Skip to content

Static welfare cost of full set-aside is ~28.9% of open-regime price (non-pharma)

Intuition (plain-language)

The set-aside costs society in two ways at once. The contract goes to a higher-cost supplier (real resource waste), and the taxpayer foots a larger bill financed through distortionary taxes. Together the static loss is about 29% of what the item would cost under open bidding — roughly R$38–89 million a year on this one product group. For the policy to still be worth it, a planner would have to value a real of SME producer surplus at about 2.42 reais of ordinary public money.

🟡 The full SME-only set-aside generates a static welfare loss of ~28.9% of the open-regime price \(p^{S_1}\) in standardized non-pharmaceutical procurement at λ=0.30 (AN-011). The loss combines real allocative waste (DWL\(_{\mathrm{alloc}}\) — the contract is assigned to a higher-cost supplier) with the fiscal distortion from financing the increased payment (\(\lambda \cdot \Delta_{\mathrm{gov}}\) at the Ballard-Shoven-Whalley benchmark). Translated to public-finance units on Group 65 alone, the welfare cost spans R\(38–89 million per year** (US\)11–25M) across the 30–70%** SME-eligible adherence range, against a Group-65 annual reference outlay of ~R\(345M (non-pharma). This is one product group of São Paulo's R\)13 billion procurement platform.

The implied SME welfare weight required for a planner to prefer the full set-aside over open auctions is ~2.42 in non-pharma — a planner must value SME producer surplus at $2.42 of welfare for the exclusionary regime to be rationalized as optimal (see Implied welfare weight).

Caveat. The welfare arithmetic inherits the maintained IPV-clock restriction (H:ipv-clock-admissible). The loss number is static — it holds entry and recovered willingness-to-supply primitives fixed; it does not solve a counterfactual legal regime in equilibrium. The pharma analog is 44.8% but more model-sensitive (AN-016); pharma is treated as a boundary case, not a second headline. The λ=0.30 baseline is the Ballard-Shoven-Whalley convention; the implied weight scales inversely with λ.

Sources.

  • Own analysis: AN-011 (welfare arithmetic table); supporting AN-016 (pharma boundary), AN-012 (preference benchmark comparison).
  • Reports: none direct.
  • News anchors: none direct.
  • Cross-refs: H:static-welfare-loss-large; H:implied-welfare-weight-large; paper §5 (Static welfare and policy design); docs/extensions.md Welfare cost panel.
  • Validation: v7-jpube-tight/scripts/55_welfare.R and 56_welfare_bootstrap.R produce the welfare-loss arithmetic; 57_welfare_adherence_sensitivity.R produces the R$38–89M annual range.