(The contents of this post arose from close readings of the case and discussions with several folks--my thanks for inspiring the examination.)

Justice Alito's majority opinion in Burwell v. Hobby Lobby (PDF) includes two extended analyses that ultimately read as dicta, because the majority opinion refuses to address the issue--and they are both issues from which Justice Kennedy's concurring opinion appears to distance itself. Further, Justice Kennedy's concurring opinion appears to be a direct response to Justice Ginsburg's dissent. So it may be worth asking the question: did Justice Alito once have a much broader opinion, one that was later restrained in two aspects because Justice Kennedy refused to join them after reading Justice Ginsburg's dissent?

Compelling Government Interest

Here's how Justice Alito's majority opinion addresses the compelling government interest (with some citations omitted or abbreviated):

HHS asserts that the contraceptive mandate serves a variety of important interests, but many of these are couched in very broad terms, such as promoting “public health” and “gender equality.” RFRA, however, contemplates a “more focused” inquiry: It “requires the Government to demonstrate that the compelling interest test is satisfied through application of the challenged law ‘to the person’—the particular claimant whose sincere exercise of religion is being substantially burdened.” O Centro. This requires us to “loo[k] beyond broadly formulated interests” and to “scrutiniz[e] the asserted harm of granting specific exemptions to particular religious claimants”—in other words, to look to the marginal interest in enforcing the contraceptive mandate in these cases. O Centro.

In addition to asserting these very broadly framed interests, HHS maintains that the mandate serves a compelling interest in ensuring that all women have access to all FDA-approved contraceptives without cost sharing. Under our cases, women (and men) have a constitutional right to obtain contraceptives, see Griswold v. Connecticut, and HHS tells us that “[s]tudies have demonstrated that even moderate copayments for preventive services can deter patients from receiving those services.”

The objecting parties contend that HHS has not shown that the mandate serves a compelling government interest, and it is arguable that there are features of ACA that support that view. As we have noted, many employees—those covered by grandfathered plans and those who work for employers with fewer than 50 employees—may have nocontraceptive coverage without cost sharing at all.

HHS responds that many legal requirements have exceptions and the existence of exceptions does not in itself indicate that the principal interest served by a law is not compelling. Even a compelling interest may be outweighed in some circumstances by another even weightier consideration. In these cases, however, the interest served by one of the biggest exceptions, the exception for grandfathered plans, is simply the interest of employers in avoiding the inconvenience of amending an existing plan. Grandfathered plans are required “to comply with a subset of the Affordable Care Act’s health reform provisions” that provide what HHS has described as “particularly significant protections.” But the contraceptive mandate is expressly excluded from this subset.

We find it unnecessary to adjudicate this issue.

Well, that's an unusual conclusion! First, a brief note on Justice Alito's history on this issue: when it comes to asserting a "compelling government interest," Justice Alito has been no friend of the government when it has found the ability to carve out exceptions for some reasons, but not religious reasons.

Consider the "police-beard case" he decided on the Third Circuit. The government asserted a compelling interest in having Muslims shave their beards to serve on a municipal police force. But the government had carved out a medical exception to its beard policy. Justice Alito found that there was no compelling government interest where there had been some non-religious exceptions granted.

That's the logic much of this lengthy excerpt assumes. The government has created exceptions to its policy; that, to Justice Alito, seems like thin gruel as a basis for the government to turn around and assert that it really does have a compelling interest when it comes to religious exemptions to its policy. He emphasizes that the "interest" is not just generic ("women's health," etc.), but precise, citing O Centro.

But, curiously, Justice Alito, after several paragraphs of suggestively weighing against the government, turns to say that the Court need not decide the issue.

Justice Kennedy, however, appears to reject any compelling government interest.

As to RFRA’s first requirement, the Department of Health and Human Services (HHS) makes the case that the mandate serves the Government’s compelling interest in providing insurance coverage that is necessary to protect the health of female employees, coverage that is significantly more costly than for a male employee. There are many medical conditions for which pregnancy is contraindicated. It is important to confirm that a premise of the Court’s opinion is its assumption that the HHS regulation here at issue furthers a legitimate and compelling interest in the health of female employees.

