Commodities Data

This dataset is built to show how, in the PEPFAR planning process, the responsibility for funding key commodities for the HIV response is divided among major funders. The vast majority of commodities are purchased by the national government, PEPFAR, or the Global Fund to Fight AIDS, Tuberculosis and Malaria. The share of these commodities paid for by each of these funders varies by country, commodity category, and by year.

The clearest picture of the commodity funding comes from PEPFAR annual planning documents, specifically the Strategic Direction Summary document. One of the standard tables in this document (usually Table 2.2.2) shows how responsibility for commodities is divided, which is a necessary step for the PEPFAR program to decide how its resources should be spent. The “PEPFAR Commodity Dataset” contains the information included in this table for all planned PEPFAR years back to COP15, and for each country where a usable commodity table was available. See below for more notes on methodology.

The underlying source of this data varies considerably. In some cases, this is backward looking, and may show what the division was a year or more before the planning meetings took place. In other cases the data is a result of detailed forecasting exercises and negotiations with national government partners and the Global Fund. Regardless of the source, we interpret this data as the PEPFAR program’s best understanding of the commodity needs at the time of planning. If it appears in the COP19 planning documents, our assumption is that is the planned division of resources in COP19.

This project was completed with help from colleagues at the Kaiser Family Foundation, for whose careful assistance we are deeply grateful.

The full dataset is available here.


Reporting is sometimes done as an overall dollar figure for the commodity, and then the specific dollar figures for which each funder is responsible. In other cases, the responsibility is listed in percentages. In some cases, the sum of the percentages does not equal 100%, and in others the sum of the funding amounts does not match the overall funding amount. We have made no changes to resolve these discrepancies, and they remain in the final data.

A special case of this is the data for Zimbabwe, which reports a total funding need, and the percent covered by each funding source. The sum of these funds is often less than 100%, and they explicitly report the gap.

In other cases, both a dollar figure and a percentage are included, and they may not match. In these cases, we have used the dollar figure listed in the table.

Finally, we made one edit to the data in the tables, where it was clear a typo had been made. The total expenditure for ARVs in Democratic Republic of the Congo in 2015 is listed as “8,521,4828”. By comparing to ARV spending in adjacent years, we believe that $8.5 million is likely to be the correct value, so we removed the last digit, and the entered value for that cell is $8,521,482.

A more detailed list of specific decisions that have been made was kept and reviewed with partners. Please write to us at if you would like more information.