What the index measures
The name “US Rodent Infestation Index” describes this research project. Its actual measured variables are municipal rodent-related complaint and service-request totals and reports per 10,000 residents. It does not estimate the number of rats or mice, infestation prevalence, disease risk, or a property's condition.
The first edition covers New York City, Chicago and Boston because publicly accessible sources and city-population estimates can be aligned. It is not a census of US cities. No participating city is described as the country's most infested city.
Source definitions and coverage
New York uses the exact 311 complaint type “Rodent,” which is broader than rat sightings. Chicago uses “Rodent Baiting/Rat Complaint,” including baiting service requests. Boston uses the legacy 311 type “Rodent Activity.” A city may record inspections, treatment requests, duplicate contacts and rat sightings differently. These definitions are not interchangeable measures of infestation.
We use request creation dates, not completion dates. Source timestamps are interpreted as the municipal calendar date as supplied. No timezone conversion is used to move a report into another day. The import interval is half-open: the start is included, the end is excluded. In this edition, January 1, 2023 through August 31, 2026 is included. Only complete 2023–2025 calendar years are ranked.
Collection and cleaning
The importer downloads only configured fields. It checks required column names, paginates sources by stable record ID and reconciles record counts before and after pagination. Records outside the configured category or date range are excluded. The selected-field source responses are retained privately with retrieval timestamps and SHA-256 checksums. They are not sent to website visitors.
Within each city, repeated identical record IDs are counted once across sources. A conflicting duplicate ID blocks the import. Chicago records explicitly marked as duplicates by the city are also excluded. Distinct service-request IDs may still refer to the same address or rodent issue. This is not a deduplication of individual rodents, properties or underlying incidents.
Geographic handling and missing data
Reports with missing coordinates stay in city totals but are not plotted. Nonfinite or impossible coordinates are flagged for review. Coordinates outside the configured city bounding extent are omitted from the map but remain in the report total. This extent is a quality screen, not a municipal-boundary spatial join. We rely on the originating city's reporting system for city assignment.
NYC's named geographic grouping is borough. Chicago's grouping is community-area code. Boston supplies neighborhood. We do not relabel these as equivalent neighborhoods. ZIP values must be five digits. Missing ZIP or area values are explicitly counted. City maps group coordinates into 0.01-degree cells; these bins are not equal-area and are not individual property locations. Hotspots show totals, not risk. No ZIP, neighborhood or grid-cell rates are calculated without matching denominators.
Population normalization
Reports per 10,000 residents = reports ÷ city population × 10,000
Population estimates come from U.S. Census Bureau Vintage 2025 city and town files, summary level 162. The city boundaries and identifiers match New York city (state 36, place 51000), Chicago city (17, 14000) and Boston city (25, 07000). The estimate for each reporting year is used for full-year comparisons. Estimates are not exact population counts and future Census vintages may revise history.
Zero, missing or nonfinite population values block publication. Partial-year records are not annualized or placed in the annual leaderboard. We do not substitute metropolitan populations for municipal populations.
Trends, seasonality and comparisons
Year-over-year change is (current-year reports − prior-year reports) ÷ prior-year reports × 100. Missing or zero prior totals produce no percentage, not zero percent. Recent trends compare the latest 12 complete months in the snapshot with the previous 12 complete months. Peak months include every tied month. An incomplete current year is excluded from annual rankings and historical charts.
National seasonality and historical charts sum the same participating-city cohort. They are unweighted totals and are dominated by cities with more reports. Ranking defaults to reports per 10,000 residents; a user-selected total-report ranking is explicitly labeled. “Top” means within participating cities only. Three cities are shown in the first edition, not an invented top 15.
Boston's reporting-system transition
Boston's official transition guide retrieved September 29, 2026 listed Inspectional Services under the expected fall 2026 transition. This edition uses only the legacy Rodent Activity category and stops before September 2026. The city posted an August 12 warning about missing 2026 cases and marked the issue fixed on August 25. We do not combine the later category families or claim that apparent changes across a system migration are changes in rodents.
Boston's dataset and transition documentation. Later imports require a reviewed source configuration and coverage boundary; the importer will not silently extend this legacy category through a new system.
Validation and updates
The workflow is fetch, validate, deduplicate, normalize, calculate, flag anomalies, generate a candidate snapshot, review, rebuild and verify. A failed import leaves the existing approved snapshot untouched. The last-updated date is the source retrieval time, not the website build time.
Missing dates, missing required fields, truncated pages, changed pagination counts, invalid populations, conflicting duplicate IDs and source coverage gaps block the import. Impossible coordinates, greater-than-fivefold annual movements and greater-than-20% drops against a previously approved full-year count require manual review of a held candidate. Approval names the candidate's exact checksum, so it cannot silently approve a different import.
This is a manually reviewed publication, not a live incident feed. Source records may be revised retrospectively. No daily or monthly automatic publishing schedule is claimed. Weather is not used in this edition and no weather-causation claims are made.
Limitations
Reports depend on population density, municipal responsibilities, awareness of 311, reporting access, repeat reports, willingness to contact government, service capacity and category design. A campaign encouraging reports can increase the total even if the rodent population stays the same. Fewer reports do not establish fewer rodents. A per-capita rate adjusts for resident population only; it does not control for buildings, businesses, commuters, visitors or these reporting differences.
Reuse and licensing
Our original analysis and chart designs may be quoted or reproduced with attribution. This is not a relicensing of municipal source data. Each source retains its own terms. Download buttons are enabled only for source scopes whose redistribution has been reviewed. No followed link is required by embed code.
Boston identifies its dataset as Open Data Commons Public Domain Dedication and License. Chicago permits derivative analysis with the required disclaimer shown below. NYC downloads remain withheld if its applicable reuse terms are not verified. Census population estimates are U.S. government public-domain data. Basemap boundaries are Census data distributed through US Atlas (ISC package license).