dissenting in part.

Defendant argues that the charge given to the grand jury by the district court impermissibly limits the grand jurors’ discretion in two ways: by instructing them that they “should not be concerned about punishment” and that they “cannot judge the wisdom of the criminal laws enacted by Congress” in deciding whether to indict. (Emphases added.) The majority rejects both challenges.

I agree that United States v. Marcucci, 299 F.3d 1156 (9th Cir.2002), requires us to resolve the first issue against the defendant. Marcucci involved a slightly different part of the same grand jury instruction, one that instructs grand jurors that they “should” indict if they find probable cause. Id. at 1159 . Over Judge Hawkins’s vigorous dissent, the Marcucci majority held that “should” is different from “shall” or “must” in that it does not preclude the grand jury from exercising its discretion not to indict, even if it does find there is probable cause. Id. at 1164 . While, as a matter of first impression, I might find more persuasive Judge Hawkins’s view on this point, id. at 1170 (Hawkins, J., dissenting), I accept Marcucci as binding, not merely as to the precise question presented there but as to the more general question whether a “should” or “should not” grand jury instruction imper-missibly constrains the grand jury’s discretion. Marcucci says it does not, and I accept that as controlling in our case.

But Marcucci’s logic works precisely in reverse as to the other part of the challenged instruction. Quite clearly, the district court told the grand jurors that they were not permitted to take the wisdom of the laws into account in deciding whether to return an indictment. This prohibition is mandatory, not hortatory; it seeks to constrain the grand jury’s discretion by limiting the matters it may consider in deciding whether to indict, and it does so with respect to an issue that could be highly relevant to a grand jury’s decision whether to indict at all and, if so, whether to indict for a more or less serious offense. See Vasquez v. Hillery, 474 U.S. 254, 263 , 106 S.Ct. 617 , 88 L.Ed.2d 598 (1986). Unlike the majority, I can find no source of authority for the district court to impose such a limitation on the grand jury. The instruction thus violates United States v. Williams, 504 U.S. 36, 50 , 112 S.Ct. 1735 , 118 L.Ed.2d 352 (1992), in that it imposes an impermissible rule of procedure for the operation of the grand jury. Alternatively, I believe the instruction constitutes an unauthorized substantive interference by the district court with the grand jury’s traditional discretion.

I will not rehearse the arguments about our constitutional responsibility to maintain the independence of the grand jury that are eloquently and forcefully marshaled by Judge Hawkins in his Marcucci dissent. Marcucci, 299 F.3d at 1166 (Hawkins, J., dissenting). Judge Hawkins says all that can or need be said on the subject and, for want of anything useful to *900 add, I simply incorporate his discussion by reference. *

I do want to emphasize two points that are particularly germane to the specific question presented to us — whether the grand jury may consider the wisdom of the law in deciding whether to indict. While the grand jury is an independent entity, not part of any branch of government, Williams, 504 U.S. at 47 , 112 S.Ct. 1735 , the function it performs is most accurately described as prosecutorial. The grand jury usually acts as a check on prose-cutorial discretion by occasionally refusing to return an indictment that the prosecutor seeks, Vasquez, 474 U.S. at 263 , 106 S.Ct. 617 , although it can also investigate, see Williams, 504 U.S. at 48 , 112 S.Ct. 1735 , and bring charges not presented to it by a prosecutor, see Vasquez, 474 U.S. at 263 , 106 S.Ct. 617 . Prosecutorial discretion — the decision whether to bring charges against a particular defendant — is widely recognized as having an important political component. Not every potential crime can (or should) be investigated or prosecuted, and an important part of the prosecutorial function is deciding which potential defendants to select for criminal prosecution, and how serious the charges should be. Prosecutors can, and often do, make such decisions based on their judgment as to how wise and important certain laws may be. See William Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L.Rev. 505, 599 (2001) (“[Prosecutors have the discretion not to enforce when the laws are too harsh.”). A prosecutor who believes that securities fraud is a particularly serious offense (or perhaps one that will gain him a lot of press coverage) may choose to devote more resources to investigating and prosecuting that crime, particularly as to high-profile defendants. A recent example is New York Attorney General Eliot Spitzer, who has made a name for himself “attaek[ing] the [mutual] fund industry with a vengeance.” Mara Der Hovanesian & Paula Dwyer, Where Will Eliot Spitzer Strike Next?, Business Week Online, Feb. 26, 2004 (internal quotation marks omitted) (Spitzer’s crusade against financial services crimes is intended to “bring[] cases to light and push[ ] Washington to tighten the laws.”). See also Bob Egelko, Medical Pot Advocate Found Guilty, S.F. Chron., Feb. 1, 2003, at A1 (calling the conviction of a prominent medical marijuana advocate “a triumph for federal prosecutors seeking to override California’s endorsement of pot as medicine” and reporting a DEA official’s *901 statement that “ ‘[t]here is no such thing as medical marijuana’ ”); Larry Neumeis-ter, Unprecedented Spate of Trials Set to Start Before Spring, Assoc. Press State & Local Wire, Jan. 20, 2004 (reporting prosecutors’ focus on punishing the “excesses of corporate America” and explaining that “the government wants to create a perception that every time a Martha Stewart comes along, she’ll be prosecuted” (internal quotation marks omitted)).

