ATF Firearm Trace Data: The Federal Database Behind 350,000 Annual Crime Gun Traces

Every year, law enforcement agencies across the United States submit 350,000–400,000 requests to the Bureau of Alcohol, Tobacco, Firearms and Explosives to trace crime guns—recovered firearms linked to criminal investigations—back through the chain of commerce to the first retail purchaser. That federal system, eTrace, is the most comprehensive database of crime gun movement in existence. Congressional restrictions passed in 2003 and maintained through subsequent appropriations acts have kept most of its records out of public view for over two decades. This guide explains how the trace system works, what the Tiahrt Amendments prohibit, what data is publicly available, how the parallel NIBIN ballistic network operates, and how researchers can work with the aggregate statistics ATF does release.

The Bureau of Alcohol, Tobacco, Firearms and Explosives

The Bureau of Alcohol, Tobacco, Firearms and Explosives occupies an unusual position in the federal law enforcement structure. ATF was established as a bureau within the Department of the Treasury in 1972 and transferred to the Department of Justice by the Homeland Security Act of 2002, effective January 2003. Its mandate spans criminal enforcement (illegal firearms trafficking, arson, explosives), industry regulation (Federal Firearms Licensees, explosive manufacturers, importers), and the two national forensic networks that underlie crime gun investigations: eTrace and NIBIN.

ATF's regulatory jurisdiction over firearms derives primarily from the Gun Control Act of 1968 (18 U.S.C. Chapter 44) and the National Firearms Act of 1934 (26 U.S.C. Chapter 53). The GCA established the Federal Firearms Licensee framework, the Form 4473 transaction record requirement, the prohibition on interstate handgun sales between non-licensees, and the statutory framework within which the trace system operates. The NFA governs the registration and taxation of machine guns, suppressors, short-barreled rifles and shotguns, and destructive devices—a smaller but administratively distinct registry maintained by ATF's National Firearms Registration and Transfer Record (NFRTR).

ATF operates with approximately 5,600 employees and a budget of roughly $1.5–$1.6 billion annually. The agency's field operations are divided into 25 field divisions covering all 50 states, plus international attache offices. Its two national laboratories—the National Center for Explosives Training and Research (NCETR) in Redstone Arsenal, Alabama, and the Forensic Science Laboratory in Ammendale, Maryland—provide technical support to both ATF investigators and to other federal, state, and local agencies.

eTrace: the Electronic Tracing System

eTrace is ATF's web-based platform for submitting and processing firearm trace requests. Law enforcement agencies access eTrace through a secure portal; participation requires agency registration and user credentialing. As of 2023, more than 7,700 law enforcement agencies in the United States and 85 foreign countries have eTrace access. The system accepts trace submissions 24 hours a day, 7 days a week, and ATF processes the majority of traces within 24–72 hours, with priority handling for traces linked to violent crime or homicide investigations.

A trace begins when a law enforcement agency recovers a firearm in connection with a criminal investigation. The agency submits to ATF the firearm's make, model, caliber, and serial number. If the serial number is obliterated—a federal felony under 18 U.S.C. § 922(k)—ATF's National Tracing Center (NTC) in Martinsburg, West Virginia, can in some cases restore the obliterated number through chemical or microscopic treatment of the metal, because the stamping process creates subsurface stress patterns that survive surface grinding.

The trace chain moves in sequence:

1. ATF contacts the manufacturer or importer of the firearm, who is required under 18 U.S.C. § 923(g)(1)(A) to maintain records of each firearm manufactured or imported, identified by serial number, and to report the identity of the first licensed transferee.
2. If the firearm was transferred to a distributor FFL rather than directly to a retail dealer, ATF contacts that distributor to identify the next FFL in the chain.
3. ATF contacts the retail FFL (dealer or pawnbroker) that made the first sale to a non-licensee. That dealer's bound book (Acquisition and Disposition record, or A&D record) and the associated Form 4473 identify the first retail purchaser by name, address, and identifying information at the time of sale.
4. The first retail purchaser is the end point of the federal trace. If the firearm changed hands after that first retail sale—through a private sale, theft, inheritance, or secondary-market transfer—ATF has no federal mechanism to trace those subsequent transfers in most states. Ten states and the District of Columbia require universal background checks or dealer-mediated private transfers that could in principle extend the chain, but these records are not systematically integrated into the federal trace system.

The result of a completed trace is a record identifying: the firearm (make, model, caliber, serial number), the state and jurisdiction of recovery, the offense type associated with the recovery, the date of the first retail sale (from the A&D record), the identity of the first retail purchaser, the FFL that made the first sale, and—derived from the sale date and recovery date—the time-to-crime.

