I am pleased to announce the publication of my article, “Obergefell, Fisher, and the Inversion of Tiers,” in the University of Pennsylvania Journal of Constitutional Law.[*] The full article is available on Maryland Carey Law School's Digital Commons, see here, and on SSRN, see here.
The post might be a bit technical for readers without a legal background, so here's some context. The article explains a very important set of Supreme Court doctrines known as tiers of scrutiny. These tiers often control how high profile cases, involving race, gender, sexual orientation, and fundamental rights, are resolved. For lawyers, the key takeaway from many such cases is the answer to the question: "Which tier has the Supreme Court chosen to apply to this category of case?"
Here I offer a novel framing that emphasizes the importance of dimensionality and multicriterial decision making, a theme that has informed several prior blog posts. My thesis is that the apparent doctrinal inconsistencies in this area arise from the Supreme Court's failure to match its tiers of scrutiny to the underlying dimensionality of the caselaw to which the tiers apply. I explain why the system of tiers cannot be, and thus should not be, abandoned. Instead, the existing framework can and should be adapted to better match the dimensionality of the caselaw. Doing so would add much needed clarity to these heavily criticized and often misunderstood doctrines.
One last comment: Until now, my constitutional law scholarship has largely focused on what you might think of as structural constitutional law, or what we law professors refer to as con law I (structure and governance). I've written fairly extensively on standing and justiciability, stare decisis (precedent), the commerce clause (affirmative and negative), the narrowest grounds rule, issue and outcome voting, and the like. This is my first major venture into the part of constitutional law often identified as Individual Rights, or what we law profs call con law II (equal protection and due process). As I've told my students, my reticence reflects the simple fact that it has taken me 25 years to figure this out, and thus actually have something worthwhile to say.
This article offers a comprehensive framework that explains the pervasive doctrinal anomalies associated with the Supreme Court's tiers of scrutiny doctrines. It also offers a simpler path forward that would produce more consistent and predictable applications within the general framework of existing tiers. The article does not recommend abandoning the tiers’ doctrines or replacing them with a single tier or a more finely grained menu of tiers. Instead, it demonstrates why a system of tiers approximating ours is an inevitable feature of our constitutional jurisprudence, and it explains how to avoid the doctrinal anomalies that have generated much-deserved critical commentary.
The tiers doctrine historically began with a binary division, rational basis or strict scrutiny. The former, a more relaxed standard, was the baseline, and the latter, a more stringent standard, required a justificatory trigger, typically a suspect class, such as race, or a fundamental right, such as contraceptive access. The later-introduced third tier, intermediate scrutiny, was principally developed for cases involving gender-based distinctions, and it has not been applied in a consistent manner. The most significant anomalies in the system of tiers involve applying strict scrutiny to sustain race-based affirmative action, and applying rational basis scrutiny to strike laws adversely affecting non-suspect (or quasi-suspect), yet politically vulnerable groups. The result has been to transform the once-binary scheme into a system with the following five categories:
As a matter of black letter law, it is commonplace to express the tiers along a simple linear scale, as shown in Table 1, ranking the tiers from lax to stringent. In terms of predicting case outcomes, however, the Table 1 presentation fails. When the Court applies rational basis plus, it strikes the challenged law, and when it employs strict scrutiny lite, it sustains the challenged laws. In terms of prediction, therefore, strict scrutiny lite abuts traditional rational basis, with both used to sustain the challenged classification, and rational basis plus abuts strict scrutiny, with both used to strike down the challenged classification. On first principles, it is hard to imagine that anyone would devise our present inverted system of tiers, generating the sequence 14325, as reflected in Table 2. The analytical difficulty involves the failure to match the scheme of tiers with the underlying dimensionality of the case law.
Dimensions are normative scales of measurement used to evaluate virtually anything being compared. Some dimensions involve simple binaries—black versus white, male versus female—although such binaries oversimplify as applied to some individuals. Other dimensions present more nuanced scales of measurement, for example, continuous gradations of height or weight. Single dimensions often combine multiple criteria. Larger objects tend to be heavier, allowing us to rank modes of transportation—a scooter, a bicycle, a car—in a manner that captures both size and weight. And yet, such alignments sometimes break down. Adding an aloft hot air balloon—larger than a car yet lighter than a scooter—forces the need to split the dimensions of size (scooter, bicycle, car, then hot air balloon) and weight (hot air balloon then scooter, bicycle, car).
