« More Retirement Plan Lawsuits | Main | Research Assistants and Comment Letters »

Saturday, August 20, 2016

Baseline Hell and the Religion Clauses

Perhaps because religions have a special interest in the diabolical, the fires of baseline hell seem to burn particularly hot in the context the Religion clauses. As I noted years ago, "baseline hell" refers to the futility of arguing about whether some burden is the imposition of a "penalty" or the withholding of a special "benefit" in the absence of a theory of distributive justice. As a salient recent example of baseline hell, one need look no further than the recent debates over whether employers' immunities' from the so-called Contraception Mandate impose a "harm" on employees or protect employers from the constitutional "harm" of being dragooned into violating their conscience. The answer, of course, is "both or either, depending on how one defines the relevant baselines." The contestants nevertheless draw on ideas of "harm"" and "coercion" as if these bare words will settle their fight absent a consensus about how to measure baselines -- a consensus in short supply, especially in the weird twilight zone between the public and private that is the "private" workplace.

Consider, for instance, Mark Gedicks' and Rebecca van Tassell's argument that, because RFRA as construed by Hobby Lobby deprives female employees of a statutory benefit, it "harms" them and thereby constitutes an unconstitutional accommodation of regulation. Gedicks' and Van Tassell's argument turns on a particular choice of status quo baselines under which loss of a statutory right that existed before judicial enforcement of the religious accommodation constitutes a "harm" forbidden by the Establishment clause. One does not need to be an ultra-Crit to see that status quo baselines are not the only criterion available by which to assess whether a loss is a "harm." Why not instead use a predictive baseline that asks whether the entitlement conditioned by a religious accommodation would exist at all absent that accommodation? (Einer Elhauge has a nice recent formulation of predictive baselines that one could use to fine-tune the analysis). Such an inquiry would require one to ask whether the Mikulski Amendment that forms the statutory basis for ACA's Contraception Mandate would have been enacted had it carved out an exception to RFRA for contraception and other medical procedures that might offend employers' religious scruples. Given that the Mikulski Amendment was silent about contraception and RFRA, it is difficult to say for sure, but the firestorm of controversy over the Stupak-Pitt Amendment suggests that critical Blue Dog votes might have jumped ship had ACA expressly attempted to waive RFRA entitlements, dooming the entire legislation. (Indeed, the prospect of such ship-jumping might explain the Mikulski Amendment's coyness about mentioning contraception and focusing instead on "preventative tests" like mammograms).

There is a perfectly plausible argument, in short, that, far from depriving female employees of an employment benefit, the preservation of RFRA's protection for employers was the necessary price for the ACA's passage. If so, where's the "harm" from a limit on a statutory right that exist only because that limit was not repealed? I do not endorse either predictive baselines in general or their particular application to ACA and RFRA. I note only that trying to distinguish between permissible and forbidden accommodations by invoking notions like "harm" invites a Critical response for which I see no easy rejoinder. As I argue elsewhere, the same complaints can be registered against the Eisgruber-Sager "Equal Liberty" baseline (defining religious entitlements by what "relevantly similar" non-religious actors receive) or the Laycock "Substantive Neutrality" baseline (defining religious neutrality by whether a religious exemption increases the attractiveness of the religion to non-believers more than the absence of that exemption deters religious practice).

Welcome to Baseline Hell! I would suggest that route to the exit door is some form of decentralization that sends these baseline fights to subnational governments -- but that's a subject for a different post.

Posted by Rick Hills on August 20, 2016 at 03:57 AM | Permalink

Comments

If you have a law that requires employers to provide an employment benefit, except to (say) Jews, it doesn't solve the constitutional problem to point out that it wouldn't have been passed if it hadn't excluded Jews. The outcome might be that fewer laws are passed that provide general benefits, but that's just the price of the non-discrimination rule. It seems to me that it will be usually true that the excluded group is better off at the pre-enactment baseline relative to post-enactment with the exclusion, but I don't think the constitutional argument even requires that; the failure to give equal treatment is enough. These aren't the same constitutional problems as those posed by an unconstitutional accommodation of religion, but I don't see any obvious reason to come to a different conclusion here. If conditioning one person's entitlement to a general benefit on another person's personal religious views amounts to establishing religion, I don't see why the analysis differs based on which baseline we pick.