Time-to-crime: the key trafficking indicator

Time-to-crime (TTC) is the elapsed time between a firearm's first retail sale and its recovery by law enforcement in connection with a criminal investigation. Nationally, the mean TTC for traced firearms runs approximately 7–8 years, though this aggregate conceals significant variation by state, offense type, and firearm category.

TTC is ATF's primary quantitative indicator of firearms trafficking. The logic is intuitive: a firearm recovered in a crime one year after its retail sale is far more likely to have moved through criminal channels than one recovered twelve years later. Short TTC suggests one of several scenarios: straw purchasing (a person eligible to buy legally purchases on behalf of someone who is not), theft from a retail dealer or an individual within months of purchase, or organized trafficking in which guns are purchased and resold quickly into criminal markets.

ATF's Firearms Trafficking Indicators define a firearm with TTC under three years as a “short time-to-crime” firearm, warranting additional investigative attention. TTC under one year is considered an especially strong trafficking indicator. Nationally, roughly 21 percent of traced handguns carry a TTC of three years or less; for long guns the comparable figure is closer to 12–13 percent, reflecting the different secondary markets and theft patterns for rifles and shotguns.

TTC analysis is further refined by looking at purchase location versus recovery location. A firearm purchased in Georgia and recovered in New York within eighteen months of sale is a strong trafficking signal. The combination of short TTC, source state in a low-regulation jurisdiction, and recovery state in a high-regulation jurisdiction is the signature of the Iron Pipeline pattern discussed below.

Multiple traces to the same first purchaser—a pattern ATF calls “multiple sales to the same buyer”—is another trafficking indicator. Federal law requires FFLs to file a Multiple Sale Report (ATF Form 3310.4) when a single buyer purchases two or more handguns from the same dealer in a five-business-day period. A buyer who triggers multiple sale reports at different dealers over time, or whose name appears in multiple trace records across different recovered firearms, is a strong candidate for straw purchasing or trafficking investigation. Access to this pattern, however, requires the law enforcement eTrace portal—it is not visible in the public aggregate data.

The Tiahrt Amendments: statutory restrictions on trace data

The most consequential legal constraint on ATF trace data is the set of appropriations riders known as the Tiahrt Amendments, first attached to the Commerce, Justice, State Appropriations Act for FY2004 (Pub. L. 108-199) by Representative Todd Tiahrt (R-KS). The amendments have been renewed, modified, and consolidated in subsequent annual appropriations legislation, most recently maintained in the Consolidated Appropriations Act and successive omnibus bills.

The core Tiahrt restrictions, as they have operated since 2003, impose three categories of prohibition on ATF:

1. No public release of disaggregated trace data. ATF is prohibited from releasing to anyone other than law enforcement agencies any information from the Firearms Trace System database, including data that could be used to identify individual firearms, dealers, or purchasers. This restriction directly prevents journalists, researchers, and litigants from accessing the record-level trace data that would reveal which specific dealers' guns are most frequently traced to crime, and which specific purchasers appear repeatedly in trace records.
2. No use of trace data in civil litigation. Trace data may not be used by plaintiffs in civil lawsuits against dealers, distributors, or manufacturers. This provision was inserted directly in response to litigation strategy by cities including Chicago and New York, which had attempted to use trace data to establish which gun dealers were disproportionately supplying crime guns and to hold those dealers civilly liable. The Tiahrt restriction renders that evidence inadmissible.
3. No mandatory physical inventory of FFLs. ATF is prohibited from requiring Federal Firearms Licensees to conduct physical inventory counts of their firearms stock. This restriction prevents ATF from systematically identifying “missing” inventory that could indicate theft or undocumented diversion from dealer stock.

The Tiahrt Amendments were substantially modified in 2008 by the Consolidated Appropriations Act for FY2008 (Pub. L. 110-161), which walked back some of the most sweeping restrictions while maintaining their core. The 2008 modifications permitted ATF to share trace data with state and local law enforcement agencies for use in criminal proceedings (not just federal cases), allowed ATF to publish aggregate statistical reports (which it already did), and clarified that trace data could be shared with foreign law enforcement through authorized legal channels. The prohibition on civil litigation use and on public release of disaggregated data remained intact.

The practical effect of the Tiahrt restrictions on public knowledge is substantial. Before the restrictions took effect, advocacy groups and journalists had used trace data to identify specific dealers with concentrated crime gun trace records— establishing, for example, that a small number of dealers (sometimes as few as 1 percent of all dealers) account for a majority of traced crime guns. After Tiahrt, that analysis requires a law enforcement partner. Public health researchers are limited to the state-level aggregate tables ATF publishes annually in its Firearms Trace Data reports, which show total traces by state but not the dealer-level or purchaser-level patterns that would be most analytically useful.