Justice Kennedy, then, seems to reject all of the points raised arguendo (and in dicta) by Justice Alito--he does not construe "compelling government interest" narrowly, and he does not view the government's granting of some non-religious exemptions as fatal to a compelling government interest in its refusal to grant religious exemptions.

But Justice Ginsburg notes in her dissent (which Justice Kennedy characterizes as "powerful"):

Perhaps the gravity of the interests at stake has led the Court to assume, for purposes of its RFRA analysis, that the compelling interest criterion is met in these cases. [Footnote 23: Although the Court’s opinion makes this assumption grudgingly, one Member of the majority recognizes, without reservation, that “the [contraceptive coverage] mandate serves the Government’s compelling interest in providing insurance coverage that is necessary to protect the health of female employees.”]

...

Stepping back from its assumption that compelling interests support the contraceptive coverage requirement, the Court notes that small employers and grandfathered plans are not subject to the requirement. If there is a compelling interest in contraceptive coverage, the Court suggests, Congress would not have created these exclusions.

Federal statutes often include exemptions for small employers, and such provisions have never been held to undermine the interests served by these statutes. See, e.g., (FMLA, ADEA, ADA, and Title VII).

The ACA’s grandfathering provision, allows a phasing-in period for compliance with a number of the Act’s requirements (not just the contraceptive coverage or other preventive services provisions). Once specified changes are made, grandfathered status ceases. Hobby Lobby’s own situation is illustrative. By the time this litigation commenced, Hobby Lobby did not have grandfathered status. Asked why by the District Court, Hobby Lobby’s counsel explained that the “grandfathering requirements mean that you can’t make a whole menu of changes to your plan that involve things like the amount of co-pays, the amount of coinsurance, deductibles, that sort of thing.” Counsel acknowledged that, “just because of economic realities, our plan has to shift over time. I mean, insurance plans, as everyone knows, shif[t] over time.” The percentage of employees in grandfathered plans is steadily declining, having dropped from 56% in 2011 to 48% in 2012 to 36% in 2013. In short, far from ranking as a categorical exemption, the grandfathering provision is “temporary, intended to be a means for gradually transitioning employers into mandatory coverage.”

So, too, why does Justice Ginsburg exert so many paragraphs to refuting an argument that the majority raises arguendo and ultimately assumes without deciding? And why relegate the tip to Justice Kennedy in a quick footnote after speculating what "[p]erhaps" motivated the majority?

Least Restrictive Means

Justice Alito's opinion similarly opines at length that the least restrictive means would be government purchase of the contraceptives in dispute:

The least-restrictive-means standard is exceptionally demanding, see City of Boerne, and it is not satisfied here. HHS has not shown that it lacks other means of achieving its desired goal without imposing a substantial burden on the exercise of religion by the objecting parties in these cases. See [RFRA] (requiring the Government to “demonstrat[e] that application of [a substantial] burden to the person . . . is the least restrictive means of furthering [a] compelling governmental interest” (emphasis added)).

The most straightforward way of doing this would be for the Government to assume the cost of providing the four contraceptives at issue to any women who are unable to obtain them under their health-insurance policies due to their employers’ religious objections. This would certainly be less restrictive of the plaintiffs’ religious liberty, and HHS has not shown that this is not a viable alternative. HHS has not provided any estimate of the average cost per employee of providing access to these contraceptives, two of which, according to the FDA,are designed primarily for emergency use. Nor has HHS provided any statistics regarding the number of employees who might be affected because they work for corporations like Hobby Lobby, Conestoga, and Mardel. Nor has HHS told us that it is unable to provide such statistics. It seems likely,however, that the cost of providing the forms of contraceptives at issue in these cases (if not all FDA-approved contraceptives) would be minor when compared with the overall cost of ACA. According to one of the Congressional Budget Office’s most recent forecasts, ACA’s insurance coverage provisions will cost the Federal Government more than $1.3 trillion through the next decade. If, as HHS tells us, providing all women with cost-free access to all FDA-approved methods of contraception is a Government interest of the highest order, it is hard to understand HHS’s argument that it cannot be required under RFRA to pay anything in order to achieve this important goal.

