Friday, September 13, 2019

Methodological Quibbles and Their Non-Quibbly Implications

Guest Blogger

For the symposium on Andrew Coan, Rationing the Constitution: How Judicial Capacity Shapes Supreme Court Decision-Making (Harvard University Press 2019).

Frederick Schauer

Andrew Coan’s important book[i] on the limits of the Supreme Court’s decision-making capacity is built on three foundational points, all of which are correct. First is that the Supreme Court’s overall policy and political importance and salience are often overestimated.[ii] The offhand comments of political pundits notwithstanding, the nation’s most highly salient policy questions are located some distance away from the Supreme Court’s docket. Even the most important health care questions, for example, are being resolved (or not) legislatively, politically, and administratively rather than judicially. The same applies to immigration, where what the Supreme Court has does is not nearly as consequential as a host of executive, political, and administrative decisions. And when we move to tariffs, climate change, relations with Israel and Iran and China, treatment of victims of sexual violence and sexual harassment, monetary policy, opioids, and the size of the American military presence abroad, for example, the gap is even greater. And just as most of the important political and policy issues overlap only loosely with the Court’s business, so too is most of what the Court does – even on questions of abortion, sexual orientation, and religion – located well down on the list of what the sophisticated polling data indicate are the public issues that Americans think most important.

Coan’s second point is that crisply formulated rules, including those emanating from the courts, can forestall litigation. Jeremy Bentham recognized the litigation-stifling character of clear rules more than two centuries ago,[iii] and although Bentham was motivated far more by revulsion of the courts than recognition of their limited capacity, he, like Coan, understood that litigation is typically prompted by the indeterminacy of the governing law. And thus if a legislature (for Bentham, and for the Napoleon of the Napoleonic Code) or the Supreme Court (for Coan) sets forth precise and easily understood conduct rules, the domain of dispute and thus of judicial dispute resolution is contracted.
Third, and relatedly, Coan demonstrates that doctrines of judicial deference to decisions of other branches or institutions may also reduce the scope of judicial power and the degree of judicial involvement with questions of policy. Whenever a court adopts highly deferential standards of review – rational basis and abuse of discretion being obvious examples, and Coan usefully catalogs others – it makes its own job smaller.

But although Coan is correct in identifying and documenting these phenomena, the heart of his book lies in his claim that the limitations of judicial capacity in general, and Supreme Court capacity in particular, play a substantial role in explaining why the Court leaves so many important issues aside, why it formulates sharp-edged rules of conduct, and why it so often adopts deferential standard of review. But although Coan accurately describes the limited capacity of the Supreme Court, a capacity that would remain highly limited even if the Court reverted to its earlier practice of deciding roughly 150 cases a year rather than the current 70, the soundness of his causal hypothesis seems less clear. Assuming (correctly) that the Court’s agenda overlaps only poorly with the nation’s policy and political agenda, assuming (correctly) that the Court often defers to other branches and other institutions, and assuming (less obviously correctly[iv]) that the Court often formulates crisp rules of conduct, the question is whether these practices are the consequence of the Court’s limited decision-making capacity.

The aforesaid causal question is rendered difficult by the availability of some number of alternative explanations – explanations for these three phenomena that might not involve the constraints of a limited capacity Supreme Court. For example, it might be that a majority of the Justices a majority of the time believed that deferring to Congress, to administrative agencies, or to the states was simply the right thing to do, perhaps because of theories of federalism, democracy, separation of powers, or judicial competence. Or perhaps for some of the Justices some of the time, and thus perhaps for a majority of the Justices some of the time, first order policy or ideological preferences are doing much of the work. For example, Chief Justice Warren’s deferential approach to Congress in United States v. O’Brien[v] might have been caused by his worry that aggressive judicial patrolling of congressional motivation or of incidental effects on speech would produce too many challenges and thus strain judicial capacity, as Coan posits, but it might instead have been the product of an aversion to non-linguistic anti-war protests, an aversion that would explain Warren’s dissent in Street v. New York.[vi] And then-Justice Rehnquist might have joined the majority in Washington v. Davis,[vii] a decision to which I shall return presently, not because he was concerned about the strain on judicial capacity that would be produced by judicial scrutiny of unintentional discriminatory effect, but because of his narrow view of equal protection more generally.

