Book Search:  


Google full text of our books:

Boundaries of Contagion:
How Ethnic Politics Have Shaped Government Responses to AIDS
Evan S. Lieberman

Book Description | Endorsements | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2009, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format


“AIDS knows no boundaries.” For much of the history of the AIDS epidemic, activists and officials around the world have repeated this incantation. As one diplomat described AIDS, “It discriminates against no ethnicity, no gender, no age, no race, no religion. It is a global problem that threatens us all.”1 From a strictly biomedical standpoint the claim is accurate. In 2006, UNAIDS, the lead international organization for AIDS control, estimated that approximately 65 million people—men and women, young and old, rich and poor, black and white, Christian and Muslim, from every continent and virtually every country—had been infected with the human immunodeficiency virus (HIV) between 1981 and 2006. An estimated 25 million people had died from the constellation of infections and ailments that comprise the syndrome known as AIDS. If ever there was a crisis that revealed the shared vulnerability of humanity, this was it.2

And yet, because the transmission of HIV is a social phenomenon as well as a biological one, boundaries have proven to be incredibly important. At almost every level, leaders and ordinary citizens have interpreted the deaths and illnesses associated with the global pandemic in terms of ethnic and national groups. In 1988, the Panos Institute produced a small volume entitled Blaming Others: Prejudice, Race, and Worldwide AIDS,3 which contained news reports and short essays identifying the worldwide prejudice associated with AIDS. In a thoroughgoing scholarly analysis, Cathy Cohen has documented how race has influenced the politics of AIDS in the United States, arguing that the historical marginalization of African-Americans produced a deafening quiet in political reactions to this stigmatized disease.4 The renowned scholar-activist-medical practitioner Paul Farmer has chronicled the dynamics of blame associated with prevalence of AIDS among Haitians, which in myriad ways has been associated with the politics of race.5 Throughout the history of the epidemic, even as millions fell ill and died, leaders and ordinary citizens have claimed that AIDS was a “foreign” scourge and someone else’s problem. Although evidence of global contagion was reported soon after HIV was isolated, many government leaders and citizens around the world clung to the idea that geographic, social, and political boundaries would insulate them from the contagion. Such ideas often stifled aggressive action against the disease. Unlike microbes, people are keenly aware of boundaries, and successful political leaders—elected and otherwise—manipulate group boundaries to maintain support and stay in power. Tragically, the will to political survival has often led to the underprovision of public policies that might have meant human survival in the face of a viral pandemic. Since AIDS was perhaps the single most important threat to human development in low- and middle-income countries at the start of the twenty-fi rst century,6 the substantive implications are enormous.

To argue that AIDS has been associated with a politics of blame and group prejudice is hardly a novel proposition.7 This book breaks new ground by offering a more careful analysis of how and why the boundaries that divide groups from one another have affected patterns of policymaking around the world. In so doing, it provides more general insights about the relationship between ethnic politics, policymaking, and development. Specifically, it helps to explain why some governments have been more ag gressive in responding to AIDS than others, and it estimates the effects of boundaries on policies relative to other influences. My analyses reveal that the relationship between ethnicity and public policy is channeled largely through political competition over the social status of ethnic groups and the propensity to view risks as pooled, a departure from the existing literature on the effects of ethnic diversity and ethnic competition, which has emphasized problems of coordination; exogenously determined, heterogeneous preferences; rent-seeking behavior; and patterns of distrust.

Although the next chapter will more fully elaborate propositions linking boundaries to policies, a brief preview is in order. Boundaries are institutions that separate groups of people from one another. These include the internationally recognized borders that give shape to more than two hundred states and a few dozen territories around the world, and also the formal and informal practices and markers that reinforce a sense of group difference within and across countries. Boundaries vary tremendously in the degree to which they are recognized and enforced in their various manifestations. In some contexts, boundaries make group membership relevant and clear to almost everyone involved, whereas in other contexts, boundaries are shifting, ambiguous, and permeable. In the former, group identities are more fixed, while in the latter, they are more fluid and undifferentiated.

In a wide-ranging set of investigations, I have found that when countries have strong, internal boundaries dividing societies into substantial and recognizable ethnic groups, the epidemic is also likely to be understood in ethnic terms. In turn, this frame of reference becomes a political constraint on national policies to combat AIDS. In countries so divided, discourses about the risk of being infected and affected by AIDS are infused with ideas about ethnic difference. Ethnic conflicts intensify the near-universal political dynamic of assigning blame and attaching shame to information about the epidemic. As citizens and political leaders seek to avoid the group shame associated with a stigmatized problem, the effect is a dampening of potential support for AIDS policies, leading to weaker and slower responses. While AIDS may aptly be labeled the first disease of the era of globalization,8 the persistence of state-level boundaries has meant that even within regions sharing common social and economic characteristics, resources and responses have been profoundly mediated by those in charge of national states. And those leaders have been sensitive to the politics that surround any major policy issue.