HHS contends that RFRA does not permit us to take this option into account because “RFRA cannot be used to require creation of entirely new programs.” [Footnote 37: here the Court examines burdens that benefit third parties in a lengthy footnote. -ed] But we see nothing in RFRA that supports this argument, and drawing the line between the “creation of an entirely new program” and the modification of an existing program (which RFRA surely allows) would be fraught with problems. We do not doubt that cost may be an important factor in the least-restrictive-means analysis, but both RFRA and its sister statute, RLUIPA, may in some circumstances require the Government to expend additional funds to accommodate citizens’ religious beliefs. HHS’s view that RFRA can never require the Government to spend even a small amount reflects a judgment about the importance of religious liberty that was not shared by the Congress that enacted that law.

In the end, however, we need not rely on the option of a new, government-funded program in order to conclude that the HHS regulations fail the least-restrictive-means test.

Another curious conclusion! Why not? Justice Alito's opinion continues:

HHS itself has demonstrated that it has at its disposal an approach that is less restrictive than requiring employers to fund contraceptive methods that violate their religious beliefs. As we explained above, HHS has already established an accommodation for nonprofit organizations with religious objections. Under that accommodation, the organization can self certify that it opposes providing coverage for particular contraceptive services. If the organization makes such a certification, the organization’s insurance issuer or third-party administrator must “[e]xpressly exclude contraceptive coverage from the group health insurance coverage provided in connection with the group health plan” and “[p]rovide separate payments for any contraceptive services required to be covered” without imposing “any cost-sharing requirements . . . on the eligible organization, the group health plan, or plan participants or beneficiaries.”

We do not decide today whether an approach of this type complies with RFRA for purposes of all religious claims. [Footnote 39: the Court notes that the Little Sisters of the Poor case challenges this accommodation in a separate RFRA claim. -ed] At a minimum, however, it does not impinge on the plaintiffs’ religious belief that providing insurance coverage for the contraceptives at issue here violates their religion, and it serves HHS’s stated interests equally well.

Note first that this conclusion on "least restrictive means" leaves open a question--the means may still violate RFRA for other claims! So, the stronger basis for the majority's opinion--the one that would emphatically not result in any religious objections--is set aside in favor of the potentially more problematic option (in the majority's own terms).

Second, consider what Paul Clement argued at oral argument (PDF) in response to a question from Justice Sotomayor:

JUSTICE SOTOMAYOR: Will your clients claim that filling out the form, if you're saying they would claim an exemption like the churches have already?

MR. CLEMENT: We haven't been offered that accommodation, so we haven't had to decide what kind of objection, if any, we would make to that. But it's important to recognize that as I understand that litigation, the objection is not to the fact that the insurance or the provider pays for the contraception coverage. The whole debate is about how much complicity there has to be from the employer in order to trigger that coverage. And whatever the answer is for Little Sisters of the Poor, presumably you can extend the same thing to my clients and there wouldn't be a problem with that.

So because of the narrow objection of Hobby Lobby in this case--at least as articulated by their counsel--the bigger problem of the self certify provision is reserved for another day. (But, it is curious that Justice Alito did not cite to this concession at oral argument.)

Justice Kennedy praises this result:

But the Government has not made the second showing required by RFRA, that the means it uses to regulate is the least restrictive way to further its interest. As the Court’s opinion explains, the record in these cases shows that there is an existing, recognized, workable, and already-implemented framework to provide coverage. That framework is one that HHS has itself devised, that the plaintiffs have not criticized with a specific objection that has been considered in detail by the courts in this litigation, and that is less restrictive than the means challenged by the plaintiffs in these cases.

The means the Government chose is the imposition of a direct mandate on the employers in these cases. But in other instances the Government has allowed the same contraception coverage in issue here to be provided to employees of nonprofit religious organizations, as an accommodation to the religious objections of those entities. The accommodation works by requiring insurance companies to cover, without cost sharing, contraception coverage for female employees who wish it. That accommodation equally furthers the Government’s interest but does not impinge on the plaintiffs’ religious beliefs.

On this record and as explained by the Court, the Government has not met its burden of showing that it cannot accommodate the plaintiffs’ similar religious objections under this established framework. RFRA is inconsistent with the insistence of an agency such as HHS on distinguishing between different religious believers—burdening one while accommodating the other—when it may treat both equally by offering both of them the same accommodation.

