As amended April 27, 2000
APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MASSACHUSETTS [Hon. Morton A. Brody,* U.S. District Judge] *Of the District of Maine, sitting by designation.
This appeal tests the limits of a federal district court's authority to promulgate local rules. The court below upheld a rule constraining the issuance of subpoenas seeking client-related information from lawyers in criminal cases. Because we find that the rule falls outside the permissible scope of local rulemaking authority, we reverse.
Before assaying a preliminary question of justiciability and addressing the controversy's merits, we trace the origins of the challenged rule and chart the travel of the case.
A. The Evolution of Local Rule 3.8(f).
The 1980s witnessed a dramatic increase in the number of subpoenas served on defense attorneys by federal prosecutors. The reasons for this trend are difficult to pinpoint, but some commentators have linked it with heightened efforts to fight organized crime and drug-trafficking, new forfeiture laws, and an unprecedented expansion of the Department of Justice (DOJ). See 1 Geoffrey C. Hazard, Jr. & W. William Hodes, The Law of Lawyering § 3.8:701, at 700 (Supp. 1996); Frank O. Bowman, III, A Bludgeon by Any Other Name, 9 Geo. J. Legal Ethics 665, 686 n.74 (1996). In 1985, mindful that forcing a lawyer to offer evidence against her client may sow seeds of mistrust and increase the incidence of conflicted interests, the DOJ introduced guidelines for the issuance of attorney subpoenas, including an internal preapproval process. See United States v. Perry, 847 F.2d 1346, 1347-48 (9th Cir. 1988) (citing United States Attorneys' Manual § 9-2.161(a) (July 18, 1985)). Responding to the prodding of bar leaders, the Massachusetts Supreme Judicial Court (the SJC) also took prophylactic action. It adopted an ethics rule, known as Prosecutorial Function 15 (PF 15), which stated that:
It is unprofessional conduct for a prosecutor to subpoena an attorney to a grand jury without prior judicial approval in circumstances where the prosecutor seeks to compel the attorney/witness to provide evidence concerning a person who is represented by the attorney/witness. S.J.C. R. 3:07, PF 15 (effective Jan. 1, 1986).
The United States District Court for the District of Massachusetts thereafter incorporated PF 15 into its local rules. PF 15 withstood the United States Attorney's ensuing challenge by the narrowest of margins. See United States v. Klubock, 639 F. Supp. 117 (D. Mass. 1986), aff'd, 832 F.2d 664 (1st Cir. 1987) (equally divided en banc). Crucial to the district court's holding was the fact that PF 15 imposed no substantive limitations on the right to issue subpoenas. See id. at 120 & n.7, 124.
For the next twelve years, PF 15 held sway in the District of Massachusetts. In the interim, the Rhode Island Supreme Court adopted a rule of conduct requiring prosecutors to obtain judicial approval prior to issuing attorney subpoenas. See R.I. Sup. Ct. Rules, Art. V., R. Prof'l Conduct 3.8 & cmt. (adopted Nov. 1, 1988). This rule closely resembled PF 15, with two significant deviations: its reach extended to subpoenas outside the grand jury context, and its text included a comment outlining substantive standards to be applied by a court in determining whether to sanction an attorney subpoena request. See id. Specifically, the comment suggested that judicial approbation should be withheld unless, inter alia, the information sought was (a) not privileged, (b) "essential" to the government's investigation, and (c) unobtainable from any "other feasible alternative." Id. When the United States District Court for the District of Rhode Island incorporated the state standard into its local rules on April 20, 1989, the United States Attorney for that district challenged it. In Whitehouse v. United States District Court, 53 F.3d 1349 (1st Cir. 1995), a panel of this court upheld the local rule, albeit strongly suggesting that a different result would obtain were the criteria limned in the comment embedded in the text of the rule itself (and, thus, made mandatory rather than precatory). See id. at 1357-58 & n.12.
We temporarily shift our focus to the national stage. In 1990, the American Bar Association (the ABA) amended Rule 3.8 of the Model Rules of Professional Conduct by adding a new paragraph (f) and comment:
The prosecutor in a criminal case shall:
(f) not subpoena a lawyer in a grand jury or other criminal proceeding to present evidence about a past or present client unless:
(1) the prosecutor reasonably believes:
(i) the information sought is not protected from disclosure by any applicable privilege;
(ii) the evidence sought is essential to the successful completion of an ongoing investigation or prosecution; [and]
(iii) there is no other feasible alternative to obtain the information; and
(2) the prosecutor obtains prior judicial approval after an opportunity for an adversarial proceeding.
Paragraph (f) is intended to limit the issuance of lawyer subpoenas in grand jury and other criminal proceedings to those situations in which there is a genuine need to intrude into the client-lawyer relationship. The prosecutor is required to obtain court approval for the issuance of the subpoena after an opportunity for an adversarial hearing is afforded in order to assure an independent determination that the applicable standards are met. ABA Standing Comm. on Ethics and Prof'l Responsibility and Section of Criminal Justice, Report 118, at 1 (Feb. 1990). After the Third Circuit struck down a bar rule patterned on Model Rule 3.8(f), see Baylson v. Disciplinary Bd., 975 F.2d 102 (3d Cir. 1992), the ABA retreated; it removed the judicial preapproval requirement by deleting both subparagraph (2) and the second sentence of the comment. See ABA Standing Comm. on Ethics and Prof'l Responsibility, Report 101, at 1 (Aug. 1995).
On June 9, 1997, the SJC amended the Massachusetts Code of Professional Conduct to replace PF 15 with the (discarded) 1990 version of Model Rule 3.8(f), omitting the second sentence of the comment, but including the rescinded subparagraph (2). See S.J.C. R. 3:07, Rule 3.8(f) & cmt. 4. The result was to alter PF 15 in three important respects: extending it to trial as well as grand jury subpoenas; providing for an "adversarial proceeding" in advance of the issuance of an attorney subpoena; and promulgating three substantive standards for judicial preapproval.
