The Originalism Blog

Constitutional Authors, Howard and the Joint Committee Chris Green

Mike Ramsey noted yesterday Garrett Epps's generally useful comments on birthright citizenship in the Atlantic, responding to Rand Paul. Epps is right that there was clear evidence that the Fourteenth Amendment's citizenship declaration, which emerged out of the Republicans' meetings between the Senate's initial discussion of the  Amendment on May 23, 1866 and its discussion a week later, applied beyond the freedmen.  As Wong Kim Ark itself noted, John Conness of California explained the declaration--which applied to anyone subject to the jurisdiction of the United States, going beyond the citizenship declaration of the Civil Rights Act of 1866, which merely applied to those not subject to foreign powers--in terms of its effect on a few people in California who were subject both to the United States and to the emperor of China. Rand Paul is thus wrong to confine the citizenship declaration to the freedmen and also wrong to speak of its "author." 

In the constitutional sense, the Fourteenth Amendment's author was a group of groups: the bicameral Congress, together with the bicameral ratifying state legislatures (20 on my view, more on others'). These many individuals spoke collectively only through the constitutional language itself. As I see it, any particular framer's views of constitutional text are not controlling when they are idiosyncratic, i.e., out of line with the meaning conveyed by that text at the time in virtue of the rules of the language in which was written. Even setting aside this theoretical issue aside, we do not know who exactly drafted the language which Jacob Howard introduced to the Senate on May 30. In the intervening week, it was clear that the Republicans had caucused and agreed to add the citizenship declaration and modify sections 2 and 3. Howard introduced the amendments to the Senate, but did not claim to have composed the language himself. Epps is right about all this.

This is not to say, though, that individual congressmen were never described as authors. The Cincinnati Commercial described an August 24, 1866 speech by John Bingham as "The Constitutional Amendment Discussed by its Author" (see page 19 of the collection here). I take this, however, as hyperbole; Bingham was not the author in the constitutional sense of having authority to adopt it, nor did he even compose the words of the citizenship declaration, as he did the rest of Section One (proposing it to the Joint Committee on April 21).

It is also important to correct Epps's serious error in stating that "Howard had not been a member of the Joint Committee; no one regarded him as an 'author' of any part of the amendment." This is wrong; Howard was one of only six senators on the Joint Committee (see page iii here), and, moreover, the one chosen to explain the Fourteenth Amendment to the Senate in the place of chairman William Pitt Fessenden, who was ill, as well as the Senator who presented the Republican caucus's changes a week later.  The Amendment was, moreover, actually known at times as the "Howard Amendment." Howard's explanation of the Privileges or Immunities Clause, though not perfectly free from interpretive difficulty, was by far the fullest account in Congress in 1866, and his excellent work on the Recess Appointments Clause (see here) and the nature of congressional reconstruction power (see here) both testify to his very high status among the constitutional lawyers of the time.  His views on anything during Reconstruction are never to be lightly disregarded.

The Horror of Morrison v. Olson – Part III: There is No General Separation of PowersMike Rappaport

I want to conclude my posts on Morrison v. Olson by discussing the majority’s treatment of the separation of powers issue generally.  As I mentioned in my prior posts, the majority approved the judicial appointment of the Independent Counsel and the restriction on the President’s removal of the IC through specific doctrines that involved appointment and removal.  But then after approving these aspects of the statute (as well as some others), the Court examined what it regarded as the overall separation of powers issue: the Court asked whether these provisions together were consistent with the separation of powers generally.

This is a peculiar way to analyze the separation of powers.  The Court seemed to believe that the different provisions could pass muster under the particular appointment and removal doctrines, but still violate the separation of powers cumulatively.  Given the Court’s premises – a belief in lenient appointment and removal doctrines and a belief that the separation of powers was about some kind of balance between the branches – I suppose the Court’s analysis made some sense.  But these premises are extremely problematic.  The language of the Constitution neither adopts such lenient doctrines nor requires an independent judicial balancing between the branches.

Instead, the correct way to analyze these matters was to apply – in an unbiased way – the Appointments Clause and the Removal Doctrine (implementing the Vesting Clause).  And that is it.  There is no additional separation of powers test.  The separation of powers is simply the relevant provisions, not some other judicial principle that the justices divine.  Thus, if the IC statute passed the appointment and removal issues, then it was constitutional.  If it did not, then it was unconstitutional.

