Methodology
Every number on PropiTax is sourced from official government data. This page explains exactly where the data comes from, how we calculate effective rates, and what the limitations are.
Last updated: May 2026
1. Data Source
All property tax data is sourced from the U.S. Census Bureau American Community Survey (ACS) 5-year estimates for 2023 (released December 2024). The ACS is the most comprehensive annual source of U.S. housing and income data, based on approximately 3.5 million household interviews per year.
The specific Census variables used for each county are:
- B25077_001E — Median home value (owner-occupied units)
- B25103_001E — Median real estate taxes paid (owner-occupied units)
- B19013_001E — Median household income
- B01003_001E — Total population
Census API documentation: api.census.gov/data/2023/acs/acs5/variables.html
2. Effective Tax Rate Formula
The effective property tax rate for each county is calculated as:
Example — Harris County, TX: Median taxes paid = $4,382. Median home value = $255,000. $4,382 ÷ $255,000 = 1.72%
This is different from the nominal or mill rate listed on your tax bill. Nominal rates apply to assessed value, which may be a fraction of market value. The effective rate cuts through all of that — it represents the actual percentage of home value paid in taxes, making counties directly comparable.
3. State Aggregation (Population-Weighted)
State-level effective rates are not simple averages of county rates. We use a population-weighted formula:
Why weighted? A county with 4 million residents has far more impact on the state average than one with 50,000. A simple average would give equal weight to Loving County, TX (population ~64) and Harris County, TX (population 4.7 million) — producing a misleading state figure. The population-weighted formula reflects the actual experience of residents in that state.
4. Coverage
PropiTax covers 3,222 county-equivalents across 52 states and territories, including:
- 3,143 standard U.S. counties
- 38 Virginia independent cities (treated as county-equivalents by Census)
- Alaska boroughs and census areas
- Louisiana parishes
- Puerto Rico municipios
- Washington, D.C. (treated as both state and county)
Approximately 14 county-equivalents have insufficient Census data (primarily very small or newly formed jurisdictions) and are excluded from rankings but still appear in county lists with a "data unavailable" note.
5. Update Frequency
The Census Bureau releases new ACS 5-year estimates each December. PropiTax updates all county and state data within 30 days of each release.
Current dataset: ACS 2023 5-year estimates (released December 2024, reflecting survey data from 2019–2023). Next update expected: January 2026.
6. Limitations
PropiTax provides estimates based on historical data. Be aware of the following limitations:
- Effective rates are historical averages. They reflect past tax years, not a prediction of future bills. Tax rates change when local governments adjust levies or reassess property values.
- Sub-county taxing districts vary. School districts, municipal utility districts, and special assessment districts can add significant taxes on top of county-level rates. These sub-district rates are not included in our estimates.
- Exemptions are not yet applied. Homestead, senior, veteran, and disability exemptions can substantially reduce your actual tax bill. Exemption calculators are on our roadmap.
- ACS data has a margin of error. 5-year estimates smooth out year-to-year variation but are still subject to sampling uncertainty, particularly for small counties.
Always consult your county assessor's office for your exact assessed value and current tax bill.
7. Open Methodology
PropiTax is built by BAMEUR AGENCY. The data collection scripts are currently in a private repository (github.com/Bameuragency/propitax) with plans to open-source the Census data pipeline in 2026.
Questions about our methodology? contact@propitax.com