To be certain, other factors have been critically important for policy-making. The extent of the epidemic in countries and regions; the resources available to the government; and the overall pressure to adopt policies from international actors and networks of activists have all been infl uential and are components of a general model of AIDS policymaking. Individual personalities and accidental historical circumstances have also shaped responses. But I approach the analysis of government responses to AIDS as a social scientist, in the sense that I identify general patterns and relationships that go beyond the particularities of any individual country’s circumstances. In this regard, boundaries have been a central and underexplained influence on AIDS politics and policymaking, with dramatic implications for the course of human development. While I consider other drivers of politics and policymaking, I put a spotlight on the effect of boundaries.

This is a book about AIDS, but it is also about global politics and the interlayered process by which political actors attempt to govern and to transform people’s lives through large-scale public policies. While the focus is on AIDS, this study is not an exhaustive account of the epidemic or of AIDS-related policies. It is not a book about why HIV/AIDS has hit some countries harder than others or which practices have been most effective for stopping it. In fact, I do not even assume that the most “aggressive” responses were the best responses from the perspective of maximizing human welfare. Generally speaking, for any given country more AIDS control was probably better than less, but my investigation does not rely on this assumption.

The more narrowly defined question I take on is this: Why have some national governments responded to AIDS more quickly and more broadly than others? More fundamentally, this is a question about the political origins of government effort, and the conditions under which a country’s leaders are willing to take a stand on a politically sensitive issue for the sake of the longer-term development and well-being of its population. By focusing on boundary politics, I take up long-standing questions in the study of comparative politics concerning the effects of ethnic politics on the provision of public goods, and on development more generally. I attempt to wed political science theory and method to the study of public health. In so doing, I am inevitably drawn in to debates about the origins of authority and societal transformation.

In political science a puzzle exists when outcomes diverge from expectations. In this sense AIDS policy is a puzzle, for there is reason to have expected much more convergence in government responses to AIDS than we fi nd.9 While cross-country differences in policies on issues ranging from national health insurance to industrial development can be ascribed to accidents of history, including the timing of certain problems in particular countries,10 AIDS has confronted countries worldwide more or less simultaneously, smack in the middle of an era of heightened global integration. In the relatively short span of two decades, HIV reached virtually every country in the world. Knowledge about how to prevent the transmission of HIV, and how to improve and to extend the lives of HIV-positive individuals, was identified early on, and has been disseminated around the world. Nonetheless, national governments have adopted a wide range of responses to the AIDS crisis, with dramatic consequences for the affected societies. To be certain, rates of infection and the resources available explain some of the cross-country variation. But even in a relatively rich, technically competent, and high-prevalence country like South Africa, the government responded slowly, and life expectancy has declined to levels not seen since the 1950s, largely because of AIDS-related deaths. In other countries, such as Brazil, the epidemic has been contained, and its effects are managed as well as, and in some cases better than, they are in the world’s richest countries. Why have governments equipped with similar resources and similar information respond to the same biomedical phenomenon in such different ways?

The remainder of this chapter provides greater context for this and related questions by situating the problems of AIDS and AIDS policy in broader scholarly and policy-oriented concerns about states, governance, and globalization. I also detail the design of the research and provide an overview of the remainder of the book.

The Puzzle of Explaining Government Policy

The challenge of explaining responses to AIDS forces us to think more generally about why governments differ in their provision of public policies, public goods, and public resources that address the general welfare. What governments do, often in coordination with other partners, can have a huge impact on human development and well-being, particularly in the case of threats to public health. In the face of highly infectious and deadly diseases, such as the fever caused by the Ebola virus, governments in even the poorest of countries have been critical to rapid containment. But most public health threats do not receive such immediate and deliberate attention, particularly in resource-poor settings. In the absence of authorities who take responsibility for providing information and resources, disease can ravage societies, as markets and voluntary action alone may fail to provide the required coverage and action. By examining how and why different governments address a common problem in different ways, we can investigate seminal questions about the political dynamics of resource allocation and social control.

Some readers will question my focus on states, particularly in a book about a problem for which nongovernment actors have been so important. To be certain, other actors, ranging from transnational organizations to small communities, also affect health and well-being in the context of a viral epidemic, and many of them are considered in the pages that follow. And yet in the contemporary era, national states are uniquely positioned to broadcast information and to affect the behavior of people: they control the lion’s share of public resources, and they are a site of negotiation and competition over a society’s priorities. So long as states play a central role in societies and economies, we ought to investigate how and why they respond to important problems.

