# Methodology and data dictionary

Version 1.2, published August 27, 2026 and updated August 27, 2026. Data through December 31, 2025.

## Scope

This report measures activity reported to the FBI Internet Crime Complaint Center. It does not estimate total fraud incidence, unique victims, offender location, household wealth, or unreported loss. The 50 states and Washington, D.C. are included in the downloadable master table. D.C. is excluded from 50-state correlations and income quintiles and is labelled separately.

## Methodology and data-source roles

The state ranking inputs are FBI IC3 complaint reports and reported adjusted loss for January 1 through December 31, 2025, divided by Census Vintage 2025 population estimates for July 1, 2025. The income analysis adds 2024 ACS median household income in 2024 inflation-adjusted dollars.

FTC Consumer Sentinel, the CFPB Consumer Complaint Database, and FinCEN SAR Filings by Industry are cited as authoritative but separate comparison systems. They are not merged into the IC3 rankings because they have different submitters, definitions, coverage, and update schedules. The analysis does not use a Q1 2025 through Q1 2026 window or 2026 state population estimates.

- FTC Consumer Sentinel Data Book: calendar year 2024, consumer reports from the FTC and data contributors; reports are unverified.
- CFPB Consumer Complaint Database: ongoing public product-complaint database; not a statistical sample and not a ranking input.
- FinCEN SAR Filings by Industry: January 1, 2014 through December 31, 2024; financial-institution filings, not consumer complaints or confirmed crimes.

## Calculations

- Complaint reports per 100,000 = IC3 complaint reports / Census 2025 population * 100,000.
- Reported losses per 100,000 = IC3 adjusted loss / Census 2025 population * 100,000.
- Average reported loss per complaint = adjusted loss / complaint reports. This is not a median.
- Pearson and Spearman correlations compare 2024 ACS median household income with 2025 state IC3 metrics for the 50 states. Spearman 95% confidence intervals use 10,000 deterministic paired bootstrap resamples. Leave-one-state-out Pearson ranges are reported as a sensitivity check. These analyses describe association only and cannot establish cause.
- Income quintiles contain 10 states each. Median household income measures income, not wealth.
- The fraud exposure matrix uses the 50-state medians for complaint reports per 100,000 and average reported loss per complaint as descriptive thresholds. D.C. is excluded and no composite risk score is calculated.
- Cryptocurrency is an overlapping IC3 descriptor. Cryptocurrency values are never added to investment or other crime-type totals.
- Cryptocurrency descriptor values are not presented as a share of total state loss. The FBI's published Oregon cryptocurrency descriptor loss exceeds its published overall Oregon state loss, demonstrating that the tables cannot be treated as mutually exclusive components without additional source clarification.
- FTC Consumer Sentinel, CFPB complaints, FinCEN SAR statistics, and FBI IC3 are separate systems and are never added together.

## How to report financial fraud

If money is moving, contact the bank, card issuer, credit union, payment app, or exchange first and ask whether the transaction can be stopped, recalled, or frozen. Preserve transaction IDs, account details, messages, screenshots, and case numbers.

- Identity theft: https://www.identitytheft.gov/
- Internet-enabled crime: https://complaint.ic3.gov/
- General scams and consumer fraud: https://reportfraud.ftc.gov/
- State consumer protection: https://www.usa.gov/state-attorney-general
- Problems with covered financial companies: https://www.consumerfinance.gov/complaint/

Filing creates a record but does not guarantee an individual investigation, recovery, or reimbursement. Do not pay an upfront fee to anyone promising guaranteed recovery.

## Principal results reproduced by the script

- Credit Card/Check Fraud complaint reports changed 45.8% from 2024 to 2025; reported loss changed 41.4%.
- Investment complaint reports changed 52.3%; reported loss changed 31.6%.
- Pearson r between state median household income and reported loss per 100,000 residents is 0.634 for the 50 states.
- Spearman rho for the same relationship is 0.745, with a 95% paired-bootstrap interval of 0.577 to 0.849.
- The leave-one-state-out Pearson range is 0.609 to 0.688.

## Review status

This publication was released without claimed approval from a Certified Fraud Examiner, payments-risk professional, attorney, cybersecurity specialist, or other external reviewer. Automated consistency checks and source reconciliation are not a substitute for human expert review. Corrections can be submitted through the FinanceFirst corrections process.

## Files

- `financefirst-financial-fraud-by-state-2026.csv`: 51-jurisdiction master table.
- `financefirst-financial-fraud-by-state-2026.json`: master table, national category and age series, quintiles, definitions, and correlations.
- `source-ledger.csv`: source provenance and limitations.
- `data-dictionary.csv`: field definitions and calculations.
- `state-rankings-and-newsroom-angles.csv`: 50-state ranks, exposure groups and publication-safe local angles.
- `claim-to-source-audit.csv`: headline claim provenance, calculations and required qualifications.
- `quotable-findings.csv`: concise findings with source IDs and wording safeguards.
- `corrections-and-update-log.csv`: public version and maintenance history.
- `CITATION.txt`, `CITATION.cff` and `financefirst-financial-fraud-2026.bib`: citation formats.
- `validation-report.json`: machine-readable quality-control results and software versions.