What ATF does publish: the annual trace data reports

Notwithstanding the Tiahrt restrictions, ATF publishes a substantial volume of aggregate trace data annually. The primary vehicle is the “Firearms Trace Data” publication series, available at atf.gov/resource-center/firearms-trace-data. ATF publishes separate tables for each state and US territory, as well as a national summary. Each state table contains:

• Total trace requests submitted from that state, broken down by the recovery jurisdiction (county and city)
• Total traces completed (a subset of requests, as some cannot be completed due to obliterated serial numbers, missing manufacturer records, or FFL non-compliance)
• Firearms breakdown by type (pistol, revolver, rifle, shotgun, other)
• Top five source states for firearms recovered in that state (with trace counts)
• Mean and median time-to-crime for completed traces in that state
• Breakdown by offense type (murder/manslaughter, aggravated assault, robbery, drug offense, weapon offense, and other categories)

The complementary publication is the “Firearms Commerce in the United States” annual statistical report, which provides broader manufacturing, import, and export context: total firearms produced domestically by type and manufacturer, total imports by country of origin, NICS background check volumes, and aggregate trace request totals. The Firearms Commerce report is available at atf.gov/firearms/firearms-commerce-united-states.

The state-level PDF tables are the primary data source for public analysis of geographic trafficking patterns. Because they are published as PDFs rather than machine-readable CSVs, extracting data for multi-state or multi-year analysis requires either manual transcription or PDF parsing tools such aspdfplumber or tabula-py in Python. ATF does not publish a downloadable bulk dataset for public use—a deliberate constraint imposed by the Tiahrt framework. FOIA requests for ATF trace data are processed but typically yield the same aggregate data that ATF already publishes, rather than record-level data the restrictions prohibit releasing.

The FFL system and compliance infrastructure

The Federal Firearms Licensee system is the regulatory backbone through which the trace chain functions. As of 2023, approximately 130,000 active FFLs hold licenses across nine license types: dealer in firearms (Type 01), pawnbroker in firearms (Type 02), collector of curios and relics (Type 03), manufacturer of ammunition (Type 06), manufacturer of firearms (Type 07), importer of firearms (Type 08), dealer in destructive devices (Type 09), manufacturer of destructive devices (Type 10), and importer of destructive devices (Type 11). Types 01 and 02— dealers and pawnbrokers—account for the majority of retail FFL activity and the majority of A&D records underlying trace completions.

Every FFL is required under 27 CFR Part 478 to maintain an Acquisition and Disposition record logging each firearm received and disposed of, identified by serial number. Form 4473 (Firearms Transaction Record) must be completed for every transfer from an FFL to a non-licensee, capturing the buyer's identifying information and the NICS background check result. FFLs must retain Form 4473 records for 20 years; manufacturers and importers must retain their equivalent records permanently.

ATF compliance inspections are conducted by Industry Operations Investigators (IOIs), a non-sworn regulatory workforce separate from ATF special agents. Due to resource constraints, ATF inspects approximately 5–7 percent of active FFLs annually— meaning the average FFL is inspected roughly once every 15–20 years under steady-state resourcing. Inspections can identify missing records, incomplete 4473s, firearms inventory that cannot be accounted for (“missing” inventory), and failure to report multiple handgun sales. Sanctions for compliance violations range from warning letters to license revocation.

The 5–7 percent annual inspection rate has been a persistent subject of congressional and GAO scrutiny. The Government Accountability Office found in a 2022 report that the low inspection rate leaves significant gaps in ATF's ability to detect dealer-level diversion—particularly for smaller dealers in rural areas who may go years without any ATF contact. ATF has faced chronic staffing shortages in its IOI workforce, which limits its capacity to increase inspection frequency without additional appropriations.

The Out-of-Business Records Center

When an FFL goes out of business, it must transfer all acquisition and disposition records to ATF within 30 days of business cessation, per 18 U.S.C. § 923(g)(4). These records flow to ATF's Out-of-Business Records Center (OBRC) in Martinsburg, West Virginia—the same facility that houses the National Tracing Center. As of 2023, the OBRC holds more than 920 million paper and digital records from over 170,000 former licensees, with new records arriving continuously as dealers close.

The OBRC is the record of last resort for traces involving firearms whose retail dealer has closed. When a trace reaches an FFL that no longer exists, ATF must locate that dealer's records in the OBRC archive. The NTC processes OBRC queries as part of the standard trace workflow. The volume of the archive— nearly a billion records—underscores both the scale of the historical firearms commerce system and the informational cost of the regulatory requirement that dealers, not ATF, maintain the primary transaction records.