Advocates of an array of tiers, including former Justice Thurgood Marshall, fail to recognize that even a single dimension is capable of sorting infinite data provided the dimension captures the relevant normative stakes. And yet, even a small number of data will thwart a single dimension if that dimension fails to capture those stakes. Conversely, advocates of a single tier, such as retired Justice John Paul Stevens, fail to recognize that new descriptors along a single dimension, marking which laws are or are not permissible, will necessarily emerge. Because lower courts will come to associate those articulated characteristics as bases for striking or sustaining challenged laws, the new terminology will, over time, replicate the system of tiers, albeit with less guidance, at least until the system sorts itself out.
To illustrate, we can sort infinite integers along the odd/even dimension or along the prime/non-prime dimension. And yet, we need both dimensions—odd/even and prime/non-prime—to sort the deceptively simple sequence 2,3,4 when the two sets of criteria are combined. The number 2, the sole even prime, forces a split of these two dimensions, just as the hot air balloon forced a split over the dimensions of size and weight. By analogy, the inversion of tiers has arisen because the category of benign racial classifications is the hot air balloon (or number 2) in our equal protection jurisprudence. Benign racial preferences force a split in the dimensionality of tiers that the Supreme Court has refused to allow. The result has been to contort strict scrutiny to sustain the narrow set of relevant challenged laws, as seen in Regents of the University of California v. Bakke, 438 U.S. 265 (1978), and Grutter v. Bollinger, 539 U.S. 306 (2003). This is evident in Justice Kennedy’s changed approach as between the first round involving race-based affirmative action at the University of Texas, in Fisher v. Univ. of Texas at Austin, 133 S. Ct. 2411 (2013) (Fisher 1), and the second round, in Fisher v. University of Texas at Austin, 136 S. Ct. 2198 (2016) (Fisher II). In Fisher I, Kennedy criticized Justice O’Connor’s relaxed version of strict scrutiny applied in Grutter, stating that although strict scrutiny need not be fatal, nor should it be feeble. Fisher I, 133 U.S. at 2421. And yet, in Fisher II, when his vote proved decisive as to whether race-based affirmative action would survive equal protection scrutiny, Kennedy adopted O'Connor's deferential approach, thereby sustaining the race-advertant overlay atop the top 10 percent ensured-admissions plan as a means of providing the elite university ongoing access to among the state's most highly qualified African American students.
Table 3, which presents the jurisprudence of race in two dimensions, illustrates the analytical difficulty:
Consider what the two intuitively opposite extreme positions concerning race nonetheless hold in common: Although modern liberals condone benign racial classifications, and Jim Crow condones adverse racial classifications, both groups permit some express reliance on race. By contrast, the color-blind position rejects any express reliance on race. The valence of color blindness, a once liberal position (as shown in Justice Harlan’s dissent in Plessy v. Ferguson, 163 U.S. 537 (1896)), that is now a core element of conservative race jurisprudence, arises as a unique consequence of this country’s tragic historical treatment of race. Indeed, both modern liberals and modern conservatives claim that their competing normative views are a necessary consequence of the tragic lessons of Jim Crow.
Until the modern era, reliance on race was almost invariably coupled with the intent and effect of disadvantaging oppressed minorities, most notably African Americans. Whereas in the era of Jim Crow, the categories of antidiscrimination and anti-subordination went hand in hand, benign racial classifications now force a dimensionality split in which modern conservatives insist upon antidiscrimination, and modern liberals insist instead upon anti-subordination. The dimensionality of race is demonstrated by the peculiar fact that although modern liberals and Jim Crow resolve each of the two core inquiries reflected in Table 3 in opposite fashion, both allow some use of race, whereas the color blind position resolves in favor of each camp on one issue (failing to condone adverse reliance on race, along with modern liberals, and failing to condone benign reliance on race, along with Jim Crow), yet insists upon an opposite outcome respecting the constitutional permissibility of employing race.