Posted by: JHW | Aug 20, 2016 9:41:59 AM

JHW: in responding to Rick's post about the relevant baseline for assessing "harm," your hypo involving a law that discriminates in favor of a particular religious group isn't analogous to laws like RFRA that accommodate all religions and thus at most discriminate against individuals with a non-religious desire for accommodation. The reason is that the relevant "harm" from laws that discriminate in favor of a particular religious group is the *lack of equal treatment*, not the *lack of a substantive benefit*: the inequality harm is eliminated *even if* the legislature levels down by depriving everyone of the substantive benefit. By contrast, the alleged "harm" from laws like RFRA that accommodate all religions is not that non-religious individuals are deprived of *equal treatment*, because the SCt has squarely held that it's permissible to accommodate religious objections without accommodating non-religious objections, see Cutter v. Wilkinson, 544 U.S. 709, 724-25 (2005). Rather, the alleged "harm" from laws like RFRA is the *narrower* claim that the accommodation may sometimes (but not always) deprive non-religious third parties of *substantive benefits.* Cf. id. at 720. And *that* harm thus requires determining the relevant baseline for measuring what "benefit" has been lost and what "burden" has been imposed, which is the point of Rick's post.

Posted by: Hash | Aug 20, 2016 3:16:30 PM

PS, JHW: Perhaps an easier way of explaining the distinction is this: in your hypo, the persons "harmed" are the non-Jews who *also* want the accommodation but are denied it *based on their religion*, which is why *that* harm can be redressed by either expanding or eliminating the accommodation so that all individuals are treated *equally* regardless of their religion; whereas in RFRA cases, the persons allegedly "harmed" are not individuals who *also* want the accommodation (since non-religious individuals can't object to the fact that they don't get accommodations too), but rather third-parties who *don't want* the plaintiff to get the accommodation because they're supposedly losing a "substantive benefit" -- yet *that" harm exists only if they would have gotten the benefit but for the accommodation, which isn't the case if the legislature wouldn't have created the benefit at all if accommodations were prohibited.

Posted by: Hash | Aug 20, 2016 3:33:06 PM

Hash, I could not (and did not) say it better myself.

Of course, RFRA and Employment Div. v Smith both also provide for an equality-based baseline, in which religious actors can argue that they were denied an exemption given to relevantly similar secular actors. That sort of baseline invites JHW's argument, but it also invites endless and unresolvable debates over whether religious reasons are analogous to secular ones. The argument that ACA's small business exemption "discriminates against" religious businesses because the former but not the latter are exempt from the Contraception Mandate has this futile flavor.

Posted by: Rick Hills | Aug 20, 2016 5:46:32 PM

I'm a bit confused about the "predictive" baseline. Surely, not a single member of Congress even imagined a RFRA exemption to a contraception regulation when deciding whether to vote for the ACA, let alone relied upon that prospect as a condition for his or her affirmative vote. Indeed, I imagine none of those who voted for the ACA even imagined the accommodation that the agencies have afforded the nonprofits.

Posted by: Marty Lederman | Aug 21, 2016 5:40:38 AM

Marty, of the 29 states that had some sort of contraception mandate in their laws regulating insurance policies prior to the Congress's enacting ACA, at least 20 twenty specifically included a religious exemption in those contraception mandates. So I am a bit mystified as to why you assert that "not a single member of Congress even imagined a RFRA exemption to a contraception regulation when deciding whether to vote for the ACA." To the contrary, I imagine that any member of Congress that thought at all about a possible contraception mandate was advised by staff that RFRA could very well impose an exemption of uncertain scope on that mandate.

For predictive purposes, why is not the relevant question whether the Mikulski Amendment would have been enacted had it contained a clause expressly repealing RFRA? If such language would have doomed the Amendment, then why cannot an opponent of the contraception mandate argue that the price of the Amendment's passage was tacit acquiescence to RFRA's limitation of the contraception mandate?

By the way, those two questions in the last paragraph are not rhetorical: I really am curious about how one would select a baseline in this context. FWIW I have no policy dog in the fight over the contraception mandate. As a matter of religious policy, it seems to me (as I believe it seemed to Doug Laycock) to be dubious that an employer's having to fill out a form somehow encroaches on his or her right of conscientious objection. On the other hand, as a libertarian type of guy, I generally am unimpressed by the idea that mandating lots of benefits in insurance policies is a great idea as a matter of policy. So please take me to be in earnest when I say that I really do not care who prevails in Zubik or related cases.

