Skip to content

Pharmaceutical procurement is a boundary case, not a second headline

Intuition (plain-language)

Pharmaceuticals show even bigger numbers — welfare loss near 45% of the open price — but we deliberately refuse to headline them. The pharma estimates lean harder on modeling assumptions, and the welfare ranking between set-aside and open can flip depending on how the post-policy SME pool is modeled. So pharma earns its keep only as qualitative confirmation that the ordering points the same way, never as a load-bearing magnitude.

🟡 The structural decomposition and welfare arithmetic in pharmaceutical procurement produce larger point estimates than in standardized non-pharma — exclusion magnitudes higher, welfare loss ~44.8% of the open-regime price at λ=0.30 (AN-011). But the pharma estimates inherit additional model sensitivity that the non-pharma estimates do not, and the welfare ranking between the full set-aside and the open auction is sensitive to how the post-policy SME pool is modeled (AN-016). The non-pharma ranking is stable across the main and strict-invariance specifications; the pharma ranking is not. Pharma is therefore reported in the paper as a boundary case, not as a second headline.

The pharma reading is held to a different evidentiary standard: it is useful as qualitative confirmation (the dominance ordering points the same way) but not as quantitative claim (the headline magnitudes are not load-bearing). Anyone citing the pharma magnitudes should note the scope conditions of paper §4 (sec:pharma_scope).

Caveat. AN-016 now documents the four-margin diagnosis of why pharma is fragile: (i) thinner pre-policy SME pool (0.55 Pregão bidders/auction vs 0.94 in non-pharma); (ii) heavier turnover — 61.9% of post-policy pharma SME firms are new entrants (vs 23% in non-pharma); (iii) primitive-invariance failure — KS distance 0.141 after UH correction (vs passing in non-pharma); (iv) strict-invariance ranking flip — implied SME welfare weight drops to 0.7 (below unity) under strict invariance, flipping \(V_3 \succ V_0\) to \(V_0 \succ V_3\) in pharma but not non-pharma. Pharma is a qualitative confirmation of direction, not a headline-citable magnitude.

Sources.

  • Own analysis: AN-016 (pharma decomposition + sensitivity); AN-011 (pharma welfare arithmetic table); AN-017 (strict-invariance sensitivity).
  • Reports: none direct.
  • News anchors: none direct.
  • Cross-refs: H:static-welfare-loss-large; paper §4.5 (Pharma scope statement); paper §5 (Welfare ranking stability).
  • Validation: v7-jpube-tight/scripts/24_pharma.R and 33_pharma_flag.R for the pharma cell definition; 57_strict_invariance.R for the strict-invariance sensitivity.