B. The Instant Litigation.
By local rule, attorneys practicing in the United States District Court for the District of Massachusetts must adhere to the ethical rules adopted by the SJC, unless a specific exception obtains. See D. Mass. R. 83.6(4)(B). On November 17, 1997, the United States Attorney for the District of Massachusetts, Donald K. Stern, wrote to the chief judge of the district court, urging that such an exception be made for State Rule 3.8(f). By letter dated December 3, 1997, the court declined to craft a specific exception, indicating that Stern's concerns would be resolved on a case-by-case basis. On January 1, 1998, Rule 3.8(f) went into effect. At that point, Stern directed his staff to hold in abeyance any applications to the district court for attorney subpoenas.
Stern and Craig C. Donsanto (a senior DOJ lawyer based in Washington and a member of the Massachusetts bar), both acting in their official capacities, filed suit in the district court on May 13, 1998. They sought to have both State Rule 3.8(f) and Local Rule 3.8(f) declared invalid as applied to members of the Massachusetts bar practicing before federal courts. The complaint alleged in substance that Rule 3.8(f), in both its federal and state incarnations, exceeded the district court's rulemaking authority, contradicted the federal common law anent grand jury practice, conflicted with the federal rules of criminal procedure and evidence, and violated the Supremacy Clause of the United States Constitution.
Initially, the named defendants were the SJC, the Massachusetts Board of Bar Overseers (the Board), Bar Counsel, and the United States District Court itself. 1 The complaint was later amended to add as defendants the individual judges of the district court (hereinafter, along with the court itself, the Judicial Defendants).
The contours of the case were narrowed when Bar Counsel filed an affidavit in which he vouchsafed that he would not wield State Rule 3.8(f) against federal prosecutors, but, rather, would refer any alleged violations to the federal district court for discipline under Local Rule 3.8(f). This concession rendered the Supremacy Clause issue moot and led the plaintiffs to dismiss their claims against the SJC and the Board. Rosenfeld then agreed not to attempt to enforce Local Rule 3.8(f) against members of the Massachusetts bar practicing outside the commonwealth, and Donsanto dropped out of the picture.
Stern had moved at the outset for preliminary injunctive relief. All the defendants opposed this motion, and the Judicial Defendants filed a motion to dismiss the complaint as unripe. The district court heard argument on October 7, 1998, and took both motions under advisement. While the court was pondering the matter, the President signed into law a bill, now codified at 28 U.S.C. § 530B, that made certain state standards directly applicable to federal prosecutors. The parties debated the relevance and effect of this provision in a series of subsequent submissions.
The district court ultimately determined the issues presented to be ripe for review and rejected the claim that section 530B rendered the controversy moot. See Stern v. SJC, 184 F.R.D. 10, 13-14 & n.5 (D. Mass. 1999). Then, reading subparagraphs (1) and (2) independently, the court found Whitehouse controlling in part and upheld both the prior judicial approval requirement and the provision for an adversarial hearing. See id. at 17-19. Finally, the court approved the substantive requirements of subparagraph (1) as ethical precepts not inconsistent with federal law. See id. at 19. Following the entry of a final judgment, Stern launched this appeal.
We consider de novo the Judicial Defendants' contention that the controversy is not ripe for adjudication. See Riva v. Massachusetts, 61 F.3d 1003, 1007 (1st Cir. 1995).
The ripeness doctrine has both constitutional and prudential dimensions. See Public Serv. Comm'n v. Wycoff Co., 344 U.S. 237, 242-44 (1952); Rhode Island Ass'n of Realtors v. Whitehouse, 199 F.3d 26, 33 (1st Cir. 1999). Its basic rationale is "to prevent the courts, through avoidance of premature adjudication, from entangling themselves in abstract disagreements." Abbott Labs. v. Gardner, 387 U.S. 136, 148 (1967). Courts must apply a two-part test to assess ripeness. See id. at 149. First, it is necessary to determine whether the issue presented is fit for judicial review - an inquiry that "typically involves subsidiary queries concerning finality, definiteness, and the extent to which resolution of the challenge depends on facts that may not yet be sufficiently developed." Ernst & Young v. Depositors Econ. Protection Corp., 45 F.3d 530, 535 (1st Cir. 1995). Second, it is necessary to evaluate the extent to which withholding judgment will impose hardship -an inquiry that typically "turns upon whether the challenged action creates a 'direct and immediate' dilemma for the parties." W.R. Grace & Co. v. EPA, 959 F.2d 360, 364 (1st Cir. 1992) (quoting Abbott Labs., 387 U.S. at 152). Both prongs of the test ordinarily must be satisfied, although a very strong showing on one axis may compensate for a relatively weak showing on the other. See Ernst & Young, 45 F.3d at 535.
This case, which questions whether the federal district court has the power to incorporate State Rule 3.8(f) into the armamentarium of its local rules, passes the first prong of the test. The issue presented can be finally resolved by declaratory judgment, its contours are sharply defined, and additional facts will not affect its resolution. The conclusion that the issue is fit for review is bolstered, moreover, by a realization that deciding the question appears unavoidable: Local Rule 3.8(f) imposes new substantive and procedural requirements on prosecutors who request attorney subpoenas, and Bar Counsel has stated unequivocally that he will enforce those requirements. Since bringing this suit, the United States Attorney has delayed requests for no fewer than twenty-six such subpoenas, and it seems reasonable to suppose that in some of these instances the prosecutors would have had difficulty in satisfying the "essentiality" and/or "no feasible alternative" criteria. Short of a continuation of Stern's self-imposed moratorium - and the public has a right to expect that such a state of affairs will not last indefinitely - a test of Local Rule 3.8(f)'s facial validity appears inevitable.