But I really do care about baselines and, in particular about whether rhetoric about "harms" (to employers' conscience, to employees' rights to benefits, etc.) is anything more than hot air. So if you've got some convincing argument that ACA's passage did not depend on the Democrats' studious silence about RFRA's application (or non-application) to contraception mandates, I'd love to hear it.

Posted by: Rick Hills | Aug 21, 2016 6:00:17 AM

I have read some on the contraceptive mandate and "bare words" is not quite how I'd phrase things in this context. The writers spell out what "harm" and "coercion" etc. means to them particularly. Since "harm" and "coercion" will matter somehow (e.g., the Free Exercise Clause itself is in a basic way concerned about "free" exercise), the concepts will surely arise.

It is complicated to provide a "theory of distributive justice" when discussing these matter, but at the end of the day, this too does seem to arise. We at times do have to look at the arguments as a whole, not specific moments. It still involves a bit less too -- fleshing out what "harm" and "coercion" means in somewhat less broad terms and what too much means.

There is of course a lot of debate there. Thus, a "theory of distributive justice" specifically might not be necessary to pick the proper "baseline." Unless any baseline (set by precedent, practicality etc.) requires some overall "theory." For instance, a question of "if" is raised. This sounds like a question of fact, not exactly "a theory of distributive justice."

Let's say the Democrats "studiously" remained silent on certain tricky issues, which sounds like something both sides do a lot when dealing with legislation. This doesn't really tell me if there is only "hot air" even if it tells me something about legislative compromise. RFRA itself, however, put a lot of responsibility into the hands of the courts, which would individually handle questions instead of having Congress provide specific opt-outs.

There is nothing SPECIAL about ACA here as compared to any number of laws where tricky religious liberty issues are left to later. RFRA seems to me in place as a catchall to deal with that. So, I don't think saying Democrats in some fashion (I use the term loosely) "waived" the issue for sake of passing the law ("the compelling state interest") works. I don't see the "hot air" even if the whole thing was pushed to another day like many a legislation.

Plus, to the degree different religious debates (e.g., over mammograms v. contraceptives) will threaten legislation differently is telling. More to me on the First Amendment problems of selective exemptions and how realistically [if unfortunately] we let that pass when "abortion" or the like somehow is involved but not sure how much this alters the "coercion" and "harm" involved overall.

And, what exactly does the "RFRA's protection for employers" even entail? We had 5-4 split on that, so what the Democrats were accepting (whatever some "single member" thought about -- citing state practice here is of limited help -- state practice itself isn't standard) is unclear. This doesn't really seem "hellish" to me though. It's complicated line-drawing. Not quite Lucifer.

Anyway, we are talking federal legislation here involving national insurance policies and gender equality. I think "decentralization" is of limited value there. A case involving local education policy seems like an easier call.

Posted by: Joe | Aug 21, 2016 10:53:48 AM

I'll leave it there but would like to underline this concerns me:

"Why not instead use a predictive baseline that asks whether the entitlement conditioned by a religious accommodation would exist at all absent that accommodation?"

That's a tricky thing to do, determining in effect a hypothetical, especially if you are a judge (given special responsibility under RFRA). Plus what WAS the fair understanding of the "religious accommodation"? That's a major debate. For instance, the abstract that leads the article by Gedicks and Van Tassell cites "the existing regulatory regime" of exemptions ... is "that" the accommodation necessary for passage?

Is the claim made that if the final votes knew that Hobby Lobby would have went the way the dissent wished that the law might not have passed?

Posted by: Joe | Aug 21, 2016 11:15:55 AM

Let's assume that the Mikulski Amendment wouldn't have passed had it repealed the RFRA as applied to the contraceptive mandate. How much follows from this? It seems to me that we could also plausibly speculate that a version of the Mikulski Amendment with an express religious exception wouldn't pass either for want of Democratic votes. Even if that's not true (though especially if it is), what this thought experiment shows isn't whether "the entitlement conditioned by a religious accommodation would exist at all absent that accommodation" (as that formulation assumes there is some accommodation), but rather whether the entitlement would exist at all absent silence regarding an accommodation that would leave the question open to judicial interpretation of the RFRA.

Posted by: Asher Steinberg | Aug 21, 2016 7:21:20 PM

Asher, I think that you frame the question very well -- and the complexity of that last sentence in your comment illustrates why I regard the problem here of defining entitlements as hellishly tricky and likely futile.