A study of governments and AIDS provides a lens onto the relationships between states and societies that lie at the core of the study of comparative politics.11 We should assume neither that states will always take the lead in public health epidemics or other major social problems, nor that they will be inconsequential, corrupt, or absent. Instead, we need to consider how and why states adopt varied approaches for particular concerns. AIDS has implied a dramatic role for governments, just at a moment in time when state power appeared to be in retreat.12 Many of the strategies for curbing the pandemic require that states intervene in the most private of matters: sexual conduct, drug use, childbirth and breast-feeding. The nature of the problem requires that political authority focus on bodily fluids, especially blood and semen, substances more amenable to deep metaphor than economic calculation. In the case of AIDS, state authorities have literally asked for blood13 in order to test for the presence of the virus. Just as Charles Tilly explained war-making as the basis for the extension of state authority,14 AIDS has provided a new exigency for state power, as well as for the powers of global governance. Some have described HIV as a security concern in the face of an “invisible enemy,” and analogies to war have persisted among policymakers and scholars alike.15

By focusing on the politics, policies, and actions of national states, I risk contributing to a reified notion of state authority, which in practice is contested and varied across time and space.16 Yet I acknowledge this governmental weakness and unevenness, because it is central to the questions I pose. The AIDS pandemic has provided states an opportunity to assert authority in their own territories and on the world stage. In an international environment in which national states remain the presumptive sources of authority, even if such authority fails to materialize in practice, we require an understanding of what shapes the exercise of power. Heads of state recognized by the international community have been pressured to act and cajoled with resources, and my concern is to understand their responses to these pressures. Within the state, infl uential actors vary in their inclination to adopt policies, depending upon their domain of authority and other personal and professional infl uences. I assume a partial coherence and autonomy in the state as policy-maker, in the sense that we can characterize a “state response” for a period of time, one that reflects the decisions of those at the top of the pyramid of political power. I consider divergent or confl icting responses (more or less aggressive than that of the central government) by particular ministries or localities, as independent political actors with a potential influence on government decisions.

Understanding the politics of AIDS requires that we think about the challenge of gaining compliance and consent more generally. Most AIDS-related policies involve asking citizens to do things that they fi nd undesirable, including wearing condoms, refraining from sex, getting their blood tested, and so on, and the benefit is an uncertain future nonoccurrence (i.e., protection from possible infection). Only late in the epidemic has treatment been part of the policy menu for most low- and middle-income countries. As a result of this ratio of effort to reward, publicizing and implementing AIDS-related policies created by the state in the private lives of citizens can be costly in monetary and political terms. Governments introduce policies related to sex, sexuality, and drugs at their peril. If citizens resent the message, they can challenge the messenger. In this sense, the political bargain over establishing an aggressive policy on AIDS is an instance of the state’s attempt to gain conformity and sacrifice more generally, as in taxation, military conscription,17 and other areas of social transformation, where states seek conformity through the imposition of new ideas, norms, and practices in order to promote “development.” Vaccinations require succumbing to a shot from a uniformed medical worker, which may seem frightening.18 Any major social policy, including the introduction of universal education, demands trade-offs and the reallocation of resources that might have been available for immediate consumption. Transformation often requires sacrifice, and in the development of public policy, actors may strike back when their moral or material interests are jeopardized.

A relatively simple explanation for cross-country differences in policy regimes is the catchall category of culture. That is, we could simply conclude that countries with “modern” cultures and values are more likely to embrace germ theories of disease and to use related technologies to combat viral spread. But if we were to classify cultures as modern according to their responses to AIDS, the proposition would be true by defi nition. It would be a tautology. Another approach would be to assess cultures in terms of potentially relevant beliefs, such as orientation toward “secularmodern” values,19 but as it turns out, measures of those values don’t do a good job of predicting which countries will take up AIDS policies aggressively and which will not.

As a political scientist, I am predisposed to look for the political underpinnings of choices in policy, especially in domains that might otherwise seem to be well guided by technical or biomedical science. I was therefore surprised to find that the comparative study of health and health policy in developing countries has received little attention from political scientists. Apart from a handful of important contributions, the study of AIDS and AIDS policy has been largely ignored.20 The enormity of the HIV/AIDS pandemic is suffi cient justification for anyone to devote time and attention to it. But AIDS is a social and political phenomenon, one of the most profound of the modern era, and deserves greater attention if we have a sincere interest in understanding how and why governments act, or ignore opportunities to act, to improve the human condition in their territories.

Explaining government responses to AIDS requires a comparative approach. Politics and policymaking rarely take place in a national vacuum, in isolation from the rest of the world. For a global problem, in an increasingly integrated world, with expansive global governance, external pressures shape the menu of possible responses and help determine which policies are selected from it. And yet transnational actors walk a delicate tightrope in exerting direct authority over peoples across the globe in a postcolonial era, when norms of sovereignty remain strong, even while many governments exercise little effective authority, and unmet needs are enormous. Any model of national government policymaking for an issue such as HIV/AIDS must take into account the role of transnational infl uences.

