By Stephen E. Smith*
Introduction
Legal readers have expectations for the writing we consume. This is especially true of judicial writing, the opinions that take up so much of our attention. We expect a certain format. We also expect rigorous honesty, attention to detail, and other qualities.[1] Among these expectations is consideration of the state of the law, both within and without the jurisdiction. We expect an opinion to manifest not only the bare act of law-making, but also sufficient attention to an issue’s context and history. We expect the whole story.
As I read the Second Circuit’s opinion in Glatt v. Fox Searchlight, Inc.,[2] I was struck by its “say-so” approach to the resolution of a legal issue. Reading it with the eyes of an experienced legal reader, but without much substantive knowledge of the issues before the court, I thought it odd that the issues were appearing with so little background to guide the court’s determination. The case reads as though the topics it addresses have never come up before. Like Athena springing from Zeus’s head, I was witnessing the test for determining whether an individual is a properly unpaid intern, or an improperly unpaid employee, spring from the panel’s pen.
I decided to investigate further, in order to discover whether the intern/employee legal field was as barren as the opinion led me to believe. After all, a court’s say-so may be all it has to work with. Perhaps I was witnessing the best effort that could be made in the circumstances. It turns out, however, that much was left on the table. The Second Circuit omitted significant extra-circuit authority on multiple issues.[3]
While there is no ethical requirement for courts to canvass their sister circuits for support and disagreement with their positions, a court can construct a more convincing argument by incorporating the history of the issues before it. Providing a provenance to ideas can both bolster them and demonstrate good faith in attempts to resolve them. Without a complete legal context, an experienced legal reader may be suspicious of the legitimacy of the claims. This Essay is agnostic on the correctness of the outcome in Glatt. Rather, it opines on a part of the process by which that outcome was reached.
- Judicial Opinions, Norms of Completeness, and the Persuasive Power of Including Previous Authority
The Federal Judicial Council describes judicial opinions as serving three functions: (1) to “communicate a court’s conclusions and the reasons for them,” (2) to “announce the law,” and (3) to “impose[] intellectual discipline on the author, requiring the judge to clarify his or her reasoning and assess the sufficiency of precedential support.”[4]
The functions of providing reasons and clarifying reasoning emphasize adequate justification. “[O]ne of the purposes of an opinion is to legitimate courts’ decisions.”[5] A reasoned opinion “assures the public that the decision is the product of reasoned judgment and thoughtful analysis, rather than an arbitrary exercise of judicial authority.”[6]
As a result, there is a persuasive aspect to opinion writing. Beyond the “announcement of the law,” a court’s opinion tries to earn the adherence of those to whom it is addressed, primarily, other judges and lawyers.[7] “The judge’s goal is to motivate the reader to agree with the opinion and to give the reader grounds to do so.”[8]
There are almost no limits on the contents of the opinion of a federal circuit court. Of course, one obvious limitation is that at least two judges must agree on its outcome. Beyond that, there are few rules. In fact, an opinion may be omitted entirely with a summary disposition simply concluding a matter.[9] The existence of a Supreme Court that may correct a ruling later is a minimal restraint. Accordingly, a court may range relatively freely, from jest, to verse, to glibness. It is commonplace, however, that a circuit court need not follow previous dictates of sister circuits.[10]
Judicial freedom does not mean there are not professional norms for opinions and reader expectations arising out of those norms. Those norms include considering previous discussions of the matter under review.[11] It is standard practice for a court to examine precedents, including those from outside the jurisdiction.[12] Readers expect analogies to be made and distinctions to be drawn in comparisons to pertinent authority.[13] Precedents, of any sort, may provide guidance for determining the outcome of the present case. The considered judgments of other judges are—it should be too obvious to say—the very basis of common law development. A prior case is part of the prior art of decision making. An earlier case may, of course, be the root of a new decision. Or, if it is poorly rendered, it may give rise to a better-reasoned resolution, either supporting or contradicting the outcome of the previous case. Different cases, with different judges and different facts, serve to add perspective. The common law develops in dialogue.[14]
The need to participate in this dialogue is especially true in an area of law a court has not previously examined. The Federal Judicial Center advises judges that “[i]f an opinion breaks new ground . . . the court should marshal existing authority and analyze the evolution of the law sufficiently to support the new rule.”[15] Lawyers know that they not only “should,” they will.[16] The expectation is of more than simple acknowledgment. Many courts describe an obligation to consider, and even give some degree of weight, to sister circuit determinations on similar issues.[17] A Ninth Circuit case, Hart v. Massanari,[18] goes so far as to write that “common law responsibilities” require “earlier authority [to be] acknowledged and considered,” and that it is “bad form to ignore contrary authority by failing even to acknowledge its existence.”[19]
With this background knowledge, held by every lawyer, it is jarring to encounter an opinion lacking these references. It is even more jarring to discover that they were available to the court, but forgone.
