In 1985, New York’s top judge, Sol Wachtler, told the Daily News that prosecutors have so much influence over grand juries they could get them to “indict a ham sandwich.” Four decades later, the data still backs him up. In 2010, federal grand juries declined to indict in just 11 out of 162,000 cases. That is a 99.99% indictment rate.
But that number, dramatic as it is, actually obscures how the system works. Grand juries are not rubber stamps in the way most people assume. The real story is more interesting: prosecutors have built a machine so tightly controlled that it almost never produces an unwanted result. And when the machine does break down, the consequences reveal everything about who holds the power and who does not.
What a Grand Jury Actually Is
A grand jury is a panel of ordinary citizens who hear evidence from prosecutors and decide whether someone should be formally charged with a crime. It is not a trial. There is no verdict. The question is simple: is there enough evidence to justify putting someone through the trial process?
The concept is old. An ancestor of the modern grand jury appeared in the Magna Carta in 1215, and the institution traces its roots to 12th-century England, where it served as a shield against the Crown. The American founders enshrined it in the Fifth Amendment: “No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentmentA formal accusation initiated by a grand jury on its own motion, without a prosecutor's request; distinct from an indictment and rarely used in modern US practice. or indictment of a Grand Jury.”
The original idea was straightforward. Before the government could drag you into court to face serious criminal charges, a group of your fellow citizens had to agree there was reason to do so. It was supposed to be a check on state power.
Federal grand juries have 16 to 23 members, and at least 12 must agree to indict. They serve for 18 to 36 months, meeting a few times a month. They can question witnesses and issue subpoenas. The United States is the only country in the world that still uses them.
Why Prosecutors Almost Always Win
The reason prosecutors rarely lose before a grand jury is not mysterious. They control virtually every aspect of the process.
Grand jury proceedings are closed to the public and to the media, as well as to the targets of the investigation and their counsel. There is no judge in the room. The prosecutor presides, presents the evidence, calls the witnesses, and instructs the jury on the law. Defense attorneys are not allowed inside.
In a typical case, the prosecutor calls only one or two witnesses, usually the arresting officer, and tries to limit testimony as much as possible. The legal bar they need to clear is low. The Supreme Court has stated that probable causeThe legal standard requiring police to have reasonable, factual grounds to believe a specific person committed a crime before making an arrest or obtaining a warrant. “requires only the kind of ‘fair probability’ on which reasonable and prudent people, not legal technicians, act.”
Federal prosecutors are not even required to show the grand jury evidence that might prove the accused is innocent. The result is a proceeding where one side presents its case, the other side is not represented, no one challenges the evidence, and the standard for approval is low. As former federal prosecutor Bruce Green put it: “There’s a pretty significant risk that, if the prosecutor gets it in their head that somebody’s guilty, they can achieve an indictment whether the person is guilty or not.”
The Numbers Behind the Machine
The Bureau of Justice Statistics has tracked federal grand jury outcomes for years. The results are consistent and stark:
- 2010: 11 no-billsA grand jury's formal refusal to indict; jurors found insufficient evidence to justify bringing criminal charges against the accused. out of 162,351 cases
- 2013: 5 no-bills out of 196,969 cases (0.003%)
- 2014: 14 no-bills out of 170,161 cases (0.008%)
- 2015: 19 no-bills out of 163,005 cases (0.012%)
- 2016: 6 no-bills out of 155,615 cases (0.004%)
These figures come from Bureau of Justice Statistics data compiled by Ballard Spahr. University of Illinois law professor Andrew Leipold summarized it bluntly: “If the prosecutor wants an indictment and doesn’t get one, something has gone horribly wrong.”
But there is an important caveat that most coverage misses. Those numbers reflect cases that prosecutors chose to present. In recent fiscal years, U.S. Attorneys’ Offices declined to prosecute roughly 22% to 26% of referrals before they ever reached a grand jury. The 99.99% indictment rate is, in large part, a product of survivor bias: prosecutors killed the weak cases internally.
Not All Grand Juries Are the Same
The Fifth Amendment requires grand jury indictment for federal felonies. But the Supreme Court ruled in Hurtado v. California (1884) that states are not bound by this requirement. Today, about half the states require a grand jury for felony prosecutions. The rest allow prosecutors to file charges directly through a document called an “information,” sometimes followed by a preliminary hearing before a judge.
Where states do use grand juries, the rules vary widely. Grand jury size ranges from 23 (federal) to 12 (Missouri), and they do not always need to be unanimous. In Missouri, for instance, 9 of 12 grand jurors can return an indictment.