Thus, while the primary focus of this book is the effects of internal ethnic boundaries in national states, both the spread of the epidemic and the associated global response reflect the extraordinary movement of people, ideas, and resources across external or national boundaries. Perhaps no phenomenon reveals the interconnectedness of the world’s peoples more than our very human susceptibility to disease, one that is passed on through behaviors associated with an innate desire and need for intimate contact—largely sexual contact and the process of childbirth. If national states are limited in their authority to transform people’s lives, international organizations are even further restricted. In light of the now obvious potential to transmit disease across borders, societies depend on each other to contain the spread of infections, and donors have emerged with large sums of money to support national responses to AIDS. This global consensus heightens our expectation for action and universal conformity to scientific best practices. But in a world still largely governed by national states, “tackling” the global pandemic through deliberate policies and actions still requires state action, and this returns us to the question of why some governments act more aggressively than others.

AIDS as a Laboratory for Comparison: Politics in Really Hard Times

While the interplay of global and domestic pressures on policymaking does complicate the task of providing a parsimonious explanation of government action on AIDS, the unfolding of the epidemic makes it valuable as a case study of the politics of policymaking and state development. From a social scientific perspective, the similarity of the problem for all of the potential “subjects” (national governments) in the research implies a rare degree of analytic control. A range of policies have been considered by other scholars—policies on economic reform, pensions, health insurance, railroads, and labor, among others—but the set of developing countries affected by AIDS provides a unique laboratory in which to study the possible determinants of policymaking.

Perhaps most important among the factors making it a valuable subject of research is that the AIDS epidemic reached every region and virtually every country in relatively short order. In a seminal work, Politics in Hard Times, Peter Gourevitch argues that major transnational shocks and crises are ripe for comparative social analysis:

[They] are to countries what reagents are to compounds in chemistry: they provoke changes that reveal the connections between particularities and the general. If the comparativist can find countries subject to the same stresses, it then becomes possible to see how countries differ or onverge and thereby to learn something about cause and effect.21

AIDS was a shock in the sense that it was unknown before the 1980s. Existing social, economic, and political patterns can thus be treated as largely exogenous, or independent, factors that might infl uence responses to the disease. This is a critically important property because other social scientific analyses of development outcomes—such as economic growth, infant mortality, and more general social policy—may themselves be determinants of, as much as they are caused by, other macro-level variables. While the AIDS pandemic also shapes these broader contexts, we can carry out a “first-order” analysis of government responses and safely treat these other factors as having been generated by processes independent of the one we are trying to explain.

The human immunodeficiency virus that causes the syndrome we now call AIDS existed prior to the 1980s. Blood samples taken in 1959 from the Belgian Congo (later Zaire, and more recently, the Democratic Republic of Congo) were later found to have been infected with HIV.22 There may have been other infections in other parts of the world, and of course, such patterns are important for any full understanding of the spread of the virus. For the purposes of politics, however, the manifestation of these earliest infections was simply unexplained illness and death, generating no political response. The beginning of the history of AIDS is more appropriately dated to 1981, when the Centers for Disease Control (CDC) in Washington, D.C., reported the strange outbreak of a rare pneumonia and a rare skin cancer, Kaposi’s sarcoma, among a handful of gay men in two cities in the United States. By 1985, a case had been reported in every world region, with virtually every country reporting a case by the early 1990s. The epidemic was initially reported among gay men in cosmopolitan cities in the wealthiest countries, but within a few years, the widest spread would occur in the developing countries, with most transmissions occurring through sexual contact between men and women.

Measuring and Interpreting the Size of the Epidemic

It is important to highlight at the outset how I measure and interpret the relative size of the epidemic across time and space. It is critical to clarify that I do not interpret either the share of the population (prevalence) or relative change in the numbers of new infections (incidence) as a proxy for the impact of government action. Doing so would require that we assume the consistent efficacy and immediacy of policies, neither of which I am prepared to do, given clear evidence to the contrary. Many other factors determine patterns of infection, and certain measures may have a counter-intuitive relationship to a government’s efforts to contain the disease. For example, effective treatment may boost the prevalence of AIDS relative to other countries because infected individuals live longer and remain potential sources of infection. A study of the impact of policies would require an entirely different form of research and analysis than what I have endeavored here.

Rather, the point of describing country-level trends in the epidemic is to provide some nuance to the claim that countries all faced the same shock. Regional and subregional patterns of infection have been distinctive, and any complete analysis must ultimately consider these differences. It certainly stands to reason (and I confirm empirically) that countries with bigger epidemics would respond more aggressively than countries with smaller ones.

