Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
The fight against HIV/AIDS requires higher and more sustained investments at the global level [1–4]. It is not clear, however, how those funding decisions are made, how to influence the level of investment, and what may spur future contributions to organizations such as the Global Fund to Fight AIDS, Tuberculosis and Malaria.
The economic theory of investment emphasizes the role of the expected rate of return as well as the influence of current and future expectations [5,6]. When investors consider funding an enterprise, they take into account expectations of future performance. Similarly, international organizations and donor countries may choose to invest in the Global Fund, not with the intention of gaining profit, but rather to increase the global number of infections prevented and persons treated .
Using data from the 360° Stakeholders Survey, this paper analyses 23 attribute gaps or unfulfilled expectations related to resource mobilization, impact measurement, harmonization and inclusion, effectiveness of the Global Fund partner environment and portfolio characteristics. The independent variables are personal and regional-level characteristics that may determine or at least correlate with some of the expectations.
The survey was not based on a representative sample of Global Fund stakeholders. As the respondents were self-selected, the analysis cannot be considered representative of the underlying opinions and expectations of all Global Fund stakeholders. Nevertheless, even if some biases result from self-selection, the current analysis can provide insight into critical issues. The effectiveness of the Global Fund may depend on several factors [8,9], including future levels of investments. To the best of our knowledge, this paper is the first to analyse the stakeholders' expectations in the fight against HIV/AIDS, tuberculosis and malaria.
The way the Global Fund seeks funding is similar to how companies raise money in the capital markets from individual and institutional investors. The investors choose a company based on historical rates of return, but also based on their more subjective opinions or expectations of both the company's performance and likely changes in its environment. The theory of expectations in the economic literature on investment goes back to Keynes, who said ‘that it is our innate urge to activity which makes the wheels go round, our rational selves choosing between the alternatives as best we are able, calculating where we can, but often falling back for our motive on whim or sentiment or chance’ .
The relatively newer theory of rational expectations in investment  can be consistent with a health investments model in which many investors would put their resources towards the production of certain output (e.g. HIV cases prevented, antiretroviral drugs provided) under a given production function. Constant investments by high-income nations, private sector, foundations and individuals can maintain a flow of production; therefore, increasing investment results in increased capacity. The stream of returns, determined by volume and product price in the traditional model can be replaced by the stream of disability-adjusted life years (DALY) that are not lost to disease, because interventions prevented disability and premature death . The model assumes that investors do not seek financial retributions, but rather that they take DALY averted as a return for their investments. Using a suitable discount rate, the model maximizes the net present value of a future stream of disease burden averted. The future DALY averted by an investment made today depend on many uncertain variables. Investors will make future decisions based on both objective and subjective information, including prevalence trends, efficiency of implementation and technological change.
The literature suggests that expectations related to financial decisions are also linked to the issue of institutional reputation. It is well documented that the banking business depends not only on financial solidity but also on customers' trust. Runs on banks are mainly the result of a lack of confidence in the financial institutions . These were common in the United States at the beginning of the 20th century and in Europe, particularly Germany, during the 1930s, and continue in some regions, including Latin America. Conversely, positive expectations and stronger institutional confidence lead to greater investments [12,13].
From the theory we can thus infer that expectations may also affect worldwide health investments. Individuals, foundations and governments make decisions about whether and how much of their resources to invest in global public health, including in the Global Fund. The investment decisions can be based on past performance, technological considerations and also on expectations of the future, including the environment in which the fund's three diseases occur, as well as the performance of the fund and its grantees. The crucial question is then: What determines expectations?
An empirical model
We can model expectations (Y) as a function of a vector X, which includes personal, institutional, regional and other characteristics as follows:
Equation (Uncited)Image Tools
The term α represents an intercept and we can assume that the term ϵ is a zero mean, normally distributed error term. At a first approximation, the empirical estimation of this model can be implemented using a single cross-section of individual responses.
We analysed data from an on-line survey called the 360° Stakeholder Survey. As part of the Global Fund 5-year evaluation efforts, a stakeholder opinion survey was e-mailed to approximately 5700 participants; hypertext links to the survey were also placed on the Global Fund website. The survey was conducted by a market research firm (TNS Healthcare) between 26 May and 8 June 2006; TNS also cleaned the data and produced a final dataset. The time to complete the survey was estimated to be between 15 and 20 minutes. It was administered in English, Spanish, French and Russian to all stakeholder groups: country coordinating mechanism (CCM) members; local fund agents; principal recipients; board members; private sector focal points; donor constituency representatives; UNAIDS country coordinators; and the Roll Back Malaria Partnership Board.
The analyses were performed using a constructed linear measure of expectations. A respondent's score of Global Fund performance was subtracted from the score the same respondent gave to the importance of that attribute; for each of the 23 Global Fund attributes analysed. The new variable created was taken as a measure of the gap or unfulfilled expectation and used as a dependent variable in linear regression analyses. Gaps or unfulfilled expectations are used as a proxy for actual expectations. As the measured gap or unfulfilled expectation refers to the linear difference between the stakeholders' ranking for each attribute importance minus the ranking they gave to the Global Fund for performance for that attribute, a larger number implies a larger gap and greater unfulfilled expectation.
In the empirical estimation, X was a covariate vector that included: respondent and institutional attributes (focus and level of work, region of residence, main interest, formal and general level of involvement with the Global Fund, HIV status) and regional characteristics (Global Fund US dollars disbursed per capita; prevalence rates for HIV: people living with HIV/AIDS per 1000 people; total tuberculosis incidence: all tuberculosis forms per 100 000 persons). Note that the omitted and thus reference categories for comparisons are: for level of work the ‘subnational level’; for focus of work ‘recipient government’; and for region of residence sub-Saharan Africa.
Table 1 presents the descriptive statistics from the stakeholders surveyed. Approximately a third of respondents' focus of work was at the non-governmental level. Approximately half of stakeholders surveyed reported that their work is at the national level. Almost a third of respondents lived in sub-Saharan Africa. Approximately 80% of stakeholders had a work interest in HIV/AIDS issues. More than a third (37%) were involved in Global Fund grants through membership in the CCM and approximately 7% were persons living with HIV.
Table 2 presents a summary of the outcome variables: opinions of stakeholders in terms of attributes' importance, their assessment of the Global Fund's performance to date and expectations. The expectation column (or gap) is the mean of the linear difference between importance and performance across all respondents. The table has been ordered by category or type of attribute in descending order by average size of the expectation gap for the category. The ‘resource mobilization’ category had the largest unfulfilled expectation. In particular, stakeholders' expectations of the Global Fund were least met (i.e. had the highest unmet expectations) in the areas of mobilization of private sector resources, followed by mobilization of new financial resources.
‘Impact indicators’ was the category with the second highest level of unmet expectations, including: programmes reaching their target population, health systems being strengthened and greater efficiency through performance-based funding.
The ‘harmonization and inclusion’ and ‘effectiveness of Global Fund partner environment’ categories had a similar average score. Stakeholders' expectations were met best (i.e. had the lowest unmet expectations) for the ‘portfolio characteristics’ category, which includes attributes on balance of funding among interventions and populations. Stakeholders reported that the most important attributes of the Global Fund are: transparent information sharing, mobilization of new financial resources, and ensuring that people affected by HIV/AIDS, tuberculosis and malaria are reached by programmes receiving the fund's support. They gave the Global Fund best performance scores for giving priority to the most affected regions or populations, for its focus on confirmed and effective interventions and for transparent sharing of information.
Table 3 presents, in summary form, the results from regressions in which the dependent variable is the gap or unfulfilled expectation rating given for each one of the 23 Global Fund attributes analysed. (For detailed results, see also Web Appendix Tables A1–A4). The table shows the statistically significant, positive correlations with a ‘+’ sign (i.e. that type of respondent was more likely to have more unfulfilled expectations for that category of attributes), and it shows negative correlations with a ‘−’ sign (i.e. that type of respondent was less likely to have unfulfilled expectations, or in other words, the expectations were better met).
Table 3 shows that in the ‘resource mobilization’ category, stakeholders working at the international level, at non-governmental organizations and persons living with HIV/AIDS were more likely to have unfulfilled expectations in terms of the mobilization of private sector resources. Similarly, those involved in the fund's CCM and people living with HIV/AIDS were also more likely to have unmet expectations in terms of the mobilization of new financial resources.
With regard to ‘impact indicators’, stakeholders at multilateral organizations and those involved in the Global Fund through the CCM were more likely to have unfulfilled expectations. On the other hand, the level of involvement with the Global Fund was negatively correlated with unfulfilled expectations; that is, controlling for all other characteristics, as the extent of involvement in the Global Fund increased, respondents believed their expectations were better met.
The patterns for the other attribute categories (‘harmonization and inclusion’, ‘effectiveness of Global Fund partner environment’ and ‘portfolio characteristics’) closely resembled the picture for ‘impact indicators’. Namely, those with formal involvement with the Global Fund working specifically for the CCM were more likely to have unmet expectations for 22 of 23 attributes. Similarly, stakeholders whose focus of work was at the multilateral level, as opposed to the national level, were also more likely to exhibit unfulfilled expectations in 15 of 23 attributes. The third group showing greater unfulfilled expectations for most attributes (16 of 23) were persons living with HIV/AIDS. On the other hand, a greater extent of involvement with the Global Fund was associated with better met expectations in 15 of the 23 attributes.
The respondents' main work interest (HIV/AIDS, tuberculosis or malaria) did not make a difference in terms of unfulfilled expectations. When compared with stakeholders living in sub-Saharan Africa, however, those living in other regions were more likely to have their expectations met. In particular, stakeholders in the Middle East and north Africa were more likely to exhibit better met expectations for five out of 23 attributes, and those in eastern Europe and central Asia showed better met expectations for two out of 23 gaps.
It is not surprising that the ‘resource mobilization’ attributes category has the highest level of expectations because this category is directly related to the Global Fund's main goal: ‘to attract, manage and disburse additional resources’. Nevertheless, given the large unfulfilled expectation, it is clear that the fund's work has just started, and that more work is needed to continue steady progress towards fulfilling the stated objectives. Some of the global health investment commitments were made before the Global Fund was created, but decisions about future needs and allocations are made daily. Stakeholders' opinions and expectations may be an important determinant of the outcomes in that realm, even when stakeholders do not directly control financial decisions at country or institutional levels.
The ability of the fund to meet expectations in the area of ‘impact indicators’ is a function not only of its own efforts but also of national and international efforts by other financing and implementation agencies. Global Fund stakeholders do have very high expectations on these attributes, but achieving them is only partly in their own hands. ‘Harmonization and inclusion’ attributes are related both to the Global Fund's prescribed mechanisms at country level as well as the way that individual CCM and principal recipients implement programmes in each country. Therefore, expectations may depend on Global Fund guidelines and specific implementation outcomes.
The indicators related to the ‘effectiveness of Global Fund partner environment’ are those most directly under the control of the Global Fund stakeholders, and thus may be best targeted for increased efforts to meet expectations. It is remarkable that expectations are poorly fulfilled across most categories for stakeholders working with multilateral organizations; those who have perhaps the greatest responsibility for making the Global Fund partnership function effectively.
Attributes related to ‘portfolio characteristics’ are the result of country-level planning mechanisms, over which many stakeholders would have influence. Country-level planning is the interface between the country-driven proposal drafting processes and Global Fund mechanisms for making funding decisions. These attributes are taken directly from the fund's founding principles relating to basic expectations of how it operates. It is illuminating and it speaks well of the work achieved thus far, to find under this category some of the most fulfilled expectations, such as funding based on achieving measurable results, and focus on funding confirmed and effective interventions against HIV/AIDS, tuberculosis and malaria.
Some discernible patterns emerge across attributes and categories. First, the stakeholders participating in CCM have unmet expectations of what the Global Fund should achieve across all attributes. As with the members of the multilateral organizations mentioned above, the very respondents who have the most potential influence over how well the fund performs are the least satisfied, suggesting that part of their dissatisfaction may be with how their own organization is contributing to the overall success.
Second, persons with HIV/AIDS have unfulfilled expectations for most attributes. This result may be understood in the context of need. Perhaps those individuals personally affected by a disease are most sensitive about the importance of past and future performance. Third, some of the regional differences can also be explained in terms of relative need. Not surprisingly, the region with the most need (sub-Saharan Africa) shows higher unmet expectations.
Finally, the people who seem most satisfied with the Global Fund's performance are the ones whose involvement with the organization is greater. Those who are more involved and more personally invested are less likely to view performance negatively. The Global Fund is, however, a newcomer to the global public health platform with substantial resources and a different implementation paradigm. This situation may cause others to view its actions skeptically . Stakeholders who know and understand the Global Fund may be more likely to evaluate its performance free of preconceived notions.
In terms of the limitations of our study, we already mentioned the likely selection bias introduced by the non-random web-based survey; those more likely to respond may have extreme views. Nevertheless, note that another recent effort to map the opinions of people around the world about their perceptions of health problems, priorities and donors' performance has reached strikingly similar conclusions: people in Africa are most worried about HIV/AIDS; donors need to do more; but more generally, people are optimistic and believe that their country is making progress in HIV prevention and treatment . That survey was conducted face-to-face and via telephone on nationally representative samples for several countries. Their overall conclusions about people in recipient countries support our underlying hypothesis: as larger investments in health are made, the opinions and expectations of end-users are an important factor in whether the efforts will be reduced, maintained or increased.
Other shortcomings are as follows. First, because the regression analyses show only correlations and cannot claim causation, we cannot conclude that greater involvement with the Global Fund causes respondents to believe that their expectations are better met. We simply observe that greater involvement is associated with smaller differences between attribute importance and performance. Perhaps this smaller gap results from greater familiarity with actual operations and developments on the ground, but this explanation remains a hypothesis. Similarly, we do not correct for simultaneity, or factors that may affect both expectations and level of involvement; thus there may be selection biases in which we observe an association between the level of involvement and expectations, but that is only because both are in turn determined by other measured or unmeasured factors. For example, there may be a ‘halo effect’, in which persons more involved with the Global Fund tend to have higher opinions of it simply because of their personal beliefs about the worthiness of its goals.
Our measure for unfulfilled expectations is only an imperfect proxy for actual expectations, which are by nature highly heterogeneous and inherently unobservable. Expectations, as well as overall reputation, depend on rational and ‘irrational’ or subjective components. Economic and behavioral models have only limited capability when it comes to subjective issues that are not easily quantifiable. We need to be developing further the methods to measure and understand expectations, especially donors' expectations, and their implications for future financial flows. Internet-based surveys are a relatively inexpensive and convenient way of obtaining data; by providing anonymity, therefore potentially increasing the possibility of truthful responses. To gain more focused information, however, we require improved means of selecting respondents. For example, a future exercise to map Global Fund stakeholders' expectations should include the opinions of lawmakers and parliamentarians, or other groups who control the finances at the country level. Interviewing legislative assistants may give a sufficiently accurate picture of the feelings about future global health investments, including whether the decisions will be made on impulse or based on careful cost–benefit calculations. Mixed survey techniques (e.g. telephone and web) may be required.
The results from the Stakeholders Survey may also be analysed through the lens of herd behavior theory, developed in parallel in the social psychology and economics fields . Under that perspective the stakeholders would demand more services from the Global Fund and donors will provide more investments, as long as others are doing the same; a critical mass may be reached through a ‘bandwagon effect’ . Similarly, stakeholders' responses, both favorable or unfavorable, may depend on collective behavior: a threshold model may apply, in which a certain number of other actors need to take action or express an opinion before a particular actor under analysis would also take action or express an opinion based on an individualized cost–benefit calculation .
Although there has been substantial work in analysing herd behavior and investment , there is much less, if any, research into herd behavior and charitable giving or foreign aid contributions. Further integration of the fields of social psychology and economics would be very useful to analyse the issues of herd behavior and investments in global health.
Expectations are not static but are updated as new information is received. This dynamic process cannot be captured with a single cross-sectional survey. The Global Fund is a young institution and has been adapting rapidly to the successes and failures of the new funding mechanisms with which it has experimented. The 360° Stakeholder Survey analysed here was intended only as an initial input for the fund's upcoming 5-year evaluation, which will better measure expectations and performance. The 5-year evaluation will not only inform stakeholders about objective performance measures but will also probably modify future expectations in an important way.
Some of the unfilled expectations relate to attributes over which the Global Fund stakeholders have actionable influence, whereas other characteristics may be determined further upstream in the process. Understanding the determinants of current and future expectations will be useful for advocates and for future formal analyses of the investment levels in the Global Fund and other international efforts.
The decisions to invest in the fight against HIV/AIDS and other global diseases may be driven largely by stakeholders' expectations. As programmes become more effective at prevention and treatment, and as they achieve better results, unfulfilled expectations may decrease and investments may continue to grow. The Global Fund and other global health institutions need to strive to achieve that virtuous cycle.
The authors collaborated with Bernhard Schwartländer, Beth A. Plowman and others at the Strategic Information and Evaluation Unit at the Global Fund to Fight AIDS, Tuberculosis, and Malaria on the analysis of the 360° Stakeholders Survey for presentations to the Board of the Global Fund. The current work reflects the authors' additional analyses and interpretations. The opinions in this paper do not represent the official view of the Global Fund or any of the organizations mentioned above. The authors are solely responsible for the contents.
1. Schwartländer B, Grubb I, Perriens J. The 10-year struggle to provide antiretroviral treatment to people with HIV in the developing world. Lancet 2006; 368:541–546.
2. Piot P, Feachem RG, Lee JW, Wolfensohn JD. A global response to AIDS: lessons learned, next steps. Science 2004; 304:1909–1910.
3. Kim JY, Gilks C. Scaling up treatment– why we can't wait. N Engl J Med 2005; 353:2392–2394.
4. Sachs JD. Health in the developing world: achieving the Millennium Development goals. Bull WHO 2004; 82:947–949.
5. Keynes JM. The general theory of employment, interest, and money. London: MacMillan and Co. Ltd.; 1936.
6. Lucas REJ, Prescott EC. Investment under uncertainty. Econometrica 1971; 39:659–681.
7. Feachem RG, Sabot OJ. An examination of the Global Fund at 5 years. Lancet 2006; 368:537–540.
8. Government Accountability Office. The Global Fund to Fight AIDS, Tuberculosis, and Malaria is responding to challenges but needs better information and documentation for performance-based funding. GAO-05-639. Washington, DC: Government Accountability Office; 2005.
9. Lu C, Michaud CM, Khan K, Murray CJ. Absorptive capacity and disbursements by the Global Fund to Fight AIDS, Tuberculosis and Malaria: analysis of grant implementation. Lancet 2006; 368:483–488.
10. Murray CJL, Lopez AD. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge, MA: Harvard University Press; 1996.
11. Hammond B. Banks and politics in America: from the Revolution to the Civil War. Princeton, NJ: Princeton University Press; 1991.
12. Fenton-O'Creevy M. Traders: risks, decisions, and management in financial markets. Oxford and New York: Oxford University Press; 2005.
13. Lucey BM, Dowling M. The role of feelings in investor decision-making. J Econ Surv 2005; 19:211–237.
15. Kaiser/Pew. A global look at public perceptions of health problems, priorities, and donors: the Kaiser/Pew Global Health Survey. Washington, DC: Kaiser Family Foundation and Pew Global Attitudes Project; 2007.
16. Rook L. An economic psychological approach to herd behavior. J Econ Issues 2006; 60:75–95.
17. Leibenstein H. Bandwagon, snob, and Veblen effects in the theory of consumer's demand. Q J Econ 1950; 64:183–207.
18. Granovetter MS. Threshold models of collective behavior. Am J Sociol 1978; 83:1420–1443.
© 2008 Lippincott Williams & Wilkins, Inc.