Safety and Justice: what the public prioritizes and why it is hard to measure

Safety and Justice, one of the ten provisional subjects in My Vote Forward, covers public safety, emergency services, courts and corrections; in practice most public debate about it concerns crime, policing, punishment and guns. American concern about crime swings widely depending on how the question is asked: in September 2026 about 1% of adults named crime or violence as the country’s most important problem in an open-ended Gallup question, while in April 2026, 47% told Pew that violent crime was a “very big problem” when shown a list.⁠^1,2 A high priority for safety and justice tells you people want attention paid to the subject; it does not tell you whether they want more police or fewer, longer sentences or shorter ones, or stricter gun laws or looser ones.

In plain words

Safety and justice is about two things most people want at once. Everyone wants to be safe at home, on the street and at work. Most people also want the justice system to be fair: to catch the right person, to punish in proportion and to treat people the same no matter who they are.

Those two goals usually point the same way. Sometimes they pull apart. More police patrols may stop some crimes but also mean more stops of people who did nothing wrong. Keeping someone in jail before trial may protect a neighborhood but also costs an innocent person a job.

People also disagree about the facts. Many believe crime is rising when official numbers say it is falling, and the official numbers themselves miss crimes that are never reported.

This page explains what people say they want and what choices sit underneath. Two companion pages explain who decides and what the evidence shows. All ten subjects are listed on the Subjects hub.

Defined terms

Safety and justice debates use several terms with specific meanings.

Safety and Justice (as a My Vote Forward subject). Public safety, emergency services, courts and corrections, as scoped in the demo’s provisional subject list. See Subjects.

Violent crime. In FBI estimates, murder and nonnegligent manslaughter, rape, robbery and aggravated assault.⁠⁸ The federal victimization survey uses a broader list that adds simple assault and leaves out homicide.⁠⁹ The two definitions are explained on the evidence page.

Deterrence. Preventing crime by making people expect punishment. Researchers distinguish the certainty of being caught from the severity of the punishment.

Incapacitation. Preventing crime by physically removing a person from the community, usually through jail or prison.

Pretrial detention. Holding a person who has been charged but not convicted. Jails hold most of these people.

Reentry. The return of a person from prison or jail to the community, including supervision, housing, work and the legal limits that follow a conviction.

Priority versus proposal. A priority says the subject deserves attention. A proposal is a specific action with an authority, a cost and a funding source. See Proposals.

What the public prioritizes in safety and justice

Public concern about safety and justice in the United States is real but uneven, and it depends heavily on survey design. In Gallup’s open-ended “most important problem” question, fielded September 1 to 17, 2026 with about 1,000 adults, crime and violence drew 1% and “judicial system/courts/laws” drew 3%.⁠¹ When Pew handed 5,103 adults a list in April 2026, 47% called violent crime a very big problem for the country and 49% said the same of gun violence, with a partisan gap in opposite directions: 55% of Republicans and 38% of Democrats on violent crime, 27% of Republicans and 68% of Democrats on gun violence.⁠²

Perceptions of crime trends move with events and with politics. Gallup’s October 2025 crime poll found 49% saying there was more crime in the U.S. than a year earlier and 30% saying so about their own area, with Republicans’ belief that crime was rising down 36 points from 2024.⁠³ In October 2022, by contrast, a record 56% said local crime had increased, the highest in a series that began in 1972, and 78% said national crime had risen; the record national reading was 89% in 1992.⁠⁴ In both readings, fewer people saw rising crime in their own area than in the country as a whole, a pattern worth remembering whenever a national survey is read as a local mandate.

On methods, the same October 2025 Gallup poll found 67% preferring more money and effort for “addressing social and economic problems such as drug addiction, homelessness and mental health” and 29% preferring “strengthening law enforcement,” with 63% of Republicans and 8% of Democrats choosing law enforcement.⁠⁵ Support for the death penalty for murder stood at 52%, the lowest in Gallup’s trend since 1972; 41% said 14 to 17 year olds who commit violent crimes should be treated as adults; and 73% supported holding parents legally responsible when children commit crimes with the parents’ guns, with support between 71% and 77% across partisan groups.⁠⁵ In a separate release, 62% rated the prison system excellent or good at keeping prisoners behind bars, but only 16% rated it excellent or good at rehabilitation.⁠⁶ On guns, Gallup’s October 2024 reading was 56% for stricter laws on firearm sales, 33% for keeping laws as they are and 10% for less strict laws.⁠⁷

Why safety and justice is hard to measure as a priority

Safety and justice is hard to measure as a priority for four reasons.