Or to put the point differently, the balancing of the branches is already conducted by the Appointments Clause and Executive Power Vesting Clause, not by the court under a vague test of whether the law unduly interferes with the executive power.

Let me conclude by discussing another problem with the Court’s general separation of powers test.  One aspect of the Court’s approach is to draw a distinction between 1) congressional aggrandizement and 2) Congress expanding or limiting the powers of the other branches.  The Court has indicated that when Congress expands its own powers – such as claiming an appointment power or a veto power – this aggrandizement will be subjected to relatively strict scrutiny.  By contrast, if the Congress either expands executive power (say by delegating quasi legislative power to the executive) or restricts executive power (say by restricting the President’s removal power), then it will be subject to much more lenient scrutiny.

In my view, this approach is also quite wrongheaded.  For one thing, the Constitution no more prohibits Congress from expanding its powers than it prohibits Congress from expanding or contracting the powers of the other branches.  For another thing, Congress can achieve similar results by either technique.  Congress can expand its power relative to the President either by increasing its own power or decreasing the President’s power.  Moreover, Congress can often benefit itself by giving another branch power, as when Congress avoids responsibility for controversial regulations by delegating quasi-legislative power to the executive.

In the end, Morrison v. Olson was a travesty.  The opinion was written by Chief Justice Rehnquist.  While many on the right praised Rehnquist, I was not much of a fan.  And he was never worse than in Morrison.

Noah Feldman on Textualism and Purposivism (and Wine)Michael Ramsey

I was a bit hard on a Noah Feldman column last week, so here's one I like (though I don't agree with it):  What Would Scalia Do With 2,447 Bottles of Wine? (commenting on the case Pennsylvania v. 2,447 Bottles of Wine).

Apparently Pennsylvania confiscated some wine that had been illegally brought into the state and the question was what to do with it.  As Feldman recounts:

A provision of Pennsylvania law dating back to the 1930s says that alcohol confiscated by the state can be destroyed or transferred “to a hospital for its use.”

Hoping to take advantage of this provision, the Chester County Hospital applied to the court to ask for the confiscated wine. The hospital intended to use the wine “to fund raise for charitable purposes.”

But Chester County Judge Edward Griffith refused to grant the request. In a nine-page order, he reasoned that in its original context, the law must’ve contemplated the use of confiscated liquor for medicinal purposes. Now, Griffith said, the law would not permit transferring the liquor to the hospital for sale. “Since the Liquor Code makes no provision for condemned wine to be sold for any purpose,” he reasoned, “the wine may not be delivered to a hospital for sale.”

Feldman says, and I agree, that a textualist would not reach this result.  As he describes it, Pennsylvania law just says the wine can be given to a hospital, period.  Presumably the drafters were thinking that the hospital would use it for medicinal purposes, but that intent isn't in the statute.  Another way to look at it is that intent was that the hospital could use it for any lawful purpose, which in those days was only medicinal, but now includes other things such as sale. In any event, the lawmakers could have, but did not, impose a statutory limit on use, so the court has no business creating one.

More significantly, Feldman also says that a purposivist should also find for the hospital.  This seems counterintuitive.  The judge in the case thought he could discern the statute's purpose (to allow medicinal use), and applied it to limit the text.  Why isn't that the right purposivist answer?  Feldman explains:

What should a court do when faced with a statutory purpose that is archaic, obsolete and no longer sensible? The law is still on the books, and it can’t be completely ignored. But it can be interpreted. And the interpreting court should follow Aristotle’s immortal advice: Imagine that those who drafted the law were reasonable people, and act as they would’ve wanted in the light of the new realities that now obtain.

Assuming the framers of the original law wanted confiscated liquor to help hospitals, what would they now want? Most likely, they would still want confiscated liquor to help hospitals. At one time, perhaps, that assistance came through medicinal use. Now it would come in other ways. Regardless, the statute’s true purpose is to take advantage of alcohol confiscation for the benefit of public health.

Assuming Feldman is right on purposive methodology, what he shows is that purposivism isn't really interpretation -- it's judges coming up with the result they think reasonable.  His conclusion as to what the legislature would want now is pure speculation.  First, it's an absurd counterfactual: what would a group of people who were in the state legislature in the 1930s, about whom we and Professor Feldman know nothing, what to do with confiscated wine now, 85 years later, knowing that hospitals have no use for it as medicine?  Second, his solution isn't the only available answer.  Perhaps those 1930s legislators, if informed that hospitals don't use wine for medicinal purposes anymore, would want to get rid of the wine altogether.  Or perhaps, once they understood that, post-prohibition, confiscated wine would have a market value, they would want the state to auction it off and use the proceeds for public health, or for something else.  In any event, once the medicinal link between wine and hospitals is broken, why should a hospital, as opposed to another institution, benefit?