This case also satisfies the hardship prong. Delaying adjudication until a more concrete controversy emerges (until, say, a particular attorney subpoena request reaches the judicial preapproval stage or disciplinary proceedings are instituted in the aftermath of a served subpoena) would inflict significant institutional costs with little corresponding gain. Indeed, charting such a course would put Stern on the horns of a dilemma, forcing him to decide whether to serve attorney subpoenas in cases arguably prohibited by the local rule and thus risk potential sanctions or to refrain from so doing and thus jeopardize the success of ongoing criminal investigations. It is precisely this sort of "'direct and immediate' dilemma," W.R. Grace, 959 F.2d at 364 (quoting Abbott Labs., 387 U.S. at 152), that Congress wished to ameliorate when it passed the Declaratory Judgment Act. See ANR Pipeline Co. v. Corporation Comm'n, 860 F.2d 1571, 1578 (10th Cir. 1988) ("Once the gun has been cocked and aimed and the finger is on the trigger, it is not necessary to wait until the bullet strikes to invoke the Declaratory Judgment Act.").
A determination that this case is ripe for adjudication squares with the way in which we, and other courts of appeals, have treated analogous cases (albeit without explicit discussion of ripeness). In Klubock, this court fielded a preemptive strike by federal prosecutors against a newly adopted ethics rule. In Whitehouse, we entertained a similar foray, explaining that "the proper method for mounting a facial challenge to the validity of [a local rule] . . . is through an action for declaratory and/or injunctive relief filed in the district court." 53 F.3d at 1353. Moreover, two other courts of appeals have adjudicated pre-enforcement challenges to ethics rules in parallel circumstances. See United States v. Colorado Supreme Court, 189 F.3d 1281 (10th Cir. 1999) (CSC II); Baylson, 975 F.2d 102.
Another case that sheds light on the question of ripeness is United States v. Colorado Supreme Court, 87 F.3d 1161 (10th Cir. 1996) (CSC I). There, the Tenth Circuit held that the United States had standing to mount a pre-enforcement challenge to a local rule identical to Local Rule 3.8(f), crediting the government's stated trepidation that the rule would interfere with federal subpoena practice. See id. at 1165-67. Although the CSC I court did not discuss ripeness per se, standing and ripeness often overlap. See Rhode Island Ass'n of Realtors, 199 F.3d at 33. So here: we think that the court's finding of standing necessarily implied that the controversy was ripe for adjudication.
In a pre-enforcement challenge to a law carrying significant penalties, standing exists when the plaintiff has manifested an intention to engage in conduct arguably proscribed by the statute, and there exists a "credible threat" of enforcement. New Hampshire Right to Life Political Action Comm. v. Gardner, 99 F.3d 8, 14 (1st Cir. 1996). In most situations, as here, that self-same credible threat serves to render the case fit for judicial review and to demonstrate the hardship that will result should no review ensue. See Rhode Island Ass'n of Realtors, 199 F.3d at 33 (stating that a "concrete plan to engage immediately (or nearly so) in an arguably proscribed activity . . . gives a precise shape to disobedience, posing a specific legal question fit for judicial review," and "[a] showing that the challenged statute, fairly read, thwarts implementation of the plan adds the element of hardship").
The Judicial Defendants' vigorous arguments to the contrary are ultimately unpersuasive. Citing Ernst & Young, they asseverate that Stern's claims are too contingent and uncertain to be fit for review. But the examples they give include events that are almost certain to materialize (e.g., internal DOJ approval of an application for a subpoena addressed to an attorney), and events that are completely irrelevant (e.g., destruction of evidence by a targeted attorney). Stern stands poised to request and serve attorney subpoenas, but reasonably fears disciplinary proceedings (for himself and his staff) if he does so. Thus, the only contingency likely to deter the parties from a collision course is continued self-restraint on the part of Stern's office. This is a far cry from Ernst & Young, where the relevant injury would materialize, if at all, only after a long chain of remote and speculative events, many of which involved third parties. 2 See 45 F.3d at 538.
Next, the Judicial Defendants note that the "linchpin of ripeness . . . is adverseness." Rhode Island v. Narragansett Indian Tribe, 19 F.3d 685, 692 (1st Cir. 1994). Building on this foundation, they complain that the current controversy lacks adverseness because the lawyers potentially subject to subpoena (and their clients, for that matter) are unrepresented. This rather simplistic formulation overlooks, however, that as a general rule a "conflict between state officials empowered to enforce a law and private parties subject to prosecution of that law is a classic 'case' or 'controversy' within the meaning of Art. III." Diamond v. Charles, 476 U.S. 54, 64 (1986). The defendants here are empowered to enforce the strictures of Local Rule 3.8(f) through disciplinary proceedings against Stern and his subalterns. This type of controversy is thus sufficiently adverse even though the class of persons that the law was designed to protect is not separately represented. After all, that class "lacks a judicially cognizable interest in the prosecution or non-prosecution" of Stern or the other attorneys in his office. Linda R.S. v. Richard D., 410 U.S. 614, 619 (1973). Consequently, the fact that the district court adopted Local Rule 3.8(f) out of a desire to shield the attorney-client relationship does not mean that particular attorneys and clients must be joined in a facial attack on the rule.
Finally, the Judicial Defendants suggest that the DOJ guidelines render this case unripe. Because the standards contained in the guidelines and those contained in Local Rule 3.8(f) overlap, this thesis goes, there is scant likelihood that Local Rule 3.8(f) will work any harm. This thesis has a plausible ring but, in the last analysis, it fails to cover the ground. We explain briefly.
There is no question that the DOJ guidelines' "reasonably needed" criterion, 3 Department of Justice Manual § 9-13.410, at 9-165 (2d ed. 2000), is markedly less demanding than the "essentiality" requirement contained in Local Rule 3.8(f). We think that the DOJ criteria are also less demanding than the "no feasible alternative" requirement; the guidelines stipulate that "all reasonable attempts shall be made to obtain the information from alternative sources . . . unless such efforts would compromise the investigation or case," id. (emphasis supplied), whereas the "no feasible alternative" provision contains no similar loophole. In all events, the guidelines are subject to change at the whim of the DOJ. Last (but not least), an internal review mechanism that specifically disavows any intent to create legally enforceable rights obviously does not burden prosecutors in the same way, or to the same extent, as does a binding rule of court that imposes substantive standards, requires prior judicial approval, and subjects government attorneys to potential disciplinary action. Cf. Whitehouse, 53 F.3d at 1362 n.18 (suggesting that DOJ guidelines did not render a prior judicial approval requirement superfluous). For these reasons, we reject the notion that the mere existence of the DOJ guidelines dissipates any hardship.
