Invisible Constitutions: Concurring Opinions and Plurality Judgments under Marks v. United States

Abstract

The Supreme Court’s decision in Marks v. United States instructs lower courts interpreting plurality judgments to follow the opinion concurring on the narrowest grounds, or the opinion closest to the dissent, creating the possibility that the position of the Court may not be one favored by the median justice. While the Marks doctrine creates a problem theoretically, it is unclear how frequently these problems materialize. In this paper, we explore how frequently the Marks doctrine actually results in non-median outcomes. We conclude with thoughts about the importance of these cases and speculate about the future of the Marks doctrine.

Keywords:

• Marks doctrine
• narrowest grounds
• non-median outcomes

Notes

1 430 U.S. 188 (1977).

2 The phrase “narrowest grounds” is itself ambiguous, as has been noted in previous research (e.g., Re 2019), but we show that the Marks doctrine still works poorly even when using what we believe to be the most workable interpretation of “narrowest grounds,” which is the opinion closest to the position of the dissenters.

3 138 S.Ct. 1765 (2018).

4 564 U.S. 522 (2011).

5 3 U.S. 308 (1796).

6 2 U.S. 402 (1792).

7 We thank an anonymous reviewer for noticing that this roughly corresponds with Warren Burger’s tenure as Chief Justice; thus, it is possible that his leadership style may have played a role in the more fractured coalitions of this period.

8 430 U.S. 188, 193 (1977).

9 332 U.S. 46, 68 (1947) (Black, J., dissenting).

10 302 U.S. 319 (1937).

11 342 U.S. 165 (1952).

12 492 U.S. 490 (1989).

13 462 U.S. 416 (1983).

14 Confirmation Hearing on the Nomination of Samuel A. Alito, Jr. to be an Associate Justice on the Supreme Court of the United States. 109^th Congress, p. 508.

15 The three dissenters in Ramos v. Louisiana, 391 U.S. 145 (2020) note how a single justice’s opinion can be controlling under Marks.

16 591 U.S. ___ (2020).

17 505 U.S. 833 (1992).

18 See Re (2019) for a discussion of success using this tactic.

19 542 U.S. 600 (2004).

20 See Boyd (2015a; 2015b) for a more general discussion of strategic opinion writing by lower court judges to insulate themselves from reversal.

21 564 U.S. 522.

22 Id. at 534 (plurality opinion).

23 Id. (Sotomayor, J., concurring in the judgment).

24 Id. at 544 (Roberts, C.J., dissenting).

25 See Steinman (2008) for an argument that Freeman falls into a general case of “biconditional rules,” which preclude one opinion from being a subset of another.

26 However, in this circumstance judges may still have trouble identifying which of the two opinions is the narrowest, as seen in Webster.

27 To clarify the “implicit consensus” approach, consider a Venn diagram with three nested circles (see Williams 2017, 809). Under this approach, the narrowest grounds would be the smallest circle that is wholly contained within the other two circles.

28 See Neuenkirchen (2013) for a defense of the “fifth vote” approach both as normatively desirable and as an interpretation of Marks. Neuenkirchen (2013, 408, n. 164) argues that the implicit consensus and fifth vote approaches are likely to prove similar in practice, although he notes a “hypothetical exception” where the implicit consensus could lead to the “sixth Justice’s position” being controlling. We show that this is not merely hypothetical, but arises in multiple cases.

29 This approach has been criticized by other courts, e.g., in King v. Palmer, 950 F.2d 771 (D.C. Cir. 1991).

30 Whether a case is unidimensional or multidimensional often depends on the framing of the issues. We do not claim that no case that we classify as unidimensional could possibly be classified as multidimensional. Rather, we claim that even if we classify these opinions as unidimensional to avoid the well-known multidimensionality problems of Marks, the Marks doctrine still creates problems in these cases.

31 E.g., Ewing v. California, 538 U.S. 11 (2003); Holder v. Hall, 512 U.S. 874 (1994); Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749 (1985).

32 E.g., LULAC v. Perry, 548 U.S. 339 (2006); Denver Area Ed. Telecommunications Consortium, Inc. v. FCC, 518 U.S. 727 (1996); Sun Oil Co. v. Wortman, 486 U.S. 717 (1988). Sun Oil is an example of a case where the principal opinion is in fact styled as an “opinion of the Court,” but which the Supreme Court Database categorizes as not receiving five votes; however, it does receive five votes on each issue, and thus does not pose a problem of interpretation under Marks.