I hope that the following correctly re-states your point. Arguably the Mikulski Amendment would not have passed had either side attempted to use some sort of perfecting amendment to clarify RFRA's application to the Mikulski Amendment. But this prediction suggests that no one has any "entitlement" in RFRA's either applying or not applying to the Amendment. The Court, therefore, cannot use the "harm" to female employees from limiting the Amendment with RFRA as a reason to distinguish the accommodation sought in Hobby Lobby from any other accommodation that some people want but other people do not want them to have -- say, an exemption from a zoning law excluding non-residential uses from a residential zone. In the zoning context, the neighbors do not want the church to have the exemption (because, say, they do not like crowded events in their neighborhood on Sunday), and the church wants the exemption (because they, perhaps, want to be located near their members in a quiet area). The burden on one group is equal and opposite to the burden on the other. The same follows in the employment context: The workers want RFRA not to apply, and the employer wants it to apply, and which side is "entitled" to their desired result turns on how the courts construe RFRA and ACA together. Neither side can claim a loss of a "legal entitlement," a "property interest," or a "vested right" as a reason to grant or withhold the exemption, without engaging in circular reasoning, because the legal status of the thing that the two sides wants turns entirely on how the courts construe RFRA.

Is that a fair restatement of your point, or have I missed your point?


Posted by: Rick Hills | Aug 21, 2016 9:30:55 PM

This whole discussion has me, quite frankly, confused. For one thing, no one, on either side of the debate, wants RFRA not to "apply," and few legislators would have voted for an amendment saying "RFRA doesn't apply" to the ACA, or to the Mikulski Amendment, or to virtually *any* other federal law (which is why Congress never votes to "un-RFRA" any legislation).

The question Rick is asking, instead (at least as I understand it) is -- *applying* RFRA (as everyone thinks we should), and *assuming* that the contraception reg, even with its accommodation, imposes a "substantial burden" on a nonprofit employer's exercise of religion (on which I'm dubious, but will assume for sake of argument here), would granting that exemption burden the women who do not receive cost-free contraception? And the answer to that question is very obviously "yes," not in some pre-legal, or metaphysical sense, but in virtually every sense we've ever considered when assessing harms.

Simply put, such a woman will either (i) pay money out of her own pocket for the same contraception; (ii) have an unintended pregnancy; or (iii) live in fear of such a pregnancy because she is not using contraception that is as effective. And, FWIW, she will be doing so in a social setting in which we have collectively concluded that women generally *should* -- and almost all women will -- have a legal entitlement to cost-free access to the most effective forms of contraception, especially IUDs. I think that virtually everyone would agree that any woman in that situation would quite justifiably conclude that the RFRA exemption has harmed her.

The RFRA question, then, is whether the government has a compelling interest in avoiding that harm to women (which I think is easily answered yes -- avoiding unplanned pregnancies is a very good thing -- but here's where the disagreements begin), and, if so, whether denying the RFRA exemption is the so-called "least restrictive means" of advancing that government interest.

I think we've been waylaid here by a nonissue, one triggered by Alito's silly suggestion in Hobby Lobby that, because women until very recently had to live with the risk of unintended pregnancies, the government must not have a very compelling interest in guaranteeing them access to effective contraception (or, at least, doing so in connnection with their existing, comprehensive health insurance plans). Once one jettisons that mistaken notion, however, I'm not sure that we need to descend into Rick's "baseline hell."

Posted by: Marty Lederman | Aug 22, 2016 6:45:44 AM

And, put that way, it occurs to me that (perhaps) what Rick is getting at is that there's no pre-legal, objective way of determining which of the myriad interests the state endeavors to advance are "compelling" and which are less-than-compelling. And that therefore any legal test, such as RFRA and many, many others we confront every day, that depends on assessing the relative value of a government interest (is it compelling? significant? important? etc.), is incoherent. Which is, of course, is some sense correct--there's no *objective* way to determine, say, whether grants to support "National Chia Seeds Week" are or are not more important than laws designed to reduce unintended pregnancies (not, at least, without agreeing on some sense of the "good"). In which case: What's new?

Posted by: Marty Lederman | Aug 22, 2016 7:02:55 AM

Marty, I apologize for confusing you. So let me try again.

