Crime Statistics — Transparent Sources

A neutral directory of 15 official and peer-reviewed U.S. crime-statistics sources — with each one's coverage of race, Hispanic origin, sex, age, and nationality/citizenship, a transparency rating, and the caveats you need to read it honestly. The goal is genuine statistics for transparent reporting — not to stratify anyone or "prove bias." We link to the primary data and make no claim about what the numbers mean; the safeguard against misuse in any direction is method, not omission. Last curated 2026-07-15.

🔎 Ask the data

Ask a specific question about U.S. crime statistics and get the sourced numbers — rates, years, and the primary source for each. This returns data only — no interpretation, no opinion; if it's asked why or what the numbers "mean," it declines and gives you the figures. If the dataset doesn't cover your question, it says so plainly.

Try:

How to read crime statistics honestly

Rates, not raw counts
A raw count of arrests or victims by group is meaningless without a per-capita rate. Divide by the group's Census/ACS population (below). Groups differ in size, age structure, and where they live.

Hispanic is an ethnicity, not a race
The FBI stores race and Hispanic origin as two SEPARATE fields. Historically ~95% of Hispanic arrestees were coded 'White,' inflating the White total — in BJS's 2018 reconstruction, reported White violent arrests (62.7%) dropped to 45.9% once Hispanic-White arrestees were reclassified. Only NIBRS microdata gives a clean race×ethnicity split.

Arrests ≠ offending ≠ convictions ≠ incarceration
Each stage has its own selection effects (who gets reported, arrested, charged, convicted, imprisoned). Incarceration shares are NOT offending shares. Useful anchor: BJS's Jan-2021 study found White and Black people were arrested roughly proportionate to their involvement in serious violent crime as measured independently by victims (NCVS).

Triangulate independent measures
Compare FBI arrests (police) against NCVS victim reports (survey) and CDC homicide deaths (medical examiner). Where they agree, confidence is high; where they diverge, that flags reporting or enforcement effects — not necessarily offending.

Confounders — raw gaps prove neither causation nor bias
Age structure, sex, poverty, and urbanicity drive crime rates. A raw demographic disparity, by itself, establishes neither a causal claim nor discrimination in either direction. Control for confounders before inferring anything.

Nationality data is thin — say so
FBI/UCR/NIBRS and virtually all local police do NOT record citizenship or immigration status. Federal data (US Sentencing Commission, BOP) records citizenship but is dominated by immigration offenses; Texas is the only state with status-linked data; and immigrant-population denominators are themselves estimates.

Mind the 2021 NIBRS break
The FBI's Jan-1-2021 switch to NIBRS-only dropped population coverage to ~65% (NYPD and LAPD reported nothing). The FBI later re-accepted the old SRS format; coverage recovered to ~94% (2023) and ~96% (2024). Do not compare across 2021 naively — treat it as a discontinuity.

Small numbers are unstable
Rates built on tiny cell counts swing wildly year to year. CDC suppresses cells of 1–9 deaths for exactly this reason; treat small-N breakdowns with caution and prefer multi-year pooling.

The numbers (rates, each cited — data only)

Curated from the primary sources below; every figure carries its source, year, and caveat. Read the guardrails above for how to interpret responsibly — these are figures, not conclusions. Updated 2026-07-27.

Overall crime levels & trends

Violent crime rate (FBI), by year

Source: FBI UCR (via Pew / FBI press releases) ↗ · Police-reported. Rose in 2020–21 (pandemic), then fell 2022–2024.

Year	per 100,000	Year
2019	366.7	2019
2020	387.8	2020
2022	380.7	2022
2023	363.8	2023

Caveat: FBI reported/estimated; revised across vintages. 2021 omitted (NIBRS-coverage break). 2024 fell a further ~4.5% — the LOWEST violent crime rate since 1969 (FBI).

Property crime rate (FBI), by year

Source: FBI UCR (via Pew) ↗ · Long-declining trend.

Year	per 100,000	Year
2020	1,958.2	2020
2022	1,954.4	2022

Caveat: 2024 property crime fell ~8.1% — the lowest recorded rate since 1961 (FBI); motor-vehicle theft dropped 18.6% (largest single-year decline). Exact 2023/2024 per-100k values not cleanly published.

Murder / homicide rate, by year (FBI vs CDC)

Source: FBI UCR + CDC (via Pew / AH Datalytics) ↗ · 2020 = largest one-year RISE on record (+29%); 2023–2024 = fastest DECLINE on record.

Year (source)	per 100,000	Year
2020 (FBI)	6.5	2020
2022 (FBI)	6.3	2022
2023 (FBI)	5.7	2023
2024 (FBI)	5.0	2024
2020 (CDC — runs higher)	7.8	2020
2023 (CDC — runs higher)	7.1	2023

Caveat: FBI is police-reported; CDC (death certificates) runs HIGHER. 2024 FBI rate (5.0) is the lowest since 2015, down ~25% from 2020.

Murder / homicide count (total), by year

Source: FBI UCR + CDC (via Pew) ↗ · FBI counts police-reported murders; CDC counts all homicides from death certificates.

Year (source)	count	Year
2019 (FBI)	16,669	2019
2020 (FBI)	21,570	2020
2024 (FBI, estimated)	16,935	2024
2024 (CDC — all homicides)	20,162	2024

Caveat: FBI 2024 estimate = 16,935; CDC 2024 total homicides = 20,162 (CDC > FBI; different methods).

Clearance rate — share of crimes solved (FBI, 2022)

Source: FBI UCR (via Pew / AH Datalytics) ↗ · Cleared by arrest or exceptional means. Most crimes go unsolved.

Crime type	% cleared	Year
Murder / non-negligent manslaughter	52.3%	2022
Violent crime	36.7%	2022
Property crime	12.1%	2022

Caveat: Violent-crime clearance has fallen (45.5% in 2019 → 36.7% in 2022). Murder clearance rose to ~57.8% in 2023.

Crime by offense type

Crime rate by offense type (FBI), 2023 → 2024

Source: FBI CDE — UCR Summary of Reported Crimes 2024 (Fig. 14–15) ↗ · National estimated rates. In 2024 EVERY category fell; in 2023 motor-vehicle theft rose (+12.6%, past 1M for the first time since 2007) while the others fell.

Offense	per 100,000	Year
Rape	39.9 → 37.5	2023–2024
Robbery	67.1 → 60.6	2023–2024
Aggravated assault	266.6 → 256.1	2023–2024
Burglary	253.3 → 229.2	2023–2024
Larceny-theft	1,359.6 → 1,272.1	2023–2024
Motor-vehicle theft	321.3 → 258.8	2023–2024
Arson	21.9 → 20.2	2023–2024
Overall violent crime	379.5 → 359.1	2023–2024
Overall property crime	1,934.1 → 1,760.1	2023–2024

Caveat: FBI SRS-estimated lane; BJS NIBRS reports slightly higher via a different method. Overall figures in this section are the FBI 2024-report vintage (it revised 2023), which differs slightly from earlier-vintage estimates in the Trends section — the FBI revises across releases.

Homicide & mortality

Firearm homicide death rate, by VICTIM race/ethnicity (CDC)

Source: CDC MMWR (mm7242a4), CDC WONDER mortality ↗ · CDC death-certificate data = the race/ethnicity of the person KILLED (the victim). For the OFFENDER/perpetrator side see 'Murder offenders, by race' (FBI SHR) and 'Perceived offender race' (NCVS).

Race/ethnicity	per 100,000	Year
Total (all)	5.9	2022 (provisional)
Black (non-Hispanic)	27.5	2022 (provisional)
American Indian/Alaska Native (non-Hispanic)	9.3	2022 (provisional)
Hispanic	5.5	2022 (provisional)
White (non-Hispanic)	2.0	2022 (provisional)
Asian/Pacific Islander (non-Hispanic)	1.1	2022 (provisional)

Caveat: Crude rates (not age-adjusted); 2022 is provisional. Firearm homicide only, not all homicide. The age-adjusted final overall figure differs (6.2, CDC NVSR).

Murder offenders, by race (FBI SHR)

Source: FBI UCR, Crime in the U.S. 2019, Expanded Homicide Data Table 3 ↗ · 16,245 murder-offender records, 2019.

Offender race	share of offender records	Year
Black	29.1%	2019
White	27.7%	2019
Unknown race	29.3%	2019

Caveat: LOAD-BEARING: race is UNKNOWN for 29.3% of offender records (unsolved/unidentified). These are shares of ALL records including unknowns — NOT 'share of murders committed by each race.' The FBI does not publish a known-offender-only recomputation, so none is shown.

Firearm homicide death rate, by sex (and by race×sex where published)

Source: CDC QuickStats (mm7226a9), NVSS 2021 ↗ · Age-adjusted, 2021 — the latest clean sex-by-race firearm split.

Group	per 100,000 (age-adjusted)	Year
Male (all races)	11.1	2021
Female (all races)	2.1	2021
Black (NH) male	52.9	2021
Black (NH) female	7.5	2021

Caveat: White/Hispanic/AIAN by-sex rates are published only inside a chart image (not machine-readable), so they are omitted here rather than guessed. Victim demographics.

Guns, gun deaths & mass shootings

Gun deaths (CDC), by intent

Source: CDC WONDER (via Pew) ↗ · Most gun deaths are SUICIDES, not homicides.

Category	Value	Year
Total firearm deaths	44,447	2024
Firearm suicides	27,593 (62%)	2024
Firearm homicides	15,364 (35%)	2024
Firearm homicides (2021 peak)	20,958	2021
National firearm death rate	12.8 per 100,000	2024
Total firearm deaths (2022)	48,204 (56% suicide / 41% homicide)	2022

Caveat: CDC death-certificate data. The split shifted as gun homicides fell ~27% from the 2021 peak — 62% suicide / 35% homicide in 2024 vs ~56% / 41% in 2022.

Firearm death rate by state (CDC, 2024, age-adjusted)

Source: CDC WONDER (via Pew / VPC) ↗ · Highest- and lowest-rate states.

State	per 100,000	Year
Mississippi (highest)	28.0	2024
New Mexico	26.6	2024
Alaska	24.4	2024
Alabama	23.7	2024
Wyoming	23.4	2024
Rhode Island	4.6	2024
New York	4.4	2024
New Jersey	4.0	2024
Massachusetts	3.8	2024
Hawaii (lowest)	3.7	2024

Caveat: TOTAL firearm deaths (suicide + homicide + other), age-adjusted — the majority are suicides. Don't blend years.