It's pretty clear that Professor Feldman doesn't really care what the 1930s legislators would think.  What matters is what he thinks, as a reasonable interpreter.  Letting the hospital sell the wine seems reasonable, to him, so that's the answer.  And thus purposivism becomes the rule of judges.

The key move (and to my mind a somewhat deceptive one) is the shift from asking what the legislators thought they were doing when they passed the statute to asking what the legislators would want if they were passing the statute today.  The first question, which is what Judge Griffith asked, is a more-or-less manageable one, and can be justified as giving effect to what the legislators meant to enact.  It's not my preferred approach for a variety of reasons (including that it can be manipulated into Feldman's approach), but it is eloquently defended by Judge Robert Katzmann in his new book.  The second question, which is what Professor Feldman asks, really asks what the judge thinks is reasonable.  Asking it depends on a completely different theory of judging -- one that can (perhaps) be justified, but not one that can be justified on the ground of giving effect to the original legislators' intent.

The Horror of Morrison v. Olson – Part II: The Supreme Court Embarrasses ItselfMike Rappaport

In my last post, I noted how the Supreme Court worked hard to hold the Independent Counsel statute to be constitutional, presumably on the ground that it was essential good government, only to find the political system rejecting it as poor policy a decade later. In this post, I criticize some of the arguments the Supreme Court used to hold the statute constitutional.

One key feature involved the appointment of the Independent Counsel. The statute provided that the IC would be appointed by a court. This appointment method would be constitutional only if the IC were an inferior officer, since the Constitution requires that superior (or non-inferior) officers must be appointed by the President with the advice and consent of the Senate.

Thus, the Supreme Court argued that the IC was an inferior officer, claiming that four factors suggested the result.  The argument here really was quite laughable. The IC is an extremely important position and the IC is not subject to the supervision of anyone in the government. The idea that the IC was an inferior officer, who was not important enough to require senatorial consent, was absurd. When Ken Starr was investigating Bill Clinton, he was arguably the second most important officer in the entire government (after the President). Yet, the Court treated his position as that of an inferior officer. If that is not bad enough, the Court actually argued that the conclusion that the IC was an inferior officer was clearly true and therefore it was unnecessary for the court to draw a more precise line. 

Some years later, the Court largely reversed its analysis in Edmond v. United States. Justice Scalia wrote the decision, largely dismissing the Court’s analysis in Morrison. Instead, Justice Scalia emphasized his position in his Morrison dissent, that an inferior officer must be subordinate to some other official (under which approach the IC would, of course, have been a superior officer).  All of the members of the Court joined his opinion. One very reasonable interpretation of the Edmond is that the Supreme Court essentially overruled Morrison’s Appointment Clause analysis by narrowing Morrison’s analysis to the facts of that case.

Another key feature of the case involved the fact that the President was deprived of the authority to remove the IC. Under prior law, the Court had made clear that agencies that engaged in quasi legislative and quasi judicial power could be made independent of the President under Humphrey’s Executor v. United States, but that officers that engaged in pure executive power had to be removable at the pleasure of the President under Myers. Thus, the IC, which engaged in the pure executive power of criminal prosecution, had to be subject to removal at the pleasure of the President. And since the IC was not removable in that way, the statute was unconstitutional. But the Supreme Court ignored its precedents and simply announced a new standard. Anything to approve the wonderful Independent Counsel statute.

What was even worse was the Court’s inadequate justification for overturning these precedents. A couple of years later, Justice O’Connor would write an opinion in Casey that emphasized the importance of sticking to precedent. But Justice O’Connor joined the majority opinion in Morrison, which justified its failure to follow the precedent simply on the ground that the Court had changed its mind. As the Court wrote: “We undoubtedly did rely on the terms ‘quasi-legislative’ and ‘quasi-judicial’ to distinguish the officials involved in Humphrey's Executor and Wiener from those in Myers, but our present considered view is that the determination of whether the Constitution allows Congress to impose a ‘good cause’-type restriction on the President’s power to remove an official cannot be made to turn on whether or not that official is classified as ‘purely executive.’”

Unbelievable.