We have said enough on this score. The threat of ethics enforcement is genuine, compliance costs are real and immediate, and the chilling effect on attorney subpoena requests constitutes an injury sufficient to support a justiciable controversy. See CSC I, 87 F.3d at 1165-67. Thus, the issue presented here - the authority of the federal district court to incorporate State Rule 3.8(f) into its local rules -is ripe for review.
III. THE MERITS
The authority of the federal district courts to adopt local rules emanates from three sources. First, Congress has empowered the Supreme Court to prescribe rules of practice and procedure for the federal courts. See 28 U.S.C. § 2072(a). In turn, the Supreme Court has authorized district courts to craft local rules to implement, or fill gaps in, national rules of practice and procedure. See Fed. R. Crim. P. 57(a)(1); Fed. R. Civ. P. 83(a)(1). Second, Congress has vested the lower federal courts with independent authority to prescribe local rules. See 28 U.S.C. § 2071(a). Third, district courts have inherent power arising from the nature of the judicial process, and this power extends to certain types of rulemaking. See Chambers v. NASCO, Inc., 501 U.S. 32, 43 (1991); Whitehouse, 53 F.3d at 1355.
Regardless of the source, local rulemaking authority is bounded. A local rule must be both constitutional and rational, and its subject matter must be within the ambit of the court's regulatory power. See Frazier v. Heebe, 482 U.S. 641, 646 (1987); Whitehouse, 53 F.3d at 1355-56. In this same vein, a local rule must be consistent with, but not duplicative of, Acts of Congress and nationally applicable rules of practice, procedure, and evidence. See 28 U.S.C. § 2071(a); Fed. R. Crim. P. 57(a)(1); Fed. R. Civ. P. 83(a)(1). Even if a local rule does not contravene the text of a national rule, the former cannot survive if it subverts the latter's purpose. See Hawes v. Club Ecuestre el Comandante, 535 F.2d 140, 144 (1st Cir. 1976). Then, too, local rules should cover only interstitial matters. See Fed. R. Crim. P. 57 advisory committee's note; see also United States v. Horn, 29 F.3d 754, 760 (1st Cir. 1994) (noting that a court's inherent power has definite limits). They may not create or affect substantive rights, see 28 U.S.C. § 2072(b), or institute "basic procedural innovations," Miner v. Atlass, 363 U.S. 641, 650 (1960).
The core issue presented by this appeal is whether the district court had the power to adopt Local Rule 3.8(f). This question of law engenders de novo review. Before answering this query, however, we pause to measure its dimensions.
A. How to Construe Local Rule 3.8(f).
At Bar Counsel's urging, the court below read subparagraphs (1) and (2) of Local Rule 3.8(f) independently. See Stern, 184 F.R.D. at 16-17. On this bifurcated reading, the substantive standards delineated in subparagraph (1) would be relevant only to potential disciplinary proceedings, and a court deciding whether to approve a subpoena request pursuant to subparagraph (2) would have no obligation to apply them. We reject this artificial construction of Local Rule 3.8(f).
The most sensible way to construe Local Rule 3.8(f) is as a unified whole. Each subparagraph is a mere sentence fragment and neither makes sense without reference to the shared introductory language. Court rules, like statutes and ordinances, generally are to be read in a holistic manner. See, e.g., King v. St. Vincent's Hosp., 502 U.S. 215, 221 (1991); Massachusetts Ass'n of HMOs v. Ruthardt, 194 F.3d 176, 180 (1st Cir. 1999); O'Connell v. Shalala, 79 F.3d 170, 176 (1st Cir. 1996). This tenet applies with special force when subdivisions are grammatically interrelated. See American Standard, Inc. v. Crane Co., 510 F.2d 1043, 1058 (2d Cir. 1974).
Here, the case for reading the rule as a whole is even stronger, for the connection between the two halves is not only grammatical but also logical. Subparagraph (2) outlines a process for determining whether a subpoena application should be approved, but gives no clue as to what standard a judge is to apply in making that determination. The obvious place to look is subparagraph (1), which sets forth such a standard. Indeed, the court below acknowledged that "as a practical matter, a court faced with a request for an attorney-subpoena is likely to engage in a similar inquiry in the course of determining whether issuance of the subpoena is appropriate." Stern, 184 F.R.D. at 16 n.12. The presence of the attorney to be subpoenaed -a feature not involved in the scheme upheld in Whitehouse - virtually guarantees such an inquiry.
On the other hand, subparagraph (2) would serve no purpose if courts evaluated subpoena applications solely on the basis of traditional motion-to-quash standards. For the adversarial hearing to be meaningful, the targeted attorney would have to be told in advance of the content of the testimony or materials sought. This notice, no less than service of the subpoena itself, would drive a wedge of distrust between lawyer and client. Thus, subparagraph (2), on the bifurcated reading, would mirror existing quashal procedure and generate no ethics benefits whatsoever.
The sockdolager is that the original drafters certainly intended that the two subparagraphs of the rule be harmonized, not balkanized. As discussed above, the text of Local Rule 3.8(f) derives from (and is identical to) a former ABA Model Rule. The comment to that rule made it transparently clear that the judicial preapproval described in subparagraph (2) was designed to ensure compliance with the substantive standards described in subparagraph (1). Absent some specific disclaimer (not present here), the district court cannot adopt verbatim the text of a model rule without accepting the drafters' unequivocal interpretation of its meaning. Cf. Felix Frankfurter, Some Reflections on the Reading of Statutes, 47 Colum. L. Rev. 527, 537 (1947) ("[I]f a word is obviously transplanted from another legal source . . . it brings the old soil with it.").
