Executive Discretion and RFRA Harmonization in the Zubik Oral Argument
by Michael Dorf
Yesterday's oral argument in Zubik v. Burwell raised a great many interesting and important questions. After setting out the background to the case, I'm going to zero in on one that is of special interest to me, which is not to say it is the most important: Could the federal executive effectively create new private contraception-only insurance programs on the exchanges, even if not directly authorized to do so by statute?
With important exceptions, the regulations implementing the Affordable Care Act (ACA) generally require employers that provide health insurance to their employees to provide plans that include contraceptive coverage with zero deductibles. In 2014, in Burwell v. Hobby Lobby, the SCOTUS held that under the federal Religious Freedom Restoration Act (RFRA), Hobby Lobby and similarly situated corporate employers whose ownership group had moral objections to providing (some kinds of contraception) to their employees could opt out using the same mechanism that the government had already established for religious nonprofit organizations, notwithstanding the fact that Hobby Lobby et al are for-profit business organized as corporations.
Zubik involves a challenge to the opt-out mechanism for nonprofits that are already entitled to opt out under the regs. The plaintiffs argue that the structure of the opt-out is itself a substantial burden on their religious exercise because notifying the government of the religious objection triggers the government's "hijacking" of the employee health insurance administrator the nonprofit uses for its non-contraception plan to provide contraceptive insurance coverage as well. The government contests this characterization of the consequences of an opt-out notice.
While acknowledging the sincerity of the plaintiffs' religious beliefs, the government argues that its use of the third-party insurance administrators to provide the employees with contraceptive coverage does not count as a substantial burden within the meaning of RFRA. The government relies in part on cases rejecting Free Exercise claims even under the pre-Employment Division v. Smith case law. In "restoring" the pre-Smith law, the government contends, RFRA did not disturb the limits found in such cases. Amici and some justices also suggested that, as a categorical rule, a burden doesn't count as substantial under RFRA if it can only be accommodated by imposing costs on third parties (here, the female employees who do not have religious objections to using contraception). If the government wins on the substantial burden point, the government wins the case.
Even if the Court finds that the ACA substantially burdens the religious exercise of the plaintiffs, the government can still win if the substantial burden is the least restrictive means of advancing a compelling interest. There appears to be some dispute over what interest the government aims to advance. In Hobby Lobby, the Court assumed for the sake of argument that providing greater access to contraception is a compelling interest, and the government invokes that interest here. But there was some pushback. For example, Chief Justice Roberts appeared to conceptualize the case as calling for a balancing (or weighing) of the relative strength of the interest in easier access to contraception against the plaintiffs' interest in avoiding (by their lights) being complicit in (what they regard as) sin. This is somewhat non-standard. Ordinarily, one asks only whether the government interest is compelling, not whether it is more compelling than the right it is asserted against. But the Chief Justice is probably correct that in practice judges weigh the competing interests against each other.
In any event, if the government does not prevail in its argument that the ACA with opt-out is not a substantial burden, the case will probably come down to the question whether the regime that the government has established is (in the language of RFRA) "the least restrictive means of furthering" the government's interest of widened contraceptive access. The plaintiffs argue that it is not, because there are other ways of providing insurance for contraception coverage. The government in turn says that all of the alternatives the plaintiffs identify would make it harder, as a practical matter, for women to obtain contraceptives.
For example, the plaintiffs suggest that the government could allow women who work for exempt non-profits to obtain stand-alone contraception insurance on the health insurance exchanges, but the government responds that this would be cumbersome: each woman would have to shop for a supplemental plan and, depending on what she got, her primary care physician for her regular plan might not take the insurance from her contraception plan. Thus, the government argues that the alternatives proposed by the plaintiffs are not "seamless" with existing insurance coverage in the way that contraceptive coverage by the plan administrator of the plan women already have through their employers is.
Part of the argument in the case is over whether seamlessness of this sort is really a compelling interest or, as the Chief Justice would have it, whether the interest in seamlessness is as compelling as the religious claims that it needs to overcome. I want to put that issue aside to focus on a different one.