Mass shootings — the definition determines the count

Source: Gun Violence Archive + The Violence Project / Mother Jones ↗ · The single most important caveat here: 'mass shooting' counts differ by ~100× depending on the definition used.

Definition — year	incidents	Year
Gun Violence Archive (4+ shot) — 2021	690	2021
Gun Violence Archive (4+ shot) — 2022	645	2022
Gun Violence Archive (4+ shot) — 2023	660	2023
Gun Violence Archive (4+ shot) — 2024	502	2024
Violence Project (4+ killed, public) — 2017–2021 TOTAL	31	2017–2021
Mother Jones (public mass shootings) — 1982–2023 TOTAL	140	1982–2023

Caveat: GVA counts 4+ SHOT (incl. wounded, any setting) → ~500–690/yr. The Violence Project / Mother Jones count 4+ KILLED in PUBLIC (excl. gang/domestic) → single digits/yr. Same year 2017: GVA 346 vs Mother Jones 11.

School shootings per year

Source: K-12 School Shooting Database + Everytown ↗ · Two non-comparable datasets (different definitions + time units).

Source — year	incidents	Year
K-12 SSDB (any gunfire) — 2023 (record)	349	2023
K-12 SSDB (any gunfire) — 2024	330	2024
Everytown (gunfire on school grounds) — 2023-24 school year	144	2023-24 SY

Caveat: K-12 SSDB counts ANY gunfire on school property (incl. accidents/suicides/afterhours), by calendar year. Everytown counts 'gunfire on school grounds' by school year. Not interchangeable.

Victimization & offending

Violent victimization rate, by victim race/ethnicity

Source: BJS, Criminal Victimization, 2023 (Table 3) ↗ · National Crime Victimization Survey — captures crime whether or not reported to police.

Victim race/ethnicity	per 1,000 (age 12+)	Year
Total	22.5	2023
Black (non-Hispanic)	26.9	2023
White (non-Hispanic)	22.5	2023
Hispanic	21.3	2023
Asian/NHOPI	10.7	2023
Other (AIAN or two+ races)	50.4	2023
Male	20.8	2023
Female	24.2	2023

Caveat: Survey of living victims; excludes homicide. 'Other' combines American Indian/Alaska Native-only and two-or-more-races (no clean standalone AIAN rate).

Perceived offender race in violent incidents

Source: BJS, Criminal Victimization, 2023 (Table 11) ↗ · The victim's PERCEPTION of the offender — the independent survey counterpart to police/arrest data.

Perceived offender group	share of incidents	Year
White (non-Hispanic)	53.7%	2023
Black (non-Hispanic)	24.0%	2023
Hispanic	14.0%	2023
Asian/NHOPI	1.3%	2023
Other (AIAN or two+ races)	2.1%	2023
Multiple offenders, various races	4.8%	2023

Caveat: Victim perception, not verified identity. Offender race/origin was UNKNOWN in ~17% of incidents (embedded in the denominator). Nonlethal violence only.

Triangulation — who victims say offended (survey) vs. who was arrested (police), violent crime, 2018

Source: BJS, Race and Ethnicity of Violent Crime Offenders and Arrestees, 2018 (NCJ 255969, Table 2) ↗ · The key honesty check: does arrest data track offending as measured independently by victims? For SERIOUS violent crime, BJS found NO statistically significant race difference between NCVS-identified offenders and UCR arrestees.

Group / measure	share	Year
White (NH) — victim-reported offenders (NCVS)	52.2%	2018
White (NH) — arrested (UCR)	45.9%	2018
Black (NH) — victim-reported offenders (NCVS)	28.9%	2018
Black (NH) — arrested (UCR)	33.0%	2018
Hispanic — victim-reported offenders (NCVS)	14.2%	2018
Hispanic — arrested (UCR)	17.6%	2018

Caveat: BJS White/Black exclude Hispanics; excludes homicide. ~9–12% of incidents had undetermined Hispanic origin (may underestimate Hispanic offenders).

Interracial violent crime — victim × perceived-offender race (NCVS)

Source: BJS — Race and Hispanic Origin of Victims and Offenders, 2012–15 (NCJ 250747) ↗ · Of violent victimizations: 51% were intraracial (same race), 41% interracial. Below = perceived offender race by victim race.

Victim → offender	share	Year
All violent crime — intraracial (same race)	51%	2012–2015
All violent crime — interracial (different race)	41%	2012–2015
White victim → White offender	56.6%	2012–2015
White victim → Black offender	14.7%	2012–2015
White victim → Hispanic offender	11.0%	2012–2015
Black victim → Black offender	63.2%	2012–2015
Black victim → White offender	10.9%	2012–2015
Black victim → Hispanic offender	6.6%	2012–2015
Hispanic victim → Hispanic offender	40.3%	2012–2015
Hispanic victim → White offender	20.0%	2012–2015
Hispanic victim → Black offender	20.5%	2012–2015

Caveat: NCVS = victim's PERCEPTION; offender race unknown in ~5% of victimizations (8% incl. unknown offender count). 2012–2015 aggregate; excludes homicide. BJS White/Black exclude Hispanics; 'interracial' includes mixed-race offender groups.

Arrests, population & incarceration

Arrest distribution (all offenses), by race

Source: FBI UCR, Crime in the U.S. 2019, Table 43 ↗ · Shares of ~6.82M arrests (agencies reporting race). 2019 = last near-universal year before the 2021 NIBRS break.

Race	share of arrests	Year
White (incl. Hispanic)	69.4%	2019
Black	26.6%	2019
American Indian/Alaska Native	2.4%	2019
Asian	1.3%	2019

Caveat: FBI 'White' INCLUDES Hispanics (Hispanic is a separate ethnicity axis: 19.1% of arrests). Do not compare directly to a non-Hispanic-White population share.

Arrest distribution (violent crime), by race

Source: FBI UCR, Crime in the U.S. 2019, Table 43 ↗ · Shares of ~355K aggregate violent-crime arrests.

Race	share of arrests	Year
White (incl. Hispanic)	59.1%	2019
Black	36.4%	2019

Caveat: FBI 'White' includes Hispanics. Arrests are not the same as offending or convictions.

U.S. resident population share, by race/ethnicity (the denominator)

Source: U.S. Census Bureau, 2020 Census (non-Hispanic basis) ↗ · You need this to turn arrest/crime SHARES into meaningful comparisons. Non-Hispanic basis (matches BJS).

Race/ethnicity	share of population	Year
White (non-Hispanic)	57.8%	2020
Hispanic (any race)	18.7%	2020
Black (non-Hispanic)	12.1%	2020

Caveat: Census 'White' here is non-Hispanic; do not compare to FBI 'White' (which includes Hispanics). Other Census universes (race-alone incl. Hispanic) give different numbers.

Imprisonment rate, by race/ethnicity and sex (all ages)

Source: BJS, Prisoners in 2023 – Statistical Tables (Table 5) ↗ · Sentenced prisoners (>1 yr) per 100,000 residents of that group, year-end 2023.

Group	per 100,000	Year
Total	360	2023
Black (non-Hispanic)	929	2023
American Indian/Alaska Native	809	2023
Hispanic	429	2023
White (non-Hispanic)	190	2023
Asian/NHOPI	70	2023
Male	676	2023
Female	43	2023

Caveat: Incarceration shares are NOT offending shares — they reflect arrest, charging, sentencing, and release. 'White'/'Black' exclude Hispanics; Asian row includes Pacific Islanders.

Recidivism of released state prisoners (BJS)

Source: Bureau of Justice Statistics ↗ · How many released prisoners are arrested again.

Metric	Value	Year
Rearrested within 3 years	67.8% of released prisoners	2005 cohort (30 states)
Rearrested within 5 years	76.6% of released prisoners	2005 cohort (30 states)
Rearrested within 9 years	83% of released prisoners	2005 cohort (30 states)
Rearrested within 10 years	82% of released prisoners	2008 cohort (24 states)
Returned to prison within 5 years	55.1% of released prisoners	2005 cohort
Returned to prison within 10 years	61% of released prisoners	2008 cohort

Caveat: Headline figures are REARREST (charged with a new crime — not guilt, reconviction, or return-to-prison). Return-to-prison is much lower. Two non-comparable cohorts (2005/30-state, 2008/24-state) — BJS warns against cross-comparing.

Wrongful convictions & exonerations

Exonerations since 1989 (National Registry)

Source: National Registry of Exonerations ↗ · The Registry calls its counts a 'basement figure' — the true number is higher.

Measure	Value	Year
Total exonerations since 1989	3,842 exonerations	1989–2026
Total years lost by exonerees	35,989+ years	1989–2026
Exonerations in 2025	97 exonerations	2025
Exonerations in 2024	158 exonerations	2024
Average years lost (2025 cohort)	14.2 years/person	2025
Compensation paid to exonerees since 1989	$5.5 billion+	1989–2025

Caveat: Totals drift by design (older cases added continually): 3,842 live vs 3,767 through end-2025. Each figure keeps its own as-of date.

Contributing factors (2025 exonerations)

Source: National Registry — 2025 Annual Report ↗ · What went wrong in the cases exonerated in 2025.

Factor	% of 2025 exonerations	Year
Official misconduct	72%	2025
Perjury or false accusation	73%	2025
False or misleading forensic evidence	40%	2025
Inadequate legal defense	33%	2025
Mistaken witness identification	23%	2025
False confession	20%	2025

Caveat: Cases usually have MULTIPLE factors, so these do NOT sum to 100%. Single-year snapshot, not the all-time share.

Race & wrongful conviction

Source: National Registry — Race and Wrongful Convictions (2017) ↗ · Black Americans are 13% of the population but ~47% of exonerees.

Measure	Value	Year
Black share of all exonerees	47% of 1,900 exonerations	2016
Murder — Black vs white wrongful conviction	~7× more likely	2017
Drug crimes — Black vs white	~12× more likely	2017
Sexual assault — Black vs white	~3.5× more likely	2017
2025 exonerees who were Black	61% of 2025 exonerations	2025

Caveat: Multipliers are the Registry's own, each hedged 'about'; based on exonerations (a proxy for wrongful convictions, not a direct measure).