Conversely, prosecutors often choose to devote fewer resources to pursuing certain crimes or punishments they believe are not as important, or perhaps even unwise. See 4 Wayne R. LaFave et al., Criminal Procedure § 13.2(a), at 13 (2d ed.1999) (explaining that prosecutors often decline to enforce criminal laws because they are outdated); see also Bowers v. Hardwick, 478 U.S. 186 , 198 n. 2, 106 S.Ct. 2841 , 92 L.Ed.2d 140 (1986) (Powell, J., concurring) (“The history of nonenforcement [of Georgia’s sodomy law] suggests the moribund character today of laws criminalizing this type of private, consensual conduct.”). For example, Manhattan District Attorney Robert Morgenthau is known for his “fervent opposition to capital punishment.” Eric Fettmann, Editorial, Morgy the ‘Extreme’; DA v. Death Penalty, N.Y. Post, Jan. 7, 2004. When New York was considering reinstating the death penalty, Mor-genthau authored an op-ed declaring that the death penalty “actually hinders the fight against crime.” Robert M. Morgen-thau, Editorial, What Prosecutors Won’t Tell You, N.Y. Times, Feb. 7, 1995, at A25. Though the death penalty was ultimately reinstated in New York, Morgenthau has never sought it. See Ronald J. Tabak, Empirical Studies of the Modem Capital Sentencing System: How Empirical Studies Can Positively Affect the Politics of the Death Penalty, 83 Cornell L.Rev. 1431, 1437 (1998); Susan Saulny, Morgenthau Rules Out Death Penalty in Triple-Murder Case, N.Y. Times, Nov. 16, 2002, at B3. In the words of one commentator, it “has long been apparent to the most casual observer ... [t]hat [Morgenthau] has no intention of carrying out a law with which he profoundly disagrees.” Fettmann, supra. On the other end of the spectrum is Philadelphia District Attorney Lynne Abraham, who seeks the death penalty as often as the law allows. In 1995, the New York Times reported that “no prosecutor in the country uses the death penalty more.” Tina Rosenberg, The Deadliest D.A., N.Y. Times, July 16, 1995, § 6, at 22. It’s therefore no surprise that Abraham was quoted as saying, “When it comes to the death penalty, I am passionate. I truly believe it is manifestly correct.” Id.

There’s no reason grand juries cannot or should not make similar political judgments about which laws deserve vigorous enforcement and which ones do not, in deciding whom to indict, and on what charges. As Judge Hawkins explains, grand jurors are traditionally viewed as the “ ‘conscience of the community’ ” — -a function that partakes far more of judgment and discretion than of the narrow ministerial role of merely weighing the evidence to determine probable cause that the challenged instruction assigns to them. Marcucci, 299 F.3d at 1168-69 (Hawkins, J., dissenting); see also Ric Simmons, Reexamining the Grand Jury: Is There Room for Democracy in the Criminal Justice System?, 82 B.U. L.Rev. 1, 39-44 (2002) (presenting cases where grand juries refused to indict despite strong evidence that a criminal law was violated). The decision whether to enforce a law enacted by a far-away legislature by commencing a prosecution in the local community seems to necessarily imply some second-guessing of the legislature’s decision to pass the law in the first place. In acting as the community’s conscience, the *902 grand jurors must decide whether conduct that appears to fall within the prohibitions of a particular statute does indeed merit criminal punishment. It seems almost impossible to make such judgments without considering the law’s wisdom in light of local realities. See Id. at 49 (finding that “[djuring times of high crime rates, grand juries are more likely to accept [a defendant’s argument that he was carrying an illegal handgun for protection] and dismiss the case, even though there are no legal grounds for doing so” but “[d]uring times of relatively low crime, grand juries will usually reject such an argument and indict the defendant”). A recent study of New York grand juries confirms the grand jury’s role as the community’s conscience. “[T]he grand jurors in [the cases studied] ... were making discretionary, non-legal judgments as to whether the defendants should be prosecuted. As the first (and usually only) means of community input into the criminal justice system, the grand jury exercises its own political, moral, and social judgment in reviewing the prosecutor’s decision to bring the case.” Id. at 46 (footnotes omitted).