Dimensionality complicates tiers analysis because by insisting upon classifying benign race cases (discriminating, but not subordinating) under strict scrutiny, the Court has inevitably contorted its strict scrutiny test to make the cases fit. By contrast, if the Court allowed a separate test (intermediate scrutiny), or if it split the dimensions across the two sets of cases—benign and adverse reliance on race each along its own permissibility spectrum—for separate treatment, the cases would naturally align within the existing framework of tiers.
The problem of dimensionality is endemic to race. It does not arise in other equal protection settings. For example, cases implicating gender sort neatly along a single dimension of anti-subordination. This does not mean that hard cases fail to arise, that the line of permissibility is unwavering, or that the Supreme Court has always gotten it right. It simply means that as a general proposition, we lack a principled normative commitment to sex-blind jurisprudence akin to that associated with race. Modern liberals and modern conservatives sometimes disagree on where to draw the line of constitutional permissibility in gender cases, but they implicitly agree that the normative inquiry in such cases involves the single dimension scale of anti-subordination. For the binary division along single dimension scale, the traditional two-tier scheme, strict or rational basis scrutiny, is adequate to the task. Indeed, reliance on intermediate scrutiny to do the work of the more traditional tests is manifest in the ongoing volley represented in such cases as United States v. Virginia, 518 U.S. 515, 533 (1996) (intermediate scrutiny as de facto strict scrutiny), Tuan Anh Nguyen v. INS, 533 U.S. 53 (2001) (intermediate scrutiny as de facto rational basis review), and Sessions v. Morales-Santana, 137 S. Ct. 1678 (2017) (intermediate scrutiny, once more, as de facto strict scrutiny).
Ironically, the Court has removed intermediate scrutiny where this alternative test has work to do, and it has imposed intermediate scrutiny where it does not. This insight helps to explain yet another tiers’ anomaly: the equal protection clause, adopted to combat the historical treatment of race, now permits greater legislative flexibility respecting benign gender-based than benign race-based classifications.
The article further explains the animus cases as providing one-time passes to strike laws adversely affecting vulnerable groups for whom legislative classifications might sometimes be appropriate, and thus without calling into question the presumptive validity of laws more generally affecting such groups. It also offers insights into Justice Kennedy’s principal reliance in Obergefell v. Hodges, 135 S. Ct. 2071 (2015), on due process, rather than equal protection, to strike the ban on same-sex marriage. While the ruling appears to avoid tiers altogether, the strategy is destined to fail in the long term. We now know that bans on same-sex marriage fall on the prohibited side of a binary divide, implicating the dimension of anti-subordination. We will eventually learn if there are any permissible bases for relying on sexual orientation in legislative classifications (for example sexual orientation preferences to eradicate past discriminatory practices), and thus where, if at all the permissibility line along that dimension will be drawn.
Ignoring the problem of tiers cannot make the need for tiers disappear. Recognizing the role of dimensionality in the application of tiers, however, provides the basis for simpler path forward, one that will improve the clarity of case presentations and analyses in this important body of case law.
This post was originally published, in an anticipation of the article's publication, in Balkinization, see here. Special thanks to Jack Balkin for permission to reprint it. This post has been updated to include two subsequently issued Supreme Court decisions: Fisher v. University of Texas, 136 S. Ct. 2198 (2016) (Fisher II), and Sessions v. Morales-Santana, 137 S. Ct. 1678 (2017). Both are discussed in greater length in the full article.
* Maxwell L. Stearns, Obergefell, Fisher, and the Inversion of Tiers, 10 U. Pa. J. Const. L. 1043 (2017).
1. This is presented as Table 5 in full article, which has additional tables.
2. The choice of label is less relevant than is the need to devise a third category, and thus to acknowledge a second analytical dimension. Indeed, the term intermediate scrutiny carries the unfortunate connotation of a mid-point along a single dimensional scale between rational basis and strict scrutiny.