DNA & death-row exonerations

Source: Innocence Project + DPIC ↗

Measure	Value	Year
DNA exonerations (Innocence Project)	375 exonerations	1989–2020
DNA cases where the real perpetrator was later identified	180 (50%) of DNA exonerees	2018
Average time served, DNA exonerees	14 years	1989–2020
Death-row exonerations since 1973	202 people	1973–2026
Death-row exoneration ratio	1 per 8 executed	2026

Caveat: The Innocence Project's 375 DNA count is FROZEN at 2020 (they stopped tracking nationwide then) — the current nationwide DNA count is higher, at the Registry.

Group / mass exonerations (police & lab scandals)

Source: National Registry + Innocence Project ↗ · Single corrupt officers or tainted labs can produce hundreds—or tens of thousands—of exonerations.

Case	Value	Year
Groups Registry — total	49 cases / 37,356 defendants	2026
Chicago Sgt. Ronald Watts (planted drugs/weapons)	200+ people	2026
Chicago Det. Reynaldo Guevara	55 exonerations	2026
Houston officer Gerald Goines	38 exonerations	2026
Massachusetts — Annie Dookhan drug-lab scandal	21,587 cases vacated	2017

Caveat: The Groups Registry counts mass drug-lab dismissals, which is why its defendant total dwarfs the individual-exoneration count.

Police use of deadly force

People fatally shot by police, per year

Source: Washington Post — Fatal Force ↗ · Fatal police SHOOTINGS only (the most complete count of these).

Year	count	Year
2024 (record high)	1,175	2024
2023	1,164	2023
2022	1,096	2022
2021	1,050	2021
Cumulative 2015–2024	10,430	2015–2024

Caveat: Journalist-compiled (not government). Excludes non-shooting police killings + in-custody deaths. Database stopped adding new cases Jan 1, 2025 (covers 2015–2024); recent-year counts are revised upward as cases verify.

All police killings (any means), per year

Source: Mapping Police Violence (Campaign Zero) ↗ · Broader than shootings — includes restraint, vehicle, taser, beating, on/off duty.

Year	count	Year
2024 (deadliest on record)	1,365	2024
2023	1,329	2023
2022	1,176	2022

Caveat: Journalist-compiled; ~92% capture. Runs several hundred above WaPo's shootings-only count. 2024 (1,365) is the highest in the series since 2013.

Fatal police shootings of UNARMED people, per year

Source: Washington Post — Fatal Force ↗ · WaPo's strict 'unarmed' category (no weapon per available evidence).

Year	count	Year
2023 (strict 'unarmed')	51	2023
2024 (strict 'unarmed')	29	2024
Decade average (2015–2024)	~57 per year	2015–2024

Caveat: Definition-sensitive: 'unarmed' is kept separate from replica/toy, vehicle, undetermined, unknown. The COUNT is more reliable than any single share (a broader framing gives ~1 in 18).

Fatal police shooting rate, by race (per million/yr, 2015–2024)

Source: Washington Post — Fatal Force (via Statista) ↗ · Per-capita rate — the fair way to compare across groups of different size.

Race	per 1,000,000 per year	Year
Black	6.1	2015–2024
Hispanic	2.8	2015–2024
White	2.4	2015–2024

Caveat: WaPo fatal shootings, 2015–2024 average. Do NOT attribute to Mapping Police Violence (which reports a ~2.9× Black-vs-White multiplier, not per-million rates).

Fatal police shootings, by race (2023 counts)

Source: Washington Post — Fatal Force ↗ · 2023 — the cleanest recent single year for by-race counts.

Race	count	Year
White	502	2023
Black	250	2023
Hispanic	165	2023
Unknown/not recorded	208	2023

Caveat: Counts, not rates — compare to per-capita rates above. 2024 by-race counts are incomplete (many cases lacked recorded race at extraction), so 2023 is used.

Legal-intervention deaths (CDC), per year

Source: CDC NVDRS (MMWR Surveillance Summaries) ↗ · 'Legal intervention' = killed by law enforcement acting officially (ICD-10 Y35). The government mortality count.

Year	count	Year
2022	1,014	2022
2021	923	2021
2020	874	2020

Caveat: CDC NVDRS covers PARTICIPATING jurisdictions, not the full nation (2022 was the first year with all 50 states — part of the rise is expanded coverage). Not a lawfulness judgment. National CDC WONDER Y35 requires an interactive query.

Official federal counts undercount police killings

Source: BJS Arrest-Related Deaths assessment + The Lancet (2021) ↗ · Why the journalist databases exist.

Measure	Value	Year
BJS: police killings its Arrest-Related-Deaths program captured	~49%	2003–2011
Lancet: police-violence deaths NOT reported by NVSS	55.5%	1980–2018
FBI 'justifiable homicide by law enforcement'	340 count	2019

Caveat: Government counts historically catch only about half of police killings. FBI 'justifiable homicide' counts (444 in 2014, 340 in 2019) are voluntary and far below the ~1,000/yr the journalist + corrected estimates show.

Hate crime (racially/bias-motivated)

Hate-crime incidents per year (FBI / DOJ)

Source: FBI Hate Crime Statistics / DOJ ↗ · Reported hate-crime incidents by year.

Year	incidents	Year
2018	7,120	2018
2019	7,314	2019
2020	8,263	2020
2021	10,840	2021
2022	11,634	2022
2023	11,862	2023
2024	11,679	2024

Caveat: 2021 is a break: the mandatory NIBRS-only transition dropped big agencies (NYPD/LAPD) — originally 7,262, revised to 10,840 in the supplemental set. Use 10,840 for trend continuity.

Hate crime by bias motivation (2023)

Source: FBI Crime Data Explorer ↗ · Which bias motivated the incident.

Bias category	incidents	Year
Race / ethnicity / ancestry	6,468 (51%)	2023
Religion	2,962 (23%)	2023
Sexual orientation	2,371 (19%)	2023
Gender identity	548 (4.3%)	2023
Disability	213 (1.7%)	2023
Gender	132 (1.0%)	2023

Caveat: Incident-level bias counts double-count multiple-bias incidents, so they sum higher than the 11,862 total. Shares shown are of the bias sum.

Racial/ethnic-bias hate crime by targeted group (2023)

Source: FBI Crime Data Explorer ↗ · Within the race/ethnicity/ancestry category (6,468 incidents), by targeted group.

Bias	incidents	Year
Anti-Black or African American	3,303	2023
Anti-White	911	2023
Anti-Hispanic or Latino	877	2023
Anti-Asian	445	2023
Anti-Multiple Races, Group	228	2023
Anti-Arab	154	2023
Anti-American Indian or Alaska Native	109	2023
Anti-Native Hawaiian / Pacific Islander	17	2023

Caveat: Counts the victim group targeted, not the offender's group.

Anti-Black & anti-Asian hate crime trend (the 2020–21 surge)

Source: FBI Crime Data Explorer ↗ · Multi-year, from the CDE live API — captures the anti-Asian spike.

Bias — year	incidents	Year
Anti-Asian — 2018	151	2018
Anti-Asian — 2020	358	2020
Anti-Asian — 2021 (peak)	781	2021
Anti-Asian — 2023	445	2023
Anti-Black — 2018	1,998	2018
Anti-Black — 2020 (peak)	3,689	2020
Anti-Black — 2023	3,303	2023
Anti-Black — 2024	3,244	2024

Caveat: Anti-Asian peaked in 2021 (781) then fell; anti-Black peaked in 2020 (3,689).

Hate crime by offense type (2023)

Source: FBI Crime Data Explorer ↗ · Most hate crimes are intimidation, vandalism, or simple assault — not homicide.

Offense	offense instances	Year
Intimidation	4,864 (33%)	2023
Destruction / damage / vandalism	3,580 (24%)	2023
Simple assault	3,151 (21%)	2023
Aggravated assault	1,656 (11%)	2023
Robbery	176 (1.2%)	2023
Arson	69 (0.5%)	2023
Murder & nonnegligent manslaughter	21	2023
Rape	24	2023

Caveat: CDE offense-instance counts (sum ~14,908) exceed the DOJ headline 13,829; shares are of the CDE breakdown.

Known hate-crime offender race (2023)

Source: FBI Crime Data Explorer ↗ · Of offenders whose race was known (8,475 of 10,595 records).

Offender race	offenders	Year
White	5,548 (66% of known)	2023
Black or African American	2,027 (24% of known)	2023
Multiple races	569 (7% of known)	2023
Asian	185 (2% of known)	2023
Unknown race	2,120 (20% of all)	2023

Caveat: 20% of offender records are 'Unknown race'; %-of-known are computed from CDE counts, not FBI-published percentages.

Religious-bias hate crime by group (2023)

Source: FBI Crime Data Explorer ↗ · Within the religion category (2,962 incidents).

Bias	incidents	Year
Anti-Jewish	2,004	2023
Anti-Islamic (Muslim)	284	2023
Anti-Sikh	164	2023
Anti-Catholic	80	2023
Anti-Eastern Orthodox	63	2023

Caveat: Anti-Jewish reached a record 2,004 (2023) / 2,092 (2024). DOJ's victims-of-single-bias count (1,832) differs — different unit, both primary.

Hate-crime VICTIMIZATION survey — includes crimes never reported to police (BJS NCVS)

Source: BJS National Crime Victimization Survey ↗ · The survey estimate is far larger than the FBI police-reported count because most hate crime is never reported (and much is victim-perceived).

Metric	Value	Year
Average annual hate-crime victimizations	246,900 per year	2005–2019
Single-year high	305,390 victimizations	2019
Not reported to police	54% of victimizations	2011–2015
Police-confirmed NCVS estimate	14,380 per year	2004–2015

Caveat: Latest survey covers 2005–2019 (no post-2019 bulletin exists). Most NCVS 'hate crimes' are victim-PERCEIVED, not police-confirmed — when limited to police-confirmed, the gap with the FBI nearly closes.

Hate-crime reporting coverage

Source: Congressional Research Service R46318 (Dec 2025) ↗ · How complete the FBI count is.

Metric	Value	Year
Agencies participating	16,419 (~85%) of UCR-enrolled	2024
Participating agencies reporting ZERO hate crimes	84% average	1996–2024
Agencies NOT participating at all	25–46% (avg ~33%) of agencies	1996–2024

Caveat: Voluntary reporting — most participating agencies report zero, and a third don't participate at all, so the true count is higher.

Hate-crime homicides per year (FBI)

Source: FBI Crime Data Explorer ↗ · Bias-motivated murders are rare but non-zero. The 2019 spike includes the El Paso Walmart anti-Latino mass shooting (23 killed).