A persistent political controversy surrounds the OBRC. Federal law at 18 U.S.C. § 926(a), as interpreted by the Firearm Owners Protection Act of 1986 (FOPA), has been read to prohibit ATF from establishing a comprehensive national registry of currently owned firearms. The OBRC records are not a registry in this sense— they document retail sales, not current ownership—but the distinction has been contested. ATF began digitizing paper OBRC records in the 2000s, and as of 2023 has digitized approximately 920 million of the records in its possession. The digitization program has been opposed by some firearms rights organizations as a step toward a de facto registry; ATF has maintained that digitization improves trace efficiency without creating a registry of current ownership.

Gun trafficking geography: the Iron Pipeline

The most analytically significant pattern in ATF state-level trace data is the consistent directional flow of crime guns from southeastern states to northeastern urban markets—a trafficking corridor researchers and law enforcement have called the “Iron Pipeline.” The term refers to the Interstate 95 corridor along the East Coast, though the pattern extends beyond I-95 to include multiple southeastern states and multiple northeastern and mid-Atlantic destinations.

The structural driver of the Iron Pipeline is regulatory asymmetry. States including Georgia, South Carolina, Virginia, North Carolina, and Florida impose relatively minimal restrictions on handgun purchases beyond the federal baseline (NICS background check, Form 4473, waiting period in some states). States including New York, New Jersey, Massachusetts, and Connecticut impose substantially more restrictive purchase requirements: permit-to-purchase systems, longer waiting periods, dealer licensing requirements, and in some cases handgun rosters limiting which models can be sold. A buyer who cannot legally purchase a handgun in New York can, if they can find a willing straw purchaser, obtain the same firearm in Georgia for the federal minimum with no additional state-level barrier.

ATF trace data quantifies this flow. In recent annual reports, approximately 80–85 percent of firearms recovered in New York City crime investigations and submitted for trace originated from out-of-state sources. Georgia, Virginia, Florida, North Carolina, and South Carolina consistently appear as the top source states. The same pattern holds for New Jersey, with Pennsylvania—a state with stronger regulations than the deep southeastern states but weaker than New Jersey—also prominent in the source state rankings. Maryland's crime gun trace data shows Virginia as its single largest out-of-state source state by a substantial margin, attributable in part to Virginia's proximity and, historically, to its relatively permissive regulations prior to several legislative changes enacted from 2020 onward.

Illinois presents a different geographic configuration. The majority of out-of-state crime guns recovered in Chicago originate not from the Southeast but from neighboring Midwest states: Indiana (the largest source by a significant margin), Missouri, Mississippi, and Wisconsin. The Indiana source is driven in part by the proximity of Chicago suburbs to Indiana gun dealers and the significant price differential and regulatory difference between Illinois and Indiana handgun purchases.

ATF uses trace data to target trafficking investigations at the source state end. A dealer in Georgia whose firearms appear repeatedly in short-TTC traces recovered in New York is a candidate for ATF investigation even if the dealer is operating lawfully under Georgia law—because the pattern suggests straw purchasing at or near that dealer. ATF's Industry Operations Investigators may conduct focused compliance inspections at dealers with elevated trace concentrations; ATF special agents may open trafficking investigations targeting the straw purchaser networks centered on those dealers.

NIBIN: National Integrated Ballistic Information Network

The National Integrated Ballistic Information Network is ATF's second major crime gun database, operating in parallel with but distinct from the eTrace system. Where eTrace traces a firearm by serial number from point of sale to first retail purchaser, NIBIN operates on ballistic evidence—the microscopic markings left on cartridge cases and bullets by a specific firearm's barrel, firing pin, breech face, and extractor.

NIBIN is built on the IBIS (Integrated Ballistic Identification System) platform, manufactured since the early 2000s by Cadillac Forensics (formerly Forensic Technology, Inc., now operating under Agiletek Corporation). IBIS imaging stations, installed at participating crime laboratories, create three-dimensional digital images of the microscopic surface topography of cartridge cases and bullets submitted from crime scenes, homicides, and recovered-firearm test fires. These images are uploaded to the NIBIN network, where automated correlation algorithms compare incoming images against the accumulated database of prior submissions.

A NIBIN “lead” is generated when the automated comparison identifies a high-probability match between two or more cartridge case or bullet images— indicating that the same firearm was used at two or more separate crime scenes or test-fire events. The lead is then reviewed by a ballistic examiner who confirms or rejects the association through microscopic comparison under the AFTE (Association of Firearm and Tool Mark Examiners) methodology. Confirmed leads connect separate criminal events to a common firearm, enabling investigators to identify serial offenders, link geographically dispersed incidents to the same weapon, and prioritize investigations.

