might be thought good policy on the ground that they provide due process for accuseds--and for that reason the ED was well within its authority in allowing schools to adopt such stricter rules--but that ensuring due process for accuseds is not part of the ED's mandate. There are, after all, lots of worthwhile values that an agency could promote, but it only has the authority to require those it regulates to follow rules that fall within the statutory delegation. I gave the following example in my column: EPA might conclude that some emissions regulation on the books from a prior administration is stricter than ideal, because it would undercut economic growth; thus, EPA could allow a power plant to pollute at a higher level than the prior regulation allowed; but, I said, EPA could not mandate that power plants emit higher levels of pollution, because it has the statutory authority to restrict pollution, not to require pollution, even if at a level that, all things considered, balances the value of clean air against the value of economic growth.
Prof. Clermont agrees with the result in my hypothetical, but he responds that the additional process that ED would require is not like requiring discrimination. He thinks my understanding of the statutory mandate--to "effectuate" Title IX's prohibition on sex discrimination (and thus, per court decisions, to respond adequately to sexual assaults)--is too stingy. A statutory delegation to an agency to effectuate an anti-discrimination provision, he says, carries with it the authority to make various ancillary rules that, while not directly combating discrimination, put into effect an all-things-considered sensible scheme.
I agree with that objection as a general matter, and to the extent that my Verdict column could be read to suggest otherwise, I disavow that reading of it. To give a mundane example, a statutory delegation to address just about any issue, whether sex discrimination, pollution, automobile safety, or whatever, would typically include agency authority to impose on regulated actors various reporting requirements, even though the delegation does not specifically refer to reporting. Well, Prof. Clermont asks, why doesn't the "effectuate" language of Title IX likewise implicitly authorize ED to require the use of fair procedures in the investigation and adjudication of claims of sexual assault?
Here too my answer is that in general a delegation to an agency to adopt substantive rules includes the implicit authority to adopt procedural rules, and one would generally want those rules to be fair. The question then becomes whether all of the procedural requirements in the proposed rulemaking really aim at procedural fairness as a mere side constraint on pursuing the anti-discrimination mandate. My suggestion in the Verdict column is that some of them cannot reasonably be understood as doing that. Put in general terms, whatever power an agency has to prescribe ancillary rules in the general service of a primary statutory mandate, it lacks the power to mandate rules that conflict with that statutory mandate.
Here's another example. Suppose that instead of forbidding the adoption of some pollution control measure--which Professor Clermont and I agree the EPA would lack the authority to do under a delegation of power to issue rules limiting pollution--the EPA prescribes procedural constraints on power plants that want to adopt some new anti-pollution technology. Perhaps EPA tells power plants that if they want to switch to some cleaner technology that will cost jobs, they can do so, but only if they first obtain consent from a supermajority of everyone who has worked at the plant in the last year. That procedural requirement serves a noble end; it is not a mandate to pollute; and it can be defended in the name of procedural fairness; and yet, I think it would be pretty clearly ultra vires, because it operates to stymie voluntary efforts to "over-comply" with the law.
So the fact that the ED's proposed new rules can be said to aim at some conception of procedural fairness does not suffice to tell us that they serve Title IX's anti-discrimination mandate. A mere assertion of "fair procedures" is not enough to ground agency authority.
To see why ED lacks the power to mandate at least some of the new rules that would protect accuseds, it's important to understand the voluntary measures by schools that they aim to displace. I'll focus on the burden of persuasion.
In certain circumstances, the ED rules would require recipient schools to use a clear-and-convincing-evidence standard for evaluating sexual assault claims even though (by hypothesis) some of those schools would, if left to their own devices on this point, employ a preponderance-of-the-evidence standard. A preponderance standard assumes that, as between an accuser and an accused who denies the accusation, the decision maker should start off neutral, resolving the case in favor of the accused if the evidence points against the accusation or is in exact equipoise. By contrast, a clear-and-convincing standard means that there will be some cases in which the evidence on balance indicates that the accused is responsible but the decision maker will nonetheless reject the charge because it falls below the clear-and-convincing threshold.
Why would one adopt a clear-and-convincing standard in any particular setting? Both the ED proposed rulemaking and various supporters of it argue that the high stakes for the accused (including reputational harm and potential expulsion from college) warrant an extra measure of confidence. As I've said repeatedly, I think this is a permissible value judgment for a college or university to make on its own. However, my point in the column is that, as a legal requirement, it cannot reasonably be said to implement Title IX's anti-discrimination mandate, because it systematically undercuts it.
