The impact of pharmaceutical tendering on prices and market concentration in South Africa over a 14-year period

Abstract

OBJECTIVE: We investigated the South African tendering system for medicines to (a) evaluate its impact on prices and market concentration over a 14-year period and (b) analyze the accuracy of government forecasts of drug demand. METHODS: We calculated Herfindahl-Hirschman indexes to measure market concentration levels based on all pharmaceutical tender contracts issued by the South African government between 2003 and 2016 (n = 8701). We estimated price indexes to track changes in medicine costs over this period. We compared prices set through tenders in the public health care system to the corresponding prices in the private system. We also analyzed government data on procurement in selected drug classes to assess the accuracy of demand forecasts. FINDINGS: Between 2003 and 2016, the prices of medicines in most tender categories in the public health care system dropped by an average of around 40% or more. The prices of medicines procured for the public system through tenders were almost always lower than those sold in the private system. Tenders generally remained moderately to highly competitive over time (i.e., Herfindahl-Hirschman indexes < 2500), although the number of different firms winning contracts decreased in many categories. There were large discrepancies between the drug need estimates by the government and the quantities it went on to procure, with estimates off by more than 50% in most drug classes (9/16 observations). CONCLUSION: Tendering may be an effective measure to lower drug costs. Because most tenders remained competitive over time, price decreases may be durable. South African government officials should monitor the availability and prices of medicines to ensure continued access to affordable medicines for patients, as it may be undermined by the decreasing number of firms winning contracts over time. Given the large discrepancy between forecasts and procurements, the government would benefit from improving the accuracy of its demand forecasts.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 56 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.