To be sure, when the SJC formulated State Rule 3.8(f), it omitted the second sentence of the comment to the former ABA rule (which made the connection between the two subparagraphs explicit). But see Whitehouse, 53 F.3d at 1358 nn.12 & 14, 1364-65 (noting that comments and commentary are not binding). In addition, the Massachusetts version lists as the sources for subparagraph (2) both former ABA Model Rule 3.8(f)(2) and PF 15, and the latter rule has been interpreted to impose no substantive restrictions beyond the prototypical motion-to-quash standards. See Klubock, 639 F. Supp. at 120 & n.7, 124. But these reeds are far too slender to shore up a construction that splits grammatically interrelated provisions, imports a standard of review from outside the ethics code, and directly contradicts the drafters' manifest intention.
Bar Counsel also argues that we should defer to his construction because he is the official charged with enforcement of ethics rules in Massachusetts. With respect, we doubt that Bar Counsel's litigation position is entitled to any deference. Cf. Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 212 (1988) (holding that an agency's litigating position is not entitled to deference); Massachusetts v. Blackstone Valley Elec. Co., 67 F.3d 981, 991 (1st Cir. 1995) (same). In all events, Bar Counsel's construction does not bind the judges of the District of Massachusetts who, as defendants here, have pointedly refused to endorse it. 3 Finally, any deference that might normally be due is overcome here by the availability of a much more logical reading and a clear statement of the drafters' intent. Cf. 12 Charles Alan Wright et al., Federal Practice & Procedure § 3153, at 533 (2d ed. 1997) (explaining that even "a district court's construction of its own rule will be reversed if the appellate court is convinced that the district court has misconstrued its own rule").
Bar Counsel next invokes the principle that courts should resist interpreting a statute in a way that provokes a constitutional problem. See United States v. Gifford, 17 F.3d 462, 473 (1st Cir. 1994). Other courts of appeals have extended this principle to the construction of local rules, thereby seeking to avoid interpretations that place local rules on a collision course with national rules. See, e.g., Marshall v. Gates, 44 F.3d 722, 725 (9th Cir. 1995); United States v. White, 980 F.2d 836, 844 (2d Cir. 1992); cf. Jaroma v. Massey, 873 F.2d 17, 20 (1st Cir. 1989) (per curiam) (explaining that ambiguous "[l]ocal district court rules cannot be construed in such a way as to render them inconsistent with applicable provisions of the Federal Rules of Civil Procedure").
There are reasons to question whether such an extension of the principle is appropriate; after all, striking down a local rule, unlike declaring a statute unconstitutional, does not implicate separation of powers or countermajoritarian concerns. But even assuming, for argument's sake, the validity of the extended principle - on the ground, say, that the local rule here derives from a state rule, and that the issue therefore has overtones of comity - it does not control the outcome in this instance. Interpretive ingenuity has its limits. The idea that judges charged with interpreting a rule should strive to do so in a way that will avoid conflicts with governing law is a useful device in doubtful cases, but it does not permit the interpreters to substitute their judgment for that of the drafters or to rewrite the rule from scratch. Here, there is both a natural, holistic reading of the local rule and a clear statement by the drafters in support of such a reading. In these circumstances, we will not embrace an implausible construction simply to minimize the potential for conflict with federal law.
For these reasons, we hold that Local Rule 3.8(f) must be read as an indivisible whole. Consequently, judicial preapproval under subparagraph (2) of the rule proceeds, if at all, on the basis of the substantive standards elucidated in subparagraph (1).
B. Rule 3.8(f) as Applied to Grand Jury Subpoenas.
Because local rulemaking authority is at a lower ebb in the grand jury context, we ease first into those relatively shallow waters.
The grand jury is deeply rooted in Anglo-American tradition and is "a constitutional fixture in its own right." United States v. Williams, 504 U.S. 36, 47 (1992) (internal quotation marks omitted). It is not a part of either the Executive or Judicial Branch, but, rather, "a kind of buffer . . . between the Government and the people." Id. Accordingly, "any power federal courts may have to fashion, on their own initiative, rules of grand jury procedure is a very limited one, not remotely comparable to the power they maintain over their own proceedings." Id. at 50. A local rule may not "substantially alter the traditional relationships between the prosecutor, the constituting court, and the grand jury itself." Id. Nor may a local rule trench upon any core attribute of the grand jury, including: "1) its independence from the court's supervision; 2) its broad investigative powers; 3) the presumption of validity accorded its subpoenas; 4) the secrecy of its proceedings; [or] 5) its general freedom from procedural detours and delays." Whitehouse, 53 F.3d at 1357. It is against this backdrop that we take the measure of Local Rule 3.8(f) as it pertains to grand jury subpoenas.
We do not write on a pristine page. As the defendants repeatedly remind us, Whitehouse held that a local rule requiring judicial preapproval for the service of an attorney subpoena neither undermined the historic role of the grand jury nor interfered with its essential attributes. See id. at 1357-62. But this holding rested squarely on the panel's determination that the particular local rule worked no substantive change in the governing law because judicial preapproval would be granted or denied under traditional motion-to-quash standards. See id. at 1357-58 & nn.12-13 (explaining that the rule "merely authorizes district courts to reject a prosecutor's attorney-subpoena application for the traditional reasons justifying the quashing of a subpoena"); see also Fed. R. Crim. P. 17(c) (authorizing courts to quash a subpoena "if compliance would be unreasonable or oppressive"). In so holding, the Whitehouse court brushed aside the seemingly more rigorous criteria delineated in the comment to the rule, on the ground that the comment was merely advisory. See 53 F.3d at 1358 n.12. So viewed, the rule imposed no additional burden on grand jury independence because courts, in theory, would apply the motion-to-quash standards that govern under Rule 17, not the comment's suggested criteria, in determining whether to approve an attorney subpoena request. 4
The Rhode Island rule's saving grace is absent here. Local Rule 3.8(f) differs significantly in that it imposes new substantive requirements for judicial preapproval of grand jury subpoenas. In so doing, the rule alters the grand jury's historic role, places it under overly intrusive court supervision, curbs its broad investigative powers, reverses the presumption of validity accorded to its subpoenas, undermines the secrecy of its proceedings, and creates procedural detours and delays. It therefore impermissibly interferes with grand jury proceedings. See 1 Hazard & Hodes, supra, § 3.8:701, at 702 (Supp. 1997) ("Rule 3.8(f) in its original form seemed clearly invalid . . . as applied to . . . federal grand jury subpoenas to criminal defense lawyers.").