How so? The Obama-era guidance rejected the clear-and-convincing (and, a fortiori, the beyond-a-reasonable-doubt) standard, because of three main sorts of reasons. First, although suspension or expulsion from college is indeed serious, it is not so serious as some other instances of the use of the clear-and-convincing standard, such as civil confinement. Defenders of the high threshold in the proposed rulemaking also point to the reputational stakes, but it is worth noting that in the paradigmatic sort of case involving alleged reputational harm--a civil lawsuit for defamation--the preponderance standard applies. Second, although the stakes for an accused can be high, the stakes are also high for an accuser. Under the clear-and-convincing standard, there will be cases in which it is more likely than not that a student's rapist is found not responsible for the offense and yet continues to go to the same classes, cafeteria, and social events as the victim, thus extending the victim's trauma. Third, a preponderance standard in sexual assault cases can be justified on the ground that for centuries and even today, people engage in unwarranted skepticism of allegations of sexual assault. Even if one thought that in general something like suspension or expulsion from college were the sort of penalty for which a clear-and-convincing standard ought to apply, one might nonetheless think that the formal standard ought to be preponderance in sexual assault cases: Given the unwarranted (and sex-stereotyped) skepticism that so many people bring to considering accusations of sexual assault, a formal preponderance standard could end up operating as a clear-and-convincing or even beyond-a-reasonable-doubt standard in practice.
How persuasive are the foregoing sorts of arguments? Different people will give different answers. But someone who rejects the arguments for the preponderance standard cannot reasonably rely on any notion of procedural fairness in the service of the sex equality mandate. Rather, he or she must rest on a judgment about either the relative frequency or the relative cost of false negatives versus false positives.
One might think that people who bring accusations of sexual assault are more often than not lying or mistaken. Donald Trump apparently thinks this. Maybe the ED and at least some supporters of its proposed rulemaking share this view. If so, they ought to say so and provide some sort of credible evidence rather than relying on rape-myth stereotypes. To their credit, with the exception of the president, backers of the ED's authority to mandate a clear-and-convincing standard for Title IX sexual assault cases have not expressly attempted to justify it on the ground that most accusers are lying hussies.
A less ugly but still ultimately unpersuasive defense of ED authority to require a heightened evidentiary standard would emphasize the stakes for accuseds, while rejecting the counter-arguments I set forth a few paragraphs up. As I've repeatedly said (and now repeat yet again), I think that's the kind of judgment that the ED can leave for a college or university to make, but for the ED to mandate it requires that the proposed clear-and-convincing rule furthers the objectives of Title IX. As my hypothetical EPA reg requiring a supermajority vote illustrates, the mere fact that it advances a conception of procedural fairness does not suffice.
That brings me to an argument made by Prof. Blackman via email, by Scott Greenfield in the blog post linked above, and by Prof. Clermont in a reply to my response to his comment on my Verdict column: A heightened evidentiary standard does advance the anti-discrimination purpose of Title IX, they say, because it protects men against the sexism of false accusations. Summarizing this argument, Greenfield pronounces that "sex discrimination is a two-way street."
In some contexts that sort of argument makes sense. For example, one could characterize race-based affirmative action as a remedy for discrimination against members of a racial minority group and as discrimination against persons of other races.
But that's because race-based affirmative action is, well, race-based. It expressly classifies based on race. By contrast, Title IX proceedings in cases of sexual assault do not classify anyone by sex. Here's an excellent explanation by Professors Joanna Grossman and Deborah Brake of why the "two-way street" idea is wrong:
What the ED probably ought to argue if it wants to advance the "two-way street" idea is that although men who are charged under Title IX are not victims of sex discrimination, an agency may, in the interest of enforcing an anti-discrimination statute, promulgate a reg that attacks disparate impact. That contention could run into trouble in the SCOTUS, of course. In 2001, in Alexander v. Sandoval, the Court assumed arguendo that a federal agency has such a power, but it suggested that in some future case the justices might invalidate disparate impact regs. Justice Scalia said for the Court that a disparate impact reg was "in considerable tension with" decisions construing the statute at issue as covering only intentional discrimination. Still, it would be interesting to watch a Republican DOJ argue to a conservative Court for a result that conservatives have long opposed on the ground that in the Trump administration, men's interests in procedures that minimize the likelihood of being held responsible for sexual assault prevail over other considerations.