He then specifies why he would not reach the initial issue raised in Justice Alito's opinion, that of government subsidy:

The parties who were the plaintiffs in the District Courts argue that the Government could pay for the methods that are found objectionable. In discussing this alternative, the Court does not address whether the proper response to a legitimate claim for freedom in the health care arena is for the Government to create an additional program. The Court properly does not resolve whether one freedom should be protected by creating incentives for additional government constraints. In these cases, it is the Court’s understanding that an accommodation may be made to the employers without imposition of a whole new program or burden on the Government. As the Court makes clear, this is not a case where it can be established that it is difficult to accommodate the government’s interest, and in fact the mechanism for doing so is already in place.

It is an intriguing gloss on RFRA--the "least restrictive means," to Justice Kennedy, at least in these terms, suggests that the government may not have to pursue the truly least restrictive means, but, perhaps, a kind of balance of a not-very-restrictive means, so long as the cost is low (i.e., "without imposition of a whole new program or burden on the Government"). (Further, some have already speculated that his approval of this regime suggests he would not be inclined to favor challenges to the self certify framework.)

Justice Ginsburg, like Justice Alito, pursues the "government subsidy" line of argument at length before acknowledging it is merely dicta:

Then let the government pay (rather than the employees who do not share their employer’s faith), the Court suggests. “The most straightforward [alternative],” the Court asserts, “would be for the Government to assume the cost of providing . . . contraceptives . . . to any women who are unable to obtain them under their health-insurance policies due to their employers’ religious objections.” The ACA, however, requires coverage of preventive services through the existing employer-based system of health insurance “so that [employees] face minimal logistical and administrative obstacles.” Impeding women’s receipt of benefits “by requiring them to take steps to learn about, and to sign up for, a new [government funded and administered] health benefit” was scarcely what Congress contemplated. Moreover, Title X of the Public Health Service Act “is the nation’s only dedicated source of federal funding for safety net family planning services.”
“Safety net programs like Title X are not designed toabsorb the unmet needs of . . . insured individuals.” Note, too, that Congress declined to write into law thepreferential treatment Hobby Lobby and Conestoga describe as a less restrictive alternative.

And where is the stopping point to the “let the government pay” alternative? Suppose an employer’s sincerely held religious belief is offended by health coverage of vaccines, or paying the minimum wage, or according women equal pay for substantially similar work? Does it rank as a less restrictive alternative to require the government to provide the money or benefit to which the employer has a religion-based objection? Because the Court cannot easily answer that question, it proposes something else: Extension to commercial enterprises of the accommodation already afforded to nonprofit religion-based organizations. “At a minimum,” according to the Court, such an approach would not “impinge on [Hobby Lobby’s and Conestoga’s] religious belief.” I have already discussed the “special solicitude” generally accorded nonprofit religion-based organizations that exist to serve a community of believers, solicitude never before accorded to commercial enterprises comprising employees of diverse faiths.

Ultimately, the Court hedges on its proposal to align for-profit enterprises with nonprofit religion-based organizations. “We do not decide today whether [the] approach [the opinion advances] complies with RFRA for purposes of all religious claims.” Counsel for Hobby Lobby was similarly noncommittal. Asked at oral argument whether the Court-proposed alternative was acceptable, counsel responded: “We haven’t been offered that accommodation, so we haven’t had to decide what kind of objection, if any, we would make to that.” Tr. of Oral Arg. 86–87.

Again, the bulk of the examination is on the government subsidy means; the self certify issue is almost an afterthought.

--

Now, perhaps all of this is nothing. Perhaps the opinions were naturally written in this way. But given some of the ostensibly superfluous discussions on materials assumed away and not deemed essential to the Court's holdings, it's at least a plausible interpretation that Justice Ginsburg's first dissent moved Justice Kennedy to distance himself from two key portions of the majority's opinion, which then was modified to ensure that there were five outright votes on the salient issues in the case.

And it's very possible, then, that Justice Kennedy's concurring opinion was a part of a bargain to push the majority to bracket two broader issues (finding no compelling government interest because of other exemptions in the overall statutory scheme, and finding that a less restrictive alternative would not be a matter of pending litigation in other RFRA claims). And it's possible that he was driven to that view after Justice Ginsburg circulated her dissent.

So, is this post rampant speculation, or a plausible interpretation?