By participating in dialogue with earlier cases, a judge demonstrates two related things. First, that the judge is rendering a decision based on complete information. It is unusual to see a decision—especially on an issue of first, or early, impression—without reference to the work of other courts. Review of previous negative and positive authority relieves the reader’s fear that the outcome was uninformed.
Second, engaging in dialogue with prior cases demonstrates that the judge is acting in good faith. By acknowledging other negative authority, it is clear that the judge is not trying to avoid challenges. It shows consideration of both sides. By addressing prior positive authority, the judge acknowledges that she is not operating on a blank slate, but is, in fact, a part of the dialogue, owing credit to other interlocutors. Conversely, a “lack of candor, when discovered, reveals a lack of integrity.”[20]
By recognizing the existence and operation of other supportive cases, a court benefits from the legitimating effect of being part of a larger body of law.[21] Indeed, it is a common belief that “the primary source of judicial legitimacy lies in reasoned appeals to appropriate legal authority.”[22] The existence of other supportive cases tends to demonstrate correctness.[23] Reference to prior cases can demonstrate consistency, a valuable currency in the world of law—how often are cases criticized as “outliers”?[24] It seems odd, then, to write a decision that does not embrace its fellows. It is rhetorically valuable for a decision to be able to say, “this is part of the mainstream of legal thought.”[25]
- Glatt v. Fox Searchlight
Glatt is a Fair Labor Standards Act (FLSA) case brought by unpaid interns, seeking their classification as “employees” under the FLSA, with the minimum wage and overtime pay that status bestows. Memorably, two of the plaintiffs had worked as interns on the film Black Swan.[26] One intern’s tasks included “purchas[ing] a non‐allergenic pillow for Director Darren Aronofsky.”[27]
The trial court, among other orders, granted partial summary judgment in favor of the plaintiff interns on the “employee” issue.[28] It reached the conclusion that they were statutory employees by applying a six-part test published by the Department of Labor’s Wage and Hour Division in its Field Operations Handbook.[29]
The Second Circuit held that the wrong analysis had been applied.[30] It first determined that the Department of Labor’s test was not due deference. It then concluded that a balancing test should be applied to determine the “primary beneficiary” of the relationship between the parties. It did not decide whether the plaintiffs were interns or employees, but remanded for that assessment.[31]
- Glatt on Deference to the Wage and Hour Division’s Field Operations Handbook
As part of its opinion, the Second Circuit had to determine whether it was bound under deference principles by the interpretive advice of the Department of Labor (“DOL”). The DOL’s Wage and Hour Division publishes an intern fact sheet that was issued in 2010 (“the DOL test”).[32] The Fact Sheet’s pedigree was considerably longer, however—the same language and test for determining whether an individual was an unpaid “trainee” (perhaps slightly different from an “intern”) or employee first appeared in the Field Operations Handbook in 1967.[33]
The six-part test contained in the Handbook and Fact Sheet was derived from a Supreme Court case, Walling v. Portland Terminal Co.[34] In that case, the Court concluded that certain railroad trainees were not “employees” for purposes of the FLSA.[35] The Second Circuit concluded it was not bound to apply the DOL test.[36] It first noted that only Skidmore[37] deference could possibly apply.[38] Unlike the better-known Chevron[39] deference, Skidmore requires courts defer to administrative interpretations of ambiguous statutes only to the extent of their “power to persuade.”[40] Skidmore deference gives courts a great deal of leeway to reject an agency’s interpretation.[41]
The Second Circuit then provided two reasons it was not persuaded to apply the DOL’s six-part test. First, “[b]ecause the DOL test attempts to fit Portland Terminal’s particular facts to all workplaces,” and second, “because the test is too rigid for our precedent to withstand.”[42]
In the course of holding that it was not persuaded by the DOL interpretation, the Second Circuit cited three cases, none of which specifically address the deference a court should accord the DOL test. The first, Skidmore, simply establishes the premise—that deference is due if the agency interpretation is persuasive.[43] The second, Velez v. Sanchez,[44] addresses the substance of the issue, noting that “employee” determinations are factorial inquiries, relying on the totality of the circumstances, but does not address this DOL interpretation.[45] Finally, a third case, New York v. Shalala,[46] also addresses a separate issue—deference to an agency’s interpretation of a judicial opinion.