Coan is admirably sensitive to such problems of ambiguous causation, and thus surveys and explores as alternative possible causes for the Court’s doctrines of deference and its occasional (Coan would say frequent) preference for comparatively rigid rules three candidates offered by the political science literature – the effect of relatively formal law (the legalist model), the first order policy preferences of the Justices (the attitudinal model), and the more sophisticated forms of attitudinal models that fall under the heading of strategic models. Coan finds all of these explanations less globally potent than his own judicial capacity model and thus concludes, to take a quote from his discussion of equal protection but one applicable to his various other doctrinal examples, that “[t]he judicial capacity model does not explain the Court’s decision to invalidate most race-based affirmative action policies or its decisions to invalidate state bans on same-sex marriage or same-sex intimacy. But it does explain why conservative justices who oppose economic regulations of business and high tax rates on the wealthy as a matter of policy have resisted the temptation to subject such laws to serious Equal Protection review. The same goes for liberal justices and laws that burden – or fail to assist – workers, consumers, and the poor. The judicial capacity model also explains why the justices have pursued their ideological goals through a fairly rigid framework of tiered scrutiny that clearly marks off various narrow categories of government action for serious review, while subjecting all others to essentially categorical deference.”

There is much that could be said about these insightful and ambitious explanatory claims, but I want to focus here on just two issues. One is the possibility of an important omitted variable, and the other is the methodological question of what domain the explanatory claim purports to explain, and what methods might justify such a conclusion.

With respect to the omitted variable, one concern, a concern that applies not only to Coan but also to most versions of attitudinal models, as my colleague Josh Fischman documents and analyzes in a forthcoming article,[viii] is just what the justices might have attitudes about. Plainly the justices, like most aware and educated Americans, have attitudes about abortion, affirmative action, same-sex marriage, and the place of religion in the public square, among many others. But the justices are likely, even more than the educated public, also to have views about, for example, federalism, separation of powers, and the role of unelected and life-tenured courts in making policy. Those views might, to be sure, themselves be the products of the justices’ views about the limited capacity of the courts, but they might instead be capacity-independent views about democracy, elections, efficient governmental organization, checks and balances, and much else. And to the extent that the justices have internalized such second-order attitudes about the extent to which they should hold back on effectuating their first-order attitudes, these second-order attitudes may offer an alternative explanation for why the Court has often developed doctrines that restrain even the first-order attitudes of those who have developed and internalized them. And thus, for example, Justice White, the author of the majority opinion in Washington v. Davis, and whose expansive views about equal protection are exemplified by his willingness in Frontiero v. Richardson[ix] to treat sex discrimination as a genuine suspect classification and in Milliken v. Bradley[x] to join the dissenters from the Court’s resistance to multi-district remedies for segregation, might have been concerned less about judicial capacity than that injecting the courts into too wide a range of legislative and administrative decisions would be inconsistent with his own conception of democracy. We do not know whether this alternative explanation is sound, but without considering the possible effect of such second-order attitudes about the role of the courts, attitudes that tend unfortunately to be neglected in most of the attitudinal literature, we cannot conclude with any confidence that worries about judicial capacity are leading the justices to stifle their own first-order policy preferences.

That such an alternative explanation might exist does not, of course, establish that this alternative explanation is sound. But acknowledging that limitation leads to another concern, one that is almost entirely methodological. Admirably, Coan seeks to offer alternative explanations for many of the Court’s outcomes and thus to test his own judicial capacity explanation against these alternatives. Coan’s testing, however, employs a methodology that is at the very least controversial. In his own words, he seeks to draw “causal inferences” about the causes of the Court’s doctrines of deference and about its seeming preference for litigation-limiting rules.  In doing so, he adopts a “qualitative case-study design,” but then a great deal turns on the selection of the cases to be studied. And here Coan says that he has selected only “a broad and diverse range of capacity-constrained domains, but no normal domains, because the causal condition on which the judicial capacity model is premised is not present in those domains. This method of case selection . . . [seeks] not to generate a representative sample of the general population of cases, but instead to identify the subset of that population in which a causal inference drawn from individual cases holds.”

