The Yale Law Journal

May 2014

Legitimacy and Federal Criminal Enforcement Power

abstract. A defining feature of criminal federalism is extreme disparities in case outcomes across state and federal forums. All else being equal, prosecution in the federal forum entails a significantly higher likelihood of conviction, and a higher penalty. But why do such disparities exist? Conventional explanations point to differences among sovereigns’ legal rules, resources, and dockets. These understandings, while valid, neglect to account for a less tangible source of federal criminal power: legitimacy. “Legitimacy” refers to the concept, refined through decades of empirical research, that citizens comply with the law, and defer to and cooperate with legal authority, when they perceive both the laws and the authorities to be fair. A legitimacy-based exploration of the federal criminal justice system significantly enriches our understanding of the sources of federal criminal enforcement power. Distilling those sources, moreover, reveals surprising and counterintuitive implications: to emulate the sources of federal legitimacy in local systems, we need more localized criminal justice.

author. Visiting Assistant Professor, Villanova University School of Law. Beginning Summer 2014, Assistant Professor, Temple University Beasley School of Law. For helpful comments on prior drafts, thanks to Todd Aagaard; Steve Chanenson; Michelle Dempsey; Jeffrey Fagan; Abbe Gluck; Rachel Harmon; Susan Klein; Judge Debra Livingston; Henry Monaghan; Geoff Moulton; participants at presentations at Villanova, Columbia, the University of Richmond, and Temple law schools; and especially to Dan Richman, who has offered invaluable guidance from the earliest stages of this project. Thanks also to Rory Eaton, Amy Spare and Karen Gause for terrific research assistance; Dan McGee at Villanova for much-appreciated assistance with statistical analysis; and Steven Kochevar and the editors of the Yale Law Journal for outstanding editorial work.


In litigation, forum matters. Nowhere is that more true than in criminal litigation, where the choice of court in which to prosecute—state or federal—is perhaps the single most significant factor influencing a case’s outcome. All else being equal, a defendant prosecuted in federal court is more likely to be convicted, and to receive a longer sentence of imprisonment, than if prosecuted for the same conduct in state court.1 The disparity has received particular attention in the area of so-called “street crimes”—the drug, gun, and violent offenses that make up the bulk of urban criminal felony dockets—because these are the sorts of crimes that many argue have historically been (and should continue to be) left largely to the states.2 For all the focus on the merits and equities of federal prosecution of street crime, though, there has been no concerted exploration of the antecedent question: why do the disparities exist in the first place?

The question is important because interpretations of forum disparities inform our understanding of the sources of federal criminal enforcement power. To date, scholars have treated these disparities largely as the product of tangible differences: in sovereigns’ legal rules (whether substantive, procedural, or constitutional), in resources, and in caseloads. These interpretations have framed our analysis of criminal federalism on issues ranging from the “federalization” of criminal law to the allocation of enforcement power and the exercise of federal enforcement discretion.3

The conventional interpretations of forum disparities are accurate, but incomplete. This Article seeks to supplement them with a different framework—one focused not on rules or resource allocations but rather on citizens’ perceptions of legal authority. The work primarily of social psychologists, criminologists and criminal law theorists, this framework—broadly-termed, “legitimacy”—posits that a system perceived as providing fair process and just laws promotes compliance with the law and respect for and deference to law enforcement authorities.4 The impressive body of empirical and theoretical work on legitimacy has not yet been mined for its potential to explain forum disparities in criminal adjudication.

This Article undertakes that effort. It offers a fresh look at the causes of forum disparities in street crimes, an alternative framework for understanding them, and a new agenda for inquiry. These disparities manifest not just differences in legal rules and resource allocations but also how actors within the respective criminal justice systems—primarily witnesses, juries, and judges—perceive the system’s legitimacy. Exploring forum disparities through the lens of legitimacy enriches our understanding of the sources of federal criminal power. And the policymaking implications are surprising and counterintuitive: by distilling the sources of federal legitimacy, we see the need for more localized criminal justice.

The need to engage these matters is urgent because the stakes are high—for defendants, victims, and the communities affected by street crime. Consider the following:

·      A defendant has been arrested forty-four times for various offenses, including narcotics trafficking, gun possession, robbery, and shooting. He has been prosecuted in his local county court for each of the forty-four arrests—sometimes facing substantial terms of imprisonment under state law—yet never once convicted. On his forty-fifth arrest, he is prosecuted for robbery and gun possession in federal court, where he is convicted and receives a sentence of thirty-two years’ imprisonment.5

·      A defendant has thirty-one convictions in state court, including two prior felony convictions for robbery. He also has numerous arrests that resulted in dismissals, including arrests for two robberies for which local grand juries refused to return indictments. A federal grand jury later returns indictments on those very same two robberies, along with five others; a federal petit jury convicts the defendant of those crimes; and a federal judge sentences him to thirty years’ imprisonment.6

·      A twenty-three-year-old defendant has eight prior criminal convictions, including separate convictions for illegal gun possession and for aggravated assault with a firearm on a police officer. He has never served more than eighty days in jail for any one offense. His ninth conviction, for possessing a firearm as a convicted felon, occurs in federal court. The federal judge sentences him to more than seven years’ imprisonment.7

These cases are extreme examples of what has aptly been called “unequal justice.”8 The phenomenon has directed a great deal of scholarly attention to the federalization of crime and the proper exercise of federal enforcement discretion.9 Nowhere in this literature, though, do scholars truly grapple with the reasons for these forum disparities. Explanations serve largely as premises for other debates, rather than starting points for independent inquiry. We are told that federal sentencing statutes and guidelines are more severe, and less malleable, than their state counterparts.10 Federal evidentiary and procedural rules are more favorable to the prosecution.11 The federal government has substantially greater resources to expend on each case.12 Federal prosecutors can be more selective in deciding which cases to bring.13

Each of these points has merit, and, collectively, they go a long way towards explaining forum disparities in street-crime cases. But a number of criminal enforcement phenomena illuminate their limitations. How, for instance, do we explain why federal prosecutors, who generally intervene in street-crime cases to secure a higher penalty, bring cases even when federal penalties are less severe than applicable state penalties?14 Why have Virginia’s firearms laws, which effectively replicate federal penalties, not replicated federal outcomes in firearms cases?15 What does it mean to say that federal procedural and evidentiary rules are more favorable to the prosecution—does lowering the evidentiary or procedural hurdles to bringing cases necessarily make it easier to win those cases? Resources and selectivity in charging are critical, to be sure; but what, exactly, do federal prosecutors select for, and when and why do resources matter? To address the disparities between federal and local prosecution of street crime, we must broaden our understanding of the sources of federal prosecutorial power.

Legitimacy encapsulates the theoretical principle that citizens comply with laws, and defer to and cooperate with legal authorities, when they perceive both the laws and the authorities to be fair.16 It is the deference and cooperation aspects of legitimacy that I engage here.17 Deference to and cooperation with law enforcement impact case outcomes.18 Case outcomes, in turn, reflect legitimacy. If witnesses will not testify; if juries do not credit the prosecution’s evidence; if sentences prescribed by a legislature or sentencing commission seem off-kilter in relation to a community’s views, then prosecutors and judges must, and will, resolve the individual cases before them in light of those realities.

A legitimacy-based account of the federal criminal justice system broadens our understanding beyond the tangible, to encompass less overt sources of federal power. It opens new avenues of inquiry: about the foundations of citizens’ trust in the law, the law’s enforcers, and the institutions of legal authority; the relationship between law and social norms; citizens’ identification with the governing authority; and the place of the jury within the larger electorate. Exploring the sources of federal legitimacy, moreover, reveals surprising implications for criminal federalism: the interactions that may matter most to legitimacy are not those between the federal government and states, but those between states and their localities.

For all the scholarly focus on legitimacy on the one hand and federal prosecution of street crime on the other, we have yet to bridge these two strains of literature. That is, we have yet to consider legitimacy’s role in federal street crime enforcement. We should. If legitimacy partly explains the outcome disparities between state and federal forums, we must reconceptualize these disparities as more than just reflections of different sovereigns’ legal rules or resource allocations. In many respects, they are that; but the disparities also reflect something else, something that cannot be remedied merely by a change in laws, or rules, or budgets. And so we need to begin asking new questions: not only about how allocations of power in the criminal justice system (between sovereigns as well as between legislators, courts, and law enforcers) affect citizens’ trust in the system, but also about how the need for citizens’ trust should inform how we allocate power.

This Article unfolds in four parts. Part I tests the limits of conventional explanations of forum disparities in street crime prosecution. Part II provides the conceptual framework of legitimacy theory in criminal justice and describes legitimacy’s significance in criminal adjudication. Part III explores federal prosecution of street crime through this framework, examining the legitimacy-enhancing features of federal criminal enforcement. These features do not make it “more legitimate” in a normative sense, and this Article makes no claim in that regard. The focus here is on perceptions of legitimacy—in the eyes of jurors, witnesses, and judges. Part IV takes on the normative implications. Criminal federalism inquiries, particularly in the context of street crime, have largely focused on how criminal power should be allocated between state and federal sovereigns. Yet if forum disparities in prosecutions of street crimes arise at least in part from a legitimacy gap, then we should also be asking an entirely different question: how can we translate the legitimacy-enhancing features of the federal system into local justice systems? Unpacking the sources of federal legitimacy reveals that, counterintuitively, the best way to emulate these sources in local systems is by enhancing localism—through greater accountability, participation, and local voice in both criminal lawmaking and law enforcement.

To be clear, the goal of this endeavor is not to emulate federal conviction rates or penalties, but rather to emulate those features of the federal system that serve, in part, to align written and applied law. Enhancing localism in urban criminal justice systems will help close the distance between written and applied law in those systems. And it will do so in a way that is more substantively just, because it will derive from the trust and cooperation of citizens most affected by street crimes and their enforcement. Attention to localism in urban criminal justice systems will thus enhance substantive justice not only in those systems, but overall: greater legitimacy in local urban justice systems will lessen the perceived need for federal prosecution of street crime and the relatively harsh penalties that come with it.19

It should be noted that this Article applies theories derived from empirical research in one area (local urban justice systems) to another (the federal justice system) in which similar empirical work has not yet been done.20 In this sense, the Article sets forth a research agenda for further work. It is my hope that this agenda encourages us to think more broadly about the sources of federal enforcement power in street crime and, perhaps, beyond.

I. conventional accounts of forum disparities and their limits

Over the last five decades, federal criminal law has vastly expanded, and the bulk of that expansion has been in the area of violent crimes and narcotics.21 In the late 1960s and early 1970s, Congress passed three pieces of criminal legislation that would prove to become staples of federal violent crime and narcotics prosecutions: the Gun Control Act of 1968, which marked the beginning of a sustained federal legislative effort at controlling gun crime;22 the Racketeer Influenced Corrupt Organizations Act, enacted to respond to mob crime but soon discovered by prosecutors as a potent tool for prosecuting violent inner city gangs;23 and the Comprehensive Drug Abuse Prevention and Control Act of 1970, which began the federal government’s foray into criminal drug enforcement.24A series of amendments to the drug laws over the next several decades saw marked increases in the types of drugs and drug-related conduct subject to federal criminal prosecution and in the penalties associated with federal drug crimes.25 Recidivist enhancements became a common feature of federal drug, gun, and violent crime penalties.26

Over this period, and with increasing intensity over the last two decades, federal prosecutors have used these and other laws to prosecute street crime.27 The enforcement effort has received sustained attention from the Department of Justice, with particularly vigorous contributions coming from those U.S. Attorney’s Offices in districts that encompass inner cities with high levels of gun and drug crime. And with rare exceptions, the courts have acceded to the federal government’s authority in this area in the face of federalism-based challenges.28

This relatively recent federal prosecutorial focus has generated much academic commentary, most of it negative, concerning the propriety of federal enforcement of what have traditionally been considered “local” crimes. The debate has largely proceeded within the familiar federalism framework with some commentators arguing for increased federal intervention and a vocal majority decrying the federal government’s infringement on local police powers.29 The participants in this particular federalism debate, though, do not line up along the typical political divisions.30 And that is because, unlike other federal interventions in the criminal justice system—for instance, habeas corpus proceedings, the procedural constraints imposed on states by the Warren Court, or § 1983 actions against law enforcement officers—this federal intervention has resulted in substantially less favorable outcomes for criminal defendants. On the whole, federal prosecution results in a more certain conviction and a likely higher sentence than a defendant would receive were he prosecuted in a local county courthouse.31 It is, in fact, this salient feature that drives much of the academic debate.32

It is also what drives federal prosecutions. Unlike large-scale white-collar or regulatory crimes, which typically require expertise and resources more available at the federal level, the default forum for street crimes is local. The corollary to this division of labor is that when the local district attorney can provide a prosecution that is equally or nearly as effective as would occur in federal court, in the form of a high likelihood of conviction and a substantial sentence, federal prosecution offers little value-added.