Year	homicide offenses	Year
2018	23	2018
2019	52	2019
2020	31	2020
2021	27	2021
2022	34	2022
2023	21	2023

Caveat: FBI CDE counts OFFENSES; news often counts VICTIMS (2019 = 52 offenses / 51 victims; 2018 = 23 offenses / 24 victims).

Sexual & domestic violence

Sexual & intimate-partner violence — lifetime prevalence (CDC NISVS)

Source: CDC National Intimate Partner and Sexual Violence Survey ↗ · Share of the population who have EVER experienced this (self-reported survey).

Measure	Value	Year
Completed/attempted rape — women	More than 1 in 5 of women (lifetime)	2016/17
Contact sexual violence — women	Nearly 1 in 2 of women (lifetime)	2016/17
Contact sexual violence — men	More than 1 in 6 of men (lifetime)	2016/17
Intimate-partner violence — women	More than 1 in 3 (43.5M) of women (lifetime)	2016/17
Intimate-partner violence — men	More than 1 in 6 (20.7M) of men (lifetime)	2016/17
Severe physical IPV — women	1 in 4 (24.3%) of women (lifetime)	2010
Severe physical IPV — men	1 in 7 (13.8%) of men (lifetime)	2010

Caveat: Lifetime prevalence — do NOT mix with the NCVS annual rate below. The widely-cited 'severe physical IPV' figures are 1 in 4 women / 1 in 7 men (NOT '1 in 9 men', which is not a CDC figure).

Rape / sexual assault — annual rate & reporting (BJS NCVS)

Source: BJS Criminal Victimization 2023 ↗ · Past-year victimization rate (age 12+), a different measure from lifetime prevalence.

Metric	Value	Year
Rape/sexual-assault rate (age 12+)	1.7 per 1,000	2023
Reported to police	46% of rape/sexual assaults	2023

Caveat: Sexual assault is heavily under-reported; the reporting rate is volatile (46% in 2023 has a large margin of error ±8 pts; it was only 21% in 2022).

Drug overdose deaths

Drug-overdose deaths per year (CDC)

Source: CDC NCHS (Data Briefs 522 & 549) ↗ · Overdose deaths rose to a 2022 peak, then fell sharply — 2024 saw the largest single-year drop of the decade.

Year	deaths	Year
2021	106,699	2021
2022 (peak)	107,941	2022
2023	105,007	2023
2024	79,384	2024

Caveat: Final figures (supersede the higher provisional numbers that made news: provisional 2024 was 80,391 / 2023 110,037).

Overdose deaths involving fentanyl / synthetic opioids

Source: CDC NCHS Data Briefs 522 & 549 ↗ · Synthetic opioids (mostly fentanyl) drove both the surge and the recent decline.

Year	deaths	Year
2022	73,838	2022
2023	72,776	2023
2024	47,735	2024

Caveat: 'Synthetic opioids other than methadone' — fentanyl, analogs, tramadol. Deaths may involve multiple drugs.

Overdose death rate by race/ethnicity (2024)

Source: CDC NCHS Data Brief 549 ↗ · Age-adjusted rates, final 2024.

Group	per 100,000 (age-adjusted)	Year
American Indian / Alaska Native (highest)	51.6	2024
Black	33.8	2024
White	24.7	2024
Native Hawaiian / Pacific Islander	20.5	2024
Hispanic (any race)	17.0	2024
Asian (lowest)	4.4	2024

Caveat: NOT adjusted for death-certificate race misclassification — CDC says AI/AN rates are underestimated ~34%, so the true AI/AN burden is higher.

Crime by major city

Murder / homicide by major city (2024)

Source: City police departments (counts) + RIT / Wirepoints (rates) ↗ · 2024 saw historic declines — 8 of 9 cities below fell (NYC roughly flat). St. Louis has the highest per-capita rate; Chicago the highest total count.

City	Value	Year
St. Louis (highest rate)	150 murders — rate 54.4	2024
Memphis	242 murders (296 homicides) — rate 38.0	2024
Baltimore	201 homicides — rate 35.6	2024
New Orleans	124 murders — rate 34.7	2024
Chicago (highest count)	573 homicides — rate 21.5	2024
Philadelphia	269 homicides — rate 17.3	2024
Houston	322 homicides — rate 13.9	2024
Los Angeles	268 homicides — rate 7.1	2024
New York City	377 murders — rate 4.7	2024
National total (all cities/US)	16,935 murders	2024

Caveat: Counts are official PD data (HIGH confidence); rates are tracker-computed (PDs publish counts, not rates). 'Murder' (FBI: murder + non-negligent manslaughter) differs from 'homicide' — clearest in Memphis (296 homicide / 242 murder). 2024 counts are preliminary.

Gangs & organized crime

Gang & member counts (estimates — frozen at 2011–2012)

Source: National Gang Center + FBI/NGIC ↗ · The two 'headline' totals measure DIFFERENT populations and are not comparable.

Measure	Value	Year
Gangs (National Gang Center)	30,700 gangs	2012
Gang members (National Gang Center)	~850,000 members	2012
Gang members (FBI/NGIC — incl. prison + biker gangs)	~1,400,000 members	2011
Gangs (FBI/NGIC)	~33,000 gangs	2011
MS-13 members in the US	more than 10,000 members	recent

Caveat: ESTIMATES from law-enforcement surveys, not censuses. The National Youth Gang Survey was DISCONTINUED after 2012, so every 'current' figure traces to 2011–2012 data. NGC counts youth/street gangs; FBI/NGIC adds prison + outlaw-motorcycle gangs.

Gang-related homicides

Source: National Gang Center / NYGS ↗

Measure	Value	Year
Gang share of all US homicides	~13% (≈15% in 2012)	2012
Gang-related homicides (NYGS sample)	2,363 homicides	2012
Chicago & LA — share of homicides that are gang-related	~50%	2011–2012

Caveat: NYGS respondent sample, last year 2012; agency recording of 'gang-related' varies widely.

Cartels & drug trafficking

Dominant cartels (DEA National Drug Threat Assessment)

Source: DEA 2024/2025 NDTA ↗ · Two Mexican cartels — Sinaloa and CJNG — dominate the US drug supply.

Finding	Value	Year
Primary threat	Sinaloa Cartel + CJNG	2025
Sinaloa reach	Operates in ≥40 countries; almost all 50 US states	2025
CJNG reach	≥40 countries; almost all 50 US states	2025
Market control	Cartels present in every US state	2024

Caveat: DEA's own characterization.

Fentanyl seized at the border (CBP, by fiscal year)

Source: CBP Nationwide Drug Seizures ↗ · Rose to a FY2023 peak, then fell sharply.

Fiscal year	pounds	Year
FY2021	11,201	FY2021
FY2022	14,700	FY2022
FY2023 (peak)	27,023	FY2023
FY2024	21,889	FY2024
FY2025	12,027	FY2025

Caveat: ~85–90% by weight is seized AT ports of entry (CBP component data), not between them. FY2026 is partial.

Drug seizures by type (CBP, FY2025)

Source: CBP Nationwide Drug Seizures ↗ · By weight, meth and marijuana dwarf fentanyl.

Drug	pounds	Year
Marijuana	185,004	FY2025
Methamphetamine	170,119	FY2025
Cocaine	70,080	FY2025
Fentanyl	12,027	FY2025
Heroin	1,258	FY2025

Caveat: Seizures, not total flow; subject to revision.

Fentanyl lethality & sourcing (DEA)

Source: DEA ↗

Fact	Value	Year
Lethal dose	2 milligrams	2022
Fake pills with a deadly dose	7 of 10	2023
Lethality per kilogram	Up to 500,000 people	2024
Precursor sourcing	China → Mexico	2024

Caveat: DEA figures; the 'deadly doses' framing multiplies seized weight by the 2 mg threshold.

Cartels designated Foreign Terrorist Organizations (2025)

Source: US State Dept / Federal Register ↗ · First-ever FTO designation of drug cartels.

Item	Value	Year
Designated FTOs (8)	Sinaloa, CJNG, Tren de Aragua, MS-13, Carteles Unidos, Cartel del Noreste (Los Zetas), Gulf Cartel, La Nueva Familia Michoacana	2025
Effective date	February 20, 2025	2025

Caveat: Effective Feb 20, 2025, under EO 14157 (Jan 20, 2025).

Cartel drug proceeds (ESTIMATES)

Source: DOJ/NDIC + Treasury/FinCEN ↗ · No official figure of total cartel drug revenue exists — DEA quantifies it only as 'billions.'

Measure	Value	Year
Wholesale drug proceeds laundered (Mexican+Colombian DTOs)	$18–39 billion/yr	2009
Suspicious transactions flagged (Chinese laundering networks, 2020–24)	~$312 billion	2025
US currency DEA seized nationwide	~$280 million	2024

Caveat: Every figure here is an ESTIMATE or a seizure/flagged-activity actual — none is a confirmed total of cartel drug revenue.

Human trafficking

Human trafficking — National Hotline (2024)

Source: National Human Trafficking Hotline (Polaris) ↗ · Situations reported to the hotline — a floor, not a prevalence estimate.

Measure	Value	Year
Trafficking cases/situations identified	11,999 cases	2024
Victims identified	21,865 victims	2024
Sex-trafficking cases	6,647 cases	2024
Labor-trafficking cases	2,220 cases	2024
Minor victims	2,666 victims	2024

Caveat: The hotline's own disclaimer: these 'do not reflect the prevalence of human trafficking… only those who were aware of and chose to contact' it. Heavily under-identified. Type categories don't sum to total (rest are unspecified).

Human trafficking — federal enforcement

Source: FBI UCR + DOJ (via CRS R47211) ↗

Measure	Value	Year
FBI UCR trafficking incidents	3,224 incidents	2023
Offenders arrested (UCR)	658 offenders	2023
DOJ federal trafficking cases filed	181 (258 traffickers) cases	2023
DOJ federal convictions	289 traffickers	2023
People in forced labour WORLDWIDE (ILO estimate)	27,600,000 people	2021

Caveat: UCR trafficking data is incomplete (agency participation within states varies). DOJ (181 cases) and BJS (1,782 prosecuted) both cover FY2023 but use DIFFERENT statutory definitions — do not merge. The global 27.6M figure is an ILO estimate, NOT a US count.

Documented government involvement (official findings)

Government & drug trafficking — DOCUMENTED findings

Source: Senate + CIA/DOJ Inspectors General ↗ · From official primary sources ONLY — Congressional investigations, Inspector General / GAO audits, and court records. Each item is labeled DOCUMENTED (an affirmative official finding) or DISPUTED (examined but not confirmed, or news-derived). Online allegations and conspiracy claims are excluded.

