Is Crime Rising in America? The Data Says No — But the Answer Is More Complicated Than Either Side Admits

The United States is experiencing what may be the fastest decline in violent crime ever recorded. FBI data released in August 2025 showed violent crime fell 4.5% in 2024 compared to 2023, reaching its lowest rate — 359.1 per 100,000 residents — since 1969 [1]. Murder dropped 14.9% in a single year, the steepest annual decline on record [1]. Preliminary data from analyst Jeff Asher's Real-Time Crime Index suggests 2025 brought an additional 20% drop in murder, potentially the lowest homicide rate since 1900 [2].

And yet, as recently as 2024, roughly 77% of Americans told Gallup that crime was increasing nationally [3]. That number has since dropped to 49% — still nearly half the country believing crime is rising while official data shows it falling off a cliff [4].

Both sides of this debate — those who insist America is in a crime crisis and those who dismiss those concerns as misinformed — are selectively reading the evidence. The truth requires holding multiple uncomfortable facts simultaneously.

What the FBI Data Actually Shows

The top-line numbers are striking. Between 2023 and 2024, every major crime category declined nationally [1]:

• Murder and non-negligent manslaughter: down 14.9%
• Robbery: down 8.9%
• Aggravated assault: down 3.0%
• Revised rape category: down 5.2%
• Overall property crime: down 8.1%, the lowest rate since 1961

Source: FBI Uniform Crime Reporting Program

Data as of Aug 6, 2025CSV

The Council on Criminal Justice's year-end 2024 report, tracking 29 major cities, found 16% fewer homicides (631 fewer deaths), 15% fewer gun assaults, a 32% decline in carjackings, and a 24% drop in motor vehicle theft [5]. The 2025 update was even more dramatic: homicide rates across study cities fell an additional 21%, with cities like Denver, Washington D.C., and Omaha seeing 40% declines in a single year [6].

By 2025, 27 of the Council's 35 study cities had lower homicide rates than in 2019 — the last pre-pandemic year [6]. Baltimore's homicide rate dropped 60% from its 2019 level. The Gun Violence Archive reported shootings down 17% in 2025, with fatal shooting victims declining 13% [2].

Source: FBI UCR / Jeff Asher Real-Time Crime Index

Data as of Jan 15, 2026CSV

The Data Nobody Trusts: FBI Reporting Gaps and Revisions

These numbers come with serious caveats that crime-is-falling advocates too often gloss over.

In 2021, the FBI retired its nearly century-old Summary Reporting System and switched to the National Incident-Based Reporting System (NIBRS). The transition was a disaster. Only 63% of the nation's roughly 18,000 law enforcement agencies submitted 2021 data — the lowest participation rate since at least 1979 [7]. The New York City Police Department and Los Angeles Police Department, representing millions of residents, were among those missing [7]. About 7,000 agencies covering roughly 35% of the U.S. population reported no data at all [8].

The FBI backtracked in 2022, allowing non-NIBRS agencies to submit data again, and coverage recovered to about 94% [9]. But the 2021 gap created a statistical hole that makes year-over-year comparisons from that period unreliable.

There is also the revision problem. Jeff Asher has documented that FBI revisions have grown larger in recent years without clear explanation. The FBI initially reported 19,252 murders in 2023, then revised that figure upward to 19,902 when it released 2024 data — a difference of 650 homicides [10]. Asher estimates the 2024 murder figure is likely undercounted by roughly 700 murders [10]. These revisions do not change the direction of the trend, but they raise questions about precision.

When Crime Surveys Contradict Crime Reports

The National Crime Victimization Survey — conducted by the Bureau of Justice Statistics, surveying roughly 240,000 people annually about crimes they experienced whether or not they reported them to police — has at times told a story that directly contradicts the FBI.

The most dramatic example: the FBI reported violent crime fell 2% between 2021 and 2022. The NCVS reported violent victimization rose 75% over the same period [11]. That is not a rounding error. It is a fundamental disagreement about whether Americans were becoming safer or less safe.

The Council on Criminal Justice identified several reasons for the divergence [11]. The surveys cover different time periods — the FBI tracks incidents during calendar years while the NCVS asks about victimizations in the prior six months during interview years, creating a misalignment that becomes acute during periods of rapid change. The NCVS underwent a redesign in 2022 that halved its sample size, increasing statistical uncertainty. And the NCVS captures crimes never reported to police, while the FBI records only what reaches law enforcement.

By 2024, the two sources converged somewhat. The NCVS reported a violent victimization rate of 23.3 per 1,000 persons age 12 or older, with only 11.2 per 1,000 reported to police [12] — meaning roughly half of violent victimizations never appear in FBI statistics. For specific crimes, the gap is wider: only 57% of aggravated assaults and 41% of simple assaults reach police records [11].