The disparity between federal and local enforcement of street crime, then, matters not just from an equity perspective but from an instrumental perspective as well: generally speaking, it is this disparity that drives federal intervention. And yet there has been relatively little attention paid to the possible explanations for it. The disparity has been the starting point for academic commentary, rather than its focus. What attention it has received has been mostly geared towards marshaling arguments in favor of, or against, federal intervention. Proponents of federal intervention argue that federal procedural rules, ample investigative resources, and enforcement discretion give prosecutors important tools for prosecuting violent crime.33 Opponents argue that the outcome disparity violates the Equal Protection Clause or, at least, offends common notions of equity, and that federal intervention constitutes an end-run around state legislatures’ sentencing preferences.34

In all of these arguments, the reasons for the disparities are largely presumed: federal sentences for gun and drug crimes are in most instances harsher and less malleable than their state counterparts; federal rules of criminal procedure are generally more favorable to prosecutors; and federal prosecutors have vastly more resources to devote to each case, as well as the luxury of choosing which cases they will bring.35 Each of these points is correct, and, collectively, they do a great deal of work in explaining criminal forum disparities. They do not, though, go the distance in explaining the extent and depth of the outcome disparities that exist. Let us examine why.

A. The Penalties Explanation

The leading explanation for the disparity between federal and local outcomes in street crime prosecution is relatively more stringent federal penalties, which we are told affect outcomes in two ways. First, harsher and more compulsory federal statutory penalties explain the substantially higher federal conviction rate: high penalties pressure defendants to plead guilty in exchange for lower sentences, resulting in more guilty pleas, fewer trials, and therefore fewer acquittals.36 Second, higher federal statutory penalties result in longer federal sentences.37

The theory is not wrong, but it has several important limitations. First, it is not always the case that penalties for prosecuted federal crimes are higher than those under applicable state law. Second, even when federal penalties are higher than their state counterparts, this difference alone does not account for the depth of the outcome disparities we see. Third, federal case disposition data reveals a somewhat more complicated relationship between penalties and conviction rates.

Begin with the first limitation. It is undoubtedly correct that in many circumstances the federal penalty for a crime is significantly harsher than the state penalty for essentially the same crime.38 But that is not always the case, and, critically, it is not the case in a number of areas in which we see robust federal enforcement. A corollary to the history of escalating federal penalties for drug and gun crimes is the history of escalating state penalties for such crimes, a history that has been largely overlooked in the violent-crime federalism debates.39

New York’s Rockefeller Drug Laws, passed in 1973, were at the time the nation’s harshest. The penalties, which remained in effect until amendments in 2004 and 2009, included mandatory minimums that exceeded those under applicable federal law.40 Other states soon followed suit. Michigan and Florida are just two examples of states whose drug penalties eclipsed their federal counterparts for decades.41 Beginning in the 1990s, many states passed “three strikes” statutes that applied to defendants with substantially less severe criminal histories than did the federal “three strikes” statute, or even the federal Armed Career Criminal statute.42 Among them were California and Louisiana, which mandated life imprisonment upon conviction of, respectively, any third felony crime43 or certain narcotics-trafficking crimes or “crimes of violence,” broadly defined to include such offenses as purse-snatching.44 California’s law was so harsh it was challenged as “cruel and unusual” before the U.S. Supreme Court, a distinction thus far not accorded to any federal sentencing law with the exception of the death penalty.45 And while many states’ penalties for gun crimes are less severe than federal penalties, not all are. Florida, Maryland, Virginia, and New York are all examples of states whose mandatory minimum penalties for illegal firearms use or possession exceed those under applicable federal law.46

The last two decades have also seen a movement in many states away from indeterminate sentencing schemes, under which offenders could secure a significantly earlier release date from a parole board. Beginning in the 1980s, and hastened in the mid-1990s in significant part by the federal “truth-in-sentencing” grant program, many states began adopting some form of determinate sentencing; requirements that prisoners (particularly those convicted of violent crimes) serve at least 85% of their imposed sentence, a ratio that substantially mirrors federal sentencing rules;47 or some combination of both.48

If more favorable federal prosecutorial outcomes were dictated by prescribed penalties, one would expect to see little federal enforcement of street crimes in states whose own drug, gun, or violent crime statutes were as least as harsh as federal statutes.49 But that is not what has happened; federal prosecutors have prosecuted thousands of cases that would be subject to an equally or more severe penalty under applicable state law. An examination of federal prosecutions of crimes covered under each of the above-discussed state laws50 reveals that federal enforcement efforts are not necessarily the product of relatively higher prescribed penalties under federal law.51

Are all these cases inadvertent misallocations of federal resources? To the contrary, they are intentioned and deliberate. Most federal street crime prosecutions are initiated at the request of local police and district attorney’s offices intimately familiar with applicable state penalties and the sentences typically imposed in actual cases.52 These cases are part of a sustained collaboration between federal and local law enforcement to reduce drug-trafficking and violent crime.53 And they are predicated on federal and local prosecutors’ belief that the defendant is more likely to be convicted and sentenced to a substantial incarceration term if prosecuted in federal court—notwithstanding a higher applicable penalty on the state’s codebooks.54

The second and more fundamental limit to federal/state penalty differences as a source of outcome disparity is that these differences do not account for the depth of the disparities that exist. The case synopses offered in the Introduction illustrate the point. For Shateek Andrews, the Bronx robber, local grand juries did not even return indictments on the armed robberies of which he was ultimately convicted in federal court—and for which he would have faced a penalty of ten to twenty-five years under state law.55 Had John Gassew, the Philadelphian arrested forty-four times for various gun and violent crimes, been convicted in local court on even one of those occasions, he would have faced a mandatory minimum sentence of ten years upon his second conviction; had he been convicted on just two occasions, he would have faced a mandatory minimum sentence of twenty-five years upon his third conviction.56 A large portion of the disparity in applicable penalties is very often a function of antecedent disparities in the ability to convict.

Penalties, of course, affect the ability to convict: the higher the penalty, the greater the inducement to reduce it through a negotiated plea bargain.57 But in this dynamic, too, case disposition data reveals a more nuanced story. For one thing, guilty plea rates do not increase (or decrease) in tandem with applicable penalties. Homicide offenses have one of the lower guilty plea rates in the federal system, but carry the highest statutory and Guidelines penalties.58 Immigration offenses have one of the highest guilty plea rates, yet most do not carry mandatory minimums or particularly high Guidelines ranges.59 And narcotics offenses, which tend to have high mandatory minimums and Guidelines ranges (and, for defendants with previous felony drug convictions, a plea-inducing penalty enhancement), have a guilty plea rate that hovers close to the mean for all offenses.60 It is difficult to ascribe this distribution to specific causal factors.61 But there are correlations, and one that stands out is ease of proof. Of federal crimes, murder is among the hardest to prove. Federal murder cases almost invariably require eyewitness testimony, usually depend upon accomplice testimony (which always carries great risk for the government), and typically involve witnesses fearful of retaliation. Immigration offenses, on the other hand, are among the easiest to prove. They are generally established with evidence of the defendant’s presence, identity, and unlawful immigration status (proven by unassailable records). Federal narcotics offenses are also relatively easy to prove. They typically involve wiretaps, undercover drug purchases, or sale-quantity amounts of drugs seized from the defendant or his residence.62 Penalties certainly incentivize plea-bargaining, but as this breakdown of guilty plea rates suggests, the outcome of that process is heavily guided by the likelihood of conviction on each potential count.63

There is more. The acquittal rate for tried cases, the overall acquittal rate, and the guilty plea rate have not changed significantly since the Supreme Court decided United States v. Booker,64 which rendered advisory the previously mandatory Sentencing Guidelines. To the extent the Sentencing Guidelines’ stiff mandatory sentences, coupled with an essentially assured reduction in return for a guilty plea,65 accounted for the high rate of guilty pleas and low rate of acquittals in the federal system, one might expect to see the guilty plea rate drop and the acquittal rate rise following Booker. Yet eight years after Booker, the federal guilty plea rate has only increased while the acquittal rate has further decreased. At the same time, non-government sponsored downward departures and variances from the Guidelines have been granted ever more frequently, and government-sponsored downward departures and variances have only slightly increased.66 This inverse trend is even more pronounced with respect to street crimes.67

There is, as well, another story that should give us pause when considering the relationship between prescribed penalties and outcomes. In 1997, the Clinton Administration piloted a firearm enforcement program in the Eastern District of Virginia, “Project Exile,” under which the U.S. Attorney adopted and prosecuted every local firearm case from the city of Richmond.68 While a debate continues as to that program’s effect on homicide and violent crime rates,69 there is no debate as to its efficacy at convicting and punishing those it prosecuted: in its inaugural year it boasted an 86% conviction rate and an average sentence of more than four-and-a-half years.70 Less attention has been paid to the relative failure of the State of Virginia’s version of Project Exile.71 On the heels of federal Exile’s touted successes, Virginia’s legislature passed a series of firearm statutes that were nearly exact replicas of their federal statutory counterparts, with the added hammer of mandatory minimum sentences that closely approximated (and in some respects exceeded) the federal sentencing guidelines.72 What’s more, the Virginia legislature also reformed its bail statute to create a presumption of pre-trial detention in firearms cases, a provision absent from the federal bail statute.73 State and federal grants were awarded to the Virginia localities selected to implement Exile, each of which assigned a full-time prosecutor dedicated to prosecuting Exile cases from arrest to sentencing, and each of which developed training teams and working partnerships with local law enforcement.74 So what happened when Virginia effectively adopted federal firearms law and dedicated intense efforts and resources to enforcing it? Virginia state prosecutors could not replicate the federal conviction or average sentence rates. Indeed, they did not even come close: of the 404 defendants prosecuted under Virginia Exile laws in the year after enactment, only 155—just 38%—were convicted of an Exile offense and received the full mandatory minimum sentence.75

If Project Exile amounts to a natural experiment, it is an imperfect one. In the first year of Federal Exile, federal prosecutors prosecuted all firearms cases in Richmond susceptible of federal jurisdiction,76 but, in the first year of Virginia Exile, local prosecutors prosecuted only firearms cases declined by the U.S. Attorney.77 For this reason, one might counter that Virginia Exile simply demonstrates another frequently advanced explanation for federal/state outcome disparities: federal prosecutors’ ability to selectively prosecute only the strongest cases, while leaving local prosecutors with cases more challenging to prove.78 But this doesn’t explain the depth of the disparity; even if every federally adopted case had been prosecuted locally and resulted in a conviction, the local conviction rate would rise to just 53%.79 n moin 2010), with a mean of 93n moin 2010), with a mean of 93n moin 2010), with a mean of 93n moin 2010), with a mean of 93n moin 2010), with a mean of 93Something additional is at play in the Exile story—something beyond penalties, resources, and case selection.

All this is not to say that prescribed penalties do not impact outcomes: of course they do. In many states and for many crimes, more stringent federal penalties indeed give federal prosecutors extraordinary leverage, particularly when accounting for guilty pleas induced via charge bargaining.80 But even when federal penalties do operate in this manner, they do not act in isolation. Other, less overt forces are at play.

B. The Procedural Rules Explanation

Another frequent explanation for better federal prosecutorial outcomes is federal procedural rules that tend to advantage prosecutors. Commentators have pointed out that in recent years many state courts (and, in some instances, legislatures) have increased defendants’ procedural protections above the floor set by federal constitutional law.81 Some states, for instance, do not recognize the “good faith” exception to the warrant requirement.82 Others rely on less flexible standards in evaluating the reliability of an informant’s tip in assessing probable cause.83 Still others prohibit convictions based only on certain types of evidence, most notably uncorroborated accomplice testimony.84 And other states have stricter requirements for grand juries to indict—for instance, requiring that grand jury evidence comport with the rules of trial admissibility.85 Procedural rules such as these undoubtedly give federal prosecutors a leg up on their state and local counterparts in certain states. But in prosecutions of street crime, how much does that leg up really matter?