Finding	Value	Year
Kerry Committee (US Senate, 1989)	Officials let the Contra war override drug enforcement; State Dept paid 4 trafficker-controlled companies from Contra-aid funds
CIA Inspector General (Hitz, 1998)	CIA was 'slow' and 'inconsistent' in cutting ties with Contra supporters alleged to traffic drugs
Operation Fast & Furious (DOJ IG, 2012)	ATF knowingly failed to interdict ~2,000 guns 'walked' to cartels; two were recovered at the murder scene of CBP Agent Brian Terry
Informant oversight (DOJ IG + GAO, 2015)	Inadequate oversight of informants' authorized 'Otherwise Illegal Activity'; 240+ long-term DEA sources reviewed <1 minute each

Caveat: Each is an official finding of tolerance/facilitation/oversight failure — NOT a finding that an agency ran drugs as policy (see the disputed table).

Government & drug trafficking — examined but NOT confirmed

Source: CIA/DOJ Inspectors General ↗ · What official bodies investigated and did NOT substantiate — included for honesty, so the documented findings aren't overread.

Claim	Value	Year
CIA drug-importation conspiracy / 'Dark Alliance' crack claims	Examined and NOT confirmed
Attorney General knowledge of Fast & Furious	Cleared
'House of Death' (Juárez informant murders)	No qualifying official finding

Caveat: OFFICIALLY DISPUTED / not established.

Federal agents convicted (drug-trafficking-related)

Source: DOJ prosecutions / court records ↗ · Individual corruption cases, fully adjudicated.

Person	Value	Year
Jose Irizarry — former DEA Special Agent	145 months (12+ years)
Chad Scott — former DEA Special Agent	160 months (13+ years)
Joseph Bongiovanni — retired DEA agent	60 months

Caveat: DOCUMENTED (court records).

Government & human trafficking — DOCUMENTED findings

Source: Senate PSI + HHS/DoD IG + GAO ↗ · From official primary sources ONLY — Congressional investigations, Inspector General / GAO audits, and court records. Each item is labeled DOCUMENTED (an affirmative official finding) or DISPUTED (examined but not confirmed, or news-derived). Online allegations and conspiracy claims are excluded.

Finding	Value	Year
HHS released children to a trafficking ring (Senate PSI, 2016)	HHS placed unaccompanied minors with a ring that forced them to work on Marion, Ohio egg farms
HHS lost contact with unaccompanied children (2018 testimony)	Whereabouts of 1,475 of 7,635 children could not be determined (Oct–Dec 2017)
Sponsor-screening gaps (HHS OIG, 2024)	16% of case files lacked documentation that required sponsor safety checks were done
DynCorp / Bosnia contractor trafficking (DoD IG, 2003)	DynCorp employees engaged in sex trafficking, incl. buying underage girls; NONE were prosecuted (jurisdictional gap)
Diplomat trafficking of domestic workers (GAO, 2008)	42 household workers alleged abuse by diplomats (2000–2008); 10 alleged trafficking incidents → 8 investigations; immunity blocked action

Caveat: Official findings of failures/gaps that enabled trafficking or abuse.

Government & human trafficking — labeled / disputed

Source: HHS + DOJ records ↗ · Widely-circulated claims that are NOT official findings — flagged so they aren't presented as fact.

Claim	Value	Year
'85,000 lost children'	NOT an official figure
Federal TSA/CBP officer convicted of trafficking	None found in the record

Caveat: DISPUTED / news-derived — do not assert as an official finding.

Traffickers convicted who held government positions (or the Marion, OH ring)

Source: DOJ prosecutions / court records ↗

Person	Value	Year
Aroldo Castillo-Serrano — Marion, OH ring leader	188 months (~15.7 years)
Larry Clay Jr. — police chief, West Virginia	25 years
Chukwuemeka Ekwonna — DC Metropolitan police officer	14 years
Xaver Boston — US Army reservist	40 years

Caveat: DOCUMENTED (court records).

Juvenile / youth crime

Juvenile arrests & long-term trend

Source: OJJDP Juvenile Arrests ↗ · Juvenile arrests fell ~74% from the 1996 peak — the fewest in ~40 years.

Measure	Value	Year
Total juvenile arrests	696,620 arrests	2019
Peak juvenile arrests	~2,700,000 arrests	1996
Decline from 1996 peak	-74%	1996–2019
Total juvenile arrests (COVID year)	424,300 arrests	2020

Caveat: Counts ARRESTS, not individuals or crimes — 'most appropriately a measure of entry into the justice system,' not the share of crime juveniles commit. 2020 reflects COVID. UCR→NIBRS break means post-2021 figures aren't comparable to earlier ones.

Juvenile share of arrests by offense (2022)

Source: BJS Crimes Involving Juveniles 1993–2022 ↗ · Juveniles (<18) as a share of all arrests for each offense.

Offense	% of arrests	Year
Violent crime	9.9%	2022
Property crime	9.3%	2022
Homicide	11.0%	2022
Robbery	19.9%	2022
Carjacking	32.9%	2022
Motor-vehicle theft	17.0%	2022

Caveat: NIBRS-based (2022 ≈77% population coverage); not comparable to earlier UCR-summary shares. Carjacking is a subtype of robbery.

Youth in residential placement

Source: OJJDP Census of Juveniles in Residential Placement ↗ · Youth confinement fell dramatically — down ~77% from the 1999 peak.

Measure	Value	Year
Youth in residential placement	24,894 youth	2021
Placement rate — White	72 per 100,000	2019
Placement rate — Black	315 per 100,000	2019
Placement rate — Hispanic	92 per 100,000	2019
Placement rate — American Indian	236 per 100,000	2019
Placement rate — Asian	19 per 100,000	2019

Caveat: By-race RATES are per 100,000 youth; racial categories exclude Hispanic ethnicity. Placement ≠ offending.

Juvenile victimization

Source: BJS Crimes Involving Juveniles 1993–2022 ↗

Measure	Value	Year
Violent victimization rate (ages 12–17)	27.4 per 1,000	2022
Homicide victims (ages 12–17)	1,410 victims	2022

Caveat: NCVS (ages 12–17). Homicide counts from FBI SHR/NIBRS.

Nationality & citizenship

Federal sentencing by citizenship (why this is NOT a national crime rate)

Source: U.S. Sentencing Commission, Quick Facts FY2025 ↗ · Federal docket only — a small slice of all U.S. prosecutions, dominated by immigration offenses.

Measure	share	Year
Non-U.S. citizens	44%	FY2025
Immigration offenses (of non-citizen cases)	79%	FY2025

Caveat: The 44% non-citizen share reflects federal jurisdiction over immigration: ~79% of non-citizen federal cases are immigration-guideline offenses (illegal reentry etc.) that ONLY a non-citizen can be convicted of. NOT a national crime-by-nationality rate. FY2024 was 34.7%; the jump reflects more illegal-reentry prosecutions, not necessarily more crime.

Texas convictions by immigration status — Cato vs. CIS (the dispute)

Source: Texas DPS data, analyzed by Cato Institute and Center for Immigration Studies ↗ · Texas is the only state linking convictions to DHS-checked immigration status. Two advocacy think tanks analyze the SAME data and disagree. Both computations are shown; the dispute is over how illegal-immigrant status is counted, and is not adjudicated here.

Group (source)	per 100,000	Year
Native-born (Cato)	3.0	2013–2022
Illegal immigrants (Cato)	2.2	2013–2022
Legal immigrants (Cato)	1.2	2013–2022
Illegal immigrants, adding prison-identified (CIS)	3.9	2012
Illegal immigrants, DHS-at-intake only (CIS)	2.7	2012

Caveat: Texas only; conviction (not arrest) rates. Cato (pro-immigration) vs. CIS (restrictionist). CIS raises the rate by adding people identified as illegal later, in prison; Cato calls that double-counting. CIS limits its 'higher' claim to homicide/sexual assault, not overall crime.

Nationwide incarceration rate, by immigration status

Source: Cato Institute (Briefing Paper 198), from 2024 ACS ↗ · The broadest national estimate of incarceration by status.

Group	per 100,000 (ages 18–54)	Year
Native-born	1,195	2024
Illegal immigrants	674	2024
Legal immigrants	303	2024

Caveat: Legal status is NOT observed in the ACS — illegal-immigrant status is statistically IMPUTED (residual method), so estimates are sensitive to assumptions. Incarceration stock, not arrests/convictions.

Non-citizens in FEDERAL prison (discontinued DOJ/DHS report)

Source: DOJ/DHS Alien Incarceration Report, FY2018 Q1 (discontinued after 2019) ↗ · Federal Bureau of Prisons custody only.

Measure	share	Year
Share of BOP inmates	~21%	FY2018 Q1
Share of confirmed aliens	94%	FY2017

Caveat: By the report's OWN statement, state prisons + local jails (~90% of the incarcerated) are excluded — so this is NOT a national measure. '21% known or suspected aliens' bundles ICE-confirmed with foreign-born still under investigation.

Texas — state detail

Texas statewide crime rate (FBI), 2023 → 2024

Source: FBI UCR / Crime Data Explorer ↗ · Texas rates fell across the board into 2024.

Metric	per 100,000	Year
Violent crime rate	394.7 → 389	2023–2024
Property crime rate	2,058.2 → 2,041	2023–2024
Murder rate	6.3 → 5.0	2023–2024

Caveat: FBI-published ESTIMATED state rates (2024 via USAFacts, rounded). CDE-derived rates from summing monthly NIBRS run ~5% higher. Texas DPS's own 'Crime in Texas' PDFs were unreachable; FBI CDE is the same data Texas submits.

Texas murder count by year (FBI CDE)

Source: FBI Crime Data Explorer ↗ · Statewide murder counts — peaked in 2022, falling since.

Year	murders	Year
2019	1,445	2019
2020	1,962	2020
2021	2,071	2021
2022 (peak)	2,100	2022
2023	1,895	2023
2024	1,656	2024

Caveat: Reported-agency NIBRS counts (high confidence).

Texas CONVICTION rate by immigration status (Cato, 2019 DPS data)

Source: Cato Institute (2019 Texas DPS data, IRPB #19) ↗ · Convictions per 100,000 of each population. Cato: illegal-immigrant rates fall below native-born on every offense; legal immigrants lowest.