Garrett Epps on Rand Paul on Birthright CitizenshipMichael Ramsey

In The Atlantic, Garrett Epps:  How Rand Paul Misunderstands the Fourteenth Amendment  (commenting on Senator Paul's statement that "The original author of the—of the Fourteenth Amendment said on the Senate floor that this was applying to slaves, and did not specifically apply to others.").  Professor Epps responds:

To begin with, Fourteenth Amendment did not have an “author.” The first draft was written by Robert Dale Owen, the radical former Congressman who served on the American Freedmen’s Inquiry Commission. From Owen, the draft passed into a closed session of the Joint Committee on Reconstruction. History doesn’t record who said what, but the draft amendment emerged much edited and was approved by a two-thirds majority of the House. From there it went to the Senate.

But at that point it had no citizenship language. That was added by a closed Senate caucus after debate began on the draft. There is no way to know who was the “author” of the citizenship clause. When it emerged with its present wording, it was shepherded to passage by Michigan Senator Jacob Howard. Howard had not been a member of the Joint Committee; no one regarded him as an “author” of any part of the amendment. Howard and the other proponents made clear that the clause would apply to the children of all aliens born in the United States, including babies born to Chinese parents (who were ineligible to citizenship) and even “Gypsies” (though it’s not clear there were any in the U.S. at that point), with the exception of children of diplomats and children of Indian tribes. I have read the entire debate on the draft amendment a number of times; I have the five volumes of the Congressional Globe on a shelf by my desk. I never saw any proponent of the amendment say that the clause, or any other part of the amendment, applied to slaves and no others.

I agree.  But as Professor Epps also says, the drafters of the Fourteenth Amendment likely weren't thinking about illegal aliens, one way or the other, because there were few restrictions on immigration at the time.  That, however, seems to me to cut somewhat in Senator Paul's favor: there probably wasn't any specific intent by the framers to grant birthright citizenship to the children of illegal aliens, because they didn't even recognize the issue.

To me, though, this illustrates the problem of getting bogged down in framers' intent.  First, as Professor Epps says, how do we identify whose intent matters?  Second, even if we could decide that, how do we establish their intent, especially about something they didn't think about.  Third, even if they had an intent, why does that matter if they didn't incorporate that intent into the words they enacted?

I therefore prefer a textual approach to this and other constitutional problems.  As I've discussed before, I think the textual approach (unlike the intent approach) yields a determinate answer.  The textual question is whether the children of undocumented aliens are "subject to the jurisdiction" of the United States when they are present in the United States.  The enactors of the Fourteenth Amendment likely didn't have an intent regarding how children of undocumented aliens should be treated for citizenship purposes, but they likely did have a shared understanding of what it meant to be "subject to the jurisdiction" of the United States.

The Horror of Morrison v. Olson -- Part I: The Supreme Court's Felt NecessitiesMike Rappaport

Back when I first read Morrison v. Olson—the case on the constitutionality of the independent counsel statute—in 1988, I thought it was a horrible opinion. Back then I was at the Office of Legal Counsel and had a pretty strong bias in favor of the executive. But even though my views on executive power have changed significantly since then, I still believe the decision is just awful. I thought I would write a couple of posts on what is so wrong with Morrison.

Morrison involved a constitutional challenge on separation of powers grounds to the independent counsel statute. Under the statute, when there was evidence of wrongdoing on the part of the President or one of his advisers that could not be clearly ruled out as insignificant, the Attorney General was required to seek the appointment of an independent counsel from a federal court. The independent counsel was essentially independent of both the Attorney General and the President, although the Attorney General could remove the IC for cause.

The Pull of the Perception of Good Government

The first problem with Morrison is that the Court seemed especially influenced by its view that the independent counsel was an essential institution of good government.  The opinion does not say this explicitly, but at the time, virtually all Democrats and many Republicans believed that the IC was necessary to police wrongdoing by the executive branch. Thus, the nonoriginalist justices on the Supreme Court were not going to strike down this essential institution that they believed was so important.

The joke, however, turned out to be on these justices (and sadly on the Constitution). As time passed, the country came to understand that the IC statute was a poor one that was not essential to good government, but instead was an engine of partisanship and bad results. Thus, the Congress let the IC statute die at the end of the Clinton presidency, with both parties saying good riddance.

Notice how this case illustrates the problem with living constitutionalism. Not only did it involve a clear mistake by the Supreme Court as to policymaking. It also involved a politically biased one.  The IC statute was popular in 1987 with legal elites—Democrats and many moderate Republicans. It was only when it was used against a Democratic President that the elites changed their mind.