In recent weeks, I wrote two blog posts (here and here) as well as a Verdict column critical of some aspects of the Dep't of Education (ED)'s notice of proposed rulemaking with regard to Title IX. I had thought I was done with that topic, but some reactions to the column (in the comments section, via an email exchange with Prof. Josh Blackman, and on criminal defense attorney Scott Greenfield's blog) raised interesting questions that I think merit further discussion. Hence, this fourth entry in the "trilogy." By way of preview, I will suggest that ED's theory on a key point relies on a very broad view of discrimination that Republican administrations and the Supreme Court have typically rejected.
My Verdict column acknowledges that the Trump/DeVos ED has the legal authority to rescind the Obama-era guidance under which colleges and universities were required to adopt certain procedures to ensure that they responded adequately to sexual assault complainants; insofar as the Republican ED opts for lighter regulation of covered institutions than a prior Democratic administration adopted, that is well within the scope of the authority delegated to the agency by Title IX. However, I went on to note that in various respects the new rules do not simply deregulate; rather, they impose procedural obligations on federally funded educational institutions that will make it more difficult for those institutions to find students responsible for sexual assault.
In a comment on the Verdict column, my colleague Prof. Kevin Clermont suggested that my view of agency authority was too narrow and would, if applied more broadly, hamstring agency action across a wide range of areas where many people (including both of us) would think agencies ought to have rulemaking authority.
My argument in the column was that rules that make it harder to obtain a finding of sexual assault might be thought good policy on the ground that they provide due process for accuseds--and for that reason the ED was well within its authority in allowing schools to adopt such stricter rules--but that ensuring due process for accuseds is not part of the ED's mandate. There are, after all, lots of worthwhile values that an agency could promote, but it only has the authority to require those it regulates to follow rules that fall within the statutory delegation. I gave the following example in my column: EPA might conclude that some emissions regulation on the books from a prior administration is stricter than ideal, because it would undercut economic growth; thus, EPA could allow a power plant to pollute at a higher level than the prior regulation allowed; but, I said, EPA could not mandate that power plants emit higher levels of pollution, because it has the statutory authority to restrict pollution, not to require pollution, even if at a level that, all things considered, balances the value of clean air against the value of economic growth.
Prof. Clermont agrees with the result in my hypothetical, but he responds that the additional process that ED would require is not like requiring discrimination. He thinks my understanding of the statutory mandate--to "effectuate" Title IX's prohibition on sex discrimination (and thus, per court decisions, to respond adequately to sexual assaults)--is too stingy. A statutory delegation to an agency to effectuate an anti-discrimination provision, he says, carries with it the authority to make various ancillary rules that, while not directly combating discrimination, put into effect an all-things-considered sensible scheme.
I agree with that objection as a general matter, and to the extent that my Verdict column could be read to suggest otherwise, I disavow that reading of it. To give a mundane example, a statutory delegation to address just about any issue, whether sex discrimination, pollution, automobile safety, or whatever, would typically include agency authority to impose on regulated actors various reporting requirements, even though the delegation does not specifically refer to reporting. Well, Prof. Clermont asks, why doesn't the "effectuate" language of Title IX likewise implicitly authorize ED to require the use of fair procedures in the investigation and adjudication of claims of sexual assault?
Here too my answer is that in general a delegation to an agency to adopt substantive rules includes the implicit authority to adopt procedural rules, and one would generally want those rules to be fair. The question then becomes whether all of the procedural requirements in the proposed rulemaking really aim at procedural fairness as a mere side constraint on pursuing the anti-discrimination mandate. My suggestion in the Verdict column is that some of them cannot reasonably be understood as doing that. Put in general terms, whatever power an agency has to prescribe ancillary rules in the general service of a primary statutory mandate, it lacks the power to mandate rules that conflict with that statutory mandate.
Here's another example. Suppose that instead of forbidding the adoption of some pollution control measure--which Professor Clermont and I agree the EPA would lack the authority to do under a delegation of power to issue rules limiting pollution--the EPA prescribes procedural constraints on power plants that want to adopt some new anti-pollution technology. Perhaps EPA tells power plants that if they want to switch to some cleaner technology that will cost jobs, they can do so, but only if they first obtain consent from a supermajority of everyone who has worked at the plant in the last year. That procedural requirement serves a noble end; it is not a mandate to pollute; and it can be defended in the name of procedural fairness; and yet, I think it would be pretty clearly ultra vires, because it operates to stymie voluntary efforts to "over-comply" with the law.
So the fact that the ED's proposed new rules can be said to aim at some conception of procedural fairness does not suffice to tell us that they serve Title IX's anti-discrimination mandate. A mere assertion of "fair procedures" is not enough to ground agency authority.