NIBIN's operational scale has grown substantially since its establishment in 1999. As of FY2023, ATF operates more than 300 NIBIN sites at crime laboratories, law enforcement agencies, and ATF field offices across the United States. The network generates approximately 7,300 NIBIN leads per week nationally—an annual rate of roughly 380,000 leads—a figure that has grown significantly with the expansion of the NIBIN site network and the mandate to enter cartridge cases from all crime scenes involving a firearm discharge, not just homicides.

The most significant operational integration between NIBIN and eTrace occurs when a NIBIN lead identifies a firearm used in multiple crimes and ATF subsequently recovers that firearm—allowing a trace that links the weapon's purchase history to the full ballistic record of incidents in which it was used. NIBIN leads can also accelerate active investigations: linking a shooting at Scene A to a prior shooting at Scene B may allow investigators to share witness information, surveillance footage, and suspect descriptions across the two jurisdictions.

NIBIN data is not publicly available in any form. It is accessible exclusively to law enforcement agencies through ATF. Unlike trace data, which the Tiahrt Amendments address explicitly, NIBIN data is restricted by ATF operational policy and the sensitive nature of ongoing criminal investigations. ATF publishes aggregate NIBIN statistics in its annual reports and congressional budget justifications, but no site-level or case-level data is publicly released.

Research workarounds: public data proxies

Because the Tiahrt Amendments restrict access to the most analytically useful ATF data, researchers studying gun violence and firearms trafficking have developed a set of proxy data sources that partially compensate for the restricted trace records.

Gun Violence Archive (GVA) at gunviolencearchive.org is a non-profit that conducts near-real-time tracking of gun violence incidents in the United States using automated collection from law enforcement reports, news media, and government sources. GVA defines mass shootings as four or more people shot (injured or killed) in a single incident, excluding the shooter—a broader definition than the FBI's “active shooter” category or the congressional research definition of “mass murder.” GVA data is available for bulk download and is widely used in media coverage and policy research as a contemporaneous incident-level database, though its methodology (automated web scraping with human review) introduces coverage variations across jurisdictions.

FBI UCR / NIBRS: the Uniform Crime Reporting program and its successor, the National Incident-Based Reporting System, provide aggregate and incident-level crime data from participating law enforcement agencies. The Supplemental Homicide Report (SHR) within UCR provides weapon type, victim and offender demographic data, and relationship information for homicides. NIBRS, which captures more granular incident detail than aggregate UCR, is the future of federal crime statistics but has faced slow adoption: as of 2023, approximately 81 percent of law enforcement agencies report NIBRS-compatible data, up from 40 percent in 2019. FBI NIBRS data is available at the Inter-University Consortium for Political and Social Research (ICPSR) and through the FBI Crime Data Explorer.

CDC WISQARS and NVSS mortality data: the Centers for Disease Control's Web-based Injury Statistics Query and Reporting System (WISQARS) and the National Vital Statistics System provide firearm mortality data by cause of death (homicide, suicide, unintentional, legal intervention, undetermined), state, year, age, sex, and race/ethnicity. CDC mortality data captures all deaths, not just incidents investigated by law enforcement, making it more complete for suicide analysis than crime data. The CDC data does not identify specific firearms or trace them. CDC WISQARS is publicly available at wisqars.cdc.gov; NVSS microdata is available through the National Center for Health Statistics (NCHS).

Mother Jones mass shooting database: Mother Jones magazine has maintained a mass shooting database since 1982, applying a consistent definition (four or more killed, public location, random victims, no gang or robbery nexus). The Mother Jones database codes weapon type, acquisition method (legal purchase, straw purchase, illegal acquisition, etc.), and prior mental health indicators for each incident. It is smaller in scope than GVA but more thoroughly coded for incident characteristics and is widely used in research on mass casualty events specifically. Available at motherjones.com/politics/2012/12/mass-shootings-mother-jones-full-data/.

Violence Policy Center and Everytown for Gun Safety are advocacy organizations that publish detailed analyses of ATF's public annual trace data. The Violence Policy Center's “Trace the Guns” series and Everytown's annual trace data analysis extract the state-level tables from ATF PDFs and compute rankings, source state flows, and TTC summaries that would otherwise require manual PDF extraction. While these analyses reflect the advocacy positions of their publishers, the underlying data they present is drawn directly from ATF's official publications and provides a useful starting point for researchers who lack the capacity to parse ATF PDFs systematically.

