Four tables use selected-state distilled-liquor production proxies rather than direct lobbying exposure. Table 2 has unusually large blended-whiskey coefficients; Table 4 uses a linear probability model after reported collinearity prevented logit/probit. These are observational results with limited causal interpretation.
45
42
PDF page numbering
Checked against the complete supplied 45-page working paper; verifies source fidelity, not independent truth or replication.
No access
Evidence
Four tables use selected-state distilled-liquor production proxies rather than direct lobbying exposure. Table 2 has unusually large blended-whiskey coefficients; Table 4 uses a linear probability model after reported collinearity prevented logit/probit. These are observational results with limited causal interpretation.
Locator and verification
PDF pages 42–45. Paraphrase checked against supplied working paper; no independently verified causal claim.