33 A list of all 23 cases where the median opinion and the narrowest opinion differ can be found in the Appendix.

34 While this data collection method was designed to find all the cases where Marks poses problems in a single dimension, two of the cases we found raise difficulties for the Marks doctrine because of multidimensionality and one (Troxel v. Granville) was difficult to classify. That the designation of an opinion as the narrowest is not always straightforward is another problem in applying Marks; see the Appendix table for further discussion of the most complex cases.

35 Consider the following example: A majority coalition consists of a plurality of 4, a special concurrence supported by 2 (concurrence A), and another special concurrence supported by 2 (concurrence B). There is a single dissenter from the right. If the plurality is clearly the left-most opinion but it is ambiguous whether concurrence A or concurrence B is narrower, either selection would pose a problem under the Marks doctrine. If concurrence A is narrower, then there would be a majority coalition that would prefer the position of concurrence B to concurrence A, since concurrence B is therefore closer to the four-justice plurality. Similarly, if concurrence B is narrower, six justices would prefer the position of concurrence A, since concurrence A is closer to the four-judge plurality. Therefore, resolving this ambiguity would not solve the Marks problem we explore in this paper.

36 All the mean comparisons are done with Welch’s unequal variances t-test. These tests cover data from the 1946-2009 terms, for which the salience measures are available.

37 U.S. v. American Library Association, 539 U.S. 194 (2003); Ashcroft v. ACLU, 535 U.S. 564 (2002); Board of Ed. of Westside Community Schools (Dist. 66) v. Mergens, 496 U.S. 226 (1990); Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555 (1980); NLRB v. Retail Store Employees, 447 U.S. 607 (1980); Lathrop v. Donohue, 367 U.S. 820 (1961).

38 138 S.Ct. 1765 (2018).

39 See Hughes v. United States, 138 S.Ct. 1765 (2018) for an example of the circuit splits resulting from attempts to interpret Freeman v. United States, 564 U.S. 522 (2011). Ramos v. Louisiana, 391 U.S. 145 (2020) provides a good example of a “Marks dispute.”

40 553 U.S. 35 (2008).

41 530 U.S. 57 (2000).

42 Planned Parenthood v. Casey, 947 F.2d 682, 694 n. 7 (3d Cir. 1991). Of course, Casey itself was not a case where the median and narrowest opinion differed, so this portion of the circuit court opinion is dicta; it has rarely been cited for this proposition by other courts. (On Google Scholar as of 9/30/2021, a search within the citing cases for the 3rd Circuit’s opinion in Casey for “necessary to secure a majority” only yields 7 results.)

43 553 U.S. 35 (2008).

44 631 F.3d 1139 (9th Cir. 2011).

45 553 U.S. at 35 (plurality opinion).

46 533 U.S. at 1556 (Thomas, J., concurring in the judgment).

47 631 F.3d at 1145.

48 Ventura v. State, 2 So.3d 194, 200 (Fla. 2009).

49 Jackson v. Danberg, 594 F.3d 210, 222–23 (3d Cir. 2010). See also Cooey v. Strickland, 610 F.Supp.2d 853, 920–21 (S.D.Ohio 2009) (noting that while there was “no clear holding from Baze,” nonetheless the instant case could be resolved because the intervening plaintiff would fail under any of the concurrences in Baze). The scholarly literature has also noted the problems with interpreting Baze under the “narrowest grounds” test. See Marceau (2009).

50 496 U.S. 226 (1990).

51 Id. at 231.

52 Id. at 263 (Marshall, J., concurring in the judgment).

53 Verbena United Methodist Church v. Chilton County, 765 F. Supp. 704 (M.D. Ala. 1991).

54 489 U.S. 288 (1989).

55 Taylor v. Louisiana, 419 U.S. 522 (1975).

56 476 U.S. 79 (1986).

57 478 U.S. 255 (1986).

58 It is worth noting that the median opinion in Teague appears to be that of White, not the plurality; White noted that he believed Court had made serious mistakes in its retroactivity jurisprudence and simply called the plurality’s approach “an acceptable application” of the line of precedent that he would presumably wish to revisit. 489 U.S. at 317 (White, J., concurring in part and concurring in the judgment). Since White’s approach essentially approves of the plurality’s standard without explicitly joining it, a judge who followed White’s opinion would have better predicted the future behavior of the Supreme Court than one who followed Stevens’s opinion.

59 501 U.S. 529 (1991).

60 468 U.S. 263 (1984).