Unfortunately, it has never been easy to identify reliable data on the number of people infected with HIV. Among other reasons, people have not wanted to get tested for the virus, they have not wanted to report their status, and governments and health facilities have not had the capacity to carry out testing. We do not know exactly how many people have been infected with the virus, but researchers and public authorities have estimated the number by extrapolating from smaller groups of people, just as in surveys of public opinion, and these estimates have been used for discussions of policy. In particular, the World Health Organization (WHO) and the United Nations AIDS agency (UNAIDS), as well as governments and ministries of health, have reported such data. While all surveys draw inferences from samples of a population, in the case of HIV/AIDS, the samples have been particularly unrepresentative. (Only recently have governments authorized nationally representative surveys that actually test for HIV.) Most HIV data have been gathered from public health facilities, voluntary reporting, or focused collections of data among groups presumed to be at high risk for HIV. Not only are such groups not representative of the population at large, but inferences about how trends in these subpopulations are related to trends in the larger population have often been based on erroneous assumptions.

Late in 2007, UNAIDS announced significant revisions to its global estimates of the magnitude of the epidemic based on new data from surveys, as well as new models extrapolating from traditional samples of HIV infection. The estimated number of people living with HIV in 2007 was 33.2 million, whereas in 2006 the number had been estimated at 39.5 million. The report went to great lengths to explain that this difference was due mostly to revisions in the estimation procedures.23 There is reason to believe that most estimates of prevalence previously reported could be improved with these recent procedures. In most cases, the relative ranking of country epidemics would not change, with the one important exception of India—a case discussed in more detail in chapter 5.

Such uncertainty about the size of the epidemic might seem to bedevil the entire social scientific enterprise of studying the AIDS pandemic in comparative perspective. In fact, this is not the case. For politics and policymaking what matters is not the actual number of infected people, but the information that is available to policymakers. For example, on the issue of environmental politics and policy, we do not know for certain what effect carbon dioxide emissions have on the ozone layer, climate change, or human development, and new evidence will almost surely suggest that scientists have under- or overestimated the precise threat. Nonetheless, contemporary political and policy action can be understood only in terms of what scientists have said up until the point that actions are taken. If we were trying to estimate a model of an outcome directly affected by the actual numbers of people infected with HIV, such as infant mortality or overall life expectancy, we would want the best available estimates, corrected retrospectively. Since politics and policy are the outcomes under investigation here, a better approach is to use estimates of HIV infections available during periods of policymaking.

As one source of data for comparative analysis, I have used time-varying estimates of infections provided in a WHO/UNAIDS map.24 These data provide four consistent periods of estimates for almost all countries across the subject regions of Africa, Asia, and Latin America/Caribbean. These estimates are reported as wide uncertainty intervals, refl ecting the low level of precise knowledge about numbers of infections, and I use the midpoints of these intervals as the basis of analysis. Again, if the goal of the analysis were to make predictions about health and epidemiological outcomes, this would be a faulty strategy, because we would want much more precise data. For the study of the politics of policymaking, however, these data provide the best estimates by technical experts on the relative size of the epidemics at different moments in time, which is the most relevant baseline for comparing policy responses. For 2003, more fi ne-grained data on HIV prevalence are available from UNAIDS 2006, and these correlate with the UNAIDS/WHO 2004 data at (Pearson’s) r = .96. This suggests that even when using broad estimates of HIV prevalence, it is possible to sort out relative differences in the size of the epidemic extremely well, giving me confidence that we can use these data as the foundation for comparing the epidemics faced by countries in the three developing regions.

These data are depicted graphically in figure 1.1. As measured by the share of the adult population 25 living with the virus (HIV prevalence), sub-Saharan Africa has been hit hardest, and by the mid-1980s, many countries in central Africa already had widespread epidemics. The African epidemic would become increasingly severe in the southern part of the continent, and by 2003, most southern African countries faced extreme epidemics.26 Latin America, and especially several Caribbean countries, have faced epidemics, but no country in that region has faced an estimated prevalence in the double digits. Overall, HIV/AIDS has come latest to Asia, and there remains a great deal of variance in levels of infection in that region. A few countries in Southeast Asia, including Thailand, Vietnam, and Cambodia, were among the earliest to identify epidemics, and HIV has continued to spread throughout the region.

Implications of the Biology and Science of HIV and AIDS

For the most part, nuanced discussions of biology and immunology are absent from these pages, and I treat the AIDS pandemic as a single biomedical phenomenon. This is a necessary simplification of a complex reality. For example, there are multiple strains of the virus, and some are more virulent than others.27 In the arenas where policies are set, however, these distinctions do not appear to have had any significant effect on the interpretations of AIDS, or on the range of prevention and treatment options available. Good clinicians know that individuals present viral loads and symptoms in a variety of ways and that best practice recommends approaches tailored to the individual. At the macro level of explaining government responses to AIDS, however, such subtleties do not fi gure prominently in the political calculus, and I do not consider them further.

