In 2025, murders in the United States fell by approximately 20 percent — the largest single-year drop ever recorded. Following declines of 13 percent in 2023 and 15 percent in 2024, the US homicide rate has now reached approximately 4.0 per 100,000 residents: the lowest figure in American public health or law enforcement data going back to 1900. This is one of the most significant public safety achievements in modern history. It is also one of the most politically contaminated stories of 2026. This article is an attempt to separate what the data says from what political actors want it to say.

WHAT ACTUALLY HAPPENED

The scale of the decline is not in dispute. The nonpartisan Council on Criminal Justice — the most authoritative source tracking US crime in real time — confirmed homicides fell 21 percent across the 35 major cities it tracks in 2025. Independent analyst Jeff Asher, whose Real-Time Crime Index is used by law enforcement agencies and researchers across the political spectrum, estimates the national murder count dropped by roughly 4,000 in 2025 alone, building on reductions of approximately 3,500 in 2023 and 4,200 in 2024.

The numbers by city are historic. Detroit recorded its lowest murder total since 1964. Baltimore fell 31 percent. San Francisco is tracking toward its lowest homicide count since 1940. Philadelphia recorded its fewest killings since 1966. New York City recorded its fewest shooting incidents in documented history through the first eleven months of 2025. The pattern is not isolated to one region, one political party, one police chief, or one policy. It is national, simultaneous, and consistent.

In human terms: approximately 12,000 fewer families lost someone to murder in 2024 and 2025 combined compared to 2020 and 2021. That number is not a statistic. It is the size of a small town — worth naming clearly before anything else.

THE TIMING QUESTION — AND WHY IT IS THE MOST IMPORTANT FACT IN THIS STORY

The credit-claiming began quickly and from multiple directions.

On December 24, 2025, Elon Musk posted on X: "Removing murderers from the streets works wonders." On December 27, White House Press Secretary Karoline Leavitt posted on X calling the trend "The Trump Effect." On January 22, 2026 — the day Axios published the CCJ data — FBI Director Kash Patel posted on X: "Axios just reported the lowest murder rate since 1900, down 20% from 2024, and somehow forgot to mention President Trump or the historic results of this FBI. Nearly 200% more arrests. Violent gangs crushed. Fugitives hunted down." On January 29, President Trump announced at a Cabinet meeting: "The murder rate in our country is the biggest drop ever recorded. It is at the lowest level in about 125 years" — attributing it to his immigration crackdown and the removal of "bad people" from the country. On February 5, Leavitt stated at a formal White House briefing: "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."

There is a structural problem with every one of these claims. The decline did not begin in 2025.

Murders fell 13 percent in 2023. They fell another 15 percent in 2024. Both years occurred entirely under the Biden administration, before a single Trump-era immigration or enforcement policy took effect. As Jeff Asher told Reason magazine in February 2026: "We are seeing the largest one-year drop in murder for the third straight year in 2025." The three-year downward trend was already well established — and already being described by criminologists as potentially historic — before January 20, 2025. Reason magazine, a libertarian publication with no partisan reason to favor either side, stated plainly: "The Trump administration will likely take credit for this feat, but this probably would have happened regardless of who was in the White House."

Any causal argument connecting Trump-era enforcement to the murder decline must first explain why the identical trend began two years earlier. The data does not currently support that explanation.

WHAT THE ICE ARREST DATA ACTUALLY SHOWS

Setting aside the timing problem, the underlying claim — that the administration is removing violent criminals from American streets — deserves direct engagement with the data.

The Cato Institute, a libertarian think tank that cannot be credibly accused of liberal bias, analyzed ICE custody data for fiscal year 2025. It found that only 5 percent of people booked into ICE custody had violent criminal convictions. Seventy-three percent had no criminal conviction of any kind.

The Brennan Center for Justice reviewed ICE enforcement data from October 2025, when ICE arrested and jailed approximately 30,000 immigrants in a single month. Of that group, 0.05 percent had murder convictions and 0.2 percent had rape convictions. More people in that cohort were detained for minor traffic offenses than for violent crimes.

