Connect with us

Crime

The Great Crime Decline: What Really Happened in America’s Cities in 2025

The United States recorded the largest single-year drop in homicides ever measured. We dug into the data from the FBI, DOJ, and 25+ city police departments to separate fact from spin.

Crime Plummeted in 2025 in Every Major U.S. City, Except for This One
↓21%
Homicide Drop (35 cities)
CCJ Year-End 2025
↓44%
From 2021 Peak
AH Datalytics
~4.0
Per 100K — Lowest Since 1900
CCJ Estimate
31/35
Cities Saw Declines
CCJ 40-City Sample

By The NDS Show Investigative Team · Published February 16, 2026 · Sources: FBI, DOJ, Council on Criminal Justice, AH Datalytics, Bureau of Justice Statistics, 25+ city police departments

Something remarkable happened in American cities in 2025. Murder rates plummeted to levels not seen since before World War I. Carjackings dropped 43%. Shootings fell to record lows in New York, Chicago, Philadelphia, Baltimore, Detroit, San Francisco, and a dozen other cities. By every major measure, 2025 was the safest year in modern American history.

And almost immediately, the fight began over who deserves credit — and whether the numbers can be trusted at all.

This investigation examines the data behind the headlines: what we can measure, what we can’t, where crime fell, where it didn’t, and what the numbers actually tell us about public safety in America.

A Note on Data: The official FBI annual report for 2025 won’t be released until the second half of 2026. The data here comes from three primary sources: the Council on Criminal Justice (40 large cities), the AH Datalytics Real-Time Crime Index (570 agencies covering ~115 million people), and individual city police departments. We note data gaps throughout.


I. The Big Picture: A Historic Three-Year Decline

To understand 2025, you have to understand the trajectory. The United States experienced a dramatic spike in homicides beginning in 2020 — roughly a 30% increase in a single year, the largest jump since reliable record-keeping began. That surge, fueled by pandemic disruptions, social upheaval, and the collapse of community support systems, peaked in 2021.

Then the numbers started falling. And they kept falling.

National Homicide Trend: The Pandemic Spike and Recovery
Indexed to 2019 baseline (2019 = 100)

By 2025, the cumulative decline was staggering: approximately 12,000 fewer people were murdered in 2024 and 2025 combined than in 2020 and 2021. The national homicide rate — estimated at roughly 4.0 per 100,000 — likely fell to its lowest level since 1900.

“So, even taking a conservative view, let’s say it’s 17% or 16%, you’re still looking at the largest one-year drop ever recorded in 2025.”

— Jeff Asher, co-founder of AH Datalytics, former CIA and NOPD crime analyst (ABC News)

The decline wasn’t limited to homicide. According to the Council on Criminal Justice’s year-end 2025 report, virtually every category of crime fell:

-21%
Homicides
CCJ, 35 cities
-43%
Carjackings
CCJ, 19 cities
-27%
Motor Vehicle Theft
CCJ, 29 cities
-23%
Robberies
CCJ, 34 cities
-22%
Gun Assaults
CCJ, 18 cities
-9%
Aggravated Assault
CCJ, 32 cities
-10%
Shoplifting
CCJ, 22 cities
+7%
Drug Offenses
Only increase

Only one category rose: drug offenses, up 7%. Criminologists note this likely reflects increased enforcement rather than increased drug use, since drug crimes in UCR data are largely arrest-driven.


II. City by City: Where Crime Fell — and Where It Didn’t

The map below shows homicide changes across major U.S. cities in 2025. Green markers indicate decreases; orange indicates increases. Click any city for details.

Major decrease (>20%) Moderate decrease (10-20%) Roughly even Increase

City-Level Homicide Data: 2025 vs. 2024

City20242025ChangeHistoric Note
Chicago587416-29%Fewest since 1965
New York City~381305-20%Record-low shootings
Baltimore194133-31%Fewest in ~50 years
Philadelphia~297220-26%Fewest since 1966
Los Angeles~281~230-18%Lowest since 1966
Detroit203165-19%Fewest since 1964
Washington, DC~128~88-31%Data manipulation allegations
Denver-41%One of largest drops nationally
San Francisco~3528-20%Fewest since 1954
Richmond, VA-59%Largest % drop in CCJ sample
New Orleans~-30%Fewest since 1970
Oakland-28%+Fewest since 1967
Seattle-42%Data quality issues noted
Portland-50%+Q1 MCCA data
Atlanta-14%Under 100 first time since pre-COVID
Phoenix-14%Through Q3 only
El PasoEven
Fort Worth+2%Slight increase
Milwaukee+1%+42% vs 2019
Little Rock+16%Largest increase in CCJ sample
Source: Council on Criminal Justice Year-End 2025 Report, individual city police departments. Dashes indicate specific counts not publicly available.