Other facts and properties of the AIDS epidemic are worth highlighting because they shape the visible manifestations of an otherwise invisible virus, and have implications for politics and policymaking. First, as compared with many other viruses, there is a long interval (as long as seven to ten years or more) between infection and onset of the symptoms, a gap during which arguments over risk and the proper role of the state can flourish. In this sense, AIDS is a problem akin to environmental degradation or cardiovascular health, in that by the time one observes the manifestation of the problem, solutions are more costly and less effective than they would have been if it had been addressed earlier. Particularly where germ theories of disease are not universally accepted, a wide range of political interpretations of increased sickness and mortality may come into play, opening up contentious discussions of the proper role for government and “modern” science.28 By contrast, in the case of fast-moving epidemics, such as SARS or Marburg, a swift government intervention, in coordination with international partners, is much easier to justify in political terms. While the time from infection to full-blown AIDS depends on factors including overall health and nutrition, in virtually any person who does not get tested, HIV is a silent killer that breaks down the immune system in an asymptomatic fashion over a very long time.

Second, the virus is contagious, but it is transmitted almost exclusively through intimate and deliberate acts. HIV gets transmitted through the blood, sexual fl uids, and breast milk, primarily through extremely intimate human contact—particularly unprotected sexual relations and between mother and child, either through the process of delivery or breast-feeding. It can also be transmitted by the direct transfer of human blood, most often through blood transfusions or through the sharing of needles in medical facilities or among drug users. By far, sexual contact has been the predominant means of spread. Of particular importance, as compared with other viruses that are transmitted through breathing or sneezing or through food, these properties have important political implications because it is relatively easy to imagine and to discuss risk of infection as a matter of lifestyle. That is not to say that every individual has a full choice about lifestyle or sexual behavior—in particular, sexual violence against women is a leading route for infection—but the transmission routes allow citizens and political leaders to imagine infection as a choice (as contrasted, again, with SARS, for which infection may seem beyond one’s control). Again, such characteristics do not make AIDS unique—other social problems such as poverty and unemployment are often understood in particular contexts to be choices, which can dampen political support for redistributive policymaking.

Because HIV is primarily transmitted through sexual contacts, and because it was fi rst identified with homosexual contact, it has remained a stigmatized problem that elites and citizens have often preferred to ignore.29 For that reason, government policymakers—particularly decision-making chief executives and ministers of health and fi nance—may view international best practices (what I describe in chapter 3 as the “Geneva Consensus”) as a high-cost, low-benefit proposition, and may choose inaction as their preferred strategy. Increasing prevalence, along with domestic pressure from organized activists, physicians, and other groups, and external pressure from a wide variety of sources increases the likelihood of government action, but the impact of such pressures is likely mediated by particular social and institutional environments.

Third, there is neither a vaccination nor a cure for HIV or AIDS. Rather, as chapter 3 discusses, there are strategies for preventing the transmission of the virus, drug treatment therapies that improve and extend the lives of those infected, and policies for mitigating the impact of disease-related disability on the infected and their dependents. Despite ongoing scientific research, to date there is no viable plan to eradicate HIV or AIDS, as was the case for smallpox and polio. As a result, the effects of the pandemic can be mitigated, but HIV and AIDS will not be wiped off the planet until technologies derived from scientifi c breakthroughs are implemented. In fact, the availability of treatment options implies that people infected with HIV will live longer lives, increasing the prevalence of infection, while presumably increasing the quality of life of those infected. Thus, existing technologies allow only the possibility of managing the epidemic, not ending it.

Outline of the Book

In the chapters that follow, I identify general theories and testable implications, and I attempt to analyze available evidence in a rigorous, skeptical, and dispassionate manner. The organization of the book refl ects my attempt to parse out theoretical claims from empirical analyses, and international dynamics from domestic ones. A brief overview will highlight how the analysis and evidence fit together, and provide guidance for those who may choose to skip ahead to a chapter of particular interest.

In chapter 2, I draw upon general theories of politics to develop propositions about how and why governments have responded to AIDS in different ways. I develop more fully the theoretical explanation linking boundary institutions to aggressive policies, through consideration of theories of ethnic politics, social identity, and the construction of risk. I also identify alternative hypotheses, including the effects of state capacities, regime types, and international influences, which help to guide the analyses, particularly by informing the choice of comparisons and control variables in the empirical analyses. While adjudicating among competing explanations of policy outcomes, I avoid investigation of the most proximate explanations. For example, it would be difficult to challenge the conclusion that a government adopted aggressive policies because those in charge of government policy acted aggressively, or that a particular president made AIDS a top priority. The relationship between cause and effect is so close that they are almost indistinguishable. “Leadership” cannot be understood as the “cause” since it is part of the outcome I am trying to explain. Instead, I attempt to make sense of key outcomes and seemingly anomalous events in terms of broader structures and factors, particularly ones that vary across space and time.