Because any one of these vices would suffice to invalidate the rule as applied to grand jury subpoenas, we confine our discussion to two of the most glaring defects: Local Rule 3.8(f)'s impact on grand jury secrecy and its potential as an incubator for delay. In United States v. R. Enterprises, Inc., 498 U.S. 292 (1991), the Court held that the government could not be required to demonstrate that the materials sought by a grand jury subpoena were relevant. See id. at 298-99. The Court explained that such a requirement "'would saddle a grand jury with minitrials and preliminary showings,'" id. (quoting United States v. Dionisio, 410 U.S. 1, 17 (1973)), and would "threaten to compromise 'the indispensable secrecy of grand jury proceedings,'" id. at 299 (quoting United States v. Johnson, 319 U.S. 503, 513 (1943)); see also Fed. R. Crim. P. 6(e) (cloaking grand jury proceedings in secrecy). Requiring a prosecutor to show that subpoenaed evidence is essential and not otherwise feasibly obtainable would have the same two impermissible effects. Consequently, we hold that Local Rule 3.8(f), as it pertains to grand jury subpoenas, encroaches unduly upon grand jury prerogatives and, therefore, is ultra vires.
C. Rule 3.8(f) as Applied Outside the Grand Jury Context.
Outside the grand jury context, Stern asseverates that Local Rule 3.8(f) is beyond the district court's competency because it goes past the "matters of detail" appropriate for local rulemaking, Fed. R. Crim. P. 57 advisory committee's note, and works a fundamental procedural change. This argument depends heavily on Miner v. Atlass, in which the Court held that a local admiralty rule authorizing oral discovery depositions exceeded the district court's rulemaking authority. See 363 U.S. at 650. The Court reasoned that this innovation, "though concededly 'procedural,' may be of as great importance to litigants as many a 'substantive' doctrine." Id. Accordingly, it was too basic to be effectuated through local rulemaking. See id.
In Colgrove v. Battin, 413 U.S. 149 (1973), the Court elaborated on this concept, explaining that "[t]he 'basic procedural innovations' to which Miner referred are those aspects of the litigatory process which bear upon the ultimate outcome of the litigation." Id. at 163 n.23. Applying this taxonomy, the Court upheld a local rule providing for six-member civil juries in lieu of traditional twelve-member juries because the party challenging the rule had not shown any discernible difference in the results reached. See id. Synthesizing Miner and Colgrove, we conclude that the relevant inquiry is whether Local Rule 3.8(f), applied outside the grand jury context, is apt to affect the ultimate outcome of criminal proceedings. See Eash v. Riggins Trucking Inc., 757 F.2d 557, 569 (3d Cir. 1985) (en banc).
The answer to this inquiry depends, in the first instance, on current practice under Fed. R. Crim. P. 17. Rule 17(c) authorizes subpoenas for the production of documentary evidence, objects, and the like. That rule prescribes a procedure and a standard for challenging such subpoenas: "The court on motion made promptly may quash or modify the subpoena if compliance would be unreasonable or oppressive." The Supreme Court elucidated the meaning of this standard in United States v. Nixon, 418 U.S. 683 (1974), holding that a subpoena duces tecum is not "unreasonable or oppressive" if the proponent establishes relevancy, admissibility, and specificity. See id. at 700.
Roughly the same standard applies to subpoenas compelling the attendance of witnesses, i.e., subpoenas ad testificandum. Although Rule 17(a), which governs such subpoenas, does not provide explicitly for quashal or modification, courts routinely have entertained motions seeking such relief and decided them by reference to comparable principles. Specifically, a subpoena ad testificandum survives scrutiny if the party serving it can show that the testimony sought is both relevant and material. See United States v. Valenzuela-Bernal, 458 U.S. 858, 867 (1982); United States v. Campbell, 874 F.2d 838, 850-51 (1st Cir. 1989).
These traditional standards for quashal of trial subpoenas form the template for further inquiry here. 5 As we have said, Local Rule 3.8(f) requires a prosecutor, before serving an attorney subpoena, to demonstrate that the information sought is essential, not privileged, and not otherwise feasibly available. These are significant departures from prior practice, and they raise the bar for obtaining relevant and material evidence. Collectively, they work changes too fundamental to be accomplished under the aegis of the district courts' local rulemaking power.
In particular, the "essentiality" and "no feasible alternative" requirements are substantially more onerous (and, thus, more restrictive) than the traditional motion-to-quash standards. Essentiality is obviously a more demanding criterion than relevancy or materiality. By like token, Rule 17 jurisprudence contains no corollary to the principle that a subpoena issued to one source cannot stand if the information sought is (or may be) available from some other source.
Two examples illustrate these points. Suppose, in a robbery case, that a defense lawyer received a lump-sum advance payment for services in the precise amount of the purloined funds from a client with no visible means of support. There is other evidence linking the client to the robbery, so the billing information could not fairly be described as "essential" to the prosecution. Hence, Local Rule 3.8(f) would prohibit the prosecutor from serving a subpoena on the defense attorney, notwithstanding the unarguable materiality and relevancy of the retainer information. Next, consider unprivileged documents in a lawyer's file relating to a complex, and possibly fraudulent, international real estate transaction. These documents may be obtainable without a subpoena duces tecum directed to the lawyer, but only through time-consuming, relatively expensive (but still feasible) alternative means. Local Rule 3.8(f) would prohibit an attorney subpoena, even though the situation easily satisfies standards of relevancy, admissibility, and specificity.