In practice, probably not a great deal. As an initial matter, the distinctions between federal and state constitutional criminal procedure are not so stark. In a number of states, state constitutional search and seizure doctrine has not diverged at all from federal search and seizure doctrine as set forth by the federal courts.86 And even in states where it has, those differences operate largely at the margins; most of the time, constitutional infirmities in street crime cases, if present, will exist under both state and federal law.87

More to the point, though, federal prosecution of street crime is rarely a stand-alone enterprise. Rather, it operates as a complement to what is almost entirely a state and local law enforcement effort.88 This has two ramifications for constitutional criminal procedure. First, because many federal narcotics and violent crime cases begin in state court, those cases have already arguably satisfied state procedural law: local prosecutions would not have been initiated absent a likelihood of prevailing on a suppression motion. Second, even cases filed in federal court at the outset will have generally relied at least to some extent on assistance from local law enforcement officers, who must operate within state constitutional constraints.89

It is difficult to overstate this second point. Enforcement of street crime, even at the federal level, requires local policing knowledge and expertise.90 Even the DEA and ATF, which investigate almost exclusively narcotics and gun cases, have relatively little pre-investigation knowledge of neighborhood intricacies and interpersonal relationships among an investigation’s subjects. A DEA agent might identify a particular target as a member of a violent drug gang, but the officers in the local precinct already know that target, potentially very well. They may have arrested the target or his accomplices on prior occasions; they probably know what bars or clubs the target and his gang frequent; or they might already have an informant who can “make buys” (purchase narcotics in an undercover capacity) from the target or his gang. Such intimate knowledge is critical in federal street crime prosecutions.

The practical upshot of this dynamic is that narcotics and violent crime investigations do not happen without local law enforcement involvement.91 And, because local law enforcement can never be sure in the end whether the case will “go federal,” they must operate under state procedural requirements. That means that if, for instance, state law requires a warrant when federal law does not, the local police will often err on the side of getting a warrant. This is not always the case; sometimes, it is clear from the outset that the federal government intends to devote its resources to the case and will prosecute no matter what. But given the absence of advance knowledge as to what an investigation will ultimately uncover, it is generally difficult to make that prediction.92 More often than one might think, federal investigations do not ultimately result in federal prosecutions.93 And even when a federal investigation does result in federal prosecution, it is not always the case that every subject ensnared will be prosecuted federally; some will be prosecuted in state court, where state rules of procedure will govern.

And what of the evidentiary limitations imposed by some states, such as prohibitions on the use of uncorroborated accomplice testimony94 or stricter evidentiary requirements in the grand jury? These rules may go a good deal of the way toward explaining why federal prosecutors are able to bring cases that state and local prosecutors cannot, but they don’t fully explain why federal prosecutors win those cases. More liberal evidentiary licenses are a double-edged sword when it comes to proving a case before a trial jury. Such rules may permit indictments or convictions based on less evidence, but the prosecutor’s burden of proof to the jury remains. Federal prosecutors, of course, know this; it is why they do not hew to the bare minimum of evidentiary requirements. In the vast majority of street-crime prosecutions, federal prosecutors, like their state counterparts, will hesitate to proceed absent corroboration (and, ideally, substantial corroboration) of accomplice witnesses through other evidence.95 And when they cannot do so, the question remains: what explains federal prosecutorial successes even in the face of relatively flimsy proof?

Professor John Jeffries and Judge John Gleeson touched on this issue when, in explaining how federal prosecutors obtain convictions in organized crime cases based on accomplice testimony, they observed that “one of the most dramatic advantages of the federal system” is that, by insulating accomplice witnesses from the other evidence in the case, “an accomplice-based case can be made to hinge not on the credibility of the inherently unreliable accomplices, but on the jury’s assessment of the integrity of the prosecutors. Federal prosecutors usually win such cases.”96 The question that has yet to be answered is: why?

C. The Selectivity and Resources Explanations

Two other leading explanations for the disparities between federal and local street crime prosecution outcomes are interrelated. The first is that, unlike state and local prosecutors who bear ultimate accountability for criminal prosecutions, federal prosecutors have the luxury of choosing which cases they bring.97 The second, a corollary of the first, is that federal prosecutors have a lower caseload and therefore substantially greater time and resources to devote to their cases than local prosecutors.98 In short, a great deal of federal prosecutorial success can be attributed to the political economy of prosecution in our federalist system.

These points have strong intuitive appeal. Being able to cherry-pick one’s cases, and having fewer cases and more resources to spend on them, seems a recipe for prosecutorial success99—and, to a large degree, it is. However, in assessing these federal advantages, we must also consider to what extent state and local prosecutors lack them.

Start with case selection. It is true that state attorneys general and county district attorneys are directly accountable to their electorate in a way that appointed U.S. Attorneys are not. It is also true that state and local prosecutors (particularly local prosecutors) are responsible for prosecuting all readily provable crimes committed within their jurisdiction. Yet notwithstanding these pressures, local prosecutors frequently decline cases.100 Political accountability imposes pressure to bring cases,101 but it also imposes pressure to win them.102 How these competing forces impact local district attorneys’ dismissal patterns, and how those patterns compare to evidentiary-based federal declinations,103 are complex questions worthy of further study. But we should not presume federal prosecutors are always more willing and able than their local counterparts to decline weak cases.

Moreover, given the nature of gang and drug-related street crime, in which victims of one crime are often also perpetrators of another, it is not necessarily the case that political accountability breeds aggressive enforcement. Indeed, the contrary may be true. It is no accident that much of the federal street-crimes docket consists of crimes against other criminals—robberies of drug dealers, murders of rival gang members, and the like. It is not merely the jurisdictional hook that lures the feds, who could easily fill their Hobbs Act docket with far more sympathetic victims of, for instance, convenience store robberies.104 It is the absence of political pressure on local district attorneys to bring cases involving unsympathetic (and usually disenfranchised) victims. These cases are often the most challenging to bring, and to win.

This raises another point. We should not presume that in selecting cases, federal prosecutors’ interest in winning convictions outweighs other interests. To the contrary, the history of federal criminal enforcement is peppered with hard-fought, high-profile cases that have generated both wins and losses. Federal prosecutors certainly “cherry pick” their cases—but often, the cherries are not the quick or easy cases.105

Whatever can be said of federal prosecutors’ selection criteria, though, the fact remains that federal prosecutors have substantially fewer cases to prosecute than their local counterparts.106 This translates into more per-case resources in the form of time, manpower, and money. The claim that federal prosecutors’ resources contribute to their success is unassailable. But how far can it take us?

Resources are critical; prosecutors cannot bring cases without them. Moreover, in a system that depends almost entirely on pretrial resolutions for its effective functioning, resources help define the “going rate” for the punishment of offenses. The greater the pressure prosecutors face to dispose of cases without the expense and time of a trial, the weaker their leverage in plea negotiations. Because much of a federal prosecutor’s docket is discretionary,107 particularly in the street crime context,she can afford to take to trial those defendants who will not agree to an appropriate (in the prosecutor’s view) plea bargain. This willingness and ability in turn reduces the discount rate offered for a guilty plea, at least relative to offers that might be extended by her local, overburdened counterparts.

Yet this is only part of the story. Willingness to proceed to trial means very little without the proven ability to secure a guilty verdict. Resources undoubtedly help in that regard, but it is worth asking how and why this is so. Federal prosecutors do not have unlimited time and money (particularly in low-profile street crime cases) to pursue every evidentiary lead and every possible target, to make unlimited undercover drug buys, or to wiretap every phone for which they could seek court approval. They, too, must make choices. And, almost uniformly, federal prosecutors choose to devote their resources to preparing cases for trial. Federal prosecutors and agents scrupulously track down and prepare witnesses (including law enforcement witnesses, who sometimes require substantial preparation), and devote time and funds to securing witnesses’ safety. These choices matter not merely because they improve trial outcomes in individual cases, but also because they leave a lasting impact on participants’ (and ultimately the community’s) perceptions of federal law enforcement—perceptions that reap benefits in future cases.108

It is worth thinking, as well, about those high-profile cases in which the federal government has expended seemingly limitless resources, and lost. John Edwards,109 Roger Clemens,110 and the “African sting” prosecution under the Foreign Corrupt Practices Act111 are all recent examples of cases in which resources could not overcome the challenge of prosecuting conduct on the margins of what the public deems imprisonment-worthy.112 These sorts of cases present evidentiary burdens quite different from street crimes, but the lesson translates: even abundant resources cannot cure a deficit of trust in the government’s case.113

The foregoing discussion is not meant to discredit conventional explanations for forum disparities but, rather, to expose the complexities within them. Each explanation, and all of them together, takes us quite far in understanding criminal forum disparities. But there are other factors we must explore too. The balance of this Article starts that project.

II. legitimacy and criminal enforcement

The concept of “legitimacy” in governance, and in criminal enforcement in particular, has received sustained and deep engagement over the last several decades by social psychologists, criminologists, and legal theorists. As a result, there is now a rich body of theoretical and empirical literature that helps us to understand what factors influence perceptions of legitimacy in law enforcement.

The work of Tom Tyler and others demonstrates that people comply with the law in large part out of a sense of obligation to the authorities enforcing the law (“legitimacy”), as well as because the law comports with their own personal conceptions of right and wrong (“personal morality”).114 In studies that have compared the influence on compliance of various attitudinal and background factors, legitimacy and personal morality proved to be among the most significant.115 By contrast, the threat of sanction was among the least influential factors.116 The relative unimportance of instrumental motivations might explain why people continue to commit crimes despite the threat of harsh punitive sanctions.117

These two aspects of compliance—“legitimacy” and “personal morality”—have spawned two largely separate bodies of literature.118 The former asks what factors influence people’s support for and obligation to the governing authorities, while the latter asks which crimes, theories of liability, theories of defense, and penalties most comport with public conceptions of right and wrong. Research has shown that people’s perceptions of an authority’s legitimacy are influenced most by their perceptions of the fairness of the process and procedures by which it enforces the law.119 This theory of compliance has thus come to be known as “procedural legitimacy.” The extent to which substantive law aligns with personal morality has been described alternatively as “moral credibility,” “moral alignment,” or “empirical desert.”120 Both theories concern the ability of governing authorities to harness voluntary compliance, cooperation, and deference, as opposed to coercing it through threat of sanction.

A. Procedural Legitimacy

What factors influence perceptions that an authority is legitimate? Substantial qualitative research suggests that, in interactions with the courts and police, people place greater value on the process of their interaction with criminal justice authorities and the motives of those authorities than on whether the outcome of the interaction is ultimately fair or favorable to them.121 These two factors—termed “procedural justice” and “motive-based trust”—have been shown to significantly shape citizens’ perceptions of the legitimacy of legal authorities.122 Procedural justice and motive-based trust influence people’s overall views of the legitimacy of criminal justice authorities (courts and police), as well as their willingness to defer to and cooperate with those authorities in a particular case.123

What personal traits influence perceptions of legitimacy? Age, race or ethnicity, gender, education, income, and political affiliation, among other things, each have an impact; yet when isolated, the influence of each of these background characteristics is relatively minor.124 Research further suggests that across racial and ethnic groups, procedural justice, rather than outcomes, is the key predictor of perceived legitimacy;125 lower perceptions of legitimacy among minorities can be traced to feelings of unfair treatment rather than outcomes.126 In addition, personal identification with the group or entity through which authority is exercised (“superordinate identification”) has been shown to have significant impact on evaluations of authority and perceptions of legitimacy.127

Procedural legitimacy has tremendous promise for governance in general and criminal justice in particular because it enables governance based on the public’s trust. A legitimate authority has the diffuse support of those whose behavior it regulates, a construct often described as a “reservoir” of trust or goodwill.128 Such diffuse support, in turn, allows the authority to impose and effectuate unfavorable outcomes on those adjudged guilty of wrongdoing. Diffuse support also gives the authority discretionary latitude to regulate within a range of behavior, even when the regulation is at odds with the moral beliefs of some (or even many) members of the community.129 Procedural legitimacy is thus particularly important in a pluralistic society with diverse moral beliefs.