Chapter 3 details the globalized environment in which states have faced pressures to respond to AIDS. An elaborate international apparatus has developed to partner with and to cajole governments to respond to the epidemic aggressively and within guidelines. While the ultimate focus of the book is on politics and policymaking in developing countries, we cannot lose sight of the fact that most of the technology, resources, and push for AIDS policies has come from above, through massive efforts at global governance. My prior expectations for similarity in national responses is founded on research pointing to extensive consensus on best practice, as well as far-reaching efforts to disseminate that consensus to every country in the world.

In chapters 4, 5, and 6, I turn to the politics of policymaking in countries, through wide-ranging empirical analyses. The work was completed with a healthy awareness of the potential for bias or spurious fi ndings in any particular approach, and with skepticism about how factors are measured and causal inferences drawn. I have tried to highlight the degree to which analytic findings are sensitive to particular interpretations. The research presented herein was a mixed-method enterprise, combining case study and statistical analyses. This approach allowed leverage over an eclectic array of data sources, and an ability to assess the robustness of my results. Only the most narrowly construed hypotheses are likely to have implications that can be decisively tested by any single research design or set of designs. And yet to the extent that the central explanation can help us to tell compelling narratives of particular countries, while providing a guide for estimating a robust statistical model of variance in AIDS policies, we can conclude that the general model provides real insights into the social and political dynamics of the problem.

Chapters 4 and 5 deploy case studies and structured comparisons 30 for generating and testing theoretical claims. While cross-country statistical analyses (discussed below) may be better suited to the estimation of the magnitude of effects, the generalizability of findings, and the analysis of multiple rival explanatory factors, case studies, conducted with attention to measurement in context, and to exploration of feasible causal pathways, also provide opportunities for evaluating general claims. We ought to be able to tell compelling narratives that link cause to effect, particularly for cases with extreme scores on the central explanatory variable. If we cannot identify a logic in the particularities of a case or set of cases that resonates with the more general theory, and if we cannot conclude that history would have unfolded differently with different scores on the explanatory variable, we lose confidence in our general claims.

On the other hand, we should not expect policymakers to specifi cally identify the political incentives and constraints on their actions in the terms presented in this book. Policymakers who have successfully addressed a problem such as AIDS are likely to highlight their own foresight and problem-solving skills, not the fact that they were unfettered by constraints faced by policymakers in ethnically divided countries. Most policymakers either defend their policies as appropriate and suffi cient, or point to competing demands, or perhaps the morals and values of their culture, as constraints on further action. It is not appropriate to demand, as a test of a social scientific theory, that citizens or elites describe their behaviors with the theoretical framework.31 If everyone were cognizant of the determinants of their behavior, there would be little need for social science. In fact, the policymaking literature has relied too much on actors as the source of explanations, when, of course, their perspectives are biased. My approach has been to look for traces of the consequences of ethnic-based confl ict for AIDS policies, in the form of otherwise unexpected political resistance to them, comparing episodes where boundaries were weaker or stronger.

I examine boundary politics and AIDS policy in Brazil, India, and South Africa, cases that provide a wide range of variation in hypothesized cause and effect. Brazil has earned a reputation as an aggressive (and effective) responder to AIDS, whereas India and South Africa have been criticized for lackluster action. These countries are economic and political leaders in their respective world regions. Each has worn the dubious mantle of containing the largest numbers of infected citizens in its region, attracting a great deal of international attention. There would have been reason to expect aggressive responses to AIDS in all three countries because each had a significant viral outbreak, substantial industrial or governmental capacity, and a relatively open and democratic political regime. Compared with other countries at analogous levels of development and with similar epidemics, Brazil emerges as a relative overperformer and South Africa and India as underperformers. The goal of the analysis in this book is to see whether the location of these cases can be explained through a consistent political model of policymaking, or if they refl ect unexplained “noise” associated with the random nature of social and political life.

The comparative analysis of AIDS policymaking in Brazil and South Africa presented in chapter 4 is a model-building analysis, in the sense that the explanation of differences in the outcome of interest was developed using “backward induction,” subjecting the comparison to theoretically driven hypotheses about why Brazil developed a more aggressive response to AIDS than did South Africa. This puzzle was of particular interest because in my own prior research,32 I was impressed by the similarities between the two countries in levels of economic development, social inequality, and racial diversity. These similarities formed the basis of fruitful comparative analyses. The importance of race politics in the discussions of AIDS in South Africa was evident to me prior to formal engagement with the materials in comparative perspective, and sustained comparative analysis suggested the mechanism linking “race” politics to aggressiveness on AIDS through the transmission of information and notions of risk. While the outstanding Brazilian response has been linked to many factors, including its democratic transition and vibrant civil society, those factors also appeared to be present in South Africa. What is striking in Brazil is that the weakly institutionalized category of race was not politicized for most of the history of AIDS, making possible a politics of national solidarity. In South Africa, discussions of risk, and a politics of blame and denial, broke down along racial lines, lowering demand for and supply of aggressive AIDS policies.