1. The issue raised by my post and Gedicks' and Van Tassell's article has nothing whatsoever to do with the interpretation of RFRA -- for instance, whether enforcing the Contraception Mandate against religious employers constitutes a substantial burden on religion or whether the federal government has a compelling interest in imposing any such burden.

2. The topic here is instead whether RFRA's exemption from the Contraception Mandate for religious employers would violate the Establishment clause, as an unconstitutional accommodation of religion. Nothing in your comment speaks to that issue, because you are focused on the litigation of Hobby Lobby -- a topic irrelevant to this post and to Gedicks and Van Tassell's article.

3. The argument that exempting religious employers from ACA's contraception mandate might constitute such an unconstitutional establishment of religion turns on the idea that, by allowing employers to withhold a statutory entitlement from their employees, such an accommodation of employers' religious beliefs would "harm" those employees in ways that ordinary exemptions from generally applicable laws do not harm third parties. On this reasoning, exempting native groups from narcotics laws for their ceremonial use of peyote would not be an unconstitutional accommodation of religion, because no one has any entitlement to control the ingestion of peyote by third parties and, thus, no one is "harmed" by the exemption. By contrast, employees DO have an entitlement under ACA to health insurance coverage, so they ARE "harmed" by the exemption conferred by RFRA (assuming indeed that RFRA confers such an exemption -- an issue about which I express no opinion: see (1) above).

4. THIS distinction between "harmful" and "harmless" accommodations, however, requires a question-begging assumption about baselines the relevance, I hope, is now plain.

Posted by: Rick Hills | Aug 22, 2016 7:27:13 AM

Oh, in that case, Rick, I think I might agree with you, at least in some abstract sense. For one thing, while I am sympathetic to the idea that the permissibility of religious exemptions, under the EC, turns in part on the nature and degree of harm to third parties (see Caldor, etc.), I am not persuaded by Gedicks/van Tassel/Schragger/Schwartzman/Tebbe that the any harm to third parties makes an exemption necessarily unconstitutional. (Among other things, that idea can't be reconciled with Amos and, more indirectly, Hosanna-Tabor.) I also agree that, whether we're talking the EC or RFRA itself, we should not treat all legal benefits or entitlements alike -- the question can't possibly be answered simply by asking whether third parties will be denied something to which the law would otherwise entitle them.

On the other hand, I do think that, like countless laws and constitutional doctrines, RFRA itself -- the compelling interest test -- requires us to assess the importance of various third-party interests. And as far as that goes, the contraceptive cases are not close to whatever line you think should be drawn: I would hazard to guess, for example, that you and virtually everyone you know and respect has long treated the avoidance of unintended pregnancies as a very important objective, indeed. And for good reason . . . .

And I continue to think that what you describe as the "predictive baseline" question is singularly unhelpful here, *especially* if it turns on the question of whether a "RFRA does not apply here" amendment would have doomed the prospect of enacting this, or any other, bill.

Posted by: Marty Lederman | Aug 22, 2016 7:42:53 AM

Marty: I don't understand your final point. While I happen to agree with your first point that harm to third parties doesn't necessarily render an accommodation unconstitutional under the EC, Rick's post is showing why there's arguably no "harm" here *even if* one thinks "harm" is relevant to the EC analysis.

And, in that regard, it seems exceedingly relevant if the Mikulski Amendment would not have passed if it had made clear that the accommodation sought by the Zubik petrs would not be provided, because that means that, regardless of the accommodation, the petrs' employees were never going to get contraception through petrs' plans, and so there's no "harm" from the accommodation.

In assessing the relevance of the predictive baseline, I think you've gotten hung up on the inartful phrasing "rfra doesn't apply", when what Rick simply meant was "the requested accommodation will not be provided".

Posted by: Hash | Aug 22, 2016 8:41:12 AM

Hash: I'm unpersuaded. To be clear, there are at least two ways you can see an equal treatment problem in cases where RFRA results in selective denial of a general benefit:

1. Only certain objectors get protected, i.e. the law favors religious objections over secular objections. I'm not relying on this argument and I agree that it is foreclosed by Cutter.

2. Certain people (with respect to the contraceptive mandate, the employees of religiously-objecting employers) are denied a benefit generally afforded to others, because of religious beliefs they do not share. Cutter does not foreclose this argument.