Legitimacy is self-reinforcing, as is its absence. Research has shown that preexisting conceptions of an authority’s legitimacy influence compliance, deference, and cooperation, which in turn enhance the authority’s ability to govern fairly, further reinforcing its legitimacy.130 By contrast, preexisting perceptions of illegitimacy result in noncompliance and refusal to cooperate and defer, breeding more crime and hindering the criminal justice system’s ability to enforce the law—often resulting in resort to force or other methods that further erode legitimacy.131 Legitimacy is self-reinforcing in another sense too: people who perceive the criminal justice system as legitimate are more likely to be influenced by process concerns than instrumental concerns in their interactions with the justice system.132 Put differently, legitimacy moves people to value process over outcome, while illegitimacy moves people to value outcome over process. In a system in which unfavorable outcomes are inevitable, illegitimacy thus creates “a spiral of increasing conflict and decreasing legitimacy.”133

B.Moral Credibility

Empirical research indicates that people are more likely to comply with laws with which they agree134 and more willing to defer to and assist authorities they view as enforcing just laws.135 Substantive law that does not reflect the community’s norms is ultimately unsustainable because effective law enforcement depends on the cooperation and respect of those within its ambit (witnesses, victims, jurors, judges, prosecutors, and defendants).136

Moral credibility’s limitations are obvious. What if the moral convictions of a society’s members are dissonant?137 What if social norms depart from philosophical and ethical conceptions of justice?138 For moral legitimists, these limitations arise from the law’s own inability to shift community norms. While the criminal law “sometimes nurtures the norm[s]” through prosecution and enforcement, “[t]he criminal law is not an independent player in that process,” but merely “a contributing mechanism.”139 Accordingly, for a society to function effectively, it must derive both moral and procedural legitimacy. That is, it must strive to enact laws with which the public largely agrees and, in the absence of societal agreement, it must be able to draw on the public’s support for its institutions as the final arbiters of what the law should be.140

C. Legitimacy and Criminal Adjudication

In the context of criminal investigation and adjudication, work on procedural legitimacy has primarily focused on how the criminal process affects perceptions of legitimacy.141 In fact, though, criminal adjudication has a dual function: it both affects and reflects legitimacy.

Begin with the criminal jury. Empirically assessing the factors motivating jurors’ decisionmaking is exceedingly difficult; defendants’ due process rights limit the means and methods by which researchers can investigate jurors’ pre-trial attitudes and predilections,142 jurors themselves may be unaware of the factors influencing their reasoning,143 and the interaction of both known and unknown variables obscures attempts to isolate and control for them.144 Nevertheless, the empirical work on juries, for all its limitations, tends to support what criminal practitioners intuit: jurors’ assessment of the evidence, and predisposition to the prosecution, is guided in part by their perceptions of the legitimacy of the pertinent laws and legal authorities. On the whole, jurors who report trust and confidence in the police are more likely to be predisposed to the prosecution in a local criminal case.145 So, too, are jurors who believe the applicable laws, and the consequences of conviction, to be fair.146

The criminal jury, though, is only the beginning of legitimacy’s measure. As others have pointed out, the adjudication of criminal cases serves an expressive function.147 It tells us what conduct is actually prohibited, and what level of opprobrium we accord that conduct. The legislature defines crimes, but prosecutors decide whether to charge those crimes. Likewise, the legislature ascribes penalties for crimes; but prosecutors decide, by virtue of the charges they bring, which penalties will apply, and judges decide what specific sentence will be imposed in a given case. And in a system that functions almost entirely through plea-bargaining, prosecutors’ decisions—which cases to charge, what charges to bring, and what charges to dismiss—reflect the realities of how juries or judges decide tried cases, and what sentences judges will impose. In short, the criminal law in the statute books can be quite different from the criminal law in practice. The adjudication of cases, through trials, guilty pleas, and sentencings, tells us what “the law” in each jurisdiction really is.

Prosecutors dismissing cases or charges because witnesses refuse to cooperate or because juries will not credit the testimony of those who do; juries acquitting defendants, or grand juries refusing to charge them, because of a lack of trust in the prosecution’s evidence; judges sentencing far below mandated or advisory terms because of a fundamental disagreement with those penalties; and plea bargaining practices that reflect these realities—all of these phenomena are expressive. And what they express, at least in part, is a disagreement between the community and the criminal justice authority over how the criminal law should be written, how it should be enforced, or both. They are symptoms of the criminal justice system’s illegitimacy. And the symptoms feed the disease. When the public sees that certain conduct is rarely policed or penalized, it internalizes those norms. The newly internalized norms result in the perception by some in the community that the law is under-enforced and by others that the law “as written” is over-punitive relative to community norms. These perceptions further erode legitimacy.148

In contrast, when we see a criminal justice system characterized by relatively high rates of convictions, adherence to mandated or advisory penalties, and pre-trial resolutions that reflect these realities, we must ask whether legitimacy might have something to do with that. As I have acknowledged, these aspects of the federal criminal justice system149 can all be explained by other factors, up to a point. Let us consider whether legitimacy might, at least in part, fill the explanatory gap.150

III. legitimacy, street crime, and the federal justice system

A legitimacy inquiry focuses on the public’s belief that the laws and criminal justice authorities are entitled to deference. Accordingly, to assess a justice system’s legitimacy, we must examine the relevant authority, the relevant public, and the laws. This Part addresses each of these three elements and their interactions with one another.

A. Authority

In every prosecution there are two relevant authorities: the authority that investigates and the authority that prosecutes. In state and local prosecutions, both authorities are typically arms of the locality: the police and the district attorney’s office.151 In federal prosecutions of violent crimes and narcotics, the prosecuting authority is the United States, specifically the United States Department of Justice (typically operating though the United States Attorney’s Office for the district in which the crime is being prosecuted). The investigating authority may be federal (for instance, the FBI, DEA, or ATF), may consist of only the local police, or might include (as is usually the case in violent crimes or narcotics cases) a combination of the two. But even when only the local police were involved up to the point of arrest—as is often the case when the federal government “adopts” a local case already underway in state court—there will almost always be a federal agent assigned to shepherd the case through the federal system or, at the very least, to offer assistance with evidence evaluation or collection.152 The federal agent assigned to the case sits with the prosecutor at counsel table and is often introduced to the jury either during voir dire or in the government’s opening statement as one of the representatives of the United States responsible for bringing the case on behalf of the federal government. From the standpoint of public perception, the “face” of any federal prosecution will always be the federal government.

The significance of federal authority is not lost on federal prosecutors. Often, they can be heard in courtrooms telling the jury in opening statements that it is their privilege to “represent the United States.” In cases with little federal investigatory involvement, prosecutors may make strategic choices specifically to ensure that a federal law enforcement agent appears before a jury. For instance, in felon-in-possession cases in which local police apprehended the defendant and seized the firearm, federal prosecutors sometimes decline to stipulate to the fact that the gun was manufactured in another state (the requisite interstate element of the crime) in order to call an ATF agent as an expert witness on the firearm’s site of manufacture. In a drug case investigated by local police, federal prosecutors might call a DEA agent as an expert witness to testify that the type of packaging and amount of drugs seized is consistent with distribution rather than personal use.

Federal prosecutors do these things because they intuitively understand the power of the federal authority. But what accounts for that power? It is not enough to say that the federal authority garners legitimacy simply because it is “federal.” History gives us examples of how quickly that construct can break down.153 Nor can one generalize about “federal” authority, a concept that can have very different meanings depending on the context.154 We must instead examine what it is about the nature of street crime, and the nature of federal law enforcement, that might enhance the legitimacy of the federal authority in street crime prosecutions.

Street crime prosecutions have three salient features. First, they typically rely on the testimony of local police officers. Second, the credibility of law enforcement witnesses (whether local police or federal agents) is often critical to a disputed issue in the case. Third, violence, or the threat of it, lurks within every case.

Consider the first feature. Street crime is so-called because it takes place, or at least originates, in the street: the open-air drug market, the corner store robbery, the drive-by shooting. Local police protect the streets. Even in cases arising from long-term federal investigations, federal prosecutors will inevitably draw upon the testimony of local police officers who, over the course of the investigation or preceding it, responded to reports of episodic crime related to the government’s case.155

The second feature—the enhanced importance of law enforcement’s credibility—is perhaps the single biggest evidentiary difference between street crime and white collar prosecutions. In white collar cases, law enforcement plays a supporting role. Agents might seize evidence or make an arrest, but very little depends on their word. It is difficult for a defendant to argue that a law enforcement agent planted evidence when the evidence consists of reams of detailed financial statements or emails on the defendant’s own computer. And because a defendant will often have retained counsel prior to having been arrested or even approached by law enforcement, he ordinarily will not make any statements to the agents. In violent crime and narcotics cases, in contrast, law enforcement officers in many ways are the evidence—or, at least, a substantial part of it. It is they who respond to the scene, where they witness first-hand either the crime in progress or its immediate aftermath. If the defendant attempts to flee, they give chase. When they apprehend the defendant, they search him and question him. If he has contraband, it is they who seize it; if he speaks, it is they who hear it. The testimony of law enforcement officers in street-crime cases is critical. Sometimes, it is the only evidence that bears on guilt.156

The third feature, violence, is critical because it imposes a uniquely onerous burden on victims and witnesses (including cooperating defendants). Violence is ever-present in street-crime cases. Witnesses fear retaliation even if no overt threat against them is ever made.157 Even in cases in which no violent act is charged, such as drug cases or gun possession cases, the threat of violence against witnesses remains.158 The near inseparability of drugs and violence in the inner city is so well-recognized that it has spawned permitted evidentiary inferences in gun and drug cases.159 And in actual violent crime cases (murders, shootings, robberies), the threat of witness retaliation is palpably real. In some cities, witness intimidation in violent crime cases has become so prevalent that it has crippled local prosecutors’ ability to bring successful cases.160

If local policing, law enforcement credibility, and violence lie at the heart of street crime cases, we must ask which aspects of the federal authority bear on these issues. Three critical features stand out. First, federal law enforcement does not police citizens. Second, federal prosecutors have demonstrated a robust commitment to policing local police departments’ interactions with citizens. Third, federal law enforcement has cultivated a sterling reputation for witness protection. Each of these three aspects of federal law enforcement authority implicates procedural legitimacy’s key components: procedural justice and motive-based trust.161

Begin with the absence of policing. This matters to public perception because, from the standpoint of perceptions, policing in the inner city is an enterprise with great downside and little upside. Urban police forces suffer from high levels of community dissatisfaction.162 Victims of crime perceive the police as unresponsive;163 suspects and defendants perceive the police as overbearing and abusive.164 Because both street crime and policing tend to be more concentrated in poor, minority communities, minorities, and African-Americans in particular, perceive these failures as by-products of pervasive racial discrimination by local police forces.165 High crime levels and underfunding exacerbate the police practices that fuel these perceptions. Overburdened police forces have little time to investigate episodic street crime beyond the immediate aftermath and even less time to follow up with victims and apprise them of the status of their case. In a number of cities, political pressure to clamp down on violent crime has resulted in police departments adopting order maintenance policing, a technique that has come to be equated with aggressive enforcement of minor misdemeanor laws and pervasive use of stop-and-frisk.166

Federal law enforcement agents, in contrast, have all of the benefits of being law enforcers without the attendant burdens of policing. The feds do not walk a beat, which means that they do not have occasion to stop and frisk people. They are not the ones who show up in response to a 911 call. They do not interact with the community at all, in fact, save for the relatively rare situations in which they are called in to investigate a crime after it has already been reported. The lack of community interaction with federal law enforcement matters a great deal, because research has shown that people’s perceptions of and satisfaction with police is grounded heavily in their personal experience.167 That personal experience, in turn, significantly influences perceptions of the justice system’s legitimacy.168

Now consider the federal authority’s sustained and visible role in enforcing the rights of individuals against local police misconduct. Rodney King,169 Abner Louima170—the federal criminal prosecutions of local police officers in cases such as these evoke positive images of the federal government as much as they evoke negative images of the local police forces involved.171 And the federal government has prosecuted less notorious cases of police misconduct, securing convictions that might lack national notoriety but nevertheless matter greatly to local communities.172 The federal government’s use of civil injunctions to remedy pervasive and systemic civil rights violations has also burnished its reputational capital when it comes to policing.173

In short, the federal government doesn’t police the community, and it does police the police. In cases that depend both on the local police and law enforcement credibility generally, then, the federal authority has a uniquely powerful voice. When a federal agent takes the witness stand, he is, in a very real sense, a physical embodiment of the federal authority. And when a federal prosecutor calls a local police officer to the stand, that officer, too, becomes imbued with federal authority. Implicit in the act of calling the officer to testify is the judgment—by the federal entity with the proven ability to prosecute him—that the officer has acted appropriately in the discharge of his duties and that he is telling the truth.