These examples are not eccentric hypotheticals, but, rather, fairly typical of the sort of situation in which a prosecutor might wish to serve an attorney subpoena (others easily can be conceived). We think that they demonstrate convincingly that Local Rule 3.8(f) imposes novel requirements that threaten to preclude the service of otherwise unimpeachable subpoenas and thus restrict the flow of relevant, material evidence to the factfinder. Considering that compelling the production of evidence is of "great importance" to litigants, Miner, 363 U.S. at 650, this is a fairly drastic alteration of the rules - too basic to be effected through local rulemaking. 6 See Klubock, 832 F.2d at 673 (equally divided en banc) (opinion of Breyer, J.) (suggesting that a local rule imposing standards of review for attorney subpoenas stricter than traditional motion-to-quash standards would fall outside the district court's rulemaking authority).
Moreover, unlike in Colgrove, 413 U.S. at 163 n.23, there has been no showing here that Local Rule 3.8(f) will not affect the outcome of criminal proceedings. Indeed, there is every reason to believe that the opposite is true. Local Rule 3.8(f) imposes new substantive requirements applicable only to prosecutors and creates a novel procedural device to ensure that these requirements are enforced. The likely result will be fewer attorney subpoenas served by the government; as the comment to the rule makes clear, the goal is "to limit the issuance of lawyer subpoenas." Certain evidence, as long as a court finds that it is not "essential," will never reach the trier of fact. So, too, when the government decides that the "feasible alternative" to an attorney subpoena is not worth the additional effort. In short, Local Rule 3.8(f), if permitted to stand, will make it measurably more difficult for prosecutors to secure convictions. The magnitude of this new burden is simply too large to be imposed by local rule. Accordingly, the rule cannot stand.
The Third Circuit reached the same conclusion with respect to a closely analogous rule. At issue there was Pennsylvania's analog to Local Rule 3.8(f), which required judicial preapproval for attorney subpoenas in criminal proceedings. See Baylson, 975 F.2d at 104. The comment to the rule 7 stipulated that approval normally would be denied unless the court found that the information sought was relevant and not confidential or privileged, that compliance would not be unreasonable or oppressive, that the primary purpose of the subpoena was not harassment, and that there were no feasible alternative means of obtaining the information sought. See id. Although only the "no feasible alternative" requirement was substantively new, the court held that the rule went beyond the "matters of detail" contemplated by Fed. R. Crim. P. 57 and thus transcended local rulemaking authority. See id. at 108-09. Because it requires essentiality as well, Local Rule 3.8(f) represents an even greater encroachment than the rule that confronted the Baylson court.
D. Section 530B.
Shortly after the hearing in the district court, Congress passed an omnibus budget bill. The bill contained a provision entitled "Ethical standards for attorneys for the Government," now codified as section 530B, which provides that:
An attorney for the Government shall be subject to State laws and rules, and local Federal court rules, governing attorneys in each State where such attorney engages in that attorney's duties, to the same extent and in the same manner as other attorneys in that State. 28 U.S.C. § 530B(a).
Bar Counsel asserts that this passage cures any conflict between Local Rule 3.8(f) and other federal law. We do not agree.
"Because of the fundamental importance of the principles shielding federal installations and activities from regulation by the States, an authorization of state regulation is found only when and to the extent there is a clear congressional mandate, specific congressional action that makes this authorization of state regulation clear and unambiguous." Hancock v. Train, 426 U.S. 167, 179 (1976) (footnotes and internal quotation marks omitted). We believe it is reasonable to require comparable specificity before inferring congressional intent to abandon the goal of national consistency anent rules of federal practice. Moreover, that benchmark is not attainable here: it simply cannot be said that Congress, by enacting section 530B, meant to empower states (or federal district courts, for that matter) to regulate government attorneys in a manner inconsistent with federal law. We explain briefly.
The federal rules of procedure, unlike state laws, are nationally uniform. See 28 U.S.C. § 2071(a); Fed. R. Crim. P. 57(a)(1); Fed. R. Civ. P. 83(a)(1). In the area of ethics, however, federal district courts generally have adopted their own rules, often modeled on the rules of the state in which they sit, but sometimes with significant deviations. See, e.g., D. Mass. R. 83.6(4)(B) (stating that SJC ethics rules govern "except as otherwise provided by specific rule of this court"). The potential for conflict between state and federal law therefore should have been obvious, but section 530B does not speak to the issue. Instead, Congress directed the Attorney General to fill out the details of enforcement by regulation. See 28 U.S.C. § 530B(b) (empowering the Attorney General to promulgate implementing regulations). These regulations dispel the notion that section 530B grants states and lower federal courts the power, in the guise of regulating ethics, to impose strictures that are inconsistent with federal law. See 28 C.F.R. § 77.1(b) (directing that section 530B "should not be construed in any way to alter federal substantive, procedural, or evidentiary law").
The Eleventh Circuit recently had the opportunity to consider the effect of section 530B. That court rejected the argument that the statute cured an incipient conflict between a state ethics rule (which had been incorporated into the federal district court's local rules) and the federal rules of evidence. See United States v. Lowery, 166 F.3d 1119, 1124-25 (11th Cir.), cert. denied, 120 S. Ct. 212 (1999). The court wrote: "If Congress wants to give state courts and legislatures veto power over the admission of evidence in federal court, it will have to tell us that in plain language using clear terms." Id. at 1125. We agree with this pronouncement and apply it here: because Local Rule 3.8(f) impermissibly interferes with federal grand jury practice and transcends district court rulemaking authority, section 530B cannot salvage it.
