Data sources and scoring methodology
Page reviewed October 25, 2026. Data observation periods are shown separately.
Model public-data-v1-minmax. All observations use free public downloads; no paid data feed or per-visitor data request is needed for comparisons.
Four preferences
- Owners: lower historical median real estate taxes paid (ACS 2020–2024). Renters: lower median gross rent (same period). Neither predicts your bill.
- Higher active physician density (HRSA AHRF, 2023 numerator and population). This measures supply, not quality.
- Lower 2024 BEA all-items regional price index, US=100. 31 counties use their metropolitan area; Kerr County and Fayette County use the Texas nonmetropolitan portion because each is outside an MSA. That value covers a broad part of Texas, not a specific county. Housing is already included, so a separate housing priority adds emphasis.
- Fewer annual days at or above 100°F, NOAA station normals 1991–2020. This does not measure overall climate comfort.
Scoring and ties
Each factor uses min/max among available observations in the fixed 33-county dataset to produce a 0–100 relative score, reversing direction where lower is preferred. If all observations are equal, the factor scores 50. We multiply unrounded scores by weights 0–3 and divide by total weight. Overall scores display one decimal. Equal displayed totals share a rank; county ID determines their display order. Adding a county can change the range and existing scores even when their figures stay the same. Scores are not suitability grades or statistically significant differences.
Missing data
A county missing a positively weighted input is listed as insufficient data. We do not fill with zero, borrow another county’s value, or redistribute its weights. Ignored metrics are not required. Hospital ratings are context only for every county; missing ratings remain unavailable.
Station selection
Nearest Texas USC/USW station to Census county internal point within 50 km, S or R completeness and at least 20 years for temperature, >=100 F days and precipitation; blank/X measurement flags only. Haversine distance; station ID breaks ties. Project proxy rule, not a NOAA standard or county-wide average. Station distance and elevation can limit applicability. A county internal point is not a population center. Individual towns require their own geographic review. No storm score, humidity estimate or unsupported effective tax rate is used.
2025 property tax worked examples
The separate tax examples are not Finder ranking inputs or property-specific estimates. They use hypothetical $500,000 appraised values, general homestead and age-65 exemptions, and the named 2025 taxing-unit combinations for Austin/Travis and Fort Worth/Tarrant. For each unit, we subtract only that unit's applicable exemptions, floor taxable value at zero, multiply by its adopted rate per $100, round to cents, and sum the unit taxes. The final Austin city rate is $0.524017 per $100 after the November 2025 election, rather than the earlier conditional $0.574017 rate. The examples do not apply a certified school ceiling or transfer percentage, local non-school ceilings, appraisal caps, additional districts or assessments, other exemptions, prorations, or later tax-year changes. The Tarrant County adopted rates, final Austin city rate, and school ceiling explanation are primary sources. Check a parcel's actual taxing units, appraised value, exemptions, and certified ceilings with its appraisal district and tax office before relying on a tax figure.
Sources and reproducibility
- CENSUS: Official Census statistical tables and geography; government works. Published reuse guidance.
- BEA: BEA public-domain RPP observations; FRED is the retained retrieval mirror, not a blanket license for its other series. Published reuse guidance.
- CMS: Public hospital data; government works. Published reuse guidance.
- HRSA: Published AHRF county aggregate release: usage limitations none; does not cover separate commercial source products. Published reuse guidance.
- NOAA: Public NOAA-generated normals CSV downloads; cite dataset and subset; no paid custom/certification service. Published reuse guidance.
Download the observation dataset · Download the source policy. County profiles link the source observations and periods. Retained raw/extracted files are hashed in the project evidence archive.
ACS estimates retain margins of error; close estimates may not meaningfully differ. A relative margin over 30% is flagged for caution as a project rule, not a Census classification. Age does not change rankings or apply tax exemptions. Hospital averages exclude unrated facilities and do not establish travel access.