- What Glatt Omits from its Deference Discussion
To read Glatt, you would think no court had ever passed on the words of the DOL’s six-part test. You would think it operated on a blank slate. It did not. There is, in fact, not only a relatively robust jurisprudence on the question, there is also what might be considered a circuit split.
Coming down on the side of deferring to the DOL test is Atkins v. General Motors Corp.,[47] which states, “[w]e recently cited these criteria with approval . . . and the Administrator’s interpretation is entitled to substantial deference by this court.”[48] Similarly, Kaplan v. Code Blue Billing & Coding, Inc.,[49] concluded that it may “properly resort for guidance” to the test.[50] On the other side of the deference ledger is McLaughlin v. Ensley[51], which disclaims the DOL test, though without mentioning the deference doctrine in coming to its conclusion. Somewhere in the middle lies Reich v. Parker Fire Protection District,[52] a case that quotes the DOL test verbatim and discusses at length the question of the deference due that test, before concluding that the factors “are relevant but not conclusive.”
None of these cases appear in Glatt. It is hard to believe, however, that they were simply missed by law clerks in the course of their research. A case cited in passing elsewhere in Glatt, Solis v. Laurelbrook Sanitarium & School, Inc., [53] itself collects most of the authority on whether the DOL guidance deserves deference.
- Glatt on the Intern/Employee Distinction
After the court declined to defer to the DOL’s test, it went on to hold that “the proper question is whether the intern or the employer is the primary beneficiary of the relationship, and we propose the above list of non‐exhaustive factors to aid courts in answering that question.”[54] It asserted this “primary beneficiary” test, again, as though it were of the court’s own making.[55]
In reaching its decision, the court cites to two intra-circuit cases addressing the existence of an employment relationship, but neither is an intern/trainee case. The first, Velez v. Sanchez, called on the court to “determine the ‘economic reality’ in a domestic service context.”[56] The second, Brock v. Superior Care, Inc., addressed the well-trod legal landscape of the independent contractor versus employee inquiry.[57] Both are certainly pertinent, and worth discussion in the course of a case arising within the same circuit, but neither engages in anything like an assessment of the “primary beneficiary” of the individual/employer relationship.
- What Glatt Omits from Its Intern/Employee Discussion
As in its deference discussion, Glatt’s substantive analysis of the intern/employee distinction omits relevant extra-circuit authority. From the text of the court’s decision, it appears that its “primary beneficiary test” is a creature entirely of its own making, perhaps inspired by the imaginative lawyers for the defendants.
Other courts, however, have proposed and applied a similar test. Glatt is not a rugged individual, elbowing its way into the Federal Reporter. The first case to apply this test was McLaughlin v. Ensley.[58] It announced a test that evaluates “whether the employee or the employer is the primary beneficiary of the trainees’ labor.”[59]
The second case to explicitly apply a “primary beneficiary” test is the Sixth Circuit’s Solis decision.[60] Glatt cites Solis to support the proposition that the appropriate test is a balancing test that considers factors, but does not require a particular showing on any of them, rather than a test that requires that a number of elements all be satisfied.[61] It does not cite Solis, however, as a source of the precise test it announces—a “primary beneficiary” test.
Besides these two sister circuit opinions, both directly on point, Glatt omits other pertinent information engaging the concept of “primary beneficiary.” The inquiry into the “primary beneficiary” of an individual’s labor arises in other areas of employment law, including the “borrowed servant” rule of vicarious liability,[62] and the question of wage credits for housing provided to migrant workers.[63]
Moreover, at least one commenter has opined on the propriety of the test.[64] While it seems far more discretionary to engage secondary sources touching on the issue than the decisions of other circuits, when a secondary source directly addresses the issue, it seems worth the minimal effort of noting. Indeed, a plausible counter to the application of a primary benefit test is provided in that article.[65]
III. Glatt’s Omissions Fall Short of Reader Expectations
There is too much law on the issues Glatt addresses for the court to ignore. It appears disingenuous to draft an opinion that leaves out prior thinking on an issue, both positive and negative. There is no question whether the Second Circuit may come to its own conclusion on any issue before it. It is unusual, however, for it to ignore previous conclusions on those very issues.