In order to show that the current accommodation is not the least restrictive means, the plaintiffs point to the fact that the government already exempts some employers from even the opt-out regime for contraceptive coverage: E.g., churches and other houses of worship; certain "grandfathered" employers; etc. The plaintiffs say that the fact that the government can make these exceptions shows that it can achieve the goals it deems compelling without resorting to the objectionably structured opt-out. The government, some amici, and some justices in turn respond that those exceptions are different, and, in any event, the existence of a few exceptions does not undermine the compelling nature of the government's goal in general. If it did, they note, the government would have perverse incentives: It would not provide exemptions from general obligations even to houses of worship, lest those exemptions be taken as evidence that its general purpose is not so compelling that it can't survive some additional religious exemptions.
I want to put that broad debate to one side as well, to focus on a subsidiary point that arose during the argument. As I have indicated, Solicitor General Don Verrilli resisted the alternatives proposed by the plaintiffs and various justices chiefly on the ground that these alternatives would not serve the government's interest in seamlessness, and thus would result in less access to contraception. But he also said that the proposed alternatives would require new legislation. For example, a health insurance company could not offer contraception-only insurance on an exchange because that is not authorized by the ACA. Justice Alito pushed back by asking:
I'm not confident what the right answer to that question is, but I would note how remarkable it is that Justice Alito apparently thinks the answer is yes. If that is his sincere view, and if he holds it consistently--two big IFs--then think of the implications for United States v. Texas, due to be argued next month. There the parties appear to agree that the president has prosecutorial discretion to withhold deportation and other enforcement actions against undocumented immigrants, while they disagree about how to characterize other aspects of the programs of Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA) and expanded Deferred Action for Childhood Arrivals (DACA). Texas and its amici argue that DAPA and expanded DACA provide status and benefits to undocumented immigrants that the law does not authorize. The government argues that DAPA and expanded DACA do no such thing, although other laws do kick in when an undocumented immigrant is eligible for DAPA or expanded DACA. What's notable about Justice Alito's question is that he seems to assume that the president has the authority--in the exercise of prosecutorial discretion under the ACA--to expand the scope of insurance available on exchanges and to spend money to subsidize that expanded scope. An analogous view in the immigration context would be that even if Texas is right in its characterization of DAPA and expanded DACA as granting undocumented immigrants lawful status unauthorized by Congress, that would be legal as an exercise of prosecutorial discretion.
Now perhaps I've misread Justice Alito's question, or perhaps there is some provision of the ACA that gives the executive branch greater power to tinker with the exchanges than the immigration laws give the executive to tinker, but at the very least, if I were SG Verrilli, I would take heart that there's at least a chance that Justice Alito has a very expansive view of executive discretion, so that when Verrilli returns to the Court in April to argue United States v. Texas, he can remind Justice Alito of his question in Zubik.
Okay, obviously the previous paragraph is wishful thinking. I'm enough of a legal realist to know that judges and justices can come up with distinctions that enable them to vote in ways that initially look inconsistent. Is there such a distinction available here? My answer is yes. The distinction is RFRA.
Here's what I mean. Let's suppose that SG Verrilli's answer is correct as a reading of the ACA standing alone: It doesn't allow the executive to use its prosecutorial discretion and money appropriated for the ACA to encourage the creation of contraception-only plans on exchanges. But the ACA doesn't stand alone. A RFRA runs through it. By its terms, RFRA says only that government can't substantially burden religious exercise, unless it satisfies strict scrutiny. RFRA does not say what happens if a court determines that a generally applicable law fails the RFRA test.
In many cases, the answer is easy: If the law imposes a primary obligation--let's say it requires that people remove their hats upon entering courthouses--the religious objectors are exempt from the obligation. But sometimes an exception is not so simple. Take a case like Smith, which generated RFRA in the first place. The finding that Native Americans wishing to use peyote in a religious ritual are exempt from the general prohibition against peyote use is easily implemented if there is a peyote prosecution: RFRA means the indictment is quashed. But what if, as in Smith itself, the case arises in a benefits context? The finding that the general peyote obligation violates RFRA would, in a case like Smith, trigger a government obligation to pay the plaintiffs money because it would mean that the state cannot treat them as voluntarily unemployed. And if RFRA can impose financial obligations on government in a case like Smith, then maybe it can do so in a case like Zubik as well.
Put differently, maybe Justice Alito was suggesting something like the following:
This case requires the government to harmonize the ACA and RFRA. Simply exempting the plaintiffs from the obligation to provide contraception-covering insurance to their employees would serve RFRA's goal of protecting religious freedom by completely sacrificing the goal of the ACA (as construed through the regs) of widening access to contraceptive health care. Accordingly, we should read RFRA as authorizing the executive to take actions regarding exchanges that are not literally authorized by the ACA.