Concern and crime move separately. In 2025 the FBI estimated a 9.3% drop in violent crime and a murder rate of 4.1 per 100,000 inhabitants, tied for the lowest since national estimates began in 1936,⁠⁸ while about half of adults still told Gallup crime was rising.⁠³ A priority ranking captures how people feel, which matters in itself, but it is not a crime statistic.

Much crime is never reported. In the 2024 federal victimization survey, about 48% of violent victimizations were reported to police.⁠⁹ A community where people have stopped calling the police can look safer on paper than it feels on the ground. See Measurement.

One word, opposite meanings. “Justice” can mean punishing offenders, protecting the wrongly accused, or ending unequal treatment. A participant who ranks this subject first may want more enforcement, less enforcement or different enforcement.

Prompting changes the answer. The gap between 1% in an open-ended question and 47% on a list.⁠^1,2 shows that the question format can matter more than the respondent. See Question design.

The real tradeoffs in safety and justice

Safety and justice policy in the United States involves tradeoffs that no ranking resolves. Each is stated here as a choice, not a conclusion.

Certainty versus severity. The National Institute of Justice summarizes the research this way: “The certainty of being caught is a vastly more powerful deterrent than the punishment,” and “Increasing the severity of punishment does little to deter crime”⁠¹¹ Certainty usually requires investigators and patrol; severity requires prison capacity. Money spent on one is not spent on the other.

More police versus fewer stops. A 2022 study estimated that each additional police officer “abates approximately 0.1 homicides,” with per capita effects twice as large for Black victims as for White victims, while larger forces also make more arrests for low-level “quality-of-life” offenses that fall disproportionately on Black Americans.⁠¹² Both findings come from the same paper, which is why the policing debate rarely ends.

Pretrial liberty versus pretrial risk. Seventy percent of the people in local jails at midyear 2023, about 467,600, were unconvicted.⁠¹⁹ A study using randomly assigned bail judges found that pretrial detention increases the chance of conviction, mainly through guilty pleas, has “no net effect on future crime,” and reduces later formal employment.⁠¹³ Others answer that a single serious crime by a released defendant is a cost that averages do not capture.

Incapacitation versus the cost of prison. A 2014 National Research Council report found that “Most studies conclude that rising incarceration rates reduced crime, but the evidence does not clearly show by how much,” and that the crime-reducing effect shrinks as incarceration grows.⁠¹⁰ State and local governments spent $87 billion on corrections in 2021, alongside $135 billion on police and $52 billion on courts.⁠¹⁷

Second chances versus repeat offending. Of people released from state prisons in 24 states in 2008, 82% were arrested within 10 years.⁠¹⁶ One side reads that as proof prison does not rehabilitate; the other as proof that release decisions carry real risk.

Gun rights versus gun restrictions. Gun policy pits a constitutional right and self-defense against firearm injury, and the research on many specific laws remains thin, as the evidence page explains.⁠¹⁸

The strongest arguments across perspectives

Safety and justice debates in the United States feature several serious positions. Each is presented here in its strongest form, attributed to those who hold it.

Public safety and victims first. Supporters of firm enforcement argue that the first duty of government is protecting people from violence, that the burden of violence falls unevenly, with a 2025 violent victimization rate of 769.7 per 100,000 Black residents compared with 209.9 per 100,000 White residents in police-reported data,⁠²⁰ and that the 2025 decline in murder followed years in which the murder rate reached 6.6 per 100,000 (2020 and 2022).⁠⁸ They point to research that police reduce homicide.⁠¹² and to the finding that incarceration does reduce some crime⁠¹⁰ In their view, under-enforcement is itself a form of injustice to victims.

Deterrence and incapacitation, properly targeted. A related position accepts that severity deters less than certainty,⁠¹¹ but argues that serious violence is concentrated among relatively few people and places, so that reliably catching and holding those people, while investing in clearance rates, protects the most potential victims. The FBI reported that police cleared 47.4% of reported violent crimes in 2025⁠⁸ which this view treats as the number most worth raising.