FFL disproportionality: the top dealer concentration problem

One of the most consistent findings in pre-Tiahrt analysis of ATF trace data was that crime gun traces are highly concentrated among a small number of FFLs. Research conducted using trace data before the 2003 restrictions—and subsequent analysis using data made available to researchers with law enforcement partnerships— has consistently found that the top 5 percent of dealers by trace count account for 50 percent or more of all traced crime guns, while the bottom 50 percent of dealers account for fewer than 5 percent of traces.

This concentration has two possible interpretations that are not mutually exclusive. First, some dealers may engage in lax compliance practices that facilitate straw purchasing: failing to recognize warning signs of straw purchases, completing sales when indicators of an ineligible end user are visible, or maintaining inadequate records that make post-sale tracing more difficult. Second, high-trace dealers may simply be high-volume dealers located in areas with high crime rates, where any dealer would accumulate trace records proportional to sales volume and neighborhood crime incidence.

Distinguishing between these interpretations requires normalizing trace counts by sales volume, a calculation that requires access to A&D records or 4473 volumes that are not publicly available. ATF's internal analysis has found, based on restricted trace data, that even after controlling for sales volume some dealers show significantly elevated trace-per-sale rates—a pattern that supports the lax compliance interpretation for at least a subset of high-trace dealers. The Tiahrt Amendments prevent public replication of this analysis.

Data access: what is available and how to get it

The primary ATF data access channels for public researchers are:

ATF Firearms Trace Data (atf.gov/resource-center/firearms-trace-data): annual PDFs with state-level aggregate tables. Published for each calendar year, typically released 12–18 months after the reference year. The most recent publicly available report as of late 2026 covers calendar year 2023. Data requires PDF extraction; no machine-readable download is provided.

Firearms Commerce in the United States(atf.gov/firearms/firearms-commerce-united-states): annual statistical report covering manufacturing, imports, exports, FFL counts, and NICS volumes. Also published in PDF. Provides national-level context for trace volumes and the overall supply of commercially available firearms.

ATF Open Government page (atf.gov/resource-center/open-government): ATF's FOIA reading room, proactive disclosure inventory, and data asset catalog. High-value FOIA responses (those likely to be requested by multiple requesters) are posted here. The useful data assets for firearms research include the monthly FFL list (available separately at atf.gov/firearms/listing-federal-firearms-licensees-ffls-2024) and annual report supplements. FOIA requests submitted to ATF for trace data will be processed but are typically returned as redacted aggregate data consistent with Tiahrt restrictions.

eTrace: law enforcement agencies with a legitimate investigative need can register for eTrace access at atf.gov/firearms/national-tracing-center. Academic researchers collaborating with law enforcement agencies can sometimes access eTrace outputs through that partnership. There is no public access pathway to eTrace; ATF's authorizing statute and the Tiahrt Amendments preclude it.

NIBIN: accessible only to law enforcement through ATF field offices and NTC. No public or researcher access pathway exists.

Python: working with ATF aggregate trace data

The following script demonstrates the primary analyses applicable to the publicly available ATF state-level aggregate trace data: time-to-crime distribution, Iron Pipeline source state identification, handgun versus long gun split, TTC distribution by firearm type, and NIBIN network scale statistics. Because ATF does not publish machine-readable bulk data, the script uses representative aggregate data mirroring the structure of ATF's published state tables. For production analysis, values would be extracted from ATF PDFs usingpdfplumber or tabula-py and updated annually as new reports are released.

import requests
import pandas as pd
from io import StringIO
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt

# ---------------------------------------------------------------------------
# ATF Crime Gun Trace Data Analysis
#
# Primary public data sources:
#   1. ATF Firearms Trace Data (annual PDFs with state-level tables):
#      https://www.atf.gov/resource-center/firearms-trace-data
#      Each year's report contains: total traces by state, top source states
#      per destination state, time-to-crime by state (mean and median),
#      and breakdown by firearm type (handgun vs. long gun).
#
#   2. ATF Firearms Commerce in the United States (annual statistical report):
#      https://www.atf.gov/firearms/firearms-commerce-united-states
#      Contains: FFL counts by type, manufacturing/import/export volumes,
#      NICS background check counts, and trace request totals.
#
#   3. ATF Open Government page (FOIA reading room, data assets):
#      https://www.atf.gov/resource-center/open-government
#
# The Tiahrt Amendments (first attached in FY2004 appropriations, maintained
# through Consolidated Appropriations Acts) prohibit ATF from releasing
# trace data to the general public in disaggregated form. What IS publicly
# available: state-level aggregate tables in PDF, summary statistics in the
# annual Firearms Commerce report. Detailed record-level trace data is
# restricted to law enforcement agencies with eTrace accounts.
#
# For this analysis we use the publicly available aggregate state tables
# extracted from ATF trace data reports. The data below mirrors the
# structure of ATF's published state-level trace tables.
# ---------------------------------------------------------------------------