To be sure, Justice Alito didn't say anything like that. But CJ Roberts did--kind of. When SG Verrilli objected that current law does not permit the executive to create the alternatives that the plaintiffs suggested, the Chief replied: "Well, the way constitutional objections work is you might have to change current law."
There's an obvious error there, of course. RFRA is a statute, not a constitutional provision. But we know what the Chief meant: RFRA works like a constitutional objection. And there is a long line of cases that say that even if a government entity needs to violate statute to respect a constitutional right, the government must respect the constitutional right. To be sure, most of those cases involve state and local governments, but the principle is the same for the federal government: the Constitution trumps any statute.
To be sure, as just noted, RFRA is not the Constitution, and so it doesn't trump the ACA. But again, it works like a constitutional right. What the Chief might have been suggesting is that RFRA should be read as a delegation of authority from Congress to the executive (and, where the executive doesn't comply on its own, to the courts) to fashion otherwise unauthorized legal regimes where RFRA is violated. In my view, that would be a sound reading of RFRA.
Is that what the Chief really meant? Probably not. If I had to bet, I'd say that the Chief meant that if the government can't accommodate the Zubik plaintiffs under the ACA as written, then it should simply exempt them, unless and until Congress changes the law. But even though that's what he probably meant, the alternative reading I'm proposing makes at least as much sense. It takes account of the fact that a reasonable Congress would want the executive and the courts to harmonize the goals of RFRA and whatever statutes RFRA limits, where possible. True, we don't actually have a reasonable Congress, but sound statutory construction has long proceeded on the assumption that, as Hart & Sacks put it, the legislature consists of reasonable people pursuing reasonable goals, reasonably--even if that's not actually true.
Yesterday's oral argument in Zubik v. Burwell raised a great many interesting and important questions. After setting out the background to the case, I'm going to zero in on one that is of special interest to me, which is not to say it is the most important: Could the federal executive effectively create new private contraception-only insurance programs on the exchanges, even if not directly authorized to do so by statute?
With important exceptions, the regulations implementing the Affordable Care Act (ACA) generally require employers that provide health insurance to their employees to provide plans that include contraceptive coverage with zero deductibles. In 2014, in Burwell v. Hobby Lobby, the SCOTUS held that under the federal Religious Freedom Restoration Act (RFRA), Hobby Lobby and similarly situated corporate employers whose ownership group had moral objections to providing (some kinds of contraception) to their employees could opt out using the same mechanism that the government had already established for religious nonprofit organizations, notwithstanding the fact that Hobby Lobby et al are for-profit business organized as corporations.
Zubik involves a challenge to the opt-out mechanism for nonprofits that are already entitled to opt out under the regs. The plaintiffs argue that the structure of the opt-out is itself a substantial burden on their religious exercise because notifying the government of the religious objection triggers the government's "hijacking" of the employee health insurance administrator the nonprofit uses for its non-contraception plan to provide contraceptive insurance coverage as well. The government contests this characterization of the consequences of an opt-out notice.
While acknowledging the sincerity of the plaintiffs' religious beliefs, the government argues that its use of the third-party insurance administrators to provide the employees with contraceptive coverage does not count as a substantial burden within the meaning of RFRA. The government relies in part on cases rejecting Free Exercise claims even under the pre-Employment Division v. Smith case law. In "restoring" the pre-Smith law, the government contends, RFRA did not disturb the limits found in such cases. Amici and some justices also suggested that, as a categorical rule, a burden doesn't count as substantial under RFRA if it can only be accommodated by imposing costs on third parties (here, the female employees who do not have religious objections to using contraception). If the government wins on the substantial burden point, the government wins the case.
Even if the Court finds that the ACA substantially burdens the religious exercise of the plaintiffs, the government can still win if the substantial burden is the least restrictive means of advancing a compelling interest. There appears to be some dispute over what interest the government aims to advance. In Hobby Lobby, the Court assumed for the sake of argument that providing greater access to contraception is a compelling interest, and the government invokes that interest here. But there was some pushback. For example, Chief Justice Roberts appeared to conceptualize the case as calling for a balancing (or weighing) of the relative strength of the interest in easier access to contraception against the plaintiffs' interest in avoiding (by their lights) being complicit in (what they regard as) sin. This is somewhat non-standard. Ordinarily, one asks only whether the government interest is compelling, not whether it is more compelling than the right it is asserted against. But the Chief Justice is probably correct that in practice judges weigh the competing interests against each other.