Finally, the federal government’s reputation for witness protection further enhances its power in street crime prosecutions by signaling a protected space to otherwise reticent witnesses. It is not that the feds can guarantee witness safety—they can’t, as any federal agent will attest. It is that witnesses perceive safety in the hands of a federal agent. Resources certainly play a role in this, but not the only one. Indeed, for witnesses in public housing (as many are), relocation is less a question of money than bureaucratic hassle. It is also about image. The federal witness protection program is the stuff of lore in popular culture.174 And although few witnesses in street crime cases need avail themselves of such stringent protection measures, their mere existence graces federal agents with credibility. Indeed, in joint federal-local investigations, it is no accident that witness security tends to be the primary responsibility of federal agencies, regardless of the extent to which local law enforcement agencies share in the expenses. When it comes to witness security, federal agents inspire trust in a way local law enforcement simply doesn’t. In urban criminal justice systems “gridlocked with fear,”175 this trust can pave the path to the witness stand.

B. The Public: Authority and Community Interaction

Legitimacy focuses on public perceptions of authority. Who, then, is “the public?” In a criminal justice system, the public is the community subject to the justice system—those who serve on its juries, elect its district attorneys (and, in many states, its judges), and are policed by its police department. In state and local justice systems, the community is the local county. In large cities, the local county is typically confined to an urban area, which may or may not include surrounding suburbs.176

In federal prosecutions, the community is the federal district, and it has two distinguishing features. First, a federal district’s community does not elect its prosecutor or its judges.177 Second, the federal district always comprises a larger geographic area than the local county.178 Depending on the district, it can comprise anything from one city and its surrounding counties to an entire state.179 Federal juries are therefore less representative of the local community in which the crime occurred. And the lack of direct political accountability by federal criminal justice authorities (both judges and prosecutors) means that these authorities will be less beholden to local constituents. In short, in the federal justice system, “the public” is both geographically broad and politically diffuse.

Scholarly attention on the jury’s role in the federal criminal justice system has focused principally on these two features and their impact on racial composition180 and prosecutorial discretion.181 Here, I focus on three additional effects of geographic breadth and political diffuseness and how those effects might bear on legitimacy. I first consider the impact on jurors’ self-identification. Next, I examine the demographic alignment between the jury pool and the larger electorate. Finally, I consider the empowerment of law over norms.

1. Juror Identification

Research has demonstrated that one’s identity with the group through which authority is exercised—that is, the “superordinate” group—significantly impacts perceptions of legitimacy.182 This phenomenon operates on two levels. First, the more one self-identifies with the superordinate group, the more likely one is to value the relational (i.e., process) rather than the instrumental (i.e., outcomes) in assessing legitimacy.183 Second, high levels of identification with a superordinate group are correlated with increased perceptions of legitimacy.184 Studies have shown, moreover, that these correlations remain regardless of the strength of an individual’s identification with his or her own subgroup—indicating that, even in societies divided into subgroups along racial, ethnic, socioeconomic, or other lines, shared identification with the superordinate group enhances perceptions of the legitimacy of authorities.185 And a number of studies have shown that strong self-identification with America in particular enhances both valuation of relational concerns and perceptions of legitimacy.186

The research on superordinate identification can enrich our thinking about the jury. A jury is tasked with overcoming differences in order to reach agreement. It must speak as a single body rather than as the collection of individuals who comprise it. And through the deliberative process, jurors’ perceptions are further refined. Uniform perceptions among jurors may be self-reinforcing; asymmetry may breed greater movement in juror views.

In federal court, the superordinate authority is the United States, and the superordinate group, Americans. In a local county court, on the other hand, the authority’s identity is less pronounced. The prosecutor is an agent of the locality, but she enforces state law. She might describe herself, variously, as representing the state or the “people” of an unarticulated community. The superordinate group, then, is also undefined. In a Bronx County court, for instance, is the superordinate group those residing in the county (the Bronx), the locality (New York City), or the state whose laws the prosecutor seeks to enforce (New York State)?

One study in particular supports the distinction between national and local identification. In that study, which polled 1,656 adults in Oakland and Los Angeles, eighty-nine percent of respondents said that being an American was important to their identity, while only sixty-two percent of respondents said that being a resident of their city was important to their identity.187 More significantly, the study tested the extent to which self-identification at the national versus local level influenced perceptions of legitimacy. The correlation between national identity and legitimacy was nearly three times greater than the correlation between local identity and legitimacy.188 In other words, people self-identifying more strongly as Americans were more likely to perceive authorities as legitimate.189

One cannot ascribe too much importance to a single study, particularly when confined to one state. Additional studies in different states and localities are needed, along with studies that assess self-identification with one’s state and the impact of blended authority on superordinate identification. But this study, and superordinate identification research more generally, has intriguing and potentially significant implications. Strong juror identification with the prosecuting authority in federal cases might, in small part, explain why, for instance, in the District of Columbia the conviction rate is substantially higher in federal court than Superior Court, even while the two forums share the same jury pool.190 It might also in part explain why the federal District of the Virgin Islands, where U.S. citizen residents cannot vote in national elections and many are apparently apathetic as to their political status,191 has the lowest conviction rate of any federal district.192 It may also in part explain why all-white federal juries in the Jim Crow south convicted white defendants accused of crimes against African-American victims far more frequently than did all-white juries in state courts at that time.193

Note the qualifier: in part. There are numerous complex factors that impact a jury’s decision in a given case or that contribute to patterns of jury decisions over time. Many of them, such as evidentiary strength or case type, surely have a greater impact than superordinate identification. And subgroup identification is also important; clearly, in criminal law and in street crime in particular, race matters.194 But to this complex mix of influences, we should add superordinate identification, and we should undertake to better understand its potential impact on juries across forums.

2. Adjudicative-Legislative Alignment

The distinguishing feature of “street crime,” as its name implies, is its locus. Unlike many other types of crime, street crimes harm not just their immediate victims, but the public life of the local community.195 Streets, parks, schoolyards, the entryways and stairwells of apartment buildings, the corner store—these are the “scene of the crime.” For residents in communities ravaged by street crime, the places of daily life become zones of fear. They also become the focus of police attention. For street crimes, more than any other, it is the local community that principally bears the burdens of the law’s violations and its enforcement.

The law-making that defines and punishes street crimes, though, occurs at the state rather than local level. Because states comprise all demographics while street crime affects primarily urban areas, laws governing street crime are made at a level of government not directly answerable to the communities the laws most affect. Adjudication, on the other hand, occurs at the local county level. To the extent, then, that local county norms of conduct or ideologies about punishment differ from state norms, those differences manifest in case adjudication. Juries might acquit more often for certain crimes. Judges might seek ways to circumvent sentences mandated by the state legislature.

This is precisely what happened in California during the time of its original “three strikes” law, a statute that was enacted pursuant to a state-wide referendum ballot.196 Juries and local judges in San Francisco—a consolidated city-county197 and the only county in the state in which a majority of voters voted against the law—frequently nullified the law; as a result, the San Francisco District Attorney declined to enforce it.198 Juries in San Diego County, on the other hand—a county comprised of both urban and suburban areas, in which approximately 76% of voters had voted in favor of the three strikes law—regularly returned convictions in three strikes cases, resulting in robust enforcement of the law in San Diego County.199 A similar dynamic occurred in Virginia when the state legislature enacted firearms laws that mirrored federal law. In Richmond County, also a consolidated city-county, and the site of the tremendously successful federal “Project Exile,”200 local prosecutors ran into roadblocks when they sought to replicate federal results in local court. Many local judges balked when told that the new laws mandated a presumption of detention in all firearms cases, and simply granted bail as if the heightened evidentiary standard did not exist.201 Some judges were reported to engage in nullification, acquitting defendants despite strong evidence of guilt simply to avoid imposing what they viewed to be an excessive punishment.202 And in New York, the 2006 law mandating a three-and-a-half-year term for illegal possession of a gun has been applied inconsistently. Defendants in the Bronx often avoid any state prison sentence, while the mandatory minimum sentence is vigorously enforced against defendants in Queens County.203

On the surface, the federal criminal justice system would seem to operate in a similar fashion: laws are made at the national level (by the U.S. Congress) and adjudicated at the local level (by juries and judges in the ninety-four federal districts). In reality, it is not at all similar. An overlooked feature of the federal criminal justice system is its alignment, at least in terms of demographics, between the legislative and adjudicative levels. Almost every federal district contains urban centers, suburbs, and rural areas.204 In some districts urban areas might be more represented, while in others suburban or rural areas might make up more of the mix. But there is not a single federal district (with the exception of the District of Columbia) that does not at least contain some portion of all three types of neighborhoods.205 In this sense, federal districts reflect the mix of communities that occurs on a national level.

This means that within the federal system there is less demographic disconnect between the constituencies that elect the lawmakers (i.e., the national electorate) and serve as jurors (i.e., the district-wide jury pool). For street crime, which primarily affects one demographic (urban dwellers), the demographic alignment between the legislative and adjudicative communities is significant. If criminal adjudication at least partly serves to give voice to “community” norms, it is unsurprising that in the federal system we see less discord between the norms of the community that makes the laws and the norms of the community that adjudicates violations of them, particularly when it comes to cases of street crime.

This is not to say this aspect of the federal system is preferable; those who think of the jury as a “check” on legislators and prosecutors might argue it is not (an argument I engage in Part IV). But before we talk about the jury’s role in constraining the criminal law, we must appreciate the jury’s relationship to the polity—not just the polity that elects district attorneys and judges, but the polity that elects lawmakers.206

3. The Empowerment of Law over Norms

If the arguments of moral legitimists are correct, criminal law can do little to move a community’s social norms when those norms are well established and widely adopted.207 But theories of moral legitimacy also recognize that, when norms within a community differ, criminal law has the power to dictate which norm will prevail—a power strengthened by the authority’s reservoir of public support.208 If the authority dictating the law is perceived as legitimate, the community will defer to the law’s enforcement. Over time, continued enforcement might even serve to shift social norms to align more closely with the law’s dictates. Even if the law fails to shift norms, enforcement will continue—people will be arrested, prosecuted, convicted, and sentenced in accordance with prescribed penalties—as a result of the larger community’s deference to the authority as the legitimate arbiter of conflicting norms.

A larger and more geographically diverse community contains a larger and more diverse set of norms. It follows, then, that the larger and more diverse the community, the greater the criminal law’s power to dictate norms. This is what we see in the federal criminal justice system: a large “community” encompassing broad swaths of a state’s population, communities, and attendant norms. Such norm diversity, in turn, empowers federal criminal law.

The empowerment is most visible in the prosecution of street crimes, where the relevant norm-setting community is uniquely delineated along geographic lines. A defendant accused of insider trading would almost certainly never face twelve jurors all within the community of Wall Street investors and traders, regardless of whether he were tried in federal or state court. But a defendant accused of selling heroin on the corner could, conceivably, face quite different communities, from the perspective of norms, in a federal versus local forum. The local forum would almost certainly include jurors who were personally affected by drugs or neighborhood drug dealing and its attendant violence; who knew someone so affected, a drug dealer, or the family members of an imprisoned drug dealer; or who were, or knew others who were, subjects of police stops. The federal forum might result in a jury bereft of any personal experience with street-level drug dealing and enforcement of the drug laws.

The jury, though, is only the start of the law-over-norm dynamic in federal court; it is at sentencing that the real work gets done. Consider one salient example. In inner-city communities, illegal gun possession, particularly by those already involved in the criminal justice system, is a norm of street life that conflicts with, and effectively lays siege to, the norms of the large majority of community residents (who are law-abiding).209 The criminal law in nearly every state aligns with the norms of the law-abiding residents and prohibits convicted felons from carrying a firearm.210 Yet in some local urban courts, the street norm overshadows the criminal law. In these courts, non-custodial sentences are not infrequent, notwithstanding the law’s applicable penalties.

For instance, in New York prior to 2006 (when the legislature decreed a three-and-a-half-year mandatory-minimum penalty for illegal gun possession), non-custodial sentences for illegal gun possession were common—even though then-existing statutes mandated a sentence of two to eight years, subject only to departures for mitigating circumstances.211 And in the Bronx, custodial sentences in gun cases remain rare, despite the new law.212 In Virginia in 2000, following the passage of mandatory sentences for certain gun crimes, some judges evaded the new sentencing laws through the use of charge reductions, bench trial acquittals, and suspension of sentences.213 Probationary sentences for gun possession are not uncommon in Philadelphia, where, in 2010, fifteen percent of defendants convicted under Pennsylvania’s felon-in-possession statute received sentences of probation and an additional twelve percent received sentences of imprisonment in county jail, for which the average minimum hovered around one year—notwithstanding sentencing guidelines ranging from eighteen months to ten years depending on the defendant’s criminal history (guidelines that are, incidentally, substantially similar to federal sentencing guidelines for the equivalent crime).214

In federal courts, in contrast, most judges tend to have little tolerance for illegal gun possession. In 2012, the median sentence for federal defendants convicted of illegal possession of a firearm was nearly four years, notwithstanding the absence of any legislatively-mandated penalty.215 Federal judges could choose, if they wished, to sentence felons-in-possession of firearms to non-custodial sentences or even sentences far below the advisory guidelines range, yet they almost never do.216 A sentence of probation for illegal firearm possession is practically unheard of in federal court.217

I use the felon-in-possession statutes in these three states218 as an illustration of a more widespread phenomenon of disparate sentencing patterns in street crime cases between the federal and local forums. In 2004, for instance (the most recent year for which data are available), twenty-two percent of all defendants convicted of violent felony offenses in state courts received a non-custodial sentence.219 In federal court, violent felony offenses almost never result in non-custodial sentences.220 These sorts of disparities are, of course, the product of many forces, including case and defendant selection and, in some jurisdictions, differing applicable penalties (whether mandated or advisory).221 In this respect, though, the felon-in-possession example offered here is particularly useful: these crimes are, by and large, the same across forums—both with respect to the nature of the crime and the applicable penalties—and yet forum-based disparities persist.