If more were needed - and we doubt that it is - Local Rule 3.8(f) clearly extends beyond the shelter that section 530B provides. Although the statutory text, which refers to "State laws and rules, and local Federal court rules, governing attorneys," is arguably susceptible to a broad interpretation, we have noted before that a statute's caption may assist in clarifying ambiguities. See, e.g., Massachusetts Ass'n of HMOs, 194 F.3d at 180; Berniger v. Meadow Green-Wildcat Corp., 945 F.2d 4, 9 (1st Cir. 1991). The title of the statute here ("Ethical standards for attorneys for the Government") removes any doubt about its scope: section 530B applies only to ethical standards. This conclusion becomes irresistible in light of the legislative history and implementing regulations. See, e.g., H.R. Conf. Rep. No. 105-825, at 1102 (1998); 144 Cong. Rec. E301 (daily ed. Mar. 5, 1998) (statement of sponsor, Rep. McDade); 28 C.F.R. §§ 77.1(b), 77.2(h). 8
That ends the matter. Local Rule 3.8(f), though doubtless motivated by ethical concerns, has outgrown those humble beginnings. Substance, not form, must control. See 28 C.F.R. § 77.2(h)(1) (explaining that the applicability of section 530B does not depend on "whether or not [the state or local] rule is included in a code of professional responsibility for attorneys"). As written, Local Rule 3.8(f) is more than an ethical standard. It adds a novel procedural step - the opportunity for a pre-service adversarial hearing - and to compound the matter, ordains that the hearing be conducted with new substantive standards in mind.
In recommending the deletion of subparagraph (2) from the former ABA Model Rule, the Standing Committee explained that the judicial preapproval provision was an anomaly: "Rather than stating a substantive ethical precept, it sets out a type of implementing requirement that is properly established by rules of criminal procedure rather than established as an ethical norm." ABA Standing Comm. on Ethics and Prof'l Responsibility, Report 101, at 7 (Aug. 1995). We agree with this assessment. Because Local Rule 3.8(f) goes beyond the realm of ethics, section 530B neither rescues it nor renders the instant case moot. 9 Cf. CSC II, 189 F.3d at 1283-89 (relying on section 530B to uphold an ethics rule consisting of subparagraph (1) but without any provision for an adversarial hearing).
We need go no further. We conclude that the authority of the district court to adopt Local Rule 3.8(f) is ripe for judicial review and that the two subparagraphs of the rule are to be read together. Thus, the substantive standards delineated in subparagraph (1) must guide a court in determining whether to approve an attorney subpoena in a pre-service hearing held pursuant to subparagraph (2). We also conclude that the adoption of Local Rule 3.8(f) exceeded the district court's lawful authority to regulate both grand jury and trial subpoenas. Finally, we conclude that 28 U.S.C. § 530B does not repair these defects. Accordingly, we hold that Local Rule 3.8(f) is without force or effect.
1. The SJC administers professional responsibility matters through the Board. In turn, Bar Counsel, currently Daniel C. Crane, serves as the Board's chief enforcement officer. See S.J.C. R. 4:01.
2. Massachusetts Ass'n of Afro-American Police v. Boston Police Department, 973 F.2d 18 (1st Cir. 1992) (per curiam), is similarly distinguishable. There, we found a pre-enforcement challenge unripe because the party with the power to inflict the feared injury had expressly disclaimed any intent to do so. See id. at 20-21.
3. Consistent with their position that the controversy is not ripe, the Judicial Defendants apparently deem it incumbent upon them to refrain from rendering what might be seen as an advisory opinion.
4. To be sure, the panel also suggested that the preapproval process did not impede grand jury independence because it did not apply to subpoenas sought by the grand jury acting independently from the prosecutor. See Whitehouse, 53 F.3d at 1357. While we are bound by the holding of Whitehouse, we eschew this component of its reasoning. As a practical matter, grand jury subpoenas are almost universally issued by and through federal prosecutors. See, e.g., In re Grand Jury Matters (United States), 751 F.2d 13, 16 (1st Cir. 1984). Moreover, the Supreme Court has explicitly rejected the notion that an otherwise impermissible rule of grand jury procedure becomes permissible if it is enforced against the prosecutor instead of the grand jury itself. See Williams, 504 U.S. at 53.
5. It bears emphasis that the motion-to-quash standards applicable to trial subpoenas historically have not been applied to grand jury subpoenas. See R. Enters., 498 U.S. at 298-99 (holding that recourse to the Nixon criteria would unduly interfere with grand jury proceedings); In re Grand Jury Proceedings (Hill), 786 F.2d 3, 5 n.2 (1st Cir. 1986) (per curiam) (declining to require a showing of "need" or "relevance" before a court may enforce a grand jury subpoena directed to an attorney). Because grand jury subpoenas are sui generis, we use the term "trial subpoenas" as a shorthand for all other subpoenas (e.g., subpoenas issued in the course of pretrial hearings).
6. In striking down a local rule in Miner, the Court repeatedly cited the fact than an analogous national rule had been considered and rejected. See Miner, 363 U.S. at 644-45, 648-51. The situation here is reminiscent of that scenario: Congress declined to enact a bill sponsored by Senator Paul Simon that was designed to "provide procedural safeguards with respect to the issuance of lawyer client subpoenas." 134 Cong. Rec. 21,589, 21,599 (1988).
7. The Third Circuit, unlike the panel in Whitehouse, see 53 F.3d at 1358 n.12, assumed that trial courts would not ignore the comment in applying the rule. See Baylson, 975 F.2d at 109.
8. Unlike Bar Counsel, we do not ascribe much weight to the dire predictions of broader applicability made by opponents of section 530B. See, e.g., 144 Cong. Rec. S12,996-97 (daily ed. Nov. 12, 1998) (statement of Sen. Abraham); 144 Cong. Rec. S12,798-99 (daily ed. Oct. 21, 1998) (statement of Sen. Hatch).
9. For essentially the same reason, Local Rule 3.8(f) is beyond the district court's inherent authority to regulate the conduct of attorneys appearing before it.