In any event, if the government does not prevail in its argument that the ACA with opt-out is not a substantial burden, the case will probably come down to the question whether the regime that the government has established is (in the language of RFRA) "the least restrictive means of furthering" the government's interest of widened contraceptive access. The plaintiffs argue that it is not, because there are other ways of providing insurance for contraception coverage. The government in turn says that all of the alternatives the plaintiffs identify would make it harder, as a practical matter, for women to obtain contraceptives.
For example, the plaintiffs suggest that the government could allow women who work for exempt non-profits to obtain stand-alone contraception insurance on the health insurance exchanges, but the government responds that this would be cumbersome: each woman would have to shop for a supplemental plan and, depending on what she got, her primary care physician for her regular plan might not take the insurance from her contraception plan. Thus, the government argues that the alternatives proposed by the plaintiffs are not "seamless" with existing insurance coverage in the way that contraceptive coverage by the plan administrator of the plan women already have through their employers is.
Part of the argument in the case is over whether seamlessness of this sort is really a compelling interest or, as the Chief Justice would have it, whether the interest in seamlessness is as compelling as the religious claims that it needs to overcome. I want to put that issue aside to focus on a different one.
In order to show that the current accommodation is not the least restrictive means, the plaintiffs point to the fact that the government already exempts some employers from even the opt-out regime for contraceptive coverage: E.g., churches and other houses of worship; certain "grandfathered" employers; etc. The plaintiffs say that the fact that the government can make these exceptions shows that it can achieve the goals it deems compelling without resorting to the objectionably structured opt-out. The government, some amici, and some justices in turn respond that those exceptions are different, and, in any event, the existence of a few exceptions does not undermine the compelling nature of the government's goal in general. If it did, they note, the government would have perverse incentives: It would not provide exemptions from general obligations even to houses of worship, lest those exemptions be taken as evidence that its general purpose is not so compelling that it can't survive some additional religious exemptions.
I want to put that broad debate to one side as well, to focus on a subsidiary point that arose during the argument. As I have indicated, Solicitor General Don Verrilli resisted the alternatives proposed by the plaintiffs and various justices chiefly on the ground that these alternatives would not serve the government's interest in seamlessness, and thus would result in less access to contraception. But he also said that the proposed alternatives would require new legislation. For example, a health insurance company could not offer contraception-only insurance on an exchange because that is not authorized by the ACA. Justice Alito pushed back by asking:
Couldn't the Executive deal with the problem of what's available on the Exchanges at the present time in this way: Policies are available that provide comprehensive coverage. Could the Executive say, as a matter of our enforcement discretion, we are not going to take any action against insurers who offer contraceptive only policies, and in fact, we are going to subsidize those insurers at 115 percent, just as we do in the situation of the self-insured plans?SG Verrilli said no, but that even if this were in the power of the executive it wouldn't sufficiently serve the interest in seamlessness. Let's set aside the second point. Is Verrilli right about the first point? Does the executive have enforcement discretion of the sort suggested by Justice Alito?
I'm not confident what the right answer to that question is, but I would note how remarkable it is that Justice Alito apparently thinks the answer is yes. If that is his sincere view, and if he holds it consistently--two big IFs--then think of the implications for United States v. Texas, due to be argued next month. There the parties appear to agree that the president has prosecutorial discretion to withhold deportation and other enforcement actions against undocumented immigrants, while they disagree about how to characterize other aspects of the programs of Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA) and expanded Deferred Action for Childhood Arrivals (DACA). Texas and its amici argue that DAPA and expanded DACA provide status and benefits to undocumented immigrants that the law does not authorize. The government argues that DAPA and expanded DACA do no such thing, although other laws do kick in when an undocumented immigrant is eligible for DAPA or expanded DACA. What's notable about Justice Alito's question is that he seems to assume that the president has the authority--in the exercise of prosecutorial discretion under the ACA--to expand the scope of insurance available on exchanges and to spend money to subsidize that expanded scope. An analogous view in the immigration context would be that even if Texas is right in its characterization of DAPA and expanded DACA as granting undocumented immigrants lawful status unauthorized by Congress, that would be legal as an exercise of prosecutorial discretion.