Offense	per 100,000	Year
All crime	Illegal 749 · Legal 510 · Native-born 1,190	2019
Homicide	Illegal 2.2 · Legal 1.3 · Native-born 3.1	2019
Sex crimes	Illegal 21.1 · Legal 11.7 · Native-born 25.1	2019
Larceny	Illegal 25 · Legal 34.9 · Native-born 111.5	2019

Caveat: Cato's analysis of Texas DPS data. The Center for Immigration Studies DISPUTES the method (identification-lag undercount) — see the CIS row. Both views are shown; the dispute is unresolved. Cato publishes per-100k cells only for all-crime, homicide, sex crimes, and larceny — not robbery/assault/burglary/drug.

Texas ARREST rate by immigration status (Cato, 2019 DPS data)

Source: Cato Institute (2019 Texas DPS data, IRPB #19, Table 1) ↗ · Arrests per 100,000 of each population.

Offense	per 100,000	Year
All crime	Illegal 1,948 · Legal 1,434 · Native-born 2,896	2019
Homicide	Illegal 2.8 · Legal 2.3 · Native-born 5.4	2019
Sex crimes	Illegal 51.6 · Legal 34.4 · Native-born 56.4	2019
Larceny	Illegal 76.6 · Legal 95.4 · Native-born 224.0	2019

Texas homicide rate by immigration status, 2013–2022 (Cato)

Source: Cato Institute (2013–2022 Texas DPS data) ↗ · Cato's most recent, homicide-focused edition (pools 10 years).

Metric	per 100,000	Year
Homicide conviction (2013–2022 pooled)	Illegal 2.2 · Legal 1.2 · Native-born 3.0	2013–2022
Homicide conviction (2022 only)	Illegal 3.1 · Legal 1.8 · Native-born 4.9	2022
Homicide arrest (2013–2022 pooled)	Illegal 3.4 · Legal 2.6 · Native-born 6.2	2013–2022
All-crime conviction (2013–2022 pooled)	Illegal 685 · Legal 551 · Native-born 1,321	2013–2022

Caveat: Cato's analysis of Texas DPS data. The Center for Immigration Studies DISPUTES the method (identification-lag undercount) — see the CIS row. Both views are shown; the dispute is unresolved. Counts individuals (the 2015 edition counted events), so editions aren't perfectly comparable.

CIS counter-analysis — Texas illegal-immigrant homicide rate

Source: Center for Immigration Studies (2022) ↗ · CIS re-estimates the rate HIGHER by adding illegal immigrants identified later, in prison.

Metric	per 100,000	Year
Illegal-immigrant homicide rate (DHS + prison-identified)	3.9	2012
Illegal-immigrant homicide rate (DHS-at-arrest only)	2.7	2012
Texas overall homicide rate (comparison)	3.0	2012

Caveat: CIS's critique: Texas takes years to identify convicts as illegal immigrants, so Cato misclassifies many as native-born. Cato's rebuttal: CIS double-counts and shrinks the population denominator. The dispute is about WHEN status is recorded + which denominator — not a coding error either side denies.

Texas major-city murders (2023)

Source: City police departments ↗ · Latest full-year city figures.

City	murders	Year
Houston	348	2023
Dallas	246	2023
San Antonio	165	2023
Austin	75	2023
El Paso	33 (FBI) / 38 (EPPD)	2023
Fort Worth	rate 8.73 per 100k (count unverified)	2023

Caveat: Murder vs homicide definitions vary by city. El Paso has two counts (FBI 33 / EPPD 38, which counted 7 intoxication-manslaughter deaths as murder).

Texas hate crime

Source: DOJ / FBI — Texas state data ↗

Metric	Value	Year
Hate-crime incidents	521 incidents	2023
Hate-crime offenses	562 offenses	2023
Texas DPS own count	549 hate crimes	2022

Caveat: Incidents ≠ offenses ≠ victims. FBI's count undercounts because Texas NIBRS agency participation dropped after 2021. Texas DPS's own methodology counted 549 for 2022 (no verified DPS 2023 figure).

Texas incarceration / prison population

Source: TDCJ + BJS Prisoners 2023 ↗ · Texas had the largest incarcerated-population increase of any state from 2022 to 2023 (+9,600).

Measure	Value	Year
TDCJ incarcerated 'On Hand' (Aug 31)	129,653	2023
BJS prisoners under jurisdiction (Dec 31)	149,264	2023
BJS prisoners in custody	132,760	2023

Caveat: Different counting dates/definitions: TDCJ On-Hand (Aug 31) vs BJS jurisdiction (Dec 31). All are 'prison' — none is jail+prison combined. 'Cases under supervision' (601,433) includes parole and is NOT incarceration.

Florida — state detail

Florida statewide crime — FDLE UCR index (this series ENDS at 2021)

Source: FDLE Crime in Florida Abstract 2021 ↗ · Florida's traditional Total Index Crime series was discontinued after 2021.

Metric	Value	Year
Total index crime rate	1,952.3 per 100,000	2021
Violent crime rate	369.1 per 100,000	2021
Property crime rate	1,583.2 per 100,000	2021
Murder (offenses)	1,110 offenses	2021

Caveat: HARD DATA BREAK: FDLE stopped publishing the summary UCR index after 2021 (FBI ended summary collection). Do NOT trend-link these to the post-2021 NIBRS figures below — different measurement basis.

Florida post-2021 (FBI NIBRS — PARTIAL coverage)

Source: FBI Crime Data Explorer (Florida) ↗

Metric	offenses	Year
Homicide offenses (reported)	939	2022
Homicide offenses (reported)	813	2023

Caveat: Only ~82–84% of Florida's population is covered by reporting agencies — these are UNDERCOUNTS, not statewide totals, and not comparable across years (coverage differs) or to the pre-2021 FDLE index.

Florida major-city murders (2024)

Source: City police departments ↗

City	Value	Year
Jacksonville	59 murders	2024
Orlando	23 homicides	2024

Caveat: Tampa, Miami, and Miami-Dade full-year 2024 counts were not cleanly quotable from a primary source and are OMITTED rather than estimated.

Florida hate crime

Source: Florida AG — Hate Crimes in Florida 2023 ↗

Metric	Value	Year
Hate crimes reported	311 incidents	2023

Caveat: Leading bias: Race/Color 38%, Religion 32%, Sexual orientation 16%.

Florida prison population (FL DOC)

Source: Florida Dept of Corrections + FL EDR ↗

Date	inmates	Year
Inmate population	87,303	2024 (Jun 30)
Inmate population	89,087	2025 (Jun 30)

Caveat: Point-in-time inmate population (June 30 of each fiscal year).

California — state detail

California statewide crime (CA DOJ 'Crime in California')

Source: California DOJ / OpenJustice ↗ · CA DOJ reports 2025 as a record-low homicide rate. Rates falling since 2023.

Metric	Value	Year
Violent crime rate	431.1 per 100,000	2025
Property crime rate	1,785.8 per 100,000	2025
Homicides (reported)	1,374 homicides	2025
Homicide rate	3.5 per 100,000	2025
Homicide arrests	1,305 arrests	2024

Caveat: CA DOJ reports homicides two ways: reported-homicide COUNT (1,666 in 2024) vs homicide ARRESTS (1,305 in 2024) — different measures, don't conflate.

California major-city homicides (2024)

Source: City police departments ↗ · Most CA cities fell in 2024; Sacramento & Long Beach rose.

City	homicides	Year
Los Angeles	268	2024
Oakland	81	2024
Sacramento	45	2024
Long Beach	38	2024
San Francisco	35	2024
San Diego	35	2024
Fresno	30	2024
San Jose	29	2024

Caveat: Municipal PD year-end figures via official statements/press (medium confidence); some are near-final, not Dec-31 final.

California hate crime (CA DOJ)

Source: CA DOJ — Hate Crime in California 2024 ↗

Metric	Value	Year
Hate-crime events	2,023 events	2024
Anti-Black bias events	494 events	2024
Anti-Jewish bias events	310 events	2024

Caveat: 'Events' ≠ offenses ≠ victims. CA publishes a detailed annual report.

California property crime & retail theft

Source: PPIC (citing CA DOJ) ↗ · Overall property crime fell in 2024 while shoplifting kept rising.

Metric	Value	Year
Property crime rate	2,084 per 100,000	2024
Shoplifting trend	+14.2% year-over-year	2024
Motor-vehicle theft trend	-16.7% year-over-year	2024

Caveat: PPIC computes off CA DOJ data; differs slightly from CA DOJ's own rate (rounding/coverage).

California prison population (CDCR)

Source: CDCR Weekly Population Report ↗

Measure	Value	Year
Institution/camps in-custody population	87,941 inmates	2026 (Jul 15)

Caveat: Institution/Camps in-custody count. Do NOT use the 'Total CDCR Population' (~131,715) as the prison count — it bundles ~34K parolees.

Data quality — how reliable is the race data?

New Jersey — documented race-data failures

Source: NJ Office of the State Comptroller (2024) + coverage ↗ · New Jersey is where this question was raised. Here is what is officially documented vs. what is claimed but unverified.

Finding	Value	Year
NJSP records system failed to record race/ethnicity	2+ years	2021–2023
NJ DOC inmate-lookup 'White' listings (viral claim)	unverified	2024–2026

Caveat: The Comptroller's finding concerns traffic-stop data and describes blank/'unknown' race, not affirmative recoding of one race as another.

When race IS mis-recorded, which way does it go?

Source: Peer-reviewed studies ↗ · The direction of the error is mixed and context-dependent — some settings push ambiguous cases toward 'White', others toward 'Black'. Both are documented.

Setting	Value	Year
Medical examiner — homicides (New Mexico)	220% as likely to record 'White'	2023
Mixed-race Black infants recorded non-Black at death	35% of racially-discordant-parent Black infants	2023
…but perceived-race data pushes the OTHER way (toward Black)	4.4x homicide victims coded Black	2011–2012
Police-killing deaths hidden — disproportionately Black (mortality data)	~60% of Black police-killing deaths misclassified/unreported	1980–2018

How often is a Black person recorded as 'White'? (direct measures)

Source: Peer-reviewed record-linkage studies (self-ID vs. agency-recorded race) ↗ · Where researchers directly compare a person's self-identified race to what agencies recorded. Single-race Black is captured accurately in most systems; the measured Black-to-'White' rate is small — but not zero.