This is norm diversity in action. Federal judges preside over a diverse class of criminal cases arising from a diverse community of litigants. In addition to street crimes, federal judges also see a fair share of white-collar crimes, immigration offenses, and interstate narcotics trafficking. They also see many more civil cases compared to their local urban counterparts (who, in many jurisdictions, preside over only criminal cases).222 And they see litigants from urban, suburban, and rural areas. This means that, when a defendant charged with possessing a gun appears before a federal judge for sentencing, he does not come on the heels of ten others like him. In all likelihood, he also does not come on the heels of ten others who have committed more serious violent crimes, such as murder, armed robbery, or rape. To the federal judge, the defendant’s gun possession appears very much aberrant, outside the norms of the large community over which the judge exercises her authority.223

In local courthouses, street norms often translate into courthouse norms: when certain criminal conduct is endemic, prosecutions of those crimes become routine. Routine cases tend to garner less outrage; the result is a courtroom culture of acceptance, in which street norms tend to dictate the “going rate” of punishment for a crime.224 Federal prosecution of street crime is largely an effort to shift that culture by leveraging the federal law’s power over entrenched local courthouse norms and the street norms that generate them.225 The norm-shifting effort is indeed a key component in many federal gun prosecution campaigns. Posters are displayed in local communities advertising the sentences given in federal gun cases, and defendants in local gun cases may be threatened with federal prosecution if they do not agree to a sentence closely approximating what they would likely receive in federal court.226

The success of the federal norm-shifting endeavor is certainly debatable. Even assuming punishment’s ability to alter norms of conduct—a highly disputed contention227—the vastly fewer numbers of federal compared to state cases present a practical challenge, to say the least, to any real norm-shifting exercise. But at least in the federal courthouse, for those few cases prosecuted, and perhaps in local forums where federal prosecution is threatened, the criminal law acquires more power than it otherwise might.

C. Moral Credibility: Penalties, Perceptions, and Communities

When it comes to street crime, moral credibility primarily revolves around penalties. There is no disagreement as to whether armed robbery should be criminalized; the debate is simply over what the penalty for such conduct should be. While some might posit that the same cannot be said for drug crimes, in fact the public overwhelmingly supports criminalization of “hard” drugs such as powder cocaine, crack, and heroin.228 The severity of the penalties associated with drug crimes and the concern that many drug offenders should receive treatment rather than imprisonment animate much of the debate over the “war on drugs.”

Moral credibility in street crime, then, turns principally on punishment, and the issue of punishment is morally fraught—all the more so for street crime because it primarily impacts the poor and minorities. What is the “right” sentence for a particular crime or a particular defendant? Is this question even answerable? And if it isn’t, then what empowers a legislature or a sentencing commission to decree punishment? In particular, what inspires fidelity to the penalty-drafters’ wishes? These questions underlie the remainder of this Part, which considers the extent to which federal penalties pertaining to street crime align with public and judicial views, the ramifications of moral credibility gaps within the federal system, and why moral credibility with respect to street crime might differ across federal and local forums.

1. Moral Credibility in the Federal System

Empirical data on public and judicial views of federal penalties reveal three salient facts. First, there is no “uniform” view as to the appropriate sentence for a crime; individual views differ, sometimes markedly. Second, despite the absence of uniformity on an individual level, with few exceptions federal penalties accord with median public (and judicial) views. And third, in the few instances where there is a significant gap between federal penalties and public and judicial views—what I refer to as a “moral credibility gap”—there are ramifications for federal enforcement power, albeit not at the extremes we see in some local county courts. This Subsection sets out data supporting the above points, and the next Subsection considers moral credibility’s potential role in forum-based outcome disparities in prosecutions of street crime.

Two important studies reveal an absence of uniform public opinion as to the appropriate sentences for federal crimes. In 1994, Peter Rossi and Richard Berk were tasked by the U.S. Sentencing Commission with undertaking the first (and, as yet, only) systematic study of public opinions of federal sentences under the Federal Sentencing Guidelines.229 The Rossi and Berk study compared the Guidelines rules for determining sentences, and the resulting sentences, with the rules and sentences preferred by a representative sample of the American public.230 It drew two principal conclusions. First, there was substantial variation among respondents with respect to appropriate penalties.231 As the authors succinctly put it: “One person’s two-year sentence may be the equivalent of another’s four-year sentence.”232 Second, despite the lack of uniformity among respondents, the Guidelines penalties tended, in most cases, to approximate the respondent median or mean.233 There was one glaring exception to this general trend: crack cocaine penalties. Median respondent views as to the appropriate penalties for crack offenses were approximately twelve years lower than existing federal penalties under statutes and the sentencing guidelines.234

In 2010, the U.S. Sentencing Commission amassed federal judges’ opinions of the Guidelines, along with statutorily-mandated penalties, through a detailed survey.235 The Sentencing Commission’s survey, while using a different methodology than the Rossi and Berk study,236 likewise revealed a wide range of opinions on the appropriate sentences for given crime categories. For any given crime category, judges expressed differing views on the appropriateness of applicable penalties. But on the whole, a slight but consistent majority of judges believed both the Guidelines and statutorily-mandated penalties were appropriate—with the exception of crack cocaine and certain child pornography penalties.237 Nearly three-quarters of surveyed judges believed that both Guidelines and mandatory minimum penalties for crack cocaine possession were too high.238 The only other crime category for which judges expressed a similar view was possession and receipt of child pornography (an offense category on which Rossi and Berk did not canvass public views and for which, in any event, Guidelines penalties were then far lower).239

What are we to draw from this data? On one level, it empirically confirms what criminal practitioners know to be true: when it comes to sentencing, moral beliefs can be wildly divergent. Indeed, the Guidelines were borne of the desire to mitigate this inescapable reality. On another level, though, the data reveals a less obvious trend: there is, to a large degree, moral alignment between federal penalties and average public and judicial views. How much might this alignment explain the relatively high degree of adherence to prescribed penalties within the federal system, at least with respect to street crimes?240 Let us consider the two documented instances of moral credibility gaps in federal penalties and their ramifications for enforcement, sentencing, and penalty reform.

Start with crack, a story that is, by now, well known. Congress’ decision to impose significantly enhanced penalties for trafficking in the base form of cocaine (crack) as opposed to the powder form (cocaine) began, after a number of years, to meet with a rising tide of disapproval from nearly all sides. Legal scholars, federal judges, the Sentencing Commission, and even law enforcement—including then-Attorney General Janet Reno—publicly criticized the crack/powder penalty disparity, both for its lack of grounding in scientific and criminological data and for the racially disparate sentences it generated.241 Anecdotal reports surfaced of jury nullification and judicial resignations.242 Though it took decades, Congress ultimately heeded the calls and amended the laws in 2010 to substantially lessen, but not eliminate, the crack-powder disparity.243

Inside the courtroom, however, the crack story weaves a considerably more complicated narrative. In some respects, the disaffection did not generate palpable results. Defendants in crack cases cooperated at least as often as defendants in other narcotics cases.244 Pre-Blakely/Booker, judges sentenced crack defendants to within-Guidelines sentences nearly as often as they did powder cocaine offenders.245 And prosecutors continued to bring large numbers of crack cases, disposing of them without significant discounts.246 The disaffection did manifest in sentencing post-Booker, when non-government-sponsored, below-Guidelines sentences for crack offenses began to increase steadily, at a rate and to a degree higher than for powder offenses.247 Moreover, this trend immediately reversed following the enactment of the FSA: in just the quarter following its enactment, the amount of non-government-sponsored, below-Guidelines sentences decreased by about one third.248

Now consider the story of child pornography penalties.249 While the crack story might be characterized as one of public views diverging from static laws, the child pornography story is one of laws steadily and quickly moving off-track from public (or at least, judicial) views.250 For more than three decades, Congress has steadily increased maximum penalties, added mandatory minimum penalties, and repeatedly directed the Sentencing Commission to impose harsher Guidelines penalties for child pornography offenses.251 As a result, penalties for child pornography are now higher than for actual child abuse, a situation that strikes many judges as unjust.252 The political story, too, is quite different. Apart from the vocal dissent of some federal judges,253 there has been far less public outcry to Congress, at least as compared to what occurred in the case of crack penalties.254 This may be in part because the divergence between penalties and views has become pronounced only fairly recently, with the enactment of the 2003 PROTECT Act.255 Or it may be in part because the disparity between child pornography and child abuse penalties doesn’t raise the disturbing specter of racially disparate outcomes, as did the crack/powder disparity. Or it may be because those who view and traffic in child pornography are simply less sympathetic as a political matter. Whatever the reason, there has been little in the way of genuine movement for penalty reform inside Congress, at least at present.

Inside the courtroom, though, the divide between applicable penalties and judicial views has real effects. Non-government-sponsored, below-Guidelines sentences for possession, receipt, or distribution of child pornography, at nearly forty percent, far eclipse any other offense category.256 And the degree of departure is substantial: sentences for such offenses are discounted by thirty-eight percent, which translates into a median discount of forty months—again, far above almost any other offense category.257 Prosecutorial practices have changed, too: as penalties gradually increased, prosecutors offered more frequent and more substantial sentencing discounts to defendants.258 And unlike in crack cases, at least some federal judges in child pornography cases have been reported to blatantly flout mandatory minimum penalties by throwing out jury verdicts and ordering retrials or threatening to advise juries of the mandatory penalties in direct disregard of permissible federal procedure.259

How, then, to explain the different trajectory of these two moral credibility gaps? There is, of course, no single answer to this question. One reason is the differences between defendants. Crack defendants generally have substantial criminal histories, while child pornography defendants typically have no criminal history, a circumstance that might make within-Guidelines sentences in many crack cases more palatable for judges.260 Child pornography defendants are also overwhelmingly white whereas crack defendants are overwhelmingly black, a difference that might lead some to conclude that the contrasts between the two sentencing patterns reflect some level of racism within the system, however unconscious.261 Still another explanation is the relationship of the Guidelines to mandatory minimum terms. In the case of crack cocaine, the Sentencing Commission gradually sought to reduce Guidelines penalties in an effort to bring the Guidelines more in line with public and judicial views, which in turn led to less distance between mandatory terms and Guidelines ranges. In the case of child pornography, Congress usurped the Commission’s role, effectively legislating Guidelines ranges that grew far higher than mandatory minimum terms.262

But perhaps the most compelling explanation stems from the dynamic underlying the Guidelines/mandatory minimum relationship: the trajectory of the moral credibility gap. In the case of crack, public and judicial views gradually parted ways from a static law. In the case of child pornography, the law continues to part ways, quite drastically, from judicial views. And so in the former scenario, there was hope and promise that the penalty-drafters would ultimately narrow the credibility gap (which they did).263 In the latter, it is primarily the penalty-drafters who have created and enlarged the credibility gap. This bodes poorly for legislative reform. Trust in the legislative system is eroded, and courts feel it incumbent upon themselves to correct the imbalance.