Now perhaps I've misread Justice Alito's question, or perhaps there is some provision of the ACA that gives the executive branch greater power to tinker with the exchanges than the immigration laws give the executive to tinker, but at the very least, if I were SG Verrilli, I would take heart that there's at least a chance that Justice Alito has a very expansive view of executive discretion, so that when Verrilli returns to the Court in April to argue United States v. Texas, he can remind Justice Alito of his question in Zubik.
Okay, obviously the previous paragraph is wishful thinking. I'm enough of a legal realist to know that judges and justices can come up with distinctions that enable them to vote in ways that initially look inconsistent. Is there such a distinction available here? My answer is yes. The distinction is RFRA.
Here's what I mean. Let's suppose that SG Verrilli's answer is correct as a reading of the ACA standing alone: It doesn't allow the executive to use its prosecutorial discretion and money appropriated for the ACA to encourage the creation of contraception-only plans on exchanges. But the ACA doesn't stand alone. A RFRA runs through it. By its terms, RFRA says only that government can't substantially burden religious exercise, unless it satisfies strict scrutiny. RFRA does not say what happens if a court determines that a generally applicable law fails the RFRA test.
In many cases, the answer is easy: If the law imposes a primary obligation--let's say it requires that people remove their hats upon entering courthouses--the religious objectors are exempt from the obligation. But sometimes an exception is not so simple. Take a case like Smith, which generated RFRA in the first place. The finding that Native Americans wishing to use peyote in a religious ritual are exempt from the general prohibition against peyote use is easily implemented if there is a peyote prosecution: RFRA means the indictment is quashed. But what if, as in Smith itself, the case arises in a benefits context? The finding that the general peyote obligation violates RFRA would, in a case like Smith, trigger a government obligation to pay the plaintiffs money because it would mean that the state cannot treat them as voluntarily unemployed. And if RFRA can impose financial obligations on government in a case like Smith, then maybe it can do so in a case like Zubik as well.
Put differently, maybe Justice Alito was suggesting something like the following:
This case requires the government to harmonize the ACA and RFRA. Simply exempting the plaintiffs from the obligation to provide contraception-covering insurance to their employees would serve RFRA's goal of protecting religious freedom by completely sacrificing the goal of the ACA (as construed through the regs) of widening access to contraceptive health care. Accordingly, we should read RFRA as authorizing the executive to take actions regarding exchanges that are not literally authorized by the ACA.
To be sure, Justice Alito didn't say anything like that. But CJ Roberts did--kind of. When SG Verrilli objected that current law does not permit the executive to create the alternatives that the plaintiffs suggested, the Chief replied: "Well, the way constitutional objections work is you might have to change current law."
There's an obvious error there, of course. RFRA is a statute, not a constitutional provision. But we know what the Chief meant: RFRA works like a constitutional objection. And there is a long line of cases that say that even if a government entity needs to violate statute to respect a constitutional right, the government must respect the constitutional right. To be sure, most of those cases involve state and local governments, but the principle is the same for the federal government: the Constitution trumps any statute.
To be sure, as just noted, RFRA is not the Constitution, and so it doesn't trump the ACA. But again, it works like a constitutional right. What the Chief might have been suggesting is that RFRA should be read as a delegation of authority from Congress to the executive (and, where the executive doesn't comply on its own, to the courts) to fashion otherwise unauthorized legal regimes where RFRA is violated. In my view, that would be a sound reading of RFRA.
Is that what the Chief really meant? Probably not. If I had to bet, I'd say that the Chief meant that if the government can't accommodate the Zubik plaintiffs under the ACA as written, then it should simply exempt them, unless and until Congress changes the law. But even though that's what he probably meant, the alternative reading I'm proposing makes at least as much sense. It takes account of the fact that a reasonable Congress would want the executive and the courts to harmonize the goals of RFRA and whatever statutes RFRA limits, where possible. True, we don't actually have a reasonable Congress, but sound statutory construction has long proceeded on the assumption that, as Hart & Sacks put it, the legislature consists of reasonable people pursuing reasonable goals, reasonably--even if that's not actually true.