This report analyzes how much money the U.S. government spent in 1998 on HIV projects in the developing world, which agencies administered these funds, how the agencies used the money, and where the U.S. was involved. To answer these questions, we analyzed budgetary and other data from USAID, NIH, and CDC. Section 8 (Other U.S. Agencies Involved in Developing Countries) gives a brief description of additional AIDS projects carried out by DOD -- from which we were unable to obtain sufficiently detailed program information -- and by other government agencies.
This report examines budget data from fiscal year 1998, the most recent year from which complete data were available. All the agencies and programs analyzed here operate on multiyear budget cycles. Some funds are carried over from year to year to meet ongoing obligations. This report did not analyze "carry-over" funds because they are expressed only in aggregate without detailed information on how they are used. We chose 1998 as our period of analysis because it immediately precedes the current expanded focus on international AIDS. The programs funded in 1998 represent the baseline international HIV infrastructure developed by the U.S. government.
The three agencies analyzed in this report have distinct roles in the world, and each operates in a different domestic political climate. The medical research budget at NIH has risen consistently over the last decade, the CDC budget has remained flat (although it is now rising), and USAID's budget experienced declines. Each agency reports budget information in a different format with varying clarity and consistency, and the two with the largest international HIV budgets, USAID and NIH, give most of their funding away to universities, non-governmental organizations (NGOs), or other government agencies in the form of contracts and grants. Synthesizing information like this into a coherent picture posed some challenges, required assumptions, and imposed limitations.
Because it is so difficult to obtain a clear and complete picture of the U.S. international AIDS program, we have tried to look at the available data in a number of different, interlocking, cross-cutting ways -- by funding agency, by contractor and grantee, by country and region, by program area, by amount of project funding when that information is available, and by number of discrete projects. Nonetheless, this report represents only a starting point for understanding the diverse AIDS activities carried out in developing countries with U.S. government funds.
The U.S. funds three kinds of programs: 1) intramural programs; 2) inter-agency programs; and 3) extramural programs. Intramural programs are funded and operated directly by U.S. government agencies (cf. Table 8). Examples of such programs include research studies conducted by NIH's intramural laboratories, the programs of USAID's missions, and all CDC activities. Inter-agency programs are funded by one U.S. agency, but operated by another agency or by the United Nations (cf. Table 8). Examples include USAID-funded programs in the Peace Corps and the Bureau of the Census, NIH-funded research projects conducted by the U.S. military, and the U.S. contribution to the United Nations AIDS Program, UNAIDS. Extramural programs are funded by a U.S. agency and carried out by a contractor or grantee, typically a U.S. university or NGO (cf. Table 7). USAID and NIH fund 48 contractors and grantees, all of which are based in the U.S. with three exceptions: the International AIDS Alliance, which is in London, and the University of West Indies and the Caribbean Epidemiology Center, both of which are in Trinidad.
The quality and form of the information varied by agency. Detailed USAID budget information is relatively inaccessible to the public. Upon request, USAID staff supplied in paper form budget data, including internal administrative allocations, reports and budgets from seven contractors, financial portfolio summaries, and other documents. We interviewed USAID staff, examined the agency web site, and reviewed the 1998 and 1999 congressional testimony of USAID officials. NIH staff provided in electronic form a description of 118 NIH grants, contracts, and internal programs that were coded by NIH staff as both "HIV-related" and "international." We interviewed NIH staff and analyzed budget reports for specific NIH programs. CDC publishes no distinct international budget, but CDC staff supplied in electronic form a detailed summary of its international programs.
Understanding USAID's budget presented the greatest challenge. USAID's allocation of HIV funds through its internal structure could be determined, and the amount it awarded in contracts could also be determined. However, the precise relationship between these two figures was never entirely clear. Internal USAID funds from all its administrative units were aggregated to support seven major global contracts, which were managed by the agency's global bureau. From the information available, we could determine how much HIV funding remained for internal USAID programs after the funding was aggregated for the contracts, but we could not determine how much remained at any specific administrative unit. The major limitation of this analysis is that the USAID budget, excluding global contracts, can be explained only in aggregate terms.
The NIH presented a different set of challenges. The Office of AIDS Research (OAR) at NIH supplied budget information on grants and contracts coded as "international" and "HIV-related." These two categories are not exclusive. A grant might be coded as 10% HIV-related and 30% international, and so on. This report analyzed a grant based on its total funded amount, and included it for analysis if it was coded as HIV-related and international at any level.