The Indian case was a model-testing analysis in the sense that investigation of the politics of AIDS largely followed the development of the theory. India shares similarities with Brazil and South Africa as a regionally dominant democracy facing an AIDS epidemic, with the capacity in the bureaucratic and private sectors to formulate a response. Like Brazil, it has a history of federalism. But like South Africa, India contains strong boundary institutions, and the national government did not respond aggressively to AIDS for most of the history of the epidemic. Beyond establishing this controlled correlational confirmation of the central hypothesis, the real test was to trace causal processes 33 and to see if I could fi nd compelling evidence linking the hypothesized cause to effect. As I report in chapter 5, I found evidence connecting the country’s ethnolinguistic, religious, and, especially, caste boundaries, to the politics of risk, and to patterns of blaming and shame avoidance, impeding a strong response. I also demonstrate with statistical evidence that variation in interethnic relations across Indian states helps to account for some of the variation in state-level responses to the pandemic in that country.

Analysis of all three countries is based on a wide range of case materials. First, I analyzed a vast library of documentation and data from governments and domestic and international nongovernmental organizations. The domestic and foreign presses have covered politics and policymaking extensively in all of these countries. Moreover, a significant secondary literature describes both politics and the AIDS epidemic in each country. I complemented these materials with targeted field research in Brazil, India, and South Africa (as well as brief stints in Uganda and Thailand) to gain additional information through open-ended, semistructured interviews and brief consultations with political elites, decision makers, activists, and bureaucrats; and to obtain additional materials not otherwise available. I attended dozens of lectures and seminars on these cases, including the two most recent international AIDS conferences (Bangkok 2004; Toronto 2006), where I listened to presentations and met with dozens of scholars, care-providers, government officials, and representatives of international organization doing work in these countries. For the most part, however, I don’t detail insider accounts of AIDS politics and policymaking. In general, insiders have good access to data and to facts, which I have gathered in volumes, but their analyses tend to support their own political or other agendas, noble or not. Instead, I have relied on nonreactive measures, guided by involved actors in the search for facts and data verifying their assessments with additional evidence. As much as possible, I have used widely available news reports, documents, and publicly available data in order to maintain transparency.

While comparative analysis of these three important countries provides insights into the dynamics of AIDS policymaking, especially the ways in which ethnic boundaries shape political dynamics, such analysis leaves key questions unanswered. To what degree does the book’s central proposition “travel” to other countries? Does it apply only to large, regional powers? Does the explanation I offer wishfully interpret the facts? In chapter 6, I present statistical analyses of a multicountry dataset, which helps to answer such questions, to adjudicate among alternative hypotheses, and to obtain estimates of the magnitude of the impact of boundaries on policy.

The domain of analysis was limited to low- and middle-income countries. To be certain, the industrialized countries have differed in their response to AIDS as well, as historians and social scientists have expertly documented.34 But on substantive, methodological, and practical grounds, there is good reason to focus on the poorer countries, where more than 90 percent of the world’s infected population lives, and where, by defi nition, the resources and institutions available for dealing with any threat to public health, especially one as diffi cult as HIV, are limited. The analyses presented in chapter 6 are analogous to a significant body of empirical work that has tested the effects of ethnic fractionalization on the provision of public goods and development across countries.35 However, because those studies have considered policies and outcomes developed over long periods of time, it is difficult to know whether they are causes or consequences of ethnic politics. Because HIV/AIDS is a new and historically specifi c phenomenon, we can be certain that the outcome of AIDS policy is not endogenous to the explanation of ethnic boundaries, even if we assume that AIDS and AIDS policy will affect ethnic demographics and politics over the longer term. Overall, the statistical results confirm the central hypothesis, while bringing to light important differences across world regions and policy areas.

In chapter 7, I conclude by highlighting the ways in which the fi ndings resonate with debates about identity politics, particularly among normative theorists. Scholars such as Amartya Sen and Anthony Appiah have emphasized the intrinsic values of combining cultural diversity with malleable institutional approaches to identity, and Appiah in particular makes an ethical case for cosmopolitanism. My positive empirical analyses emphasize the potential implications of such an approach for development outcomes. Permeable ethnic boundaries facilitate perceptions of interconnectedness and may help authorities to gain sacrifices for the public good. By contrast, the rigid institutionalization of ethnic difference impedes the pooling of risks, and encourages competition instead of cooperation. If well-intentioned international actors want to promote cooperative governance of global public problems, they would do better to promote cosmopolitanism than to strengthen ethnic boundaries.

Return to Book Description

File created: 2/24/2009

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share