Coan’s approach, one often denigrated as “selecting on the dependent variable,” and which is as problematic for qualitative as for quantitative research,[xi] does indeed have the capacity to tell us something about the subset of cases in which the Court’s limited decisional capacity plays a role. And that is an interesting and important contribution, my caveat about the omitted variable aside. But the method that Coan employs does not have, nor does it purport to have, the ability to explain the causal factors at work in the full set of which this subset is a part. The method might be useful in suggesting causal hypotheses about that full set, but actually testing those causal hypotheses, even with qualitative and interpretive methods, nevertheless requires looking at the full set, and at the very least considering those instances in which the alleged causal agent was present but for which the hypothesized effect was not present.[xii]  If, for example, there are instances both within Coan’s doctrinal groups and also of other groups in which the judicial capacity hypothesis seems not to hold, that might tell us something about its generalizability. Were the Court as concerned with judicial capacity as Coan posits, for example, it is unlikely it would have been as willing to allow judicial scrutiny of unintended effects as it has been in dormant commerce clause,[xiii] Establishment Clause,[xiv] and free speech cases,[xv] or as willing as it has been in various areas to use loose lists of factors rather than crisp rules.[xvi]

This is not to say that Coan’s judicial capacity explanations are unsound for some cases in some areas. But judicial capacity, which is indeed as limited and as important as Coan documents, may yet not have as much an effect on the Supreme Court’s doctrines as Coan concludes. Whether it does, and whether such an effect explains Supreme Court decision-making and opinion-writing more generally, remains either to be proved or disproved. Coan has offered us a genuinely novel and more than plausible hypothesis about one factor that might influence the Court’s behavior, but broad-scale testing of that hypothesis remains as a task for the future.

Frederick Schauer is David and Mary Harrison Distinguished Professor of Law at the University of Virginia. You can reach him by e-mail at schauer at

[i] Andrew Coan, Rationing the Constitution: How Judicial Capacity Shapes Supreme Court Decision-Making (2019).

[ii] See Frederick Schauer, Foreword: The Court’s Agenda – And the Nation’s, 120 Harv. L. Rev. 4 (2006).

[iii] Jeremy Bentham, Of the Limits of the Penal Branch of Jurisprudence 226-29 (Philip Schofield ed., 2010) (1789); Gerald J. Postema, Bentham and the Common Law Tradition 404-06 (1986); Dean Alfange Jr., Jeremy Bentham and the Codification of Law, 55 Cornell L. Rev. 58 (1969).

[iv] On the Court’s unfortunate reluctance to formulate broad and crisp conduct-guiding rules as often as Coan and I would prefer, see Frederick Schauer, Abandoning the Guidance Function – Morse v Frederick, 2007 Sup. Ct. Rev. 316.

[v] 391 U.S. 367 (1968).

[vi] 394 U.S. 576 (1969).

[vii] 426 U.S. 229 (1976).

[viii] Joshua Fischman, “Politics and Authority in the U.S.Supreme Court” (forthcoming 2020). Cf. Lisa A. Kloppenberg, Playing It Safe: How the Supreme Court Sidesteps Hard Cases and Stunts the Development of Law 2001).

[ix] 411 U.S. 677 (1973).

[x] 418 U.S. 717 (1974).

[xi] Gary King, Robert O. Keohane, & Robert Keohane, Designing Social Inquiry: Scientific Inference in Qualitative Research (1994).

[xii] Id. at 38.

[xiii] E.g., Hunt v. Washington Apple Advertising Comm’n, 432 U.S. 333 (1977).

[xiv] E.g., Lemon v. Kurtzman, 403 U.S. 602 (1971), which may no longer be good law after The American Legion v. American Humanist Ass’n (June 20, 2019).

[xv] E.g., Ward v. Rock Against Racism, 491 U.S. 781 (1989). Whether the Court’s often-announced effects scrutiny in free speech cases actually has any bite is hardly clear. See Clark v. Community for Creative Non-Violence, 468 U.S. 288 (1964). Nor is it clear that it can or should. See Larry Alexander, Trouble on Track Two: Incidental Regulation of Speech and Free Speech, 44 Hastings L.J. 921 (1993).

[xvi] E.g., Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). And also, Coan’s interpretation to the contrary, United States v. Lopez, 514 U.S. 549 (1995), and United States v. Morrison, 529 U.S. 598 (2000).

Older Posts
Newer Posts