The Police Exception
There is one glaring category where grand juries do not indict at nearly the same rate: police shootings. In Houston, Texas, local grand juries cleared police of shooting civilians 288 consecutive times.
The ABA Journal has reported three possible reasons for this pattern: grand jurors tend to trust police, prosecutors tend to favor officers in their case presentation, and prosecutors sometimes feel obligated to bring police shooting cases to a grand jury even when the evidence is weak, effectively using the grand jury as political cover for a decision not to prosecute.
The 2014 grand jury proceedings in the Michael Brown shooting in Ferguson, Missouri, illustrated the dynamic. About 60 witnesses were called during 75 hours of proceedings, producing almost 5,000 pages of transcript. That is wildly unusual. Most grand juries hear one witness per case. The prosecutors cross-examined their own potential prosecution witnesses, an assistant district attorney cited a Missouri statute that had been overturned by the Supreme Court in 1985, and the prosecutors did not recommend specific charges. The grand jury declined to indict.
Columbia Law professors Jeffrey Fagan and Bernard Harcourt, who analyzed the Ferguson proceedings, wrote that “everything” about the proceeding was unusual. “The proceedings resembled a trial rather than a grand jury proceeding.”
The Scholarly Critique
Academic criticism of the grand jury is not new. In 1972, Professor James Shannon wrote what remains one of the most cited descriptions of the problem:
A 2017 Harvard Law Review note titled “Restoring Legitimacy” argued that the grand jury’s decline as a meaningful check coincides with the rise of plea bargainingAn agreement between a prosecutor and defendant where the defendant pleads guilty—usually to a lesser charge or in exchange for a lighter sentence—rather than going to trial.. Up to 95% of defendants plead guilty and never see the procedural protections of a trial. That makes the grand jury the last checkpoint before the full weight of the state’s leverage comes down on a defendant. And it is a checkpoint controlled by the prosecution.
The National Association of Criminal Defense Lawyers (NACDL) has called for four key reforms: allowing defense counsel in the grand jury room, requiring prosecutors to produce witness transcripts, giving witnesses advance notice to appear, and mandating the presentation of exculpatory evidenceEvidence that tends to clear a defendant of guilt or reduce their culpability; federal prosecutors are not required to present it to grand juries..
2025-2026: The Grand Jury Pushes Back
Something unusual has been happening. After decades as a near-automatic step in the prosecution pipeline, federal grand juries have started refusing to indict in politically charged cases.
In December 2025, a federal grand jury in Norfolk, Virginia, refused to indict New York Attorney General Letitia James on mortgage fraud charges. The Department of Justice tried again with a second grand jury in Alexandria. It also refused. Prosecutors reportedly had evidence that undercut some of the allegations in their own earlier indictment.
Ballard Spahr documented a broader pattern: in Washington, D.C., three successive grand juries declined to indict a protester accused of brushing an FBI agent’s arm. In Los Angeles, multiple grand juries declined to indict protesters after line prosecutors themselves said the evidence fell short. In Virginia, a grand jury reviewing charges against former FBI Director James Comey declined to indict on one of the counts.
What changed? According to legal analysts at Ballard Spahr, the answer involves the collapse of internal DOJ gatekeeping. The Public Integrity Section, historically mandated to oversee investigations of elected officials, has reportedly been stripped of its oversight role. Cases that career prosecutors would have killed internally are now reaching grand juries unvetted. And some of those juries are doing the filtering that the DOJ no longer does.
As Ballard Spahr put it: “The high historical indictment rate relied on prosecutors killing weak cases internally. By removing the gatekeeper, the DOJ is presenting cases to jurors that likely would have been declined by career prosecutors, resulting in rare ‘no-bills.'”
What This Means
The grand jury was designed to be a shield between the citizen and the state. For most of modern history, it has functioned as something closer to a conveyor belt. Prosecutors present, juries approve, defendants face trial or accept a plea.
The recent wave of no-bills does not necessarily mean the institution is healing. What it may mean is that when the prosecutor’s internal screening fails, citizens on grand juries are sometimes willing to do the job themselves. Whether that proves to be a temporary blip driven by politically unusual cases, or the beginning of a structural shift, will depend on what happens next inside the Department of Justice.
For now, the ham sandwich maxim still holds for the overwhelming majority of cases. But the sandwich has started biting back.
In January 1985, New York Court of Appeals Chief Judge Sol Wachtler told the Daily News that district attorneys have so much influence on grand juries that “by and large” they could get them to “indict a ham sandwich.” He was proposing scrapping the system entirely. Nobody did. And for forty years, the data consistently validated his premise.
