social_context_social_capital_county_year
County × year social-capital proxy: County Business Patterns establishment counts for NAICS 813 (religious, grantmaking, civic, professional and similar membership organizations) per 10,000 residents, 2023. This is the associational-density component — it counts organizations, a proxy for the civic infrastructure / social capital they represent, NOT a direct measure of trust or cohesion. Deliberately the raw component, NOT the Rupasingha-Goetz County Social Capital Index (a composite that double-counts voter turnout and bundles constructs estimated separately — keep R-G as a validation benchmark only). Counties with no NAICS 813 establishments are 0-filled (absence in CBP means zero, not suppression). associations_coverage flags sparse-count counties as LOW_PRECISION. Tier A (administrative business count).
overview
All counties with a 2023 PEP population are covered; counties with no NAICS 813 establishments are 0-filled. associations_coverage marks counties with <= 3 establishments as LOW_PRECISION (per-10k rate is volatile at low counts) and counties without a population denominator as UNSUPPORTED.
current vintage — CBP 2023
history — Single vintage exposed (2023). Raw CBP is loaded for 2019–2023; earlier association-rate years are blocked on a multi-year population denominator (PEP/ACS).
source & licensing
fields
| name | type | definition |
|---|---|---|
| associations_coverage | string | Per-cell coverage/precision flag (context/guides/coverage-flags.md). SUPPORTED = usable; LOW_PRECISION = <= 3 establishments, volatile per-10k; UNSUPPORTED = no population denominator. Filter on this for small-county work. |
| associations_establishments | int64 | Count of NAICS 813 (membership organization) establishments in the county (CBP 2023). 0 where the county has none. |
| associations_per_10k | float64 | NAICS 813 establishments per 10,000 residents — the social-capital / associational-density proxy. |
| country_idkey | string | ISO alpha-2 country code (always 'US' for domestic tables).Part of primary key. |
| county_idkey | string | 5-character FIPS code identifying the county.Part of primary key. Joins dim.counties on county_id. |
| population | int64 | County total population, 2023 PEP estimate (per-10k denominator). |
| state_idkey | string | 2-character FIPS code identifying the state.Part of primary key. Joins dim.states on state_id. |
| yearkey | int64 | Reference year of the observation.Part of primary key. |
joins
how to use this table
associations_establishments = sum of CBP establishment counts for NAICS 813 (3-digit subsector, matched as REGEXP_REPLACE(naics_code,'[^0-9]','') = '813') per county for 2023, 0-filled for counties absent from CBP. associations_per_10k = associations_establishments / PEP 2023 population * 10,000. Computed in dbt from raw.census_cbp_county_2023 and staging.census_pep_county_year.
A direct measure of trust, cohesion, or social connection — it counts organizations, not the connection they produce (it is a proxy). Precise counts in small counties (CBP noise-ranges and the per-10k rate is volatile where establishments are few — filter on associations_coverage). As a composite social-capital index (this is the associational-density component only; the Rupasingha-Goetz index is intentionally excluded under the no-composite rule).
NAICS 813 establishment counts can be noise-ranged for confidentiality in small counties, and the per-10k rate is volatile where organizations are few — associations_coverage flags counties with <= 3 establishments as LOW_PRECISION. Single 2023 vintage; annual rate panel blocked on a multi-year population denominator. Counts organizations, not membership or connection.