Skip to content

The Paper

Bitter Pills to Swallow: The Enforcement Costs of Health Litigation

Authors: Darcio Genicolo-Martins and Paulo Furquim de Azevedo Affiliation: Insper Institute of Education and Research, Sao Paulo, Brazil Version: February 2026

Download

The latest version of the paper is available as a PDF in the manuscript directory of the replication package.

Contribution

This paper makes three main contributions to the literature on public procurement, judicial enforcement, and health policy:

  1. Quantifying the fiscal costs of judicial intervention in public procurement. While a growing literature studies health litigation in Brazil and other countries, the existing evidence focuses primarily on the legal and medical dimensions. This paper provides the first systematic estimate of how court-mandated purchases affect procurement prices, competition, and outcomes.

  2. Identifying the "under the gun" mechanism. By exploiting the institutional distinction between litigated and administrative urgent purchases---both of which bypass standard procurement timelines, but only the former expose officials to judicial sanctions---we isolate the causal effect of enforcement pressure on procurement costs. This "under the gun" effect accounts for the bulk of the observed price premium.

  3. Informing the design of judicial remedies. The findings suggest that courts should consider the procurement-side costs of their mandates and that institutional reforms reducing the sanction channel---such as centralized procurement for litigated items or pre-negotiated supply agreements---could substantially lower the fiscal burden of health litigation.

Institutional Background

Brazil's universal public health system (SUS) provides free access to pharmaceuticals listed in official formularies. When patients cannot obtain medications through regular channels, they may file lawsuits compelling the government to purchase and deliver specific drugs. These court orders create a distinct procurement channel---litigated purchases---that operates under tight judicial deadlines and the threat of sanctions (fines, contempt proceedings) against procurement officials.

The Sao Paulo state government conducts pharmaceutical procurement through its electronic platform (BEC), which records all transactions at the bid level. Purchases are classified into three types:

  • Ordinary purchases: Planned procurement following standard timelines and competitive bidding rules.
  • Administrative urgent purchases: Expedited procurement authorized by the administration itself (e.g., due to supply disruptions or inventory shortfalls), which may use simplified procedures but are not subject to judicial sanctions.
  • Litigated purchases: Procurement compelled by court orders, subject to judicial deadlines and sanctions for non-compliance.

Data and Sample

The analysis uses the BEC-G65-WORK1 dataset, a bid-level panel covering all pharmaceutical procurement transactions by the Sao Paulo State Department of Health (SES/SP) from January 2009 through December 2019.

Feature Description
Source Bolsa Eletronica de Compras (BEC), Sao Paulo state
Coverage All bids for Group 65 (medical, dental, and hospital supplies)
Period January 2009 -- December 2019
Observations 479,330 bids (full sample); 226,000 (items with both litigated and ordinary)
Unit of observation Bid (firm x item x procurement event)
Key outcomes Reference price, negotiated price, quantity, number of bidders, tender success
Treatment Purchase type: Ordinary (0), Administrative (1), Litigated (2)

Empirical Strategy

The identification strategy relies on comparing procurement outcomes across purchase types within narrowly defined cells. The preferred specification includes:

\[ Y_{ijpt} = \alpha + \beta \cdot \text{Litigated}_{ijpt} + \mathbf{X}_{ijpt}'\gamma + \mu_i + \delta_t + \phi_p + \varepsilon_{ijpt} \]

where:

  • \(Y_{ijpt}\) is the outcome (log price, log number of firms, or tender success) for item \(i\), buyer \(p\), at time \(t\)
  • \(\text{Litigated}_{ijpt}\) is an indicator for court-mandated procurement
  • \(\mu_i\) are item fixed effects (comparing the same product)
  • \(\delta_t\) are year fixed effects (absorbing time trends)
  • \(\phi_p\) are public buyer unit (PBU) fixed effects (comparing within the same procurement office)
  • Standard errors are clustered at the PBU level

Four Fixed-Effects Specifications

Spec Fixed Effects Purpose
(1) Item Baseline: within-product comparison
(2) Item + Year Controls for aggregate time trends
(3) Item + Year + PBU Preferred: controls for buyer heterogeneity
(4) Item + Year-Month + PBU Most saturated: monthly time effects

The "Under the Gun" Design

To isolate the effect of judicial sanctions from general urgency, we restrict the sample to urgent purchases only (administrative + litigated) and estimate:

\[ Y_{ijpt} = \alpha + \beta \cdot \text{Administrative}_{ijpt} + \mu_i + \delta_t + \phi_p + \varepsilon_{ijpt} \]

A negative coefficient on the administrative indicator means that litigated purchases are more expensive, isolating the sanction channel.

Software and Estimation

  • Primary analysis (v4): R 4.5 with fixest package (equivalent to Stata's reghdfe)
  • Legacy analysis (v2/v3): Stata/SE with reghdfe and ftools
  • Winsorization: 1st/99th percentiles (baseline); robustness at 0% and 5th/95th
  • Clustering: PBU (primary); item and two-way PBU x Item (robustness)