The Bureau of Justice Statistics figures tell a story of near-total prosecutorial dominance over the grand jury process. In fiscal year 2013, grand juries declined to indict in just 5 of 196,969 criminal matters (0.003%). In 2014: 14 of 170,161 (0.008%). In 2015: 19 of 163,005 (0.012%). In 2016: 6 of 155,615 (0.004%). The 2010 data, frequently cited in the wake of Ferguson, showed 11 no-billsA grand jury's formal refusal to indict; jurors found insufficient evidence to justify bringing criminal charges against the accused. in 162,000 federal cases.
But the 99.99% figure, while accurate, has always been misleading without context. It is the product of a two-stage filter. In recent fiscal years (2020-2023), U.S. Attorneys’ Offices declined to prosecute roughly 22% to 26% of referrals before presenting anything to a grand jury. As Ballard Spahr’s Henry Hockeimer and Brad Gershel put it: “High indictment rates reflected institutional selectivity, not juror passivity.”
Constitutional Framework and Structural Asymmetry
The Grand Jury Clause of the Fifth Amendment provides that “no person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentmentA formal accusation initiated by a grand jury on its own motion, without a prosecutor's request; distinct from an indictment and rarely used in modern US practice. or indictment of a Grand Jury.” The National Constitution Center notes that this is one of the few Bill of Rights provisions the Supreme Court has not incorporated against the states via the Fourteenth Amendment. In Hurtado v. California, 110 U.S. 516 (1884), the Court held that states may charge felonies by prosecutorial information rather than grand jury indictment, concluding that the grand jury is not a fundamental requirement of due process.
The result is a split system. The federal government and approximately half the states require grand jury indictment for felony charges. The remaining states use preliminary hearings before a judge, informations filed by prosecutors, or some combination. Grand jury size ranges from 23 (federal) to 12 (Missouri), and unanimity is not always required. Federal grand juries require a minimum of 12 jurors out of 23 to return an indictment.
The grand jury’s origins trace to 12th-century England, where it was conceived as a civilian barrier between the Crown’s power and individual liberty. The United States is now the only country in the world that still uses the institution.
The Mechanics of Prosecutorial Control
The structural asymmetry of grand jury proceedings is well-documented. Columbia Law professors Fagan and Harcourt detailed the dynamics: proceedings are closed to the public, to the media, and to the targets and their counsel. No judge sits in the grand jury room. The prosecutor presides over the process, controls the witness list, and instructs the jury on applicable law.
In typical cases, prosecutors call only one or two witnesses, usually the arresting officer, and deliberately limit testimony to avoid creating impeachment material. The Supreme Court has held in Kaley v. U.S., 134 S. Ct. 1090 (2014), that probable causeThe legal standard requiring police to have reasonable, factual grounds to believe a specific person committed a crime before making an arrest or obtaining a warrant. “requires only the kind of ‘fair probability’ on which reasonable and prudent [people,] not legal technicians, act.”
Critically, the Supreme Court has held that federal grand juries need not adhere to trial rules of evidence or be told of evidence exculpating the defendant. Some states require disclosure of exculpatory evidenceEvidence that tends to clear a defendant of guilt or reduce their culpability; federal prosecutors are not required to present it to grand juries.; the federal system does not. This produces what the Harvard Law Review described as a situation where “it can fairly be said that the prosecutor holds all the cards before the grand jury.”
The structural dependence runs deeper than evidentiary control. As the Harvard Law Review noted, the grand jury lacks a physical building or staff and is not a singular body but rather “a series of panels of citizens summoned for part-time service as grand jurors, who meet at the convenience of the prosecutor and the court.”
The Administrative Agency Critique
The academic literature has been scathing for decades. In 1972, Professor James Shannon described the grand jury as having become:
Justice William O. Douglas offered his own assessment: “It is, indeed, common knowledge that the grand jury, having been conceived as a bulwark between the citizen and the Government, is now a tool of the Executive.”
The Harvard Law Review’s 2017 analysis argued that this dynamic compounds the plea-bargaining crisis. Up to 95% of defendants plead guilty without ever going to trial. The grand jury is therefore the last point of civilian oversight before the full weight of prosecutorial leverage bears down. And it is a checkpoint where the prosecution controls both the evidence and the rules.
Fordham University law professor Bruce Green, a former federal prosecutor, summarized the practical consequence: “If the original idea of the Founding Fathers was, as I believe it was, to be a restraint on government power… it’s probably not a very effective tool to protect people from prosecution overreaching.”