On the deference question, at least four cases had previously addressed the precise issue before it.[66] Using these cases, the Glatt court could find support for its conclusion, or find inadequacies in the reasoning of decisions coming to a different conclusion. This sort of comparing and contrasting is standard practice. At the very least, readers expect the court to acknowledge the prior art on the subject. A complete decision tests its conclusions against those that have come before. It does not simply omit those earlier conclusions. A court’s acknowledgement that it is joining a greater battle creates credibility. It demonstrates both good faith and the possession of complete background information.
The failure to acknowledge the provenance of the “primary beneficiary” test is another failure of completeness. To present an idea as your own, without placing it in greater context, is odd to begin with, but especially so given the obvious rhetorical benefits of providing that context. Telling readers that you are adding to a body of existing law makes your lawmaking act unremarkable, standard, and obvious. It is better for the court’s legitimacy to be joining a doctrinal movement rather than presenting itself as inventing one out of whole cloth, especially while tossing to the wind the cloth of the DOL test. There is no upside to ignoring bolstering decisions.
When a reader comes upon a legal analysis without context, the reader simply assumes an isolated decision is being made. It may or may not be the correct decision, so far as the reader knows, but there is no concern that something is being hidden. Failing to engage the larger discussion misleads the reader into believing that a court has little to work with. This inappropriately adds to the opinion’s authority—it presents itself as an original attempt to resolve a unique problem. When there is a greater body of law on the issue that the court has ignored, the reader is misled.
Of course, there is value in conciseness as well. It makes sense in many cases to “abjure rote recitations of established legal principles [and] forgo superfluous citations.”[67] In a case like Glatt, however, faced with novel principles, this value seems less compelling.
A reader who perceives an absence of authority may be left with an intuition or perception of incompleteness—”this seems like an area of law that should have some forerunners.” When this intuition is confirmed, however, a sense of disappointment replaces it.[68] The reader may wonder why the court pretends there are no signposts to guide it. The trust between reader and writer is violated when expectations of inter-circuit dialogue are undermined.
* Associate Clinical Professor of Law–Santa Clara University School of Law.
[1]. See generally Gerald Lebovits et. al., Ethical Judicial Opinion Writing, 21 Geo. J. Legal Ethics 237, 238 (2008) (describing and analyzing requisites of form, substance, and ethics in judicial opinion writing).
[2]. 791 F.3d 376 (2d Cir. 2015).
[3]. See infra Parts II.B, II.D.
[4]. Judicial Writing Manual: A Pocket Guide for Judges, Fed. Judicial Ctr. 1 (2013), http://www2.fjc.gov/sites/default/files/2014/Judicial-Writing-Manual-2D-FJC-2013.pdf.
[5]. Chad M. Oldfather, Writing, Cognition, and the Nature of the Judicial Function, 96 Geo. L.J. 1283, 1317 (2008).
[6]. Lebovits, supra note 1, at 244.
[7]. Id. at 246 (“[A]ppellate opinions are mostly directed at lawyers and judges.”).
[9]. See, e.g., 9th Cir. R. 3–6, available at http://cdn.ca9.uscourts.gov/datastore/uploads/rules
/rules.htm#pID0E0Q2B0HA.
[10]. United States v. Williams, 184 F.3d 666, 671 (7th Cir. 1999) (“While we carefully and respectfully consider the opinions of our sister circuits, we are not bound by them.”).
[11]. Lebovits, supra note 1, at 285 (“[A]ll opinions should contain the sources from which the principle is derived.”); see also Oldfather, supra note 5, at 1334 (“[W]e expect courts to tell us why a given result is correct and to do so with reference to appropriate legal materials.”).