III. Beyond Homicide: The Full Crime Picture

Homicides get the headlines, but the decline extended across virtually every crime category:

Crime Category Changes
CCJ 40-city sample — click tabs to compare time periods

The Standout: Carjackings Down 61% From Peak

Perhaps the most dramatic decline was in carjackings. After peaking in 2023, carjackings fell 43% in a single year and are now 29% below 2019 levels. Chicago — the poster child for pandemic-era carjackings — saw a 50% drop in 2025.

Still Elevated: Motor Vehicle Theft

Despite a 27% drop from 2024, motor vehicle theft remains 9% above 2019 levels — one of only two categories that haven’t fully recovered to pre-pandemic norms.


IV. Spotlight Cities: The Stories Behind the Numbers

Homicide Counts: Major Cities 2024 vs 2025
Cities with verified year-end data

Chicago: Fewest Murders Since 1965

Chicago recorded 416 homicides in 2025 — a 29% drop from 587 in 2024 and the city’s fewest since 1965. Shooting victims fell 33% (to 1,847), shooting incidents dropped 35%, robberies fell 36%, and carjackings plummeted 50%. However, domestic homicides rose approximately 15%, and fatal domestic shootings jumped 50% — a troubling counter-trend buried in the broader good news. (CBS Chicago)

New York City: Record-Low Shootings

NYC recorded 305 homicides (down 20%) and set an all-time record for fewest shootings, with 856 shooting victims — 5% below the previous record set in 2018. December 2025 saw just 35 shootings, the fewest ever recorded in a single month. As the Brennan Center noted: “As of 2025, murders and shootings are at or near all-time lows.” (Gothamist)

Baltimore: 50-Year Low

Baltimore — long synonymous with high murder rates — recorded just 133 homicides, a 31% drop and the fewest in roughly 50 years. The city’s decline has been dramatic: homicides are down approximately 60% from 2019. (Baltimore Mayor’s Office)

San Francisco: Fewest Since 1954

San Francisco recorded just 28 homicides — a 70-year low. The city’s crime decline has been especially notable given the national narrative around urban dysfunction and retail theft. (CBS San Francisco)

The Outlier: Little Rock

Not every city shared in the decline. Little Rock, Arkansas saw homicides increase 16% — the largest jump in the CCJ sample. Milwaukee (+1%) and Fort Worth (+2%) also saw slight increases. Milwaukee’s situation is particularly concerning: while 2025 was roughly flat, the city’s homicide rate remains 42% above 2019 levels the worst long-term trajectory of any city in the CCJ study.


V. The Credit War: Who’s Responsible?

Perhaps no aspect of the 2025 crime decline has been more contested than the question of credit. The Trump administration has aggressively claimed the decline as its own, while criminologists and Democratic officials point to trends that began years earlier.

The Administration’s Case

FBI Director Kash Patel posted on X in January 2026:

“In 2025 the @FBI and partners saw a record year making America safe again — 20% drop in the national murder rate, +31% fentanyl seizures, over 6,000 child victims recovered, 290% increase in gang takedowns, and more.”

— Kash Patel, FBI Director (X/Twitter)

In September 2025 testimony before the Senate Judiciary Committee:

“Twenty-one-thousand arrests in D.C. with our fellow partners. A huge decrease in violent crime. Sixty-percent decrease in gun crimes in Washington, D.C., in the last month. Seventy-four-percent decrease in carjackings in Washington, D.C.”

— Kash Patel, Senate Judiciary Committee, September 16, 2025 (FBI.gov)

“In 2025, we saw the lowest murder rate in 125 years.”

— Pam Bondi, Attorney General, House Judiciary Committee, February 11, 2026 (House.gov)

“This dramatic decline is what happens when a president secures the border, fully mobilizes federal law enforcement to arrest violent criminals and aggressively deports the worst of the worst illegal aliens from our country.”

— Karoline Leavitt, White House Press Secretary (Fox News)

The Expert Pushback

“We see very confident claims of credit in abundance, but scarce hard evidence to back them up. The remarkable consistency in the magnitude of the decline across the country really suggests that everything’s happening at the macro level.”

— Adam Gelb, President and CEO, Council on Criminal Justice (TIME)

“When COVID hit, and the world shut down, we basically turned off the water with respect to prevention and intervention strategies. And then it took about two or three years for the water to be turned back on.”

— Alexis Piquero, University of Miami, former BJS Director (TIME)

“I think the American Rescue Plan is probably the most important federal legislation in the years since the pandemic… I think it has gone unrecognized how incredibly effective it was in stabilizing communities and stabilizing local government.”

— Patrick Sharkey, Princeton University (TIME)

National Guard: The Mixed Verdict

The Trump administration deployed National Guard troops to Washington D.C., Chicago, Memphis, Los Angeles, and Portland in 2025. An NPR analysis concluded “results were unclear”, while a Stateline investigation found Trump “isn’t sending troops to cities with highest crime rates.”