Andrew Schwartz on Regulatory TakingsMichael Ramsey

Andrew William Schwartz (Shute, Mihaly & Weinberger LLP) has posted No Competing Theory of Constitutional Interpretation Justifies Regulatory Takings Ideology on SSRN.  Here is the abstract:  

Compensation for excessive regulation of the use of property under the Just Compensation Clause of the Fifth Amendment has gained wide acceptance. Introduced in 1922 in Pennsylvania Coal Co. v. Mahon, and gathering considerable momentum in 1978 with Penn Central Transportation Co. v. City of New York, regulatory takings constrains government regulation protecting the environment, public health, consumer safety, affordable housing, and other community interests. Upon close examination, however, the regulatory takings doctrine does not appear to be justified by any of the competing theories of constitutional interpretation: textualism, originalism, or evolutionary document. Rather, the doctrine seems to arise from a misunderstanding of the Just Compensation Clause as guaranteeing a laissez-faire political economy. The initial parts of this article rely for the most part on existing scholarship analyzing regulatory takings under the textualist and originalists theories of interpretation. The bulk of the article is devoted to an examination of regulatory takings under the evolutionary document theory, which has received less attention in the literature of takings. The argument that the regulatory takings doctrine in its entirety is unwarranted under an evolutionary document approach is founded on the absence of precedent for granting the courts a significant role in the formulation of what is essentially economic policy, and profound conflicts between regulatory takings and core values of the Constitution, such as liberty, equality, and democracy. I also respond to claims that a broad reading of the Just Compensation Clause is necessary to balance the interests of property owners against society or that regulatory takings is a practical tool for property regulation. Finally, the article recommends an alternative system for government policy-making to control the use of property that relies almost exclusively on statutes and administrative regulations adopted by the political branches of government.

I'm sympathetic to the claim that regulatory takings doctrine isn't justified by the Constitution's original meaning.  But this paper is more interesting as an example of what I regard as the fallacy that living constitutionalism is a logical system capable of generating determinate outcomes rather than an exercise in judicial policymaking.  To be sure, the paper is a fine brief for eliminating regulatory takings doctrine.  But I can easily argue for limits on regulatory takings from a living constitution perspective:

Assuming the original Constitution did not have protection against regulatory takings, [I would say], the circumstances have changed dramatically.  The modern regulatory state and its threat to property rights could never have been imagined by the framers; but if they had imagined it, they would have seen that it poses as great a threat (if not a greater threat) to the security of private property rights than do "true" takings of the type the framers feared and which the Constitution limited.  To give property rights the degree of protection the framers would have wanted, we must evolve a regulatory takings doctrine they did not envision, to combat a threat they did not envision.  Otherwise, their core goal to protect private property rights as a foundation of republican democracy would be rendered empty.  The short of it is that, without some protection against regulatory takings, in the modern world private property is effectively at the mercy of the whim of the government, something the framers would never have tolerated.

Whether this line of argument is persuasive depends almost entirely on whether the hearer is broadly sympathetic to property rights or broadly sympathetic to the regulatory state.  That is, its resolution depends on the judge's policy intuition.  The idea that a resolution can be achieved by appeal to a logical process that transcends policy is pure fantasy.

To be clear, as I've said before, I don't necessarily see this conclusion as a conclusive argument against living constitutionalism.  Perhaps the question "should there be constitutional limits on regulatory takings?" should be resolved by appeal to the policy intuitions of whomever happens to be on the Supreme Court at the time.  My objection is to the idea that living constitutionalism contains some legalistic answer to that question, which I take to be the supposition of the paper.

Lee Strang: Original Meaning Originalism and Computer-Assisted Research TechniquesMichael Ramsey

Lee J. Strang  (Georgetown Center for the Constitution; University of Toledo College of Law) has posted Blunting the Instability Critique: Original Meaning Originalism and Computer-Assisted Research Techniques on SSRN.  Here is the abstract:      

In this Article, I bring together a widely observed phenomenon — the theoretical move toward original meaning originalism — with an unnoticed phenomenon — the use of computer-assisted research technologies and techniques (“CART”) in originalism. I argue that originalists’ conceptual move toward original meaning originalism, when coupled with their adoption of CART, will reduce the force of the Instability Critique — the claim that originalism’s reliance on history makes any resulting constitutional law unstable. Computer-assisted research techniques reduce legal instability by increasing the degree of epistemic determinacy of the foundational aspect of originalist analysis: the recovery of language conventions contemporary with ratification of the constitutional text. 