Moral credibility, then, impacts federal criminal enforcement power. Most federal penalties (and particularly those for street crime) have substantial public and judicial support. Those that don’t have been less consistently applied. And the extent of the divergence between written and applied law depends on the trajectory of written law and public views. Where the system is at least receptive (if not immediately responsive) to public views, actors within that system are more willing to trust the system to make necessary corrections. Where the system is perceived as having deliberately run off-track from community conceptions, actors within the system are more likely to flout it.264 Indeed, the Justice Department’s own recognition of this dynamic lies at the heart of its recent efforts to align its enforcement practices with evolving community views on penalties—and even, in the case of marijuana, on decriminalization.265 The Department has launched these initiatives not merely for the criminal defendants who benefit, but primarily for law enforcement agents and prosecutors, whose efforts to fight crime are significantly strengthened by the trust and support of the communities they serve.266

2. Moral Credibility and Forum Disparities

The examples of crack cocaine and child pornography teach us that moral credibility affects federal criminal justice. We might, then, conclude that federal conviction rates and sentences in street crime cases may owe, at least partially, to public and judicial agreement (on average) with federal penalties for such crimes. Yet if so, why might we see such different outcomes in local urban justice systems? How, in other words, might moral credibility, or its absence, partially explain forum disparities in street crime cases?

To the extent moral credibility bears on forum disparities, surely it is not a result of differences between applicable federal and state penalties. As noted, even when penalty differences are miniscule or non-existent, forum-based outcome disparities persist.267 Moreover, as others have observed, applicable federal penalties are, in most instances, harsher than their state counterparts.268 In such cases, one would hardly expect public or judicial agreement with federal penalties to result in greater leniency at the state level (whether through plea-bargaining, jury verdicts, or judicial sentencing discretion).

Greater moral credibility gaps in local forums, then, must derive, if at all, from differences in public perceptions. What factors influence public perceptions of punishment, and how might the different “publics” in the two systems exacerbate (or mitigate) those influences? A careful examination of the Rossi and Berk study reveals both why a representative sample of the American public agrees, at the median, with most federal street crime penalties, and why we might see less robust agreement in local systems.

Rossi and Berk sought to ascertain which demographic, experiential, and attitudinal factors most closely associated with individuals’ views on sentencing.269 Utilizing regression analyses, the researchers found a number of statistically significant factors that related to views on punishment.270 Specific to street crimes (what the study authors label as “drug trafficking,” “drug possession,” and “street crimes”),271 statistically significant factors272 were:

1.    Region of residence;273

2.    Age;274

3.    Views on the investigative freedom of police;275

4.    Views on the rights accorded an accused;276

5.    Views on civil rights accorded minorities;277

6.   The frequency of contacts with the justice system;278

7.    Social and political views;279

8.   Years of education.280

Rossi and Berk’s study did not test the extent of sentencing variance by community type (i.e., urban, suburban, or rural). Instead, they measured only the effects of community size, which, by virtue of the methodology they employed, lumped these differing residential milieus together.281 Further research, then, is needed to ascertain empirically the correlation between community type and sentencing views. But a reasonable hypothesis is that, with the potential exceptions of region (which will manifest in both the federal and local forums in a given region of the country)282 and education, the influence of the remaining factors will tend to be more pronounced in local urban courts. The young are more concentrated in cities,283 as are those who identify as having “liberal” views.284 And it is city residents—and particularly residents of the inner-city, where most street crime occurs—who are most disaffected by local policing and by the treatment accorded suspects, defendants and minorities in particular.285

On this last point, Rossi and Berk’s findings impart a particularly important message. Perceptions of procedural justice—that is, perceptions of the fairness with which law enforcement treats citizens—correlate with perceptions of the fairness of laws. And procedural justice perceptions are a particularly significant in the street crime arena, where police-citizen encounters are endemic.

The effects of such credibility gaps may also be particularly destabilizing in local courts, which deal with a narrower range of crime categories.286 If a significant number of street crime penalties and enforcement practices are perceived as unjust in a system for which such crimes are the mainstay, the effects can be profound.287 Conversely, when credibility gaps are perceived as isolated and limited, and the enforcing authority is otherwise perceived as legitimate, the authority maintains legitimacy through its “reservoir” of support.288 This “reservoir” dynamic partly explains the federal government’s relative success in enforcing even unpopular laws. Indeed, even the crack and child pornography laws, for all their criticism, have generated a relatively low level of dissent in the courthouse—at least as compared to what occurs in some local courts when there is marked disagreement between penalties and public or judicial views.289

This federal “reservoir” dynamic might also explain, at least in part, why the nearly three decades-long crack/powder differential, and for that matter the nearly five decades-long “war on drugs,” has well outlasted the federal government’s thirteen-year assault on liquor. The federal government in 1920 held an entirely different place in criminal law enforcement, and American life for that matter, than it does today. The Volstead Act came at a time when the federal government enforced very few criminal laws. The feds in the 1920s were not prosecuting swindlers and fraudsters to any great degree; they were not policing stock trades; and they were not taking on organized crime or international terrorists.290 They were also not overseeing the distribution of the many benefits and services integral to modern American life, such as Social Security, food stamps, urban housing, Medicare and Medicaid, which have in turn given rise to new arenas of federal criminal enforcement.291 There are many reasons Prohibition ultimately failed.292 But probably among them—or at least, among the reasons it failed so swiftly—is the fact that the federal government at the time had a far smaller reservoir of support, at least with respect to criminal enforcement, than it does today.293

Of course, federal and local enforcement patterns, and the trust reservoirs those patterns create (or deplete) do not remain static. A second decade of transformative national security challenges brings with it a new landscape of both federal and local enforcement,294 and with that, a change in the relationship between citizens and the institutions of authority. The direction of that change, its permutations, and its particular impact on street crime enforcement—and perhaps, forum disparities—remain to be seen.295

IV. why legitimacy matters

A legitimacy-focused exploration of the federal criminal justice system enhances our understanding of federal criminal power. It also challenges us to reconsider the significance of forum disparities in criminal cases. These disparities are not merely the result of different sovereigns’ choices of legal rules, or of resource allocations between sovereigns. They are also the result of less intentioned forces.

This observation yields a number of insights. First, it explains why states that have sought to emulate federal prosecutorial outcomes merely by emulating federal penalties have largely come up short. Virginia Exile is just one example of a state borrowing from federal penal laws;296 there are others.297 Appreciating legitimacy’s role in federal criminal justice illuminates the shortcomings of this approach. If legitimacy at least in part explains federal prosecutorial success, legislators’ attempts at emulation have been misdirected—or, at least, they have been directed at a piece of federal criminal enforcement that has questionable value in isolation. Worse, this single-minded approach can affirmatively undermine successful law enforcement if the laws lack the support of the communities they most affect.

Second, attention to legitimacy reveals a deeper and more complicated federalism. While much of the criticism of federal intervention in street crime has focused on the circumvention of state laws and procedures,298 circumvention also occurs on a more fundamental level. Federal intervention bypasses the essence of localized criminal justice: beat policing, community-delineated jury pools, local courthouse norms. Is this sort of evasion of the local desirable?299 To be sure, there are good arguments on both sides. But given the structural and political dynamics of our system, these arguments are largely futile. So long as urban crime persists and localities feel unequipped to handle it, Congress and the Justice Department will respond. And so long as there are limited federal budgets and a lack of political appetite to expand them, the federal role in street crime enforcement will remain a supporting one.

Focusing on legitimacy thus presents a different, and ultimately more answerable, question: Is there a way to achieve legitimacy in street crime enforcement without circumventing localism? This Part argues that the answer to this question is yes and that, counterintuitively, the lesson we should draw from the federal system is the need for more localism in criminal justice, not less. Because if we distill federal legitimacy to its elements—enhanced citizen trust, juror cohesion, greater norm diversity, and public agreement with written law—we can see that the best way to translate these features into local systems is to embrace localism: by enhancing community voice and participation, and local law enforcement credibility. Not only will this approach stem the impetus for federal street-crime prosecutions, it will also result in a more substantively just system overall—a system that embodies the desires and beliefs of the communities most affected by street crimes and their enforcement.

A. A Legitimacy-Based Reform Agenda

By understanding the sources of legitimacy in federal street crime enforcement, we can theorize how to incorporate these features into local systems. In the federal system, legitimacy is derived in four principal ways: (i) enhanced citizen trust, arising from federal oversight of local policing, witness protection, and an absence of citizen policing; (ii) juror cohesion, arising from the relationship between the jury and the prosecuting authority, as well as the jury and the larger electorate; (iii) an empowerment of law over norms, arising from the breadth and diversity of participants in the justice system; and (iv) moral credibility, arising from an alignment between the laws and the adjudicative audience for those laws (judges and juries) and trust in the relevant penalty-drafting institutions.

Not all of these sources can be emulated in local systems. As long as there are police-citizen encounters, there will always be some level of citizen discontent, no matter how fair the encounter’s process or the officer’s motives. And self-identification with one’s country, for example, can’t necessarily be replicated on the local level. Nevertheless, some sources of federal legitimacy can be translated to local systems. I suggest three principal areas of focus. First, improving local policing and redirecting federal intervention towards investigation and witness protection in local cases can enhance citizen trust. Second, increasing local community voice in the formulation of sentencing policy and criminal laws can contribute to both juror cohesion with the lawmaking electorate and moral credibility. Third, increasing local voice at sentencing and resorting to more, but substantially less severe, mandatory penalties can help empower law over norms.

1. Enhancing Citizen Trust

Building citizen trust in law enforcement should be done at two levels: first, by strengthening and reforming local policing and, second, by reinventing the federal role in street crime investigations.

The need for policing reform is hardly a novel cry,300 but insufficient attention has been given to the attendant law enforcement benefits. Explaining away federal prosecutorial success as a product of laws, procedures, or resources diminishes the importance of the law’s enforcers in the successful enforcement of the law. Rooting out police misconduct—in all its forms, from misuse of force to discourtesy—should not be the agenda only of inner-city communities, activists, and civil rights prosecutors. It should be a principal priority of law enforcement and, critically, of local police forces themselves. Police department focus on these issues will not only improve police-citizen relations on an individual level, but will also enhance credibility more systemically. Visible, department-wide focus on police-citizen relations will engender citizen trust, just as the federal government engenders trust by virtue of its role in local policing oversight. Some police departments are beginning to make active efforts in this regard, but far more must be done.

At the same time, the federal government should reimagine its role in street crime enforcement as one of support for local prosecutions, rather than the inverse. Over the last several decades, “joint task forces” (law enforcement teams comprised of federal, state, and local officers) have been created to investigate crimes ranging from drug trafficking to violent gangs to human trafficking to terrorism.301 These groups have been organized with an eye to making federal cases. But why operate only under that model? Joint task forces can be utilized to prosecute local cases too. Local police and prosecutors could then rely on these task forces to help them investigate and bring stronger cases. One example of this model is New York City’s Office of the Special Narcotics Prosecutor, which collaborates with and frequently brings cases investigated by federal agencies and joint task forces.302 This model could be expanded more broadly.

This new federal role would have two principal benefits. First, more cross-pollination between federal and local law enforcers might benefit local policing and thereby enhance citizen trust. Of course, there are risks to greater federal involvement in local investigations.303 As federal agents play greater community-facing roles, they risk losing some of the credibility that inures by virtue of their rarefied position. And a larger law enforcement role for any authority carries the potential for abuse.304 But currently, at least, federal agents are far from reaching that point in street crime investigations; they could play a greater role while still retaining their reputational benefits. And a greater diversity of sovereign authorities in criminal investigations might serve to mitigate the potential for abuse or misconduct that sometimes accompanies consolidated sovereign power. In this sense, dual-sovereign criminal investigations would serve, on a very grassroots level, one of federalism’s fundamental aims.305 Given the absence of effective law enforcement oversight from coordinate government branches,306 more functional oversight through a “cooperative federalism” model might do a great deal of good.

Second, a larger federal investigative presence will significantly enhance both the protection of witnesses in local cases and those witnesses’ perceptions of security. As it now stands, federal agencies protect witnesses in federal cases; witnesses in local cases virtually never receive federal protection. In line with that model, the last Congress’s proposed fix for local witness intimidation was to make it a federal crime.307 The proposed bill was well-intentioned, but its approach one-dimensional. Why engage federal protection only through federal prosecution? And why engage it only after the fact, by prosecution of past threats or harms to witnesses? By then, the harm to the local case has already been done, and federal prosecution serves merely as a deterrent against future crimes. If local law enforcement were able to leverage federal witness protection at the outset, before threats are made—by having federal agents take the lead on witness security, interface with and provide a federal point-of-contact for witnesses on security matters—the effects of witness intimidation would decrease. Would-be intimidators would be deterred at the outset simply by knowing the feds are on the case; witnesses would both be and feel safer, engendering greater witness trust and cooperation. And unlike federal prosecution, which requires a federal jurisdictional hook,308 federal law enforcement assistance requires no jurisdictional nexus, thus permitting broader federal aid.