This kind of community judgment strikes me as particularly important in federal prosecutions, and not merely because Washington is usually much farther away geographically than the state capital. State prosecutors are elected locally and must stand for re-election on a regular basis. They will, of necessity, take the local community’s values into account. United States Attorneys, by contrast, are appointed by the President and never have to stand for election. In their daily operations, they are supervised by the Department of Justice, whose prosecutorial policies they implement. Except for the tradition of senatorial courtesy, which gives the state’s senators some say in who will be the United States Attorney in a particular district, there is very little state control, and almost no local control, over federal prosecutors. See John Gleeson, Supervising Federal Capital Punishment: Why the Attorney General Should Defer when U.S. Attorneys Recommend Against the Death Penalty, 89 Va. L.Rev. 1697, 1716 (2003) (“In a federal system that rightly accords great deference to states’ prerogatives, the federalization of the death penalty should be limited to cases in which there is a heightened and demonstrable federal interest, one that justifies the imposition of a capital prosecution on communities that refuse to permit them in their own courts.”). An independent grand jury — one that interposes the local community’s values on prosecutorial decisions that are controlled by policies set in Washington as to the enforcement of laws passed in Washington — seems like an important safeguard that is entirely consistent with the grand jury’s traditional function. Yet the challenged grand jury instruction seeks to neutralize this aspect of the grand jury’s discretion.

Second, allowing (perhaps even encouraging) the grand jury to consider the wisdom of the law under which a suspect is to be prosecuted seems particularly urgent, given that we no longer permit petit juries to exercise such discretion. See, e.g., United States v. Powell, 955 F.2d 1206, 1213 (9th Cir.1991); United States v. Simpson, 460 F.2d 515, 518-20 (9th Cir.1972). See also Simmons, 82 B.U. L.Rev. at 47 (“[T]he grand jury has become the primary vehicle for members of the community to participate in and influence the criminal justice system.”). Petit juries, unlike grand juries, are an integral part of the adjudicative function, which is suffused with an array of procedural protections. Given modern conceptions of due process, it would be wholly intolerable to allow petit jurors to make up the law as they go *903 along. Petit jurors, rather, must decide guilt or innocence strictly in accordance with clearly established and scrupulously defined legal standards. Yet the cry that a member of the community should not be convicted of a crime unless an independent group of his peers believes that such a conviction would be consistent with community values, is not without historical plausibility. See, e.g., http://fija.org/ links.htm (last visited Mar. 3, 2004) (quoting statements by the Founders that support this view). Because the petit jury may not do this, it is even more important to foster this traditional function of the grand jury — a body not subject to the prohibition against double jeopardy or other procedural constraints that apply once the ease proceeds to trial. See Williams, 504 U.S. at 49 , 112 S.Ct. 1735 ; United States v. R. Enters., Inc., 498 U.S. 292, 298 , 111 S.Ct. 722 , 112 L.Ed.2d 795 (1991); Marcucci, 299 F.3d at 1171 (Hawkins, J., dissenting). The challenged instruction precludes the grand jury from exercising one of its core functions and thus, I believe, constitutes an unauthorized encroachment on the grand jury’s authority. Like Judge Hawkins, I believe the error is structural, and thus not subject to defea-sance based on harmless error analysis. Id. at 1172-73 .

For these reasons, and those articulated by Judge Hawkins in his scholarly dissent in Manned, I would vacate defendant’s conviction and remand with instructions that the district court dismiss the indictment, with leave for the government to seek a new indictment from a grand jury not infected by the erroneous instruction.

The majority also relies on the following footnote in United States v. Adams, 343 F.3d 1024 (9th Cir.2003):

Id. at 1027 n. 1.

My colleagues claim that "[t]he Adams court read Marcucci broadly as holding that the model charge did not impermissibly infringe on the grand jury's independent exercise of its discretion,” and that we are consequently bound by Adams’s broad reading of Marcucci . Maj. op. at 898. However, Adams doesn’t tell us on what basis the grand jury instruction was being challenged, so we can't tell whether the challenge raised by Navarro-Vargas was considered and rejected. A court cannot declare a grand jury instruction constitutional en gross and immunize it from all future constitutional challenges. Because Adams cited Marcucci without explication or elucidation, it can be read no more broadly than Marcucci itself. Adams adds nothing to our analysis.