[12]. United States v. Washington, 584 F.3d 693, 698 (6th Cir. 2009) (“[T]his court routinely looks to our sister circuits for guidance when we encounter a legal question that we have not previously passed upon.”); Hall v. United Ins. Co. of Am., 367 F.3d 1255, 1260 (11th Cir. 2004) (“For direction, we consider the decisions of our sister circuits.”); Aldens, Inc. v. Miller, 610 F.2d 538, 541 (8th Cir. 1979) (“Although we are not bound by another circuit’s decision, we adhere to the policy that a sister circuit’s reasoned decision deserves great weight and precedential value.”).
[13]. Clark v. United States, 289 U.S. 1, 13 (1933) (“It is then the function of a court to . . . summon[] to its aid all the distinctions and analogies that are the tools of the judicial process.”). The use of comparisons goes back to rhetoric’s beginning; see Aristotle, Rhetoric Book II, in Poetics and Rhetoric, ch. 20, 335 (W. Rhys Roberts trans., Barnes & Noble Classics 2005) (originally published 4th Century BCE) (“We will first treat of argument by Example, for it has the nature of induction, which is the foundation of reasoning.”), available at http://rhetoric.eserver.org/aristotle/twoindex.html.
[14]. Lebovits, supra note 1, at 245 (“Opinion writing helps judges structure their decisions as dialogues that consider the common law’s past and future.”).
[15]. Judicial Writing Manual: A Pocket Guide for Judges,supra note 4, at 18; see also id. at 6 “[W]hen the decision involves novel issues or a developing area of law, it is appropriate to trace the prior development of the law and develop the legal and policy rationale at some length.”); id. at 4 (“When, however, an opinion enters less developed areas of the law, laying down a new rule or modifying an old one, . . . [the writer] should discuss and analyze the precedents in the area.”).
[16]. Hart v. Massanari, 266 F.3d 1155, 1169 (9th Cir. 2001) (“When ruling on a novel issue of law, [courts] will generally consider how other courts have ruled on the same issue. This consideration will not be limited to courts at the same or higher level, or even to courts within the same system of sovereignty.”).
[17]. Am. Med. Int’l, Inc. v. Sec’y of Health, Educ. & Welfare, 677 F.2d 118, 123 (D.C. Cir. 1981) (“When cases presenting legal questions with national implications have arisen, this court has always considered itself obligated to subject the analyses of coordinate tribunals to close scrutiny, with the aim of producing a sound and well-reasoned decision.”); Washington Energy Co. v. United States, 94 F.3d 1557, 1561 (Fed. Cir. 1996) (“We thus temper the independence of the analysis in which we engage by according great weight to the decisions of the other circuits on the same question.”).
[18]. 266 F.3d 1155, 1170 (9th Cir. 2001).
[20]. Lebovits, supra note 1, at 293; see also Scott C. Idleman, A Prudential Theory of Judicial Candor, 73 Tex. L. Rev. 1307, 1321 (1995) (Candor may be defined as “full disclosure of relevant information, evaluated subjectively from the judge’s point of view.”).
[21]. Lebovits, supra note 1, at 264 (“[T]he authority an opinion cites bolsters its legitimacy); see also Chad Flanders, Toward a Theory of Persuasive Authority, 62 Okla. L. Rev. 55, 77 (2009) (“A consensus among other circuits on a relevantly similar matter is, as a matter of practice, entitled to some weight. The mere fact that other circuit courts have decided a matter one way is relevant. It exerts a pull towards that result—not an inexorable pull, but a pull nonetheless.”).
[22]. Oldfather, supra note 5, at 1334.
[23]. See Frederick Schauer, Authority and Authorities, 94 Va. L. Rev. 1931, 1951 (2008) (“The author of a brief or opinion who uses support to deny genuine novelty is asking the reader to take the supported proposition as being at least slightly more plausible because it has been said before than had it not been.”); Eric A. Posner & Cass R. Sunstein, The Law of Other States, 59 Stan. L. Rev. 131, 143 (2006) (a majority view may be “probative of what is right.”).
[24]. See, e.g., Hill v. Freedman Anselmo Lindberg, LLC, No. 14-C-10004, at 3 (N.D. Ill. May 1, 2015) (“Blakemore is a single outlier in a wave of district court opinions.”).
[25]. See In re Greene, 33 B.R. 1007, 1009 (D.R.I. 1983) (Selya, J.) (criticizing cases because they “are alien to the mainstream of judicial thought and must be regarded skeptically”).