Encouraging more federal-local collaboration at the pre-charging stage would strengthen local cases, thus reducing the need for federal prosecutions. Fewer resources would be spent on federal prosecutions and federal prisons, and the resulting savings could be devoted instead to federal law enforcement agencies. Localities would retain jurisdiction and control over their criminal prosecutions, and cases would be adjudicated by residents of the cities where the crimes, and the law enforcement response to them, occur.

2. Legislative-Adjudicative Alignment and Moral Credibility

We should also strive to enhance both juror cohesion with the lawmaking electorate (what I’ve called legislative-adjudicative alignment) and the moral credibility of laws. Greater local voice in criminal lawmaking and penalty-drafting can bring us closer to both goals. And we can strengthen local voice at two levels: in the formulation of sentencing policy and in criminal lawmaking.

States that use sentencing commissions can make more concerted efforts to incorporate the views of urban localities when formulating penalties. This could involve deliberate outreach or, better yet, reconstituting sentencing commission membership to ensure participation by local inner-city community leaders personally familiar with the tolls of both crime and punishment. As it now stands, many states’ commissions call for membership by members of the judiciary, the prosecution and defense bars, and sentencing experts from academia and elsewhere.309 A few allow a seat for non-lawyer citizens, and some for crime victims. But none expressly call for membership from a broader group of citizens from the state’s urban centers. If sentencing commissions’ recommendations were the product of input from the localities most affected by those crimes, the recommendations would be more credible (and probably more substantively just, as well). Credible recommendations are more likely to be followed, further reinforcing their credibility. A ready illustration of the potential of local voice in sentencing policy is the District of Columbia. Its sentencing commission, comprised entirely of local community members, has made concerted efforts to develop sentencing policy in line with the views of local judges and residents.310 As the commission continues to develop and amend its policies in line with local views, judicial compliance with the sentencing guidelines has increased; in 2012, Superior Court judges complied with the advisory guidelines in 98% of felony cases.311

With respect to lawmaking, localities should consider more robust use of local laws. In the mid-1990s, cities inundated with gang violence passed public order ordinances designed to reclaim public spaces for communities.312 Many of these ordinances did not survive constitutional vagueness challenges, and the exercise was largely abandoned.313 But the idea of empowering localities to legislate (or lobby state legislatures) on issues of urban crime should not be cast aside. City ordinances, or state laws that apply uniquely to certain cities, permit local prosecutors to enforce laws with strong local support. Chicago, for instance, has passed a series of firearms ordinances; while some have been struck down on Second Amendment grounds,314 the city council has retooled these laws to survive Second Amendment challenges.315 Philadelphia was able to lobby the Pennsylvania legislature to mandate stricter requirements for gun possession in that city. Thus, under Pennsylvania law, it is illegal to carry a concealed firearm in the city of Philadelphia.316 Currently, local laws like these have gone largely unenforced because they do not entail prison time: prosecutors prefer to spend their resources on “serious” crimes, and the absence of a custodial penalty sends a strong message that the crime is not serious.317 Strengthening the penalties for these laws, even by small amounts, will result in more enforcement; enforcement, in turn, might be more successful when juries and judges feel more personal allegiance to the laws.318 Of course, locally applicable laws can only do so much to combat the sorts of crimes (such as gun and drug trafficking) that easily cross city lines. But the crimes that most immediately and directly impact communities—street-level drug sales, shootings, robberies, homicides, gun brandishing—tend to be geographically confined. For these types of crimes, local laws can be effective and can impart a message to the public, judges, and juries that the penal law has not been foisted upon them by a governing body with little understanding of the issues surrounding urban street crime and its enforcement.

Greater local voice in penalty-drafting will not only strengthen trust in the laws themselves but will also engender more confidence in the institutional frameworks of criminal lawmaking. As we have seen in the federal system with the examples of crack and child pornography, sentencing courts are more likely to adhere to law when there is trust that the system will work to rectify perceived injustices.319 Where, in contrast, trust in lawmaking institutions to remedy injustice is eroded, legitimacy suffers.320 Fair process in penalty-drafting thus not only generates fairer penalties, it also generates greater support for those penalties.

Enhancing localism at the penalty-drafting level is critical. While some scholars have articulated a vision of criminal justice localism based on less federal prosecution and more narrowly-drawn local districts,321 these proposals address only the adjudicatory piece. Yet the need for localism in criminal lawmaking and penalty drafting is equally important; without it, we see the sort of dysfunction that exists in some consolidated city-counties, where local voice is robust in the courthouses but diluted in statehouses and sentencing commissions.322 It is worth noting that in these cities, reforms, when they have finally come, have arisen not as a result of local adjudicative voice (in the form of low conviction rates and sentences that depart significantly from those legislated or advised), but rather from political forces outside the criminal justice system. Such was the case in the New York, where policing and drug law reforms came not from the message sent by distrustful juries in isolated local jurisdictions,323 but rather from the political attention spurred by a federal civil rights class action324 and budgetary pressures on the state level.325 In Philadelphia, it was likewise not low conviction rates, but rather a federal civil rights class action that finally yielded reforms to policing practices.326 In New Orleans, policing reforms were spurred not by messages from the city’s courthouse, but by the intervention of the Justice Department’s Civil Rights Division.327 In California, despite decades of trial-level dissent in San Francisco courthouses, reforms to the three strikes law came only after a U.S. Supreme Court order to relieve prison overcrowding coincided with a mounting state debt crisis and plunging crime rates.328 Indeed, the recent reforms to many states’ criminal laws are the product not of local adjudicative voice but rather changing political realities at the state level, spurred by fiscal constraints on state budgets, new evidence-based research on the effects of imprisonment, and decreasing levels of crime.329

Making criminal justice systems more accountable to the local citizenry is a laudable goal. To get there, we must focus not only on localizing adjudication, but on localizing lawmaking and penalty-drafting as well.

3. Empowering Laws over Norms

Finally, we should focus on ways to empower laws over norms. This is of a piece with the prior goal—aligning the legislative and adjudicative institutions and written law with moral beliefs—because community involvement in lawmaking will result in laws that more closely align with community norms. But we must also strive to empower adherence to those community-sanctioned laws in the face of conflicting street norms (among the much smaller, non-law abiding community), and the courthouse norms that those street norms engender.

One relatively recent project tilts in that direction. Project Safe Neighborhoods (PSN), a billion-dollar program operated by the Department of Justice, is a federal-state-local collaboration in major cities across the United States. The PSN approach varies by locality, but, in some cities, violent offenders and gang members are brought to “offender notification meetings” where a range of community members—religious leaders, parents, social workers, and reformed offenders—impress upon offenders that the norms of street life will no longer be tolerated and federal law enforcers advise that future criminal conduct will be prosecuted federally.330 This particular PSN approach, which has been used in some parts of Chicago, Indianapolis, and cities in North Carolina and, most recently, Connecticut, has been lauded as highly effective.331 The model relies on the threat of federal prosecution, but its fundamental approach to norm-shifting—by giving voice to the norms of the broader community—can be translated to local forums too.332

One way to do that is to give community members and leaders a role in criminal sentencing hearings. A program called “Ceasefire PA” is an effort along those lines in local Philadelphia gun cases.333 The group, formed in response to gun violence in Pennsylvania, has initiated a “courtwatch” program, in which local community members and leaders attend and testify on behalf of the local community at sentencing hearings.334 Early anecdotal evidence indicates that the program has been successful at impressing upon judges the gravity with which the community views gun-related crimes.335 This sort of broader community involvement might serve to shift courthouse norms in courts where violent crimes and gun-related cases have become routine.

Another method of empowering law over norms is through a more effective use of mandatory minimum penalties. Legislatures have largely operated at the extremes, with most criminal penalties carrying either no mandatory term or a mandatory term of at least five or ten years. But there is a middle ground, and much to be said for a more graduated approach to shifting sentencing norms.336 Local judges used to giving probationary terms for gun crimes may balk when told the legislature has suddenly mandated a sentence of five years. But a one- or two-year mandatory penalty is a less drastic change. Washington, D.C. undertook precisely this tactic in gun cases. In 2007, the city council passed a law mandating a one-year sentence of imprisonment for any defendant in possession of a firearm who has a prior felony conviction, and a three-year sentence of imprisonment for any such defendant with a prior violent felony conviction.337 In the years following, federal prosecutions of firearms offenses in the city decreased by eighty-five percent.338 This legislative tactic could be applied to any number of street crimes for which sentencing data indicates a significant difference between the sentence suggested by statute or guidelines and the sentences most often imposed in actual cases.

Of course, this approach must go hand-in-hand with localities’ involvement in criminal lawmaking; mandatory penalties should only be used as a method of implementing localities’ desires with respect to sentencing. If the communities most affected by street crime want to ensure that judges imprison criminals for certain crimes, tempered mandatory penalties can and should be used to achieve that objective. Washington, D.C., Philadelphia, Chicago, New York, and Baltimore are all examples of localities that are seeking (and in the case of Washington, D.C., have achieved) mandatory minimum terms for certain gun crimes. When the communities most affected by crime ask for specific laws, state legislatures should listen. Similarly, if the communities affected by crime ask to repeal mandatory sentences for certain crimes—and this will probably be the case for many non-violent drug offenses—legislatures should listen to that too.

To be sure, there are costs to this strategy. Mandatory minimum penalties achieve uniformity at the expense of individualized sentencing, a tradeoff that can result in unfair outcomes.339 The penalties do not eliminate discretion but merely transfer it, from judges to prosecutors and from sentencing decisions to charging decisions.340 And in our current state of mass incarceration, increased use of mandatory penalties in many ways seems directed at precisely the wrong goal. But by accompanying mandatory minimums with other reforms, we can reduce these costs. Lessening the severity of mandatory minimum penalties mitigates the effects of overbreadth and reduces prosecutors’ incentive to use penalties as bargaining chips. Increased local voice in criminal lawmaking and administration will ensure that the penalties, and prosecutors’ charging decisions, align with local desires. And over the long term, imprisonment rates might well decrease. More mandatory minimums, but with lesser terms, will reduce the distance between written and applied law. Laws applied as written engender trust in and compliance with the written law, resulting in fewer arrests, convictions, and punishments. Moreover, to the extent that federal sentences contribute to nationwide incarceration rates,341 more, yet tempered, mandatory minimum penalties at the state or local level will help lessen penalties overall. If local law enforcement could better trust in local systems to effectively deal with street crime, requests for federal intervention—and the hefty prison sentences that come with it—would recede.

The proposals set forth here entail high up-front outlays, in that significant resources must be spent on strengthening local justice systems before a diminution of federal prosecution could be politically acceptable. Yet continuing on our current trajectory is far more costly—not just in terms of resources over the long term but, more fundamentally, in the tax on our collective trust in the justice system. The key is to refocus our approach. Rather than debating which sovereign should enforce which criminal laws, we should instead turn our efforts to thinking about how to engender legitimacy in local criminal justice systems. And rather than relying on criminal adjudication as the primary vehicle for local voice and participation, we should strive for a system that incorporates local voice elsewhere too—a system that functions more collaboratively at the outset, through the trust and confidence of the governed.


Forum disparities in street crime cases have been conceptualized primarily as the product of tangible differences among sovereigns’ legal rules (penalties, procedures, evidentiary rules) and in resources and caseloads. This understanding is accurate, but it neglects to account for the intangible features that also fuel federal prosecutorial success: enhanced citizen trust, the jury’s identification with the prosecuting authority and relation to the larger electorate, greater diversity of social norms, and moral credibility. These features arise from the unique ways that the components of legitimacy—authority, community, and personal morality—interact in the federal criminal justice system.

This understanding of federal criminal power has a number of implications. First, it puts outcome disparities in a new light. They are not necessarily (or not only) the natural outgrowth of our federalist order, but also reflections of less intended dynamics of urban crime and policing. In this respect, criminal outcome disparities should be understood less as an inevitable feature of federalism, and more as a symptom of dysfunction.

Second, the federal system achieves its legitimacy-based features largely through circumventing the local in criminal justice systems: beat policing, community-delineated jury pools, and entrenched courthouse norms. But legitimacy need not only be had in these ways. Once we appreciate the legitimacy-based features of the federal system, we can theorize means of translating these features into local systems in ways that preserve, and even enhance, localism.

It is here that we should focus our intellectual efforts. Let us put aside for now the debates on which crimes should be prosecuted by which sovereign. Instead, let us envision how we might rebuild local criminal justice systems so as to enhance legitimacy. If we do this, forum disparities will naturally recede, making questions of forum and sovereign power allocation ultimately less important. This, in turn, will enhance the legitimacy of our larger, multi-jurisdictional system of criminal justice.