To see why ED lacks the power to mandate at least some of the new rules that would protect accuseds, it's important to understand the voluntary measures by schools that they aim to displace. I'll focus on the burden of persuasion.
In certain circumstances, the ED rules would require recipient schools to use a clear-and-convincing-evidence standard for evaluating sexual assault claims even though (by hypothesis) some of those schools would, if left to their own devices on this point, employ a preponderance-of-the-evidence standard. A preponderance standard assumes that, as between an accuser and an accused who denies the accusation, the decision maker should start off neutral, resolving the case in favor of the accused if the evidence points against the accusation or is in exact equipoise. By contrast, a clear-and-convincing standard means that there will be some cases in which the evidence on balance indicates that the accused is responsible but the decision maker will nonetheless reject the charge because it falls below the clear-and-convincing threshold.
Why would one adopt a clear-and-convincing standard in any particular setting? Both the ED proposed rulemaking and various supporters of it argue that the high stakes for the accused (including reputational harm and potential expulsion from college) warrant an extra measure of confidence. As I've said repeatedly, I think this is a permissible value judgment for a college or university to make on its own. However, my point in the column is that, as a legal requirement, it cannot reasonably be said to implement Title IX's anti-discrimination mandate, because it systematically undercuts it.
How so? The Obama-era guidance rejected the clear-and-convincing (and, a fortiori, the beyond-a-reasonable-doubt) standard, because of three main sorts of reasons. First, although suspension or expulsion from college is indeed serious, it is not so serious as some other instances of the use of the clear-and-convincing standard, such as civil confinement. Defenders of the high threshold in the proposed rulemaking also point to the reputational stakes, but it is worth noting that in the paradigmatic sort of case involving alleged reputational harm--a civil lawsuit for defamation--the preponderance standard applies. Second, although the stakes for an accused can be high, the stakes are also high for an accuser. Under the clear-and-convincing standard, there will be cases in which it is more likely than not that a student's rapist is found not responsible for the offense and yet continues to go to the same classes, cafeteria, and social events as the victim, thus extending the victim's trauma. Third, a preponderance standard in sexual assault cases can be justified on the ground that for centuries and even today, people engage in unwarranted skepticism of allegations of sexual assault. Even if one thought that in general something like suspension or expulsion from college were the sort of penalty for which a clear-and-convincing standard ought to apply, one might nonetheless think that the formal standard ought to be preponderance in sexual assault cases: Given the unwarranted (and sex-stereotyped) skepticism that so many people bring to considering accusations of sexual assault, a formal preponderance standard could end up operating as a clear-and-convincing or even beyond-a-reasonable-doubt standard in practice.
How persuasive are the foregoing sorts of arguments? Different people will give different answers. But someone who rejects the arguments for the preponderance standard cannot reasonably rely on any notion of procedural fairness in the service of the sex equality mandate. Rather, he or she must rest on a judgment about either the relative frequency or the relative cost of false negatives versus false positives.
One might think that people who bring accusations of sexual assault are more often than not lying or mistaken. Donald Trump apparently thinks this. Maybe the ED and at least some supporters of its proposed rulemaking share this view. If so, they ought to say so and provide some sort of credible evidence rather than relying on rape-myth stereotypes. To their credit, with the exception of the president, backers of the ED's authority to mandate a clear-and-convincing standard for Title IX sexual assault cases have not expressly attempted to justify it on the ground that most accusers are lying hussies.
A less ugly but still ultimately unpersuasive defense of ED authority to require a heightened evidentiary standard would emphasize the stakes for accuseds, while rejecting the counter-arguments I set forth a few paragraphs up. As I've repeatedly said (and now repeat yet again), I think that's the kind of judgment that the ED can leave for a college or university to make, but for the ED to mandate it requires that the proposed clear-and-convincing rule furthers the objectives of Title IX. As my hypothetical EPA reg requiring a supermajority vote illustrates, the mere fact that it advances a conception of procedural fairness does not suffice.
That brings me to an argument made by Prof. Blackman via email, by Scott Greenfield in the blog post linked above, and by Prof. Clermont in a reply to my response to his comment on my Verdict column: A heightened evidentiary standard does advance the anti-discrimination purpose of Title IX, they say, because it protects men against the sexism of false accusations. Summarizing this argument, Greenfield pronounces that "sex discrimination is a two-way street."
In some contexts that sort of argument makes sense. For example, one could characterize race-based affirmative action as a remedy for discrimination against members of a racial minority group and as discrimination against persons of other races.