Source: Bureau of Justice Statistics — NCVS 2024

Data as of Sep 1, 2025CSV

This "dark figure" of unreported crime is not new, but it complicates any confident claim about national trends. If reporting rates shift — because communities lose trust in police, because victims believe nothing will happen, or because certain offenses become normalized — then FBI statistics can show declines that reflect changes in reporting behavior rather than changes in actual crime.

The Outlier: Shoplifting

Not every crime category followed the downward trend. While overall property crime reached its lowest level since 1961, shoplifting was the sole major offense category that increased in the Council on Criminal Justice's 2024 tracking — up 14% [5]. FBI data showed reported shoplifting nationwide rose 8.9% in 2024 [1].

The National Retail Federation has reported 18% more shoplifting incidents per year in 2024 versus 2023, with violence during theft events up 17% [13]. Industry groups describe an "organized retail crime" crisis driven by coordinated criminal networks reselling stolen goods online.

Skeptics point out that "shrink" — the retail industry's term for inventory loss — encompasses employee theft, administrative errors, vendor fraud, and product damage alongside shoplifting [13]. Some analyses of police data do not show a uniform national surge; several major cities saw shoplifting return to pre-pandemic levels or decline [13]. By 2025, the Council on Criminal Justice found shoplifting had actually decreased 10% from 2024 levels [6].

The organized retail crime debate illustrates a broader pattern: a real problem exists, but the scale and nature of that problem look different depending on whether you consult industry lobbying data, police reports, or academic analysis.

The Perception Gap: Why Americans Feel Unsafe

For years, Gallup has documented a persistent gap between Americans' perception of crime and measured crime rates. In 2024, 77% of respondents said crime was increasing nationally. By 2025, that figure had dropped to 49% — still a significant number given the data [4].

The partisan breakdown is instructive. In 2024, 90% of Republicans said crime was rising nationally; by 2025, that dropped to roughly 54% — a 36-point swing coinciding with the transition to a Republican presidential administration [4]. Democrats moved in the opposite direction, becoming more likely to report rising crime after the same transition. The 61-point gap between Republican and Democratic perceptions of crime in 2025 was the largest Gallup had recorded since it began tracking the question in 1989 [4].

Americans consistently rate crime in their own area as less serious than crime nationally. Over 25 years, the gap has averaged 43 percentage points — people believe their neighborhood is safer than "America" in the abstract [4]. This local positivity bias suggests national crime perceptions are shaped more by media coverage, political messaging, and social media than by personal experience.

But dismissing the perception gap as pure ignorance is also wrong. Quality-of-life issues that erode the feeling of safety — open-air drug markets, visible homelessness, aggressive panhandling, car break-ins, fare evasion on public transit — often do not register in major crime indices. A city can have falling murder rates and rising disorder simultaneously. The person who steps over a fentanyl user on their morning commute is not irrational for feeling less safe, even if the statistical probability of being a violent crime victim has declined.

The Progressive Prosecutor Debate

Few crime policy questions generate more heat and less clarity than whether progressive prosecutors drive crime increases. The electoral record is mixed: San Francisco voters recalled Chesa Boudin in 2022; Los Angeles voters ousted George Gascón in 2024; Chicago's Kim Foxx declined to seek reelection [14]. But Philadelphia's Larry Krasner and Manhattan's Alvin Bragg both won reelection by comfortable margins in high-crime cities where voters had direct experience with both the crime and the prosecutor [14].

The Heritage Foundation has argued that progressive prosecutors "sabotage the rule of law" by refusing to prosecute entire crime categories, dropping felony cases at elevated rates, and enabling repeat offenders through no-cash-bail policies [15]. Their case rests partly on city-level statistics: under Krasner, Philadelphia homicides rose 49% and shootings climbed 59% from his 2018 inauguration through 2020; under Boudin, San Francisco burglaries rose 57.6% during the pandemic shutdown [15].

The strongest academic counterevidence comes from a quasi-experimental study by researchers at the University of Toronto, published in Criminology & Public Policy, which examined crime rates in the 100 largest U.S. counties from 2000 to 2020. The study found "no evidence that progressive prosecutors increased violent crime" but did find a modest association with higher property crime rates [16]. A Center for American Progress analysis reached similar conclusions, noting that crime rose and fell in similar patterns in cities with and without progressive prosecutors during the pandemic period [14].