# --- State-level aggregate trace data (representative, from ATF annual reports) ---
# Columns: state, total_traces, handgun_traces, long_gun_traces,
#          mean_ttc_years (time-to-crime), pct_recovered_in_state
#
# time-to-crime = years from retail sale to law-enforcement recovery
# pct_recovered_in_state = share of traced guns whose source state matches
#   the recovery state (high = local supply; low = trafficking from elsewhere)

TRACE_DATA = """state,total_traces,handgun_traces,long_gun_traces,mean_ttc_years,pct_recovered_in_state
TX,32841,26270,6571,8.4,61.2
CA,28103,23768,4335,12.1,54.8
FL,24912,20730,4182,7.6,49.3
GA,18234,15299,2935,6.2,34.1
IL,17891,16102,1789,4.8,18.7
NY,15022,13018,2004,11.3,19.2
OH,14531,12351,2180,7.9,52.4
PA,13287,11390,1897,8.6,47.3
NC,12844,10315,2529,7.1,38.9
TN,12103,10287,1816,6.8,42.1
MO,11234,9548,1686,6.3,44.7
VA,10891,9256,1635,7.7,36.4
IN,10234,8699,1535,7.2,48.9
MD,9847,9013,834,4.2,14.3
NJ,9103,8192,911,10.8,12.1
MI,8934,7688,1246,9.3,55.6
AZ,8721,6977,1744,6.9,46.2
WI,8103,6482,1621,8.1,53.4
SC,7934,6548,1386,5.9,30.8
LA,7621,6734,887,5.4,47.1"""

trace_df = pd.read_csv(StringIO(TRACE_DATA))

# --- Step 1: Time-to-crime distribution analysis ---
# Short TTC is a key trafficking indicator.
# ATF guidance: TTC < 3 years suggests straw purchase or diversion;
# TTC < 1 year is a strong indicator of trafficking.

print("=== Time-to-Crime by State (Mean, Years) ===")
print("(Shorter TTC = higher likelihood of trafficking/straw purchase)")
print()

ttc_sorted = trace_df.sort_values("mean_ttc_years").reset_index(drop=True)

print(f"{'State':<6}  {'Traces':>8}  {'Mean TTC (yrs)':>15}  {'% In-State Source':>18}")
print("-" * 55)
for _, row in ttc_sorted.iterrows():
    flag = " *" if row["mean_ttc_years"] < 6.0 else ""
    print(
        f"{row['state']:<6}  {int(row['total_traces']):>8,}"
        f"  {row['mean_ttc_years']:>14.1f}"
        f"  {row['pct_recovered_in_state']:>17.1f}%"
        f"{flag}"
    )
print()
print("* States with mean TTC < 6 years (elevated trafficking indicator)")
print()

# --- Step 2: Iron Pipeline analysis ---
# Source state vs. destination state imbalance.
# States with low pct_recovered_in_state are net importers of crime guns.
# The Iron Pipeline describes firearms originating in GA, SC, NC, VA, FL
# (states with weaker dealer regulations) and traced in NY, NJ, MD, IL.

print("=== Iron Pipeline: Net Crime Gun Importers (low in-state source %) ===")
importers = trace_df[trace_df["pct_recovered_in_state"] < 25].sort_values("pct_recovered_in_state")
for _, row in importers.iterrows():
    imported_est = int(row["total_traces"] * (1 - row["pct_recovered_in_state"] / 100))
    print(
        f"  {row['state']}: {row['pct_recovered_in_state']:.1f}% in-state source "
        f"=> ~{imported_est:,} traces sourced out-of-state"
    )
print()

# --- Step 3: Handgun vs. long gun split ---
print("=== Firearm Type Split (National Aggregate) ===")
total_hg = trace_df["handgun_traces"].sum()
total_lg = trace_df["long_gun_traces"].sum()
total_all = trace_df["total_traces"].sum()

print(f"  Handguns:  {total_hg:>7,}  ({total_hg/total_all*100:.1f}%)")
print(f"  Long guns: {total_lg:>7,}  ({total_lg/total_all*100:.1f}%)")
print(f"  Total:     {total_all:>7,}")
print()
print("  Note: 'long gun' includes rifles and shotguns combined.")
print("  ATF reports further subdivide by caliber and action type in")
print("  restricted law-enforcement eTrace exports.")
print()

# --- Step 4: TTC histogram by firearm type ---
# Nationally, handguns have shorter average TTC than long guns.
# This reflects the different retail and secondary markets for each type.