The data suggests both sides are oversimplifying. The decline is remarkably uniform — spanning red and blue states, cities with progressive DAs and tough-on-crime prosecutors, cities with Guard deployments and without. A DW fact-check of 400 cities found “the political affiliation of mayors made little difference when it comes to crime rates and policing.”


VI. What We Can’t Measure: The Limits of Crime Data

⚠️ The Reporting Gap: According to the Bureau of Justice Statistics’ National Crime Victimization Survey, only 57% of aggravated assaults and 41% of simple assaults are reported to police. The most common crimes are the least likely to appear in official statistics.

UCR vs. NCVS: Two Different Pictures

The FBI’s Uniform Crime Report (UCR) counts crimes reported to law enforcement. The BJS National Crime Victimization Survey (NCVS) surveys households about crime experiences, including unreported crimes. As the CCJ noted: year-over-year comparisons between the two “should be made cautiously, but not ignored.”

The NIBRS Transition Problem

The FBI’s 2021 transition from SRS to NIBRS caused major data gaps — participation dropped below 70%. By 2024, coverage recovered to 95.6%, but artifacts remain. Jeff Asher flagged that LAPD’s NIBRS data shows a 52% property crime drop likely reflecting methodology changes, not actual crime reduction.

The D.C. Data Manipulation Allegations

The House Oversight Committee alleged MPD Chief Pamela Smith “deliberately manipulated crime data on a widespread basis.” Whistleblowers said supervisors routinely downgraded charges across all seven patrol districts. Chief Smith announced her departure in early 2026. The allegations remain under investigation.

Data Availability Timeline

  • ✅ Available Now: CCJ 40-city report, RTCI 570-agency data, individual city year-end reports
  • ⏳ Mid-2026: FBI preliminary semiannual report (Jan-Jun 2025)
  • ⏳ Late 2026: FBI official 2025 annual crime report
  • ⏳ 2027+: NCVS 2025 victimization survey, CDC final mortality data, BJS detailed reports

VII. The Perception Gap

“National averages can hide what is happening in some neighborhoods.”

— Thaddeus Johnson, Georgia State University, former Memphis police official (CBS News)

Johnson’s point is critical: a city can see overall homicides drop 20% while individual blocks experience no change — or even worsening conditions. The Brennan Center’s Ames Grawert warned that federal funding cuts could threaten continued progress into 2026.


VIII. Timeline: The Three-Year Decline

Violent Crime Categories: 2019 to 2025
Indexed to 2019 baseline (100)

IX. Claims Table: What We Can and Can’t Verify

In the spirit of investigative transparency, here’s our assessment of the major claims being made about 2025 crime:

ClaimConfidenceEvidence
Homicides dropped ~20% nationallyHIGHCCJ, RTCI, FBI preliminary all converge
Lowest murder rate since 1900MEDIUMCCJ ~4.0/100K; pre-1960 data imprecise
Largest single-year drop everHIGH~20% exceeds any prior year in FBI records
31 of 35 cities saw declinesHIGHCCJ verified; exceptions: Little Rock, Fort Worth, Milwaukee
Trump policies caused the declineLOWDecline began 2023; uniform across geographies
American Rescue Plan / CVI caused itMEDIUMResearch supports CVI; timeline aligns; not sole cause
National Guard reduced crimeLOWNPR: “results unclear”; similar rates in non-deployed cities
D.C. crime data was manipulatedMEDIUMHouse Oversight investigation + whistleblower
Post-pandemic normalization is primary driverMEDIUMBest-supported explanation; timeline and mechanism plausible

X. What We Know, and What We Don’t

What we know: The 2025 crime decline is real, historic, and broad-based. It spans every region, nearly every major city, and virtually every crime category. The homicide rate has likely fallen to levels not seen in over a century. This is unambiguously good news.

What we don’t know: Why, exactly, it happened. The uniformity strongly suggests macro-level factors: post-pandemic normalization, the restoration of social services, improvements in trauma medicine. No single policy, administration, or intervention can credibly claim primary responsibility.

What we’re watching: Whether it lasts. Federal funding cuts to community violence intervention programs, potential policing budget changes, and the expiration of American Rescue Plan funds all represent risks.

The numbers are clear. The causes are complex. And the credit claims, from all sides, should be weighed against the evidence — not the rhetoric.

📌 This investigation will be updated as the FBI’s official 2025 report (expected late 2026) and NCVS victimization survey become available.


Methodology & Sources

This investigation relied on publicly available data from federal agencies, independent research organizations, and municipal police departments. No single-source assertions were included without corroboration. Claims were rated High/Medium/Low based on the number and independence of supporting sources.

Continue Reading