Country-specific data patches

The source Stata pipeline hardcodes three country-list branches inside n0_1_accounts_indicators_v2.do to work around idiosyncrasies of specific national data sources — these are data-source patches, not general econometric rules (see Analysis/stata.md section 8 and Pseudocode/n0_1_accounts_indicators_v2.md steps 6, 8, 17). They are kept here as one explicit, documented table instead of scattered if (country %in% c(...)) chains, so extending them as more countries join the project only requires editing this table.

Usage

country_patches

Format

A named list with three character vectors:

rescale_to_thousands

Countries whose raw financial variables are not in the package's EUR-thousand convention and must be divided by 1000: PT, ES, RO, SI, HU.

non_euro

Countries whose turnover needs converting to national currency via exchange_rates.dta (used to detect the "20-200-20" employment/turnover jump pattern in national-currency terms): DK, HU, RO, CA, HR.

aggregate_small_regions

Countries where TL2 regions with a thin count of HGET10 scalers get coarsened up to the TL1 level: AT, CA, HU, IT, NL, ES, RO.