Population / source	Value	Year
Criminal defendants (agency vs. Census self-ID)	~2.7% of Black defendants	2000–2019
Prison inmates (agency vs. Census self-ID)	~0.95% of Black inmates	2000–2019
Veterans (VA records vs. self-report)	5.0% falsely coded another race	2002
Emergency-department records vs. self-report	98.7% match for Black (22.9% for multiracial)	2024
Death certificates vs. self-report	highly accurate for Black & White	1999–2011

Caveat: 'Not zero' is not 'systematic.' See the direction-of-error and dominant-miscoding tables below for context.

Other states — documented race-data problems

Source: State audits + investigative journalism ↗ · Beyond New Jersey, several states have documented race-recording failures. Note the pattern: where the direction is known, the miscoding runs Hispanic→'White' — Black drivers were coded Black correctly. We include a checked state that found NO miscoding, for contrast.

State / finding	Value	Year
Florida — Hispanic inmates coded non-Hispanic White	40,000+ inmates	2024
Texas — troopers record Hispanic drivers as White (DPS)	Hispanic→White 16M+ citations	2015–2016
Louisiana — Jefferson Parish codes Hispanic drivers White	0 of 167 'Lopez' coded Hispanic	2021
California (San Francisco) — SFPD reclassified Latinos as White	Hispanic→White	2012
Florida — reports NO Latino data at all	no Latino data arrests/prison/probation/parole	2016
Vermont & New Hampshire — the positive contrast	race + ethnicity distinct (arrest data)	2016
Illinois — checked, NO race-miscoding found	disparities only	2015–2023

The largest miscoding: Hispanic (and others) recorded as 'White'

Source: NBER/CJARS + others ↗ · The dominant, systematic distortion in 'White' crime counts is Hispanic (and Middle-Eastern/North-African, and AIAN) people being recorded as 'White' — which inflates the 'White' number.

Measure	Value	Year
Hispanic recorded 'White' — criminal defendants	9.16% of all defendants	2000–2019
Hispanic recorded 'White' — mugshot scrape (26 states)	29% of predicted-Hispanic	2025
Effect of correcting it (2018 prison counts vs. BJS)	+85k White / +80k Black inmates	2018

Missing race data — the reporting gaps

Source: FBI/BJS + The Marshall Project + Census ↗ · Even where race is recorded, large gaps make national race comparisons unreliable.

Gap	Value	Year
Agencies that did NOT report to the FBI (2021)	~40% of US agencies	2021
'White' includes Hispanic + (until 2024) MENA	upper-bound bucket	1997–2024
States that report ethnicity in arrest data	~15 of 50	2016–2024

Source audit — funding, reliability & criticism

Source audit — how much weight each finding carries

Source: profoundd audit (4 sourced agents, symmetric review) ↗ · We audited every source behind the race-data-reliability section: provenance, funding, method, the documented criticisms, and whether those criticisms hold up. Grades are ours; the funder and criticism rows below show the work. Symmetric standard — the Gates-funded studies and the anonymous blog were held to the same bar.

Source	Value	Year
NBER/CJARS race study (w32657)	MEDIUM	2024
Lancet/GBD police-killing study	MEDIUM-HIGH	2021
Miami Herald 'Hidden Bias'	HIGH	2024
ProPublica (Jefferson Parish, LA)	HIGH	2021
KXAN (Texas DPS)	HIGH	2015
VA veterans study (Boehmer 2002)	HIGH	2002
ER records study (Pettit 2024)	MEDIUM	2024
NM medical-examiner (Appel 2023)	MEDIUM-HIGH	2023
NCHS death-certificate validity (Arias 2016)	HIGH	2016
'White by Default' (Substack)	LOW	2025
Urban Institute (state reporting facts)	MEDIUM	2016
NJ State Comptroller audit (2024)	HIGH	2024
Census / OMB race-ethnicity standard (SPD-15)	HIGH	2024
Racial-fluidity studies (Saperstein/Penner; Noymer)	MEDIUM	2011-2012

Caveat: HIGH = primary/authoritative + corroborated; MEDIUM-HIGH = peer-reviewed, minor precision caveats; MEDIUM = solid direction, provisional magnitudes; LOW = do not rely on the specific numbers.

Source audit — funding & who paid

Source: Funding disclosures / COI statements ↗ · Who paid for each source, from its own funding/conflict-of-interest disclosure. Disclosed funding is not disqualifying — it is context. Note that Gates money sits behind both heavy quantitative sources (IHME/GBD directly, and the CJARS data platform).

Source	Value	Year
NBER/CJARS race study (w32657)	NSF + Arnold Ventures + Gates + RWJF (via the CJARS platform)	2024
Lancet/GBD police-killing study	Bill & Melinda Gates Foundation + NIH (NIMHD, NHLBI)	2021
Miami Herald 'Hidden Bias'	McClatchy (owned by Chatham Asset Mgmt) — news org, left-center lean	2024
ProPublica (Jefferson Parish, LA)	Nonprofit; Sandler Foundation-founded; left-center lean	2021
KXAN (Texas DPS)	Nexstar Media — largest US local-TV owner, right-center lean	2015
VA veterans study (Boehmer 2002)	VA Health Services R&D (grant IIR 98.161)	2002
ER records study (Pettit 2024)	Not verified (paywalled; no grants listed in Europe PMC)	2024
NM medical-examiner (Appel 2023)	National Institute of Justice (DOJ/OJP)	2023
NCHS death-certificate validity (Arias 2016)	Federal (CDC/NCHS + U.S. Census Bureau)	2016
'White by Default' (Substack)	Undisclosed — anonymous pseudonymous author	2025
Urban Institute (state reporting facts)	Foundations (Ford, Gates, RWJF, MacArthur…) + federal contracts	2016
NJ State Comptroller audit (2024)	State of New Jersey (independent statutory watchdog)	2024
Census / OMB race-ethnicity standard (SPD-15)	U.S. federal government (OMB / Census Bureau)	2024
Racial-fluidity studies (Saperstein/Penner; Noymer)	Russell Sage Foundation (Study A); no funding (Study B)	2011-2012

Source audit — main documented criticism → does it hold up?

Source: Published critiques / rebuttals ↗ · The strongest documented criticism of each source, who makes it, and our read on whether it holds — judged on merits (a biased critic can still be right; a friendly source can still be wrong).

Source	Value	Year
NBER/CJARS race study (w32657)	CONTESTED	2024
Lancet/GBD police-killing study	CONTESTED	2021
Miami Herald 'Hidden Bias'	HOLDS	2024
ProPublica (Jefferson Parish, LA)	HOLDS	2021
KXAN (Texas DPS)	REFUTED	2015
VA veterans study (Boehmer 2002)	HOLDS	2002
ER records study (Pettit 2024)	CONTESTED	2024
NM medical-examiner (Appel 2023)	CONTESTED	2023
NCHS death-certificate validity (Arias 2016)	HOLDS	2016
'White by Default' (Substack)	HOLDS	2025
Urban Institute (state reporting facts)	CONTESTED	2016
NJ State Comptroller audit (2024)	REFUTED	2024
Census / OMB race-ethnicity standard (SPD-15)	CONTESTED	2024
Racial-fluidity studies (Saperstein/Penner; Noymer)	CONTESTED	2011-2012

Caveat: HOLDS = the criticism is valid (as a limit or a refutation); CONTESTED = partly valid, bounds precision; REFUTED = the criticism does not stand.

What each source covers

✓ = recorded · label = partial/qualified (e.g. victim's perception, victim-only) · — = not recorded.

Source	Race	Hispanic	Sex	Age	Nationality	Transparency
FBI Crime Data Explorer (UCR / NIBRS) FBI	✓	✓	✓	✓	—	high
NIBRS incident microdata FBI / NACJD (ICPSR)	✓	✓	✓	✓	—	high
UCR Arrest data (by race / ethnicity / sex / age) FBI	✓	✓	✓	✓	—	high
Supplementary Homicide Reports (SHR) FBI	✓	✓	✓	✓	—	high
FBI Hate Crime Statistics FBI	✓	✓	partial	partial	—	high
BJS National Crime Victimization Survey (NCVS) Bureau of Justice Statistics	perceived	perceived	perceived	perceived	—	high
BJS — Race and Ethnicity of Violent Crime Offenders and Arrestees, 2018 Bureau of Justice Statistics	✓	✓	—	—	—	high
BJS — Victims and offenders by race/Hispanic origin Bureau of Justice Statistics	perceived	perceived	—	—	—	high
CDC WONDER — homicide mortality CDC / National Vital Statistics System	victim	victim	victim	victim	—	high
BJS Prisoners / National Prisoner Statistics Bureau of Justice Statistics	✓	✓	✓	✓	partial	high
U.S. Sentencing Commission — federal offender citizenship U.S. Sentencing Commission	✓	✓	✓	✓	✓	high
Texas DPS — convictions by immigration status Texas Dept. of Public Safety (+ DHS status checks)	partial	partial	partial	partial	✓	medium
DOJ/DHS Alien Incarceration Reports + GAO DOJ / DHS / GAO	—	—	—	—	✓	medium
Peer-reviewed syntheses (immigration & crime) National Academies / academic journals	partial	partial	partial	partial	✓	medium
U.S. Census Bureau / ACS — population denominators U.S. Census Bureau	✓	✓	✓	✓	✓	high

Tier 1 — Police-reported (FBI UCR / NIBRS) (5)

FBI Crime Data Explorer (UCR / NIBRS) ↗ high

FBI · latest: 2024 final; 2025 preliminary

The FBI's public portal for police-reported offenses and arrests, national → state → agency. Bulk CSV downloads plus a read-only REST API.

Transparency: Open bulk microdata (NIBRS incident files 1991→present) + REST API (free api.data.gov key); methodology documented.

Caveat: Reporting is voluntary; national totals are partly imputed and carry no confidence intervals. The 2021 NIBRS-only transition dropped coverage to ~65% (NYPD/LAPD absent); after SRS was re-accepted, coverage returned to ~94% (2023) and ~96% (2024). Race is officer-perceived. No citizenship field.

Crime Data Explorer ↗ Bulk downloads ↗ Open API (GitHub) ↗

NIBRS incident microdata ↗ high

FBI / NACJD (ICPSR) · latest: annual (e.g. NIBRS 2021 = ICPSR 38690)

Incident-level records — the only source that cleanly crosses race × Hispanic origin × sex × exact age for offenders, victims, and arrestees, with weapon and circumstance.

Transparency: Full incident microdata + codebooks archived and downloadable at NACJD/ICPSR; documented.