[26]. Glatt, 791 F.3d at 379–80.
[31]. Id. The court also addressed class certification, but because that part of the court’s decision seems less pertinent to the thesis of this Essay, it is omitted.
[32]. Fact Sheet #71: Internship Programs Under the Fair Labor Standards Act, U.S. Dep’t of Labor, Wage and Hour Division (2010), http:// www.dol.gov/whd/regs/compliance/whdfs71.pdf [hereinafter Fact Sheet #71].
[34]. Id. at 383 (citing Walling v. Portland Terminal, 330 U.S. 148 (1947)).
[35]. Walling v. Portland Terminal, 330 U.S. 148, 153 (1947)).
[37]. Skidmore v. Swift & Co., 323 U.S. 134 (1944).
[39]. Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984).
[40]. Skidmore, 323 U.S. at 140.
[41]. See Jim Rossi, Respecting Deference: Conceptualizing Skidmore Within the Architecture of Chevron, 42 Wm. & Mary L. Rev. 1105, 1118 (2001) (“Without a doubt, however, Skidmore affords less deference than Chevron.”).
[43]. Skidmore, 323 U.S. at 140 (cited in Glatt, 791 F.3d at 383).
[44]. 693 F.3d 308, 330 (2d Cir. 2012).
[45]. Velez, 693 F.3d 308 (cited in Glatt, 791 F.3d at 383).
[46]. 119 F.3d 175, 180 (2d Cir. 1997) (quoted in Glatt, 791 F.3d at 383).
[47]. 701 F.2d 1124 (5th Cir. 1983).
[48]. Id. at 1128 (citations omitted).
[49]. 504 F. App’x 831 (11th Cir. 2013).
[51]. 877 F. 2d 1207 (4th Cir. 1989).
[52]. 992 F. 2d 1023 (10th Cir. 1993).
[53]. 642 F.3d 518 (6th Cir. 2011) (quoted in Glatt, 791 F.3d at 385).
[56]. 693 F.3d 308, 327 (2d Cir. 2012) (cited in Glatt, 791 F.3d at 384).
[57]. 840 F.2d 1054 (2d Cir. 1988) (cited in Glatt, 791 F.3d at 384).
[58]. 877 F.2d 1207, 1209 (4th Cir. 1989). McLaughlin also addressed the deference issue.
[61]. Glatt, 791 F.3d at 385 (citing Solis).
[62]. Green v. United States, 709 F.2d 1158, 1163 (7th Cir. 1983) (“[T]he evidence simply does not support the conclusion that his work—however narrowly defined—was of primary benefit to CVSA.”).
[63]. Ramos-Barrientos v. Bland, 661 F.3d 587, 596 (11th Cir. 2011) (“employer may not receive wage credits . . . because this expense primarily benefits the employer.”).
[64]. Stephen A. Mazurak, The Unpaid Intern: Liability for the Uninformed Employer, 29 ABA J. Lab. & Emp. L. 101, 116–17 (2013).
[65]. Id. at 116–17 (“But in most of those cases, the employer also benefits by having a particular job performed, maybe not at the level of proficiency that a skilled employee would impart but still at a level that generates a profit for the employer. In most cases, balancing the benefit to the individual and the benefit to the employer will not solve the problem since each will receive some benefit.”). Additionally, a recent student note describes two approaches to the intern determination—”primary beneficiary,” and “totality of the circumstances.”; Cody Elyse Brookhouser, Whaling on Walling: A Uniform Approach to Determining Whether Interns Are “Employees” Under the Fair Labor Standards Act, 100 Iowa L. Rev. 751, 756 (2015).
[66]. See supra notes 48–54 and accompanying text.
[67]. Bruce M. Selya, In Search of Less, 74 Tex. L. Rev. 1277, 1279 (1996).
[68]. As one circuit judge wrote: “I would be less than candid if I failed to express my disappointment that, except for a passing reference by Chief Judge Feinberg, the concurring opinions of my colleagues never address, or even acknowledge, directly pertinent cases from our sister circuits which reach contrary results on the precise issues considered here. We are, of course, free to go our own way. But it would seem to me that the considered views of other circuits are at least entitled to our respectful consideration.” United States v. Monsanto, 852 F.2d 1400, 1418 (2d Cir. 1988), rev’d, 491 U.S. 600 (1989).