Falling Crime and Public Perception: Comparing the Dates Before Declaring a Gap

Editorial correction — September 20, 2026. The earlier article incorrectly reported 78% from a Gallup release that says 49%, confused a large annual decrease with a historic low level, and treated a 2025 poll as a response to data released in 2026. Expanded September 20, 2026 with additional primary-source data and a Sources list.

Two things can be true at once: reported crime can fall sharply, and many Americans can still believe crime is rising. Comparing them fairly takes more care than a headline usually allows, because official statistics, public opinion and victim surveys measure different things at different times. This article lines up what the FBI, Gallup and the Bureau of Justice Statistics (BJS) each reported, in the order the data were collected, and says what can and cannot be concluded.

What the FBI’s early look shows

On May 13, 2026, the FBI’s Uniform Crime Reporting Program released “First Look: 2025 Crime Data” on its Crime Data Explorer. The estimates rest on data from more than 17,000 agencies covering 96% of the population, up from 16,675 agencies in 2024. More than 15,000 of them reported through the National Incident-Based Reporting System, covering nearly 90% of the population, and nearly 500 more agencies used that system in 2025 than in 2024. The FBI calls the figures preliminary and says they are subject to change before its full report, Reported Crimes in the Nation, 2025, later this year.

The release also carries a statement from the FBI Director describing the 2025 decreases in violent crime and murder as the largest since 1937. That is an official’s characterization, and this article has not checked it against the historical series. It is also a statement about the size of a one-year change. A large drop tells us the direction and pace of movement, not whether the resulting level is the lowest on record, which would need the full multi-year rates to answer.

Coverage matters too. Because more agencies reported in 2025 and more switched to the incident-based system, the FBI’s national figures are estimates built on a changing set of reporters. The changing reporting base is one reason to treat the early figures as provisional.

What Gallup found, and when

Gallup’s annual Crime poll was conducted October 1 to 16, 2025, and the results were published October 30, 2025 by Megan Brenan. This was about seven months before the FBI’s first look at 2025 data. That order matters: the poll cannot show how people reacted to the May 2026 release, and no one in the October survey could have seen it.

Gallup says the two national measures are at their lowest levels since at least 2018, and that Americans now view national crime more favorably than at any point in recent years, similar to the early 2000s. It also reports that 33% said there was less crime nationally and 8% volunteered that it was about the same. So the poll does not show that Americans ignore good news: perceptions improved in each of the last two years. It shows that in October 2025 a near-half share still said crime had risen, even though Gallup notes FBI data showed a 4.5% decrease in crime for 2024.

National worry versus local experience

Gallup notes that Americans have consistently rated crime in their own area as less serious than crime in the country at large. The 2025 numbers show the gap plainly: 49% said national crime was a serious problem, and 12% said the same of their local area. This article does not claim to know why. What is measurable is that the two questions produce different answers, so quoting one as if it were the other misleads.

Politics shapes the answer

Gallup found that Republicans’ belief that crime is rising fell 36 points from 2024, while Democrats became more likely than a year before to say crime had increased. Gallup suggests this may reflect the change from a Democratic to a Republican administration, a pattern it says has occurred before: people whose party holds the White House are generally less likely to say crime is up, and people from the other party are more likely to say so. If so, part of the gap between perception and the data may reflect political mood.

What people worry about is not what the headline counts

The crimes that worry Americans most in Gallup’s survey are not the ones in the FBI’s headline list. Identity theft topped the list at 69% (worrying frequently or occasionally), followed by being tricked into giving financial information to scammers at 53%. Both were the only crimes with majority worry. Worry about a school-aged child being physically harmed at school was 43% overall and 60% among parents of school-aged children. Fewer worried about a car being stolen (39%), a home burglarized (34%), a hate crime (30%), being mugged (29%), attacked while driving (27%), murdered (22%) or sexually assaulted (21%). Gender was the largest differentiator: 38% of women and 4% of men worried about sexual assault. Fraud and identity theft are not among the categories in the FBI release described above, so a decline in the FBI’s headline categories says nothing about them.

A third measure: what victims report

Police statistics only count crimes that come to police attention and are recorded. The BJS National Crime Victimization Survey (NCVS) asks a sample of households directly, whether or not they called the police. BJS’s Criminal Victimization, 2024, published September 29, 2025, reports 23.3 violent victimizations per 1,000 people aged 12 or older, and 1.45% of people in that age group experienced at least one violent victimization, similar to 2023. The rate of violent victimizations reported to police was 11.2 per 1,000, which by simple arithmetic is roughly half (48%) of the total. Among teens 12 to 17, 1.95% experienced one or more violent crimes in 2024, compared with 1.45% in 2023.

Two cautions apply. BJS says its 2024 statistics use the legacy survey instrument, and it plans a separate release for estimates from the redesigned instrument, so those figures may not match future ones. And the NCVS covers 2024, not 2025, so it cannot be set against the FBI’s 2025 estimates without care. It also counts nonfatal crimes only and includes simple assault, so its categories do not line up with the FBI’s.

What can be concluded

• The FBI’s preliminary data show declines in every violent and property category listed, for 2025 compared with 2024.
• Gallup’s October 2025 poll shows public perceptions improving for the second straight year, though 49% still said national crime had risen.
• The two were measured at different times, and the poll predates the FBI release, so neither can be said to answer the other.
• The victim survey suggests a large share of violent crime never reaches police records, so official counts are a partial picture.
• Level and change are different claims: a big decline does not establish a record low.

How to read the next release

• Check the year each source covers and the month it was collected.
• Ask whether the source counts crimes reported to police, crimes reported by victims, or opinions.
• Treat preliminary figures as revisable, and check the final report when it arrives.
• Separate national perception from local perception, and note the party of the sitting president.

Sources

1. FBI, “FBI Releases Historic Early Look at Annual Crime Data,” May 13, 2026: fbi.gov press release
2. Gallup, Megan Brenan, “Crime in U.S. Seen as Less Serious for Second Straight Year,” October 30, 2025: news.gallup.com/poll/697124
3. Bureau of Justice Statistics, Susannah N. Tapp and Emilie J. Coen, “Criminal Victimization, 2024,” September 29, 2025: bjs.ojp.gov/library/publications/criminal-victimization-2024

Editorial standards · Report an error · Subscribe