The Deportation Data Project — using FOIA-obtained data from UC Berkeley and UCLA — found that the number of violent criminals arrested by ICE's Fugitive Operations unit actually fell during this period, from approximately 100 per week to 50–60 per week, even as total arrest numbers surged. ICE was arresting fewer actual violent criminals over time, not more, as enforcement expanded.

Even setting all of this aside: the United States has a population of 330 million people. The removal of a few thousand individuals — the vast majority of whom had no violent criminal record — cannot mathematically account for a 20 percent decline in murders spread uniformly across every region, every city size, and every political jurisdiction in the country. The CCJ data is explicit on this point: the breadth and simultaneity of the decline points to national structural forces, not targeted enforcement actions.

WHAT THE RESEARCH ACTUALLY POINTS TO

The CCJ convened independent criminologists to identify what drove the decline. The strongest evidence points to two converging forces.

The first is the American Rescue Plan Act of 2021, which allocated over $15 billion to public safety at the state and local level. A significant portion funded Community Violence Intervention programs — credible messenger outreach, violence interrupters, hospital-based intervention, and youth worker programs deployed specifically in the neighborhoods where homicides were concentrated. These programs had peer-reviewed evidence behind them before ARPA scaled them. The murder decline tracked almost precisely with their deployment timelines.

The second is post-pandemic stabilization. The 2020 murder spike — a 30 percent single-year increase, the largest in recorded American history — was driven by what researchers at Brookings identified as a specific mechanism: the sudden removal of young men and boys from school and work in low-income urban neighborhoods during COVID lockdowns in March and April of 2020. As those conditions normalized through 2022 and 2023 — schools reopened, hospitality employment returned, chronic absenteeism declined — the spike reversed. The timing matches with precision that neither policing policy nor immigration enforcement can replicate.

The Bipartisan Safer Communities Act, passed with Republican and Democratic support in 2022, added $400 million more for violence intervention programs. Its effects began appearing in the 2023 data.

THE RED STATE / BLUE STATE PICTURE — THE FULL VERSION

Both parties cherry-pick this data. Here is the unselected version.

At the state level, Republican-governed states have had consistently higher murder rates than Democratic-governed states for over two decades. The CDC's most recent data: Mississippi (19.4 per 100,000), Louisiana (19.3), Alabama (14.8) — all deep red. California: 5.1. An analysis of 2024 FBI data found that police agencies in Republican-governed states recorded murder rates 32 percent higher than in Democratic-governed states.

At the county level, blue-voting urban counties have higher murder rates than rural red counties. The reason: 13 of the 20 US cities with the highest murder rates in 2024 were located in Republican-governed states — Memphis, Indianapolis, St. Louis, Kansas City among them — but governed by Democratic mayors. The dominant geographic variable is urban density, not party affiliation at any level of government.

The decline was universal across all of these geographies. It is not a partisan achievement. It is a national one.

THE 2026 WARNING

This is the part of the story that deserves as much attention as the progress itself.

The American Rescue Plan funds that are most strongly associated with the decline had to be allocated by the end of 2024 and disbursed by the end of 2026. Most have already been spent. The community violence intervention programs they funded are now facing shrinkage or closure as that money runs out. The current federal administration has cut community-based prevention programs. The CCJ's own researchers forecast "greater variation in crime in 2026" — a careful way of saying that some of what drove this decline may not hold.

The progress is real. The risk that it does not continue is also real.

WHY THE NARRATIVE GAP MATTERS

Twelve thousand fewer families lost someone. That fact has been almost entirely displaced in public conversation by partisan credit-claiming, algorithmically amplified fear, and the structural incentive news media has to lead with what is burning rather than what is healing.

Bright Side Global exists precisely for this gap. Not to manufacture optimism, but to insist that verified progress — with all its complexity, all its caveats, and all its political contamination stripped away — deserves to be seen clearly. The data on America's murder decline is extraordinary. So is the gap between what the data says and what most Americans currently believe.

Both things are worth reporting. We just did.

Sources: