Foreign Assistance and the Struggle Against HIV/AIDS in the Developing World

Abstract

The few studies that have examined the systematic determinants of HIV/AIDS policy cross-nationally have left the possible impact of foreign aid out of the equation. At a time when developed nations are critically reassessing their foreign aid commitments a deeper understanding of the impact of HIV/AIDS foreign aid on policy outcomes in the developing world is vital. This study expands the present literature by analyzing the role of foreign funding in a nation's response to the epidemic. The authors find that while HIV/AIDS directed foreign aid has significantly positive effects on a country's treatment coverage rates, the level of traditionalism is a more important influence with regard to the proclivity of a country to adopt preventative policies centred on HIV/AIDS education.

Civil and political rights are critical, but not often the real problem for the destitute sick. My patients in Haiti can now vote but they can't get medical care or clean water. (Paul Farmer)

Acknowledgements

The authors wish to thank Michael Hillman and the Persepolis Institute for their generous help with the translation of the Iran country report. Special thanks to Jacek Kugler for reviewing an earlier version of this paper. The authors would also like to thank Mark Gose and other participants at the 2008 International Studies Association-West and 2009 International Studies Association conferences for their valuable feedback on a previous version of this paper. Any errors are the authors' exclusively.

Notes

An Online Appendix is available for this article which can be accessed via the online version of the journal available at www.informaworld.com/fjds

1. As impressive as these increases are, we can hardly call the fight against the HIV/AIDS epidemic a solved international collective action problem. Most multilateral institutions focused on HIV/AIDS publish reports of unmet needs, as illustrated by UNAIDS, which considers the gap between estimates of resource needs compared to resources available in 2007 to be at least $8bn.

2. This graph was produced with data from the OECD. Specifically the OECD reports the amount of foreign aid that was given bilaterally from DAC members and from multilateral institutions for ‘HIV/AIDS control.’ The OECD reports these numbers in current US dollars, which we converted to constant 2007 US dollars.

3. Lake and Baum (2001) is the one notable exception.

4. This is not an exhaustive list.

5. Mission statements from the USAID, and the Global Fund lead us to believe that it is likely that foreign aid flows for HIV/AIDS control are directed to the nations which are in most need. Both organisations state that flows go to nations which have widespread or specific HIV/AIDS related problems, and nations which are lower to middle income.

6. Data for these figures were gathered from the OECD. The OECD reports the total amount of foreign aid given from DAC countries to developing countries for ‘HIV/AIDS control’ which includes both treatment and prevention programmes. The OECD also reports the total given from multilateral institutions.

7. This chart includes data on a country's bilateral ‘HIV/AIDS control’ foreign aid as well as data on a country's multilateral foreign aid to two major multilateral institutions. The OECD was consulted for the total aid given to ‘all developing countries’ for ‘HIV/AIDS control’ from bilateral sources by donor country, as well as the total aid given to multilateral agencies for ‘HIV/AIDS control’. According to OECD's aid statistics in 2006, of the total aid given for ‘HIV/AIDS control’, 25 per cent flowed through multilateral agencies and UNAIDS and the Global Fund accounted for 78 per cent of all foreign aid flowing to multilateral agencies. In an effort to track a nation's bilateral and multilateral aid for HIV/AIDS control, we added a nation's bilateral funding to ‘HIV/AIDS control aid’ (from OECD) with their donation to UNAIDS for 2006 (accessed at: http://data.unaids.org/pub/BaseDocument/2007/core_2006_en.pdf) and 58 per cent of their Global Fund contribution in 2006 (accessed at: http://www.theglobalfund.org/documents/publications/annualreports/2006/AnnualReport.pdf). We only included 58 per cent of their Global Fund contribution because as a Global Fund report states, 58 per cent of their total resources go to work on HIV/AIDS treatment and prevention (accessed at: http://www.theglobalfund.org/documents/publications/annualreports/2006/2006AnnualReport.pdf, p.107). ‘Others’ indicates the total share of resources given from other nations and private sources that did not amount to 1 per cent individually.

8. Multilateral and bilateral figures were gathered from OECD statistics. Data on domestic spending was gathered from UNAIDS (2006). This is the national fund spent by governments from domestic sources on HIV/AIDS.

9. Our sample includes the following countries: Algeria, Angola, Antigua and Barbuda, Armenia, Azerbaijan, Bahamas, Bangladesh, Barbados, Belarus, Belize, Benin, Bolivia, Bosnia, Botswana, Brazil, Bulgaria, Burkina Faso, Cambodia, Cameroon, Canada, Cape Verde, CAR, Chad, Chile, China, Colombia, Comoros, Congo, Congo, DRC, Costa Rica, Cote d'Ivoire, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Ethiopia, Fiji, Finland, Gabon, Gambia, Georgia, Ghana, Greece, Grenada, Guatemala, Guinea, Guinea-Bissau, Guyana, Haiti, Honduras, India, Indonesia, Jamaica, Jordan, Kazakhstan, Kenya, Laos, Latvia, Lebanon, Lesotho, Lithuania, Macedonia FYR, Madagascar, Malawi, Malaysia, Mali, Marshall Islands, Mauritania, Mauritius, Mexico, Mongolia, Montenegro, Morocco, Mozambique, Namibia, Nepal, New Zealand, Nicaragua, Niger, Nigeria, Pakistan, Palau, Panama, Papua New Guinea, Peru, Philippines, Romania, Rwanda, Saint Kitts and Saint Nevis, Saint Lucia, Saint Vincent and the Grenadines, Sao Tome and Principle, Senegal, Seychelles, Sierra Leone, Slovenia, Somalia, South Africa, Spain, Sri Lanka, Swaziland, Sweden, Tajikistan, Tanzania, Thailand, Togo, Trinidad and Tobago, Tunisia, Turkey, Tuvalu, Uganda, Uruguay, Vietnam, Zambia, Zimbabwe.

10. The majority of countries that submitted country reports have a per capita income of less than US$10,000. In this way, our analyses are limited in their generalisability to the developed world. Furthermore, because we rely on self-reported data, the exclusion of certain countries may be a function of their capability to obtain such data and, therefore, the data availability may be non-random. While we are excited that we are able to offer this exploratory study which utilises measures assessing HIV/AIDS policy effectiveness that the field has yet to explore, it should be noted that there is a real concern for measurement reliability due to the fact that the data are self-reported by Member States. The possibility of deviation from the proposed standardised methodology as laid out by UNAIDS cannot be ignored.

11. These four were selected because they provided a sufficient number of observations to run regression analyses. Furthermore, the variables chosen are specifically related to policy and not behaviour.

12. One might be concerned that this sector of foreign aid does refer to foreign aid targeted at all sexually transmitted diseases (including HIV/AIDS), and would thus be an unfit measure to approximate HIV/AIDS control assistance. In 2004, the OECD addressed this concern by analysing all commitments made under ‘STD control including HIV/AIDS’ from 2000–2002 and found that 85 per cent of the commitments mentioned HIV/AIDS explicitly in the project title or description (OECD, 2004: 7). The authors of that report went on to argue that ‘another justification for this approach was that it was difficult to imagine a STD control programme that would not contribute to HIV/AIDS control.’ (OECD, 2004: 7).

13. Measures of the percentage of people who adhere to Islam were rarely significant in our ordinary least squares (OLS) models.

14. Other measures which we explored were Government Effectiveness (World Governance Indicators) and literacy rates. Upon a detailed examination of the underlying data used to contruct the government effectiveness indicator, it was determined that it includes measures of health related government performance. We chose not to include this indicator in the analyses, therefore, because of the conceptual equivalence between it and our dependent variables. We chose to not report models which include literacy rates due to a low level of data coverage.

15. We estimated the same specifications using OLS and logged dependent variables and received similar results, especially with regard to the significant covariates. Results are available upon request. Furthermore, the correlation matrix reported in the online appendix suggests that there is little cause for concern that the models suffer from multicollinearity.

16. The odds effect of an interquartile range shift is estimated by: , where Q₁ and Q₃ are the lower and upper quartile values of x. See Hardy and Bryman (2004: 292) for a more thorough discussion on interpreting odds ratios.

17. In unreported models we included several different regional dummy variables, namely South America/Caribbean, South-East Asia and sub-Saharan Africa. We find it encouraging that the significance of our variables is robust with the inclusion of these regional dummy variables. We ultimately chose to limit the regional dummies to only include sub-Saharan Africa in an effort to maximise degrees of freedom.