What about 2. requires asking whether the general entitlement would have come into existence in the absence of the exemption? Framing the matter as an abstract denial of substantive benefits is a bit misleading here because of course there is no general constitutional problem with failing to provide a general benefit like contraceptive coverage without copay. Instead, the constitutional problem as framed by cases like Estate of Thorton v. Caldor is that denial of the benefit is predicated on the religious views of others--some people's statutory entitlements are curbed to accommodate the religious views of others. That's true whatever story you tell about the enactment of the ACA. The only "baseline" we need is the general rule, here, that of the contraceptive mandate.

Posted by: JHW | Aug 22, 2016 8:51:55 AM

Is there *any* constitutional context where we ask this question? If a state enacts a viewpoint-based restriction on what can be said within a designated public forum, is it a defense to a free speech challenge if we're convinced the forum wouldn't have been created in the absence of the restriction? Does Missouri win Trinity Lutheran Church v. Pauley if it can convincingly demonstrate that its playground subsidy wouldn't have been enacted if people had known it might go to church schools?

Yes, these are equal treatment claims (of a sort), too. But that's because they rest on a government program (a designated public forum, state grants for rubber playground surfaces) that is not independently constitutionally required. We get around the "baseline hell" of a society where the ground rules are contested by construing substantive constitutional rights as non-discrimination rules. What's going on here is just the same move: there is a substantive constitutional right (by hypothesis) to not be deprived of a benefit on account of someone else's religious beliefs, and our "baseline" for such purposes is the rule for cases where religious beliefs play no part.

Posted by: JHW | Aug 22, 2016 9:08:50 AM


Rick: This is a helpful exchange -- thanks to Marty for alerting me to it. I don't have time to respond fully right now, but I do want to quickly point out that Schwartzman and Schragger and I have two new pieces dealing with these issues. On the baselines question, we argue that the question of whether third parties have been harmed by religious accommodations can only be answered using normative argument. Whether an "objective" answer to such questions is available is of course a familiar and deep problem, but I for one reject fundamental skepticism about it (though not in this paper):

http://ssrn.com/abstract=2812121

On the question of third party harms, we agree that the mere existence of harm to others does not automatically defeat a religious accommodation, and we propose an "undue hardship" standard. I want to emphasize, too, that the principle of avoiding harm to others is not exclusively rooted in the Establishment Clase, but is incorporated into free exercise doctrine and ought to be a part of RFRA analysis as well.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2811815

Rick, I'd be interested to hear your reactions to these arguments, which are different from the ones you address in the post and (I think) in your chapter. It's terrific that you're writing in this area.

Nelson


Posted by: Nelson Tebbe | Aug 22, 2016 9:10:52 AM

JHW: I think there's a flaw in your theory #2. I don't understand you to be disputing that, if Congress can't provide the accommodation only to religious employers, Congress could respond by denying the benefit to everyone in order to avoid burdening the religious employers. But the existence of that option contradicts the hypothesized broad constitutional right "to not be deprived of a benefit on account of someone else's religious beliefs" -- indeed, there, *all employees* would be deprived of the benefit on account of the religious beliefs of *only some employers*. This, I think, is why any such EC claim must be framed more narrowly -- not as the loss of *any* benefit where motivated by a third-party's religious beliefs, but only a benefit to which the individual *would otherwise be entitled,* thus necessitating the baseline inquiry that Rick and I have suggested. (If instead you believe that Congress *can't* eliminate the benefit for everyone *for the purpose* of avoiding burdening religious employers, then you're just disputing our premise that that would be a constitutionally valid baseline.)

Posted by: Hash | Aug 22, 2016 10:42:49 AM

Two quick thoughts beyond what Nelson has written above:

1. The crit arguments above apply equally to libertarian baseline claims, including in the context of the Hobby Lobby litigation, e.g., holding that the government never imposes burdens when it lifts obligations it has imposed. That argument assumes a baseline that also begs the question.

2. The predictive baseline described above trades normative reasoning about the interests that existing laws protect for counterfactual speculation about possible legislative outcomes. The question isn't why go for predictive baselines, but given the indeterminacy of the counterfactual inquiry, why would a crit do that?

Posted by: Micah Schwartzman | Aug 22, 2016 3:02:44 PM

Society gets a drizzle of dumb little laws when it neglects big wise laws.

Posted by: Art Deco | Aug 22, 2016 3:59:17 PM

Post a comment