The NIH definition of "international" is different from ours. In our analysis, "international" refers to programs and research taking place in or involving researchers from developing countries (for our purposes, those not belonging to the Organization for Economic Cooperation and Development, or OECD, which comprises the world's richest 29 countries). NIH codes all contracts or grants that are awarded to historically minority institutions in the U.S. or to those in Alaska, Hawaii and Puerto Rico as international. NIH also funds research projects at academic institutions, such as Oxford University, which are undeniably international but which are unrelated to the U.S.'s response to the HIV epidemic in the developing world. Some NIH projects also seem to be coded as international in error. Of 118 grants and contracts supplied by NIH, 76 met our criteria for an "international" project.
A major goal of this analysis was determining the country-level funding and activities of the U.S. government, which proved more difficult than we had anticipated for a number of reasons. USAID programs took place both through missions at the country or regional level and through large NGO contracts with multiple projects operating in multiple countries and regions. From the information we had, it was impossible to correlate country-level activities of the NGO contracts with any specific dollar figure. We know which countries they operated in, we know the total budget, but we do not know the exact budget for each project in each country.
To describe the degree of USAID involvement in any country, two different methods were used. First, all country-level financial information that was available, (i.e., original USAID administrative allocations to individual missions and bureaus before these funds were aggregated to pay the contracts) was recorded. Although these figures do not account for USAID's seven large NGO contract budgets, these funding levels were assumed to have a relationship to actual funding levels among countries. Second, all contract and grant reports were examined, and every country-level activity reported in them was recorded as an "international HIV project." For instance, a single contract might operate in 21 countries and report 105 discrete projects. Each of these projects was recorded as assigned to the appropriate program area and country. We did not assign budget information to such projects, however, since accurate data were not available.
NIH grants and contracts may also involve one country, or several countries and regions. When the NIH project took place in only one country or region, we assigned its budget to that country or region, and all reported research activities were recorded and classified. When an NIH project occurred in multiple countries or regions, we have itemized the number of projects by country, but have been unable to disaggregate funding on a country-by-country basis. We thus tabulated these funds on the "global" activities budget line, or, when the project took place in only one region (e.g., Africa) on the budget line for that region. CDC activities were all directly sponsored by CDC and they provided specific country-level project and budget information. Describing accurate and comprehensive funding allocations to individual countries was not possible in this analysis for reasons discussed above.
From all the information reviewed, we identified 516 discrete international HIV projects. Of these, 463 projects were country- or regional-level projects, and 53 were "global," which means that they occurred in more than one region, or were truly global but otherwise unidentifiable.
Although the projects identified through this method are unequal in size (i.e., some projects are larger and more extensive than others), this approach did allow a description of the geographic scope of all activities, including activities in countries that receive no direct American funding. The number of international HIV projects identified in this analysis underestimates the actual number because some contracts, such as that for the U.S. Bureau of the Census, did not identify where the activities took place.
This report groups the U.S. government-sponsored international program into nine broad thematic categories, described in How the Government Spent the International AIDS Budget in 1998. We established these categories solely for this analysis; they do not reflect internal coding schemes employed by the agencies. These categories were based on a careful reading of program descriptions and contract reports, allowing us to make descriptions across agencies. However, since information from each agency varied in detail, the categories have limitations. For example, the post-contractor budget for USAID missions was categorized as health systems support because most of the reported activities of USAID missions in congressional testimony described this type of activity.
This analysis attempts to describe the size and nature of the U.S. role in the fight against the global HIV pandemic at one point in time. The information used for this analysis permitted a detailed examination of the U.S. role, but sometimes the information was inconsistent and unclear. Budgetary information may need revision, and the number of U.S.-funded HIV projects was probably greater than that described in this report. The report is, however, the first comprehensive attempt to describe the baseline administrative and fiscal structure of the U.S.-role across all federal agencies. With luck, this can help inform current debates about new U.S.-sponsored HIV initiatives for the developing world.
|Table 2: U.S. International AIDS Funding by Agency, FY 1998|
|Latin American bureau||$15,894,000||8.5%|
|Other institutes & centers||$1,908,421||1.0%|
|Total U.S. Int'l AIDS Program||$185,119,072||100%|
USAID = U.S. Agency for International Development; NIH = National Institutes of Health; NIAID = National Institute of Allergy & Infectious Diseases; NCI = National Cancer Institute; FIC = Fogarty International Center; NICHD = National Center for Child Health & Human Development; CDC = Centers for Disease Control & Prevention; NCHSTP = National Center for HIV, STD and TB Prevention; NCID = National Center for Infectious Diseases.