61 Hulin v. Fibreboard Corp., 178 F.3d 316, 333 (5th Cir. 1999).

62 Glazner v. Glazner, 347 F. 3d 1212, 1216 (11th Cir. 2003).

63 Misciagno v. Secretary of DHHS, 786 F. Supp. 1120 (E.D. N.Y. 1992).

64 535 U.S. 564 (2002).

65 U.S. v. Kilbride, 584 F.3d 1240, 1254–55 (9th Cir. 2009). The opinion stated that Ashcroft had “no explicit holding” but noting that “five Justices concurring in the judgment, as well as the dissenting Justice, viewed the application of local community standards in defining obscenity on the Internet as generating serious constitutional concerns.”

66 The Eleventh Circuit’s misprediction of Teague v. Lane (see Thurmon 1992) in at least one panel may reflect ideological distance between the panel and the Supreme Court. Consider Hall v. Kelso, 892 F.2d 1541 (11th Cir. 1990). The panel, while majority Republican, contained two Republican judges well known for their support of civil rights in the South following Brown v. Board of Education (Elbert P. Tuttle and Frank M. Johnson, Jr.) (see Bass 1981). The Rehnquist-era Supreme Court thus likely sat to the right of this panel.

67 138 S.Ct. 1765 (2018).

68 Justice Sotomayor was the lone concurrence in a 4-1-4 decision in Freeman v. United States, 564 U.S. 522 (2011).

69 For example, in Ramos, Justice Alito writing in dissent states that if an opinion representing the views of only one justice is the narrowest, that opinion can overturn a prior Supreme Court precedent, 140 S.Ct. 1390, at 1431, while Justice Gorsuch’s plurality opinion disagrees, id. at 1403–04. Perhaps not surprisingly, given the topic, neither justice had five votes for his position.

70 Transcript of Oral Argument at 32, Hughes v. United States, 138 S.Ct. 1765 (2018) (No. 17-155).

71 Id. at 33-34.

72 Either Stevens’s opinion (casting doubt on the death penalty altogether) or Breyer’s (adopting the dissent’s test) is the narrowest.

73 It is unclear what the narrowest ground is here. Souter presents his opinion as being narrower because it avoids wading into the substantive due process issue, but in doing so he strikes down the law in a broader way (facially rather than as-applied). It is also unclear how to order the opinions in a unidimensional way. In fact, there appear to be two dimensions here: comfort with substantive due process as a rationale and decision on the constitutionality of the statute.

74 When ordering the opinions, it is not entirely clear which of Kennedy’s or Souter’s should be considered narrower; Kennedy focuses on the availability of state tort action for malicious prosecution (leaving open that the absence of such a tort would present a different case), while Souter argues that the plaintiff’s claims are actually based in the Fourth Amendment (leaving open that if the claims genuinely did not arise from a search or seizure, there may be a due process claim). Whichever is placed to the right leaves the other opinion as the median. This may well simply be a case of multidimensionality, which Marks also does not handle well.

75 However, White is vague as to whether he disagrees with O’Connor in any significant way.

76 It is not entirely clear how Blackmun and Stevens differ, but whichever one is narrower leaves the other one as the median; this is thus a problem under Marks, albeit a trivial one.

77 There are two dimensions here: whether the appropriate standard is objective or subjective and whether there was purposeful discrimination in this case. Blackmun falls into the median on both (refusing to answer the question of whether disparate impact would suffice and finding that there was purposeful discrimination but that the remedy was improper). On the second, he concurred on the narrowest grounds, but on the first, Stevens was actually closer to (at least some of) the dissenters.

78 Powell’s opinion appears to be both narrowest and the median if the justices are projected into one dimension. However, there is arguably a multiple dimensionality problem, as Powell would limit Sabbatino in a way that the plurality would not, while nonetheless being closer to the dissent on the question about the Bernstein exception. This poses its own problem under Marks.

79 In a single dimension, Brennan’s opinion would appear to be both the median and the opinion concurring on the narrowest grounds; however, as Justice Black notes, the remaining five justices believe that the constitutional issue was properly before the Court (although they split on its proper resolution), while the four members of the plurality do not. Thus, while the plurality is in some sense narrower, they are simply in the minority on that point. This shows that there are two dimensions in the case, posing a problem under the Marks rule.

80 However, in the companion cases of Turner v. Pennsylvania, 338 U.S. 62 (1949) and Harris v. South Carolina, 338 U.S. 68 (1949), Jackson dissented, making it relatively clear that his position in Watts should not be taken as the holding of the Court.