Originalism rests on the premise that it is able to ascertain the Constitution’s original meaning with reasonable accuracy. This will lead to a number of virtues, one of which is thatoriginalism leads to relative stability in constitutional law. It does so by tying constitutional interpretation, and resulting constitutional law, to the Constitution’s determinate original meaning. 

A recurring criticism of originalism is that, on the contrary, originalism leads to instability in constitutional law. Originalism leads to instability because it depends on an activity — the recovery of the Constitution’s meaning via the methods of history — that cannot bear the weight. Instead, the critics argue, the Constitution’s meaning is either unrecoverable in principle or, if it is recoverable, interpreters’ understanding of that meaning is necessarily subject to modification. Thus, even assuming good faith and diligent research, the criticism goes,originalism will inevitably lead to fluctuating constitutional meaning. 

In response to this criticism, originalists made a major conceptual move: they rearticulated originalism as original meaning originalism in place of original intent originalism. Originalists now focused on the constitutional text’s public meaning, when it was adopted. In this Article, I build on that conceptual move, and I tie it to a modification to the method of historical research for originalism that will make the process more accurate, thereby blunting the Instability Critique’s force. In particular, I argue that original meaning originalism’s focus on the text’s conventional meaning at the time of ratification, coupled with now-widely available CART, diminishes the force of the nonoriginalist Instability Critique, identified above. 

The language conventions contemporary with the Framing and Ratification are the building block of original meaning. Computer-assisted research permits — in a way unassisted techniques did not — the relatively easy and relatively accurate recovery of these language conventions. Originalism’s conceptual change, combined with this change in how originalists perform research, provides (much of) the interpretative stability claimed by originalists. However, CART will not eliminate the Instability Critique in five situations, which I describe.

Zachary Price: Politics of NonenforcementMichael Ramsey

Zachary S. Price University of California Hastings College of the Law) has posted Politics of Nonenforcement (Case Western Reserve Law Review, Vol. 65, No. 4, 2015) on SSRN.  Here is the abstract:      

Constitutional controversies over executive nonenforcement have emerged as a major theme of the Obama Presidency. Yet similar controversies arose in other recent administrations — and in past debates, the two political parties’ positions on this issue were often reversed. Building on previous work addressing constitutional principles that properly govern executive enforcement discretion, this brief symposium contribution reflects on these principles in light of our current, highly polarized politics. It does so in three ways. First, Part I provides historical perspective on current debates by describing major enforcement-related controversies from the Reagan and George W. Bush Administrations. Second, Part II proposes criteria for assessing how faithful an agency’s enforcement policy is to the agency’s underlying statutory mandate. As Part II explains, several qualities — most importantly, transparency and clarity — that are generally considered virtues in administrative law are often counterproductive in the enforcement context. Finally, Part III tentatively explores possible practical, political implications of weakening norms of executive enforcement obligation.

Is John Roberts Restrained?Michael Ramsey

At Bloomberg View, Noah Feldman: The Lonely Road Ahead for John Roberts.  He argues:

[Senator Ted] Cruz’s repudiation of Roberts, a fellow product of the conservative legal establishment, is just the latest confirmation of an astonishing process: The chief justice, a lifelong conservative who hasn't abandoned his views, is nevertheless being abandoned by conservatives -- without being embraced by liberals.

Having exercised judicial restraint in the two Obamacare cases and in the gay-marriage decision, Roberts is in the admirable and unenviable position of having a principled, across-the-board stand against activism. It's admirable because justices are supposed to have, you know, a coherent judicial philosophy. And it's unenviable because, in an era of activism, it wins you nothing but enemies on both sides.

But does Roberts really have a "principled, across-the-board stand against activism"?  In a word, no. (I use the phrase "judicial restraint," as Feldman apparently does, to mean deference to the political branches).  Consider the Chief Justice's positions in these cases:

Citizens United v. FEC (wrote 5-4 majority opinion invalidating federal campaign finance laws, over a strong dissent invoking deference).

McCutcheon v. FEC (wrote plurality opinion invalidating federal limit on campaign contributions, over a strong dissent invoking deference).

United States v. Stevens (wrote majority opinion invalidating federal law against depiction of animal cruelty)

United States v. Alvarez (joined opinion invalidating federal statute punishing lies about military service).