Caveat: Only covers agencies that report in NIBRS format (coverage grew over time); no citizenship field ('resident status' means local jurisdiction, not immigration).

NACJD (crime data archive) ↗ Decoding FBI crime data (codebook guide) ↗

UCR Arrest data (by race / ethnicity / sex / age) ↗ high

FBI · latest: 2024 (SRS Table 43 near-universal through 2019)

Arrests by offense, broken out by race, Hispanic origin, sex, and age brackets.

Transparency: Published tables + bulk arrest master files at CDE; documented.

Caveat: Published SRS tables give age×sex and (separately) age×race — a full offense×race×sex×age crosstab exists ONLY in NIBRS microdata. Ethnicity is optional and under-reported in SRS, so race×ethnicity from SRS is unreliable.

CDE arrests ↗

Supplementary Homicide Reports (SHR) ↗ high

FBI · latest: 2023 (Homicide Victimization in the US, 2023)

One record per murder: victim AND offender age/sex/race/ethnicity, victim-offender relationship, weapon, and circumstance.

Transparency: Microdata at CDE (1985→), ICPSR, the Murder Accountability Project, and Jacob Kaplan's concatenated 1976–2024 file; documented.

Caveat: Voluntary — SHR totals run below the UCR and CDC homicide counts; Florida was absent 1988–2020. Offender race/relationship are unknown for unsolved cases (~30% of offender race missing), so offender analyses skew toward SOLVED cases. No 'AR-15' code — it falls under 'Rifle.'

CDE (SHR) ↗ Murder Accountability Project ↗

FBI Hate Crime Statistics ↗ high

FBI · latest: 2023 (11,862 incidents)

Bias-motivated incidents by category (race/ethnicity, religion, sexual orientation, disability, gender, gender identity), offense, and victim/offender counts.

Transparency: Master files downloadable 1991→present; documented.

Caveat: Severe undercount — ~84% of participating agencies report ZERO hate crimes and there is no imputation; the FBI's ~11k/yr compares with NCVS's ~250k/yr estimate (roughly 40× larger). Treat as a lower bound; year-over-year moves can be participation shifts.

CDE (Hate Crime) ↗

Tier 2 — Independent cross-checks (survey, mortality, corrections) (5)

BJS National Crime Victimization Survey (NCVS) ↗ high

Bureau of Justice Statistics · latest: Criminal Victimization, 2024

A ~240,000-person annual survey that captures crimes NOT reported to police — the key independent check on police data. Records the victim's PERCEIVED offender race, Hispanic origin, sex, and age.

Transparency: Public-use microdata at NACJD/ICPSR (series 95); methodology fully documented.

Caveat: Offender traits are the victim's perception, not verified identity, and are unknown in ~19% of violent incidents. A 2006 redesign and 2012 method change create trend breaks. Survey-based, so it has sampling error.

NCVS at BJS ↗ Criminal Victimization, 2024 ↗

BJS — Race and Ethnicity of Violent Crime Offenders and Arrestees, 2018 ↗ high

Bureau of Justice Statistics · latest: 2021 (report)

The model for honest triangulation: it compares FBI UCR arrests to NCVS victim-reported offender race to test whether arrest disparities reflect offending or bias. Found White and Black people arrested roughly proportionate to serious-violent-crime involvement.

Transparency: Method and both underlying datasets are public; the report shows its work.

Caveat: Limited to serious nonfatal violent crime; does not settle questions about lesser offenses, drug enforcement, or downstream stages.

Report (PDF) ↗

BJS — Victims and offenders by race/Hispanic origin ↗ high

Bureau of Justice Statistics · latest: 2024 (Violent Victimization by Race or Hispanic Origin, 2008–2021)

The victim×offender race/ethnicity pairing tables (including interracial breakdowns) — 'Race and Hispanic Origin of Victims and Offenders, 2012–15' and its 2008–2021 update.

Transparency: Published tables + the NCVS microdata behind them.

Caveat: Same NCVS caveats (perception; unknown-offender share). These are the tables most cited in the 'suppressed then republished' discussion — see the publication-history note.

Race and Hispanic Origin of Victims and Offenders, 2012–15 ↗ Violent Victimization by Race or Hispanic Origin, 2008–2021 ↗

CDC WONDER — homicide mortality ↗ high

CDC / National Vital Statistics System · latest: 2023 final; 2024 provisional

Homicide deaths from death certificates (medical-examiner/coroner based), by VICTIM race, Hispanic origin, sex, age, and state — fully independent of police reporting.

Transparency: Queryable online with documented ICD-10 codes (homicide = X85–Y09); reproducible.

Caveat: Victim demographics only — no offender data. Suppresses cells of 1–9 deaths. Counts run above the FBI's homicide totals (more complete).

CDC WONDER ↗

BJS Prisoners / National Prisoner Statistics ↗ high

Bureau of Justice Statistics · latest: Prisoners in 2023

The incarcerated population by race, Hispanic origin, sex, and age.

Transparency: Statistical tables + NPS microdata at ICPSR (series 886).

Caveat: Incarceration shares are NOT offending shares — they reflect arrest, charging, sentencing, and release decisions. A snapshot of stock, not flow.

Prisoners in 2023 ↗

Tier 3 — Nationality / citizenship (genuinely limited) (4)

U.S. Sentencing Commission — federal offender citizenship ↗ high

U.S. Sentencing Commission · latest: 2024 Sourcebook; annual datafiles

The main FEDERAL source that records offender citizenship (150+ nationalities), plus offense type and sentence — via annual datafiles and the Sourcebook of Federal Sentencing Statistics.

Transparency: Individual-offender datafiles + Sourcebook downloadable; documented.

Caveat: FEDERAL cases only, and heavily immigration offenses (~72% of non-citizen cases are immigration-guideline). The ~44% non-citizen share of federal sentencings therefore reflects federal jurisdiction over immigration — it is NOT a national 'crime by nationality' rate.

USSC research + datafiles ↗ Quick Facts: Non-U.S. Citizen Offenders ↗

Texas DPS — convictions by immigration status ↗ medium

Texas Dept. of Public Safety (+ DHS status checks) · latest: 2013–2022 (via PIA requests)

The ONLY state that links criminal convictions/arrests to DHS-verified immigration status — the single most-analyzed dataset on citizen vs non-citizen offending.

Transparency: Raw data obtained by Public Information Act request (not a standing open dataset). We present BOTH the Cato and CIS analyses and their methodological dispute — the model transparency case.

Caveat: Cato (Nowrasteh) finds illegal-immigrant conviction rates BELOW native-born (e.g. 2013–2022 homicide: 2.2 vs 3.0 per 100k); CIS argues this undercounts because status is identified retroactively. Read both. One state's pattern may not generalize.

Cato: Illegal Immigrant Murderers in Texas, 2013–2022 ↗ CIS: Misuse of Texas Data (rebuttal) ↗

DOJ/DHS Alien Incarceration Reports + GAO ↗ medium

DOJ / DHS / GAO · latest: Alien Incarceration Reports 2017–2019 (discontinued); GAO 2024

Quarterly federal reports (2017–2019, since discontinued) on non-citizens in Bureau of Prisons custody, plus GAO 'criminal alien' analyses.

Transparency: Reports are public but the underlying BOP counts are not routinely published; the quarterly series was discontinued.

Caveat: Covers only FEDERAL inmates — roughly 10% of everyone incarcerated (state prisons and local jails, ~90%, are excluded). Not a national measure.

GAO: Noncitizens in the U.S. (2024) ↗ BJS: Non-U.S. Citizens in the Federal CJ System, 1998–2018 ↗

Peer-reviewed syntheses (immigration & crime) ↗ medium

National Academies / academic journals · latest: 2015–2020

The academic literature that works around the data gaps — the National Academies' 2015 review, Light/Massoglia/King (2014, ASR) on USSC data, and the PNAS (2020) Texas comparison.

Transparency: Methods peer-reviewed and documented, but these are analyses, not primary datasets — read them alongside the raw sources above.

Caveat: Findings converge that immigrants (esp. first-generation) have lower offending than native-born, but every study flags confounders (age/sex composition, enforcement patterns, generational effects). Note funding/framing when weighing think-tank work on any side.

PNAS (2020): undocumented, legal, native-born in Texas ↗ National Academies (2015): Integration of Immigrants ↗

Denominators — population counts you need for per-capita rates (1)

U.S. Census Bureau / ACS — population denominators ↗ high

U.S. Census Bureau · latest: 2020 census + annual estimates

Population by race, Hispanic origin, sex, age, and geography — the denominators you MUST divide by to turn crime counts into honest per-capita rates. ACS also gives citizenship/foreign-born counts.

Transparency: Fully open tables + API; documented; intercensal estimates published annually.

Caveat: Race/ethnicity categories must be matched to the crime source's categories or the rate is wrong. For unauthorized-immigrant rates you also need a modeled denominator (DHS/Pew estimates), which carries its own uncertainty.

data.census.gov ↗ Population estimates ↗

How we rate transparency

Each source is scored HIGH / MEDIUM / LOW on:

Open raw / microdata downloadable (not PDF-only tables)
Documented, reproducible methodology (codebooks, definitions, imputation disclosed)
Coverage / response rates / known gaps stated openly
Versioned + linked to the primary source
Independence — does it validate a measure it is separate from?

On “data that was suppressed and later republished”

Some crime-demographic tables were less available for a stretch and reappeared later — here is what the record actually shows, neutrally. BJS published detailed offender-race and interracial-victimization tables in its annual 'Criminal Victimization in the United States — Statistical Tables' compendium THROUGH its final 2008 edition. That entire compendium was discontinued after 2008 (not a targeted removal of race data), leaving a 2009–2016 gap in the standing tabulation; a 2012 change to how victims' perceived offender race is recorded also broke comparability. BJS then reinstated dedicated race reports in 2017, 2021, and 2024 — and the underlying NCVS microdata stayed publicly downloadable at NACJD/ICPSR the entire time, which cuts against a 'hidden data' reading. Separately on nationality: the DOJ/DHS quarterly 'Alien Incarceration Reports' were a 2017–2019 product (Trump executive order) that was discontinued afterward. Bottom line: availability shifted with format decisions, methodology changes, and administrations; a deliberate-suppression motive is asserted online but is not established by the record. We show the sources and let you judge.

Editorially curated, no paid placement. This page catalogs transparent sources and how to read them; it does not compute or endorse any particular figure. Data years and links change — verify at the primary source before you rely on a number. Missing a legitimate official dataset? Use the Submit form.