Reducing reliance on incarceration. Critics of mass incarceration cite the National Research Council’s finding of “no clear evidence that greater reliance on imprisonment achieved its intended goal of substantially reducing crime” and its documented harms to families and communities.⁠¹⁰ They note that the U.S. imprisonment rate, though down 25% from 2013, was still 360 sentenced prisoners per 100,000 residents at yearend 2023,⁠¹⁴ and that the World Prison Brief puts the overall U.S. rate at 542 per 100,000 for 2023, compared with 69 in Germany in January 2025.⁠¹⁵

Racial fairness. Reform advocates stress that the adult imprisonment rate for Black Americans, 1,218 per 100,000 in 2023, was about five times that of White Americans (231).⁠¹⁴ Others answer that disparities in imprisonment partly reflect disparities in victimization and offending in the neighborhoods most affected by violence, and that reducing enforcement there would harm the same communities. The National Research Council found both that Black Americans “had long been more likely than whites to be arrested for violence” and that drug enforcement and long sentences increased disparities.⁠¹⁰ The evidence page sets out the data each side relies on.

Pretrial and sentencing reform. Reformers argue that cash bail ties freedom to wealth and that detention itself produces convictions.⁠¹³ Defenders of detention argue that judges need the power to hold dangerous defendants, a power the Supreme Court upheld for federal courts in 1987 (see who decides).

Gun policy. Supporters of gun rights emphasize self-defense, a constitutional right and the interests of lawful owners; supporters of stricter laws emphasize that more than 48,000 people died of firearm injuries in 2022, more than half of them by suicide,⁠²¹ and point to policies with stronger evidence, such as background checks and child-access prevention laws.⁠¹⁸ Both sides can cite the RAND review, which finds supportive evidence that shall-issue concealed-carry and stand-your-ground laws increase homicides, moderate evidence that background checks decrease them, and inconclusive evidence for many other laws.⁠¹⁸

Evidence

The evidence base for safety and justice is large but uneven. Survey data on public concern come from Gallup and Pew.⁠^1,2,3 Crime counts come from two federal systems that measure different things: police reports compiled by the FBI.⁠⁸ and household interviews in the Bureau of Justice Statistics victimization survey⁠⁹ Correctional counts come from BJS.⁠^14,19 Causal questions, such as whether police or prison reduce crime, rely on quasi-experimental studies and expert reviews.⁠^10,12,13,18 The evidence page covers how these sources differ, where they conflict and what remains unknown.

Competing interpretations

A high priority for safety and justice can be read in at least three ways.

A fear reading. The priority reflects worry about victimization, which may track crime, news coverage or political cues. Gallup’s finding that perceptions shifted sharply by party after a change in administration⁠³ supports the view that some of this concern is partisan.

A fairness reading. The priority reflects concern about how the system treats people: wrongful convictions, unequal sentences, conditions in prisons. The low ratings Gallup found for rehabilitation and inmate safety⁠⁶ fit this reading.

An effectiveness reading. The priority reflects frustration that crimes go unsolved and offenders reoffend, a concern shared by people who want more enforcement and people who want less. Clearance and recidivism figures⁠^8,16 support this reading.

My Vote Forward does not choose among these readings. A subject ranking cannot distinguish them, which is why opinions on specific proposals are asked separately with exact wording. See Opinions.

Application to this framework

Safety and justice fits the My Vote Forward framework the same way every subject does, with one caution about privacy.

• Priority. A participant can rank Safety and Justice against the other nine subjects. The result reports attention, not direction. See Priorities.
• Opinion. Views on a specific proposal, such as a hypothetical county pretrial services program, are a separate, voluntary step with exact wording. See Opinions.
• Private voting. Views on crime and policing can carry social and workplace risk. The framework keeps any vote private and any opinion sharing optional. See Private voting.
• Authority. Most criminal justice decisions are made by states, counties and cities, not Washington, so the framework points each concern to the body that can act. See Who decides and Jurisdictions.
• Tradeoffs. Ranking safety above other subjects does not allocate money to police, courts or prevention. See Tradeoffs.

Hypothetical example: in a county where participants rank Safety and Justice first, the published result would state that ranking with its denominator and method. It would not say participants favor more patrols, bail reform or any other specific measure. A sheriff or county board could see the priority but would still need to consult on any specific proposal.

Unresolved questions

Safety and justice raises several questions this framework has not settled.

• Should emergency services, which are not mainly about crime, sit in the same subject as policing and corrections?
• Can a priority question separate concern about crime from concern about how the justice system treats people? See Question design.
• How should results be read when perceptions of crime diverge from official statistics?
• Should a local priority for safety be reported differently when the relevant authority is a county sheriff, a city council or a state legislature?
• How can the framework protect participants who hold unpopular views on policing in their own neighborhood? See Representation.