Both positions contain legitimate observations. Heritage is correct that some progressive jurisdictions saw crime increases — but those increases occurred during a national surge driven by COVID disruption, and comparable increases hit cities with traditional prosecutors. The academic studies are correct that progressive prosecution does not appear to be a primary driver of violent crime trends — but the property crime finding is real, and the studies' time horizons may not capture longer-term effects.

Bail Reform: What the Evidence Shows

Bail reform has been a lightning rod, with critics blaming it for repeat offenders cycling through the system. The Brennan Center for Justice studied 33 cities — 22 with bail reforms and 11 without — from 2015 through 2021 and found "no evidence that bail reform affects crime rates" across any major crime category [17]. Research on New York's bail reform found that eliminating bail for most misdemeanor and nonviolent felony charges actually reduced recidivism [18].

The picture is not entirely clean. Some analyses found limited increases in pretrial rearrests among high-risk defendants with pending cases or recent violent felony arrests [18]. But the broad claim that bail reform unleashed a crime wave does not survive contact with the data from the jurisdictions that actually implemented it.

The COVID Effect, the Ferguson Effect, and Police Staffing

The most empirically supported explanation for the 2020-2021 crime surge — and the subsequent decline — involves the cascading social disruption of COVID-19. The pandemic simultaneously dismantled routine activities that suppress crime (commuting patterns, business hours, school attendance), strained social services, flooded communities with economic anxiety, and overwhelmed hospitals and courts [19].

The "Ferguson effect" — the theory that public scrutiny of police causes officers to withdraw from proactive policing, leading to more crime — has mixed evidence. After Michael Brown's death in Ferguson, Missouri, self-initiated arrests fell 62%, foot patrols dropped 82%, and pedestrian checks declined 76% [20]. Similar patterns appeared in other cities following high-profile police killings. But whether this withdrawal caused crime increases or merely coincided with them remains debated.

What is less debatable is the police staffing crisis. A Police Executive Research Forum survey found that 70% of agencies reported greater difficulty hiring compared to five years ago [20]. New Orleans and Minneapolis saw their forces shrink by 40% over a decade [20]. Agencies that experienced the George Floyd protests saw significant excess losses of sworn personnel — between 2.2% and 16% above normal attrition [20]. Some departments have responded by lowering hiring standards, a choice whose consequences may take years to materialize.

What Happens Next

Crime forecaster James Tuttle projects the FBI-reported murder rate will fall to between 3.6 and 3.99 per 100,000 in 2026 — a 3% to 12.5% decline from 2025's estimated 4.1 per 100,000, which was already a 65-year low [21]. Supporting factors include declining youth victimization as the fentanyl crisis shows signs of receding, and Black homicide rates approaching historical lows [21].

But several forces could slow or reverse the trend. The expiration of American Rescue Plan funding — which bankrolled community violence intervention programs, additional police hiring, and social services — means cities lose a significant policy lever by the end of 2026 [19]. The National Policing Institute warned that federal immigration enforcement operations risk eroding community trust in local police, making witnesses less likely to cooperate and victims less likely to call 911 [22]. Patrick Sharkey of Princeton has cautioned that American cities remain "vulnerable to surges of violence" because of persistent gun prevalence and concentrated disadvantage [19].

The Council on Criminal Justice's expert panel expects "greater variation in crime than in 2025" — meaning the national average may continue falling while individual cities diverge sharply based on local conditions [22].

Who Deserves Credit?

Politicians on both sides have claimed the crime decline. The Biden administration pointed to the Bipartisan Safer Communities Act and American Rescue Plan investments. The Trump administration has attributed the continuing decline to its immigration enforcement and tough-on-crime messaging.

The honest answer is that neither president deserves primary credit. Homicide declined across large cities, suburbs, and rural areas simultaneously — a pattern that suggests national forces (pandemic recovery, economic stabilization, the return of routine social activity) rather than any single policy intervention [19]. The cities that invested American Rescue Plan funds in violence prevention may see their gains erode as that funding expires, which would provide a natural experiment in whether the money mattered [19].

The Bottom Line

Crime in America is falling, probably faster than at any point in modern history. The murder rate may be approaching levels not seen since before World War I. This is a genuinely important development that affects millions of lives.

It is also true that the data measuring this decline has significant holes — a botched FBI transition, surveys that sometimes contradict each other, and roughly half of violent crimes never reaching police records. It is true that shoplifting and visible disorder create real quality-of-life problems that major crime indices do not capture. It is true that certain communities, particularly low-income Black and Hispanic neighborhoods, continue to experience violence at rates far above the national average even as that average falls.

The question "is crime rising?" has a clear empirical answer: no, not by any major national measure. The better question — "are Americans safe enough?" — is harder, more local, and more political. The data can tell us what is happening. It cannot tell us what to feel about it.