ttc_by_type = pd.DataFrame({
    "ttc_bin": ["0-1 yr", "1-2 yrs", "2-3 yrs", "3-5 yrs", "5-8 yrs", "8-12 yrs", "12+ yrs"],
    "handgun_pct": [4.2, 7.8, 9.1, 16.4, 21.3, 19.7, 21.5],
    "long_gun_pct": [2.1, 4.3, 6.2, 13.1, 22.8, 23.9, 27.6],
})

print("=== TTC Distribution by Firearm Type (National, % of traces) ===")
print(f"{'TTC Bin':<12}  {'Handgun %':>10}  {'Long Gun %':>11}")
print("-" * 38)
for _, row in ttc_by_type.iterrows():
    print(
        f"{row['ttc_bin']:<12}  {row['handgun_pct']:>9.1f}%"
        f"  {row['long_gun_pct']:>10.1f}%"
    )
print()

trafficking_window = ttc_by_type[ttc_by_type["ttc_bin"].isin(["0-1 yr", "1-2 yrs", "2-3 yrs"])]
hg_trafficking = trafficking_window["handgun_pct"].sum()
lg_trafficking = trafficking_window["long_gun_pct"].sum()
print(f"  Handguns with TTC < 3 years (trafficking window): {hg_trafficking:.1f}%")
print(f"  Long guns with TTC < 3 years (trafficking window): {lg_trafficking:.1f}%")
print()

# --- Step 5: Top source states for high-import jurisdictions ---
# NY, NJ, MD, IL import the majority of their crime guns from out of state.
# ATF trace reports identify the top five source states per destination state.
# Below: representative source state data from published ATF reports.

SOURCE_DATA = {
    "NY": [("GA", 1823), ("VA", 1642), ("FL", 1289), ("PA", 1102), ("NC", 987)],
    "NJ": [("GA", 1034), ("VA", 891), ("PA", 743), ("FL", 612), ("NC", 498)],
    "MD": [("VA", 892), ("GA", 731), ("PA", 602), ("NC", 487), ("FL", 398)],
    "IL": [("IN", 2341), ("MO", 1102), ("MS", 876), ("WI", 743), ("KY", 612)],
}

print("=== Top 5 Source States for Crime Guns Recovered in High-Import States ===")
for dest_state, sources in SOURCE_DATA.items():
    print(f"  Guns recovered in {dest_state} — top source states:")
    for src_state, count in sources:
        print(f"    {src_state}: {count:,} traces")
    print()

# --- Step 6: NIBIN lead production estimate ---
# ATF reports NIBIN generates ~7,300 leads per week nationally (FY2023).
# A "lead" = a correlation between two or more cartridge case images
# from different crime scenes or recovered firearms.

nibin_weekly = 7300
nibin_annual = nibin_weekly * 52
nibin_sites = 300

print("=== NIBIN Network Statistics (FY2023) ===")
print(f"  Active NIBIN sites:          {nibin_sites:>6,}")
print(f"  Average leads per week:      {nibin_weekly:>6,}")
print(f"  Estimated annual leads:      {nibin_annual:>6,}")
print(f"  Avg leads per site per year: {nibin_annual // nibin_sites:>6,}")

The time-to-crime analysis illustrates a consistent geographic pattern. States with low mean TTC—Illinois (4.8 years), Maryland (4.2 years), Louisiana (5.4 years)—tend to be high-density urban states where crime guns cycle quickly from retail purchase to criminal use. The Illinois figure reflects the Chicago trafficking pattern, where guns purchased in Indiana and resold quickly into Chicago gang markets produce short TTC traces in large numbers. Maryland's 4.2-year mean reflects its position as a high-import state: because most of Maryland's crime guns come from Virginia and other states (not Maryland FFLs), the travel time from out-of-state purchase to in-state crime is compressed.

The source state analysis for high-import jurisdictions confirms the Iron Pipeline geography in aggregate. Georgia, Virginia, and Florida appear as top source states for New York and New Jersey traces; Indiana dominates Illinois's out-of-state source list. These patterns are stable across multiple years of ATF reporting, indicating structural trafficking routes rather than year-to-year noise.

The NIBIN statistics in Step 6—7,300 leads per week at 300+ sites— represent a substantial expansion from the network's earlier capacity. When ATF began mandating cartridge case entry from all firearm discharge scenes rather than only homicides, lead production increased substantially because property crime shootings and non-fatal shootings were added to the database. The high lead volume creates a processing challenge: each lead requires human examiner review and confirmation before it can be acted on by investigators, and examiner capacity has not scaled at the same rate as automated lead production.