Snyder v. Phelps (wrote majority opinion invalidating state law giving remedies for offensive speech).

Brown v. Entertainment Merchants Association (concurred in judgment invalidating state law restricting violent video games, over dissent invoking deference)

Reed v. Town of Gilbert (joined majority opinion invalidating municipal sign regulation).

McCullen v. Coakley (wrote majority opinion invalidating state law regarding speech near abortion clinics).

True, these are all First Amendment cases.  Feldman acknowledges Citizens United (in a footnote!) but says (based on nothing) that "the First Amendment has always been a gray area for judicial restraint."  Well, how about non-First Amendment cases?

District of Columbia v. Heller (joined 5-4 majority invalidating federal gun regulation, over strong dissent invoking deference).

McDonald v. City of Chicago (joined 5-4 majority invalidating state and local gun regulation, over strong dissent invoking deference)

Shelby County v. Holder (wrote 5-4 majority opinion invalidating part of federal voting rights act, over a strong dissent invoking deference)

Free Enterprise Fund v. PCAOB (wrote 5-4 majority opinion invalidating removal provisions of federal statute establishing PCAOB).

Parents Involved v. Seattle (wrote 5-4 majority opinion invalidating local affirmative action program, over a strong dissent invoking deference).

Arizona State Legislature v. Arizona Redistricting Commission (wrote dissent that would have invalidated state redistricting law).

Arizona v. United States (joined majority opinion invalidating key parts of state immigration statute).

Of course, Roberts has voted to uphold laws on the basis of deference (most recently in the Obergefell same-sex marriage case).  But Feldman's claim is that Roberts' pursuit of restraint is setting him apart from the other conservative-leaning Justices (a "lonely road").  As in Obergefell, almost all of Roberts' "restraint" cases find him aligned with his fellow conservatives (death penalty, criminal procedure, voter identification).

Feldman's only evidence comes from the two Obamacare cases. Even these don't seem very persuasive.  In King v. Burwell, the statutory case, Roberts' vote seems as easily explained by an interpretive approach that looks more to congressional purpose than plain text (consistent with his vote in other statutory cases such as Yates v. United States and Bond v. United States -- both cases where he ruled against the government).  NFIB v. Sebelius, the constitutional case, points in various directions.  First, Roberts invalidated part of the law (relating to Medicare expansion).  Second, he upheld the individual mandate under Congress' taxing power only after finding it beyond Congress' commerce power -- a complicated (one might say convoluted) approach that may be the basis for future rulings against Congress while leaving this particular law standing.  NFIB thus looks as much like an attempt to preserve the core of one law rather than a general principle.

What Feldman needs (but can't find) is a pattern of cases in which Roberts voted with the liberal-leaning Justices on the basis of deference to uphold laws conservatives wanted struck down. While Feldman doesn't come up with any (apart from the Obamacare cases), I can think of a couple -- Williams-Yulee v. Florida Bar (First Amendment) and Comstock v. United States (federalism).  But these seem like outliers, hardly enough to show the Chief Justice's core interpretive principles.

You might think I am cherry-picking cases.  But in fact, I'm just taking the main cases in the leading Constitutional Law casebook (which I use) -- co-authored by Professor Feldman.  I assume, therefore, that Professor Feldman is well aware of the Chief Justice's long list of non-restrained cases.  What, then, is the point of his column?

A small goal may just be to highlight and celebrate the tension between Roberts and some conservatives (though not all) over the Obamacare cases.  But one may speculate that a bigger goal is to advance a narrative in which Roberts stars as the principled man of judicial restraint, to encourage him to break with conservatives in the future.  Consider that in the upcoming term there are major cases in which conservatives seek to overturn laws much favored by liberals: Friedrich v. California Federation of Teachers (compelled union dues); Fisher v. University of Texas (affirmative action at the university level); Evenwel v. Abbott (counting non-citizens for purposes of redistricting).  It would be nice (for liberals) to have Roberts leaning their way on deference grounds in these cases.

I doubt that Feldman's narrative will take hold.  The Chief Justice is indeed cautious, often preferring incremental rulings rather than sweeping ones.  But I see no evidence of overarching deference.  More likely, he sees himself doing what he famously said in his confirmation hearings -- calling balls and strikes.  In the umpire model, there's no weight given to strikes over balls; the umpire doesn't defer to the pitcher and the catcher on the right call.