The Police Shooting Anomaly
If grand juries almost never refuse to indict ordinary defendants, they display a conspicuously different pattern in police use-of-force cases. In Houston, local grand juries cleared police of shooting civilians 288 consecutive times. In St. Louis County, from 1991 to 2014, four cases of police killings of citizens were presented to grand juries, yielding zero indictments.
The Ferguson grand jury proceedings in the Michael Brown shooting became a case study in how the same institution can be used in radically different ways depending on prosecutorial intent. About 60 witnesses were called over 75 hours, producing nearly 5,000 pages of transcript. Most grand juries hear one witness per case. The prosecutors cross-examined their own potential witnesses. Officer Darren Wilson testified for four hours without rigorous cross-examination. An assistant DA cited a Missouri statute on deadly force that had been overturned by the Supreme Court in 1985, correcting the record only weeks later. The prosecutors did not suggest which charges the grand jury should consider.
As Fagan and Harcourt concluded: “Everything” about the proceeding was unusual. “The proceedings resembled a trial rather than a grand jury proceeding.”
The 2025 Inflection Point
Beginning in late 2025, an unprecedented cluster of federal grand jury no-bills disrupted the institutional pattern. Ballard Spahr identified the trend: no-bills in Chicago (ICE protest cases), Washington, D.C. (three successive grand juries refused to indict a single protester), Los Angeles (multiple protest-related declinations), and Virginia (partial no-bill in the Comey matter).
The James case was the most dramatic. In December 2025, a federal grand jury in Norfolk, Virginia, refused to indict New York Attorney General Letitia James on mortgage fraud charges. The DOJ presented the case to a second grand jury in Alexandria, this time adding a third felony count. That grand jury also refused to indict. Prosecutors’ own investigators had found evidence that undercut the allegations in the original indictment.
The February 2026 Ballard Spahr analysis identified the structural cause: the traditional internal gatekeeping at DOJ had broken down. The Public Integrity Section, historically responsible for vetting investigations of elected officials, had reportedly been stripped of its oversight role. Politically appointed prosecutors were bringing cases that career prosecutors would have declined. The result was predictable: legally weak theories that had bypassed internal review were now reaching grand juries, and the juries were performing the screening function that DOJ no longer did.
“The grand jury was designed as a buffer between the state and the citizen,” Ballard Spahr concluded. “For much of modern federal practice, that buffer was largely theoretical — prosecutorial self-restraint performed the same function. With the internal guardrails gone, the constitutional structure is reactivating.”
Reform Proposals and Institutional Inertia
The National Association of Criminal Defense Lawyers has advocated four structural reforms based on research into grand jury practices in New York and Colorado: (1) allowing defense counsel in the grand jury room, (2) requiring production of witness transcripts for the defense, (3) giving witnesses advance notice of their appearance, and (4) mandating the presentation of exculpatory evidence.
None of these proposals have gained traction at the federal level. The institution is, as the Constitution Center noted, “written into the Fifth Amendment too clearly to be ‘interpreted’ away.” And neither side of the political spectrum has strong institutional incentives to push for reform: prosecutors benefit from the status quo, and defense advocates occasionally benefit from the rare no-bill.
The practical implication for defense practitioners, however, has shifted. Ballard Spahr now advises that “the grand jury stage is no longer merely procedural. It is substantive, contestable, and — under current conditions — potentially decisive.” For the first time in modern federal practice, pre-indictment advocacy may matter.
The Structural Tension
The grand jury sits at the intersection of two competing principles. It is, constitutionally, a citizen check on prosecutorial power. It is, operationally, a prosecutorial tool. For decades, the tension was invisible because the outcomes aligned: prosecutors screened their own cases rigorously, presented strong evidence, and grand juries approved. The 99.99% rate was not a sign of dysfunction; it was a sign of a system where the real decision happened upstream.
What the 2025-2026 no-bills have exposed is what happens when the upstream filter fails. Grand juries are not inherently passive. They were passive because they were not being tested. Present them with cases that career prosecutors would have rejected, and the constitutional safeguard reactivates.
Whether this represents a durable institutional correction or a temporary anomaly driven by extraordinary political circumstances remains to be seen. The Justice Manual permits re-presentation of declined cases to new grand juries (Section 9-11.120), and the DOJ has already used that option. If re-presentation becomes routine, it will raise serious questions about whether the grand jury’s independence is being honored or undermined.
The ham sandwich may have started biting back. How the system responds will determine whether that matters.