But that's because race-based affirmative action is, well, race-based. It expressly classifies based on race. By contrast, Title IX proceedings in cases of sexual assault do not classify anyone by sex. Here's an excellent explanation by Professors Joanna Grossman and Deborah Brake of why the "two-way street" idea is wrong:
Although male students who have claimed that they were wrongly disciplined for sexual harassment have sued their schools under Title IX, courts have generally rejected those claims without proof of animus or discriminatory intent against males as a class. Claims that male plaintiffs were denied a fair process (such as the lack of notice, an impartial decision-maker, and an opportunity to present evidence) have (sometimes rightly) succeeded in stating claims for due process violations and/or breach of contractual promises. But courts have not equated the failure to provide a fair process to students accused of sexual misconduct with sex discrimination against men.Here's another way to understand the point: Not all accusers are women and not all accuseds are men. It's true that in most cases the accuser is a woman and the accused is a man, but that only shows that providing redress to victims of sexual assault will have a negative disparate impact on men. Under longstanding principles of anti-discrimination law that conservative jurists and Republican administrations have championed vigorously, disparate impact without proof of an intent to discriminate on an illicit basis is not discrimination.
What the ED probably ought to argue if it wants to advance the "two-way street" idea is that although men who are charged under Title IX are not victims of sex discrimination, an agency may, in the interest of enforcing an anti-discrimination statute, promulgate a reg that attacks disparate impact. That contention could run into trouble in the SCOTUS, of course. In 2001, in Alexander v. Sandoval, the Court assumed arguendo that a federal agency has such a power, but it suggested that in some future case the justices might invalidate disparate impact regs. Justice Scalia said for the Court that a disparate impact reg was "in considerable tension with" decisions construing the statute at issue as covering only intentional discrimination. Still, it would be interesting to watch a Republican DOJ argue to a conservative Court for a result that conservatives have long opposed on the ground that in the Trump administration, men's interests in procedures that minimize the likelihood of being held responsible for sexual assault prevail over other considerations.
* * *
That ends the core of my response to the critics of my column, but I want to add two postscripts.
(1) In his initial comment, Prof. Clermont suggested that under my logic, any procedural requirements would be invalid. He asked whether I thought that even the ones required by the Obama-era guidance ("adequate, reliable, and impartial investigation of complaints, including the opportunity for both parties to present witnesses and other evidence") were invalid. I agree with Prof. Clermont's characterization that it would be "silly" to treat such rules as ultra vires. As I have explained above, to the extent that my column could be read to treat all procedural rules that do not directly advance the anti-discrimination mandate as ultra vires, I want to clarify that that was not my intention. I am grateful to him for pointing to the ambiguity in my column and thus for leading to the clarification here.
Rules that aim simply at accuracy and reliability do not run counter to the anti-discrimination mandate. Indeed, if a school uses procedures that produce a substantial number of inaccurate findings--whether false positives or false negatives--that will undercut faith in the system for evaluating and responding to claims of sexual assault. And for that reason, I don't object to everything that the Trump ED proposes, just those requirements that place a heavy thumb on the scale on the side opposite a claim. (And even then, I acknowledge that ED has the authority to permit schools to place such a thumb on the scale.)
(2) Greenfield's blog post says some inaccurate things about my Verdict column. For instance, he thinks it ironic that I criticize the DeVos rulemaking as a power grab but that I was fine with the Obama administration's proceeding by informal guidance that was not subject to the rigor of the Administrative Procedure Act's notice-and-comment review process. That could be ironic if accurate, but it isn't.
In my Verdict column, I said that the Obama ED's "guidance attracted criticism partly because it was not promulgated as a formal regulation." I did not say what I thought about the merits of that criticism, so I'm not sure why Greenfield thinks I rejected it. In fact, I more or less endorsed the criticism. I wrote the following in my November 19 blog post discussing another aspect of the proposed ED rulemaking: "In one important respect, the proposed regulation is a step forward: it is a proposed regulation rather than a less formal administrative action." (Emphasis in original).
Greenfield can be forgiven for not having read the prior blog post along with the Verdict column. But he misreads me in another way that is more egregious. He attributes to me the view that "[u]ndoing" the Obama administration's guidance is a "power grab." But in the very column that Greenfield was criticizing I said extremely clearly and in multiple places that the Trump ED was entitled to rescind the Obama ED's guidance, to "undo" it. The "power grab" in the title of my essay unmistakably referred to some of the ways in which the new proposed rules go beyond undoing the Obama policy.