There are actually two types of impossibility embedded in the public health executive's charge to control the spread of HIV infection. The first stems from the collective-action nature of the problem and is seemingly abstract, but in practice it stands as the most intractable obstacle to the successful control of AIDS. The second type of impossibility is more commonly observed in public management: It is the problem of meeting mandated objectives in an environment of competing goals, ambiguity, and legal and fiscal constraints-all under enormous demands on personal time, energy, and intellect.
AIDS cannot be controlled by the efforts of a public health official because at root there is a fundamental mismatch between the risks and rewards of individual behavior and the collective interests of the groups at highest risk.
To understand the impossibility of the public health executive's job in controlling the spread of HIV infection, it is necessary to sketch a model of the collective response to AIDS. This model, consistent with Mancur Olson's theory of individual and group behavior in the face of a collective interest, provides a partial but significant explanation for the apparently slow response of the high-risk groups to the threat of HIV infection.
The collective interest in controlling the spread of HIV infection may seem obvious, but it needs to be articulated for the purposes of the analysis. Because the virus is invisible and its consequences so socially and economically devastating, it would seem to everyone's advantage to promote measures that limit the spread of the infection. The diffusion of HIV infection out into the population can be viewed as a "collective bad": collective because it is effectively impossible to prevent large numbers of people from engaging in those behaviors that will bring them into contact with the virus and because many of the benefits of prevention will spill over to society at large. The imagery of a collective bad is even more powerful since many people do not know they are infected (and can pass the infection on to someone else) and may remain asymptomatic for a long time.
The epidemiological evidence suggested early on that the spread of HIV infection was concentrated among a few groups: gay men, IV drug users, and their sexual partners. Some minority communities, by virtue of the numbers of people who fell into these categories, became areas of high prevalence in their own right. The early mobilization of any of these groups to control the spread of HIV infection is a function of several variables: the degree of perceived risk (and by extension the perceived value of risk reduction), the costs of risk reduction, the efficacy of risk reduction, and the size of the group. Where the risk is perceived to be low, the group large, the efficacy of protection uncertain, and the "costs" of prevention high (in sexual freedom and in drug-taking ritual), little mobilization can be expected. This was essentially the situation during the early period of the epidemic. Alternatively, if risks were considered high, the group small, the efficacy of prevention certain, and the costs low, it would be reasonable to expect a mobilization to control the epidemic.
Many members of the gay community and even some public health officials believed that control of the spread of the epidemic would be arrived at internally. In the face of the threat of AIDS, the affected groups, especially the gay community, would organize and develop workable prevention strategies on their own. The logic of collective action, however, .suggests otherwise: As long as the perceived risk of infection remained low and the membership of the group was large, no mobilization of the group to control the spread of infection would be forthcoming.11 In Olson's terms, the groups potentially affected by AIDS could be expected to remain latent until the calculus of transmission changed dramatically. Most people would not deem it in their interest to make the changes necessary to control the spread of the infection (i.e., engaging in "safe sex") when the potential reduction in personal and group risk appeared so small.
There are several additional explanations why the "payoffs" of these risky encounters might overwhelm the perceived harms. First, the carrier or partner may be ignorant of the presence of infection, or the person at risk may be ignorant of his or her partner's other activities. Second, even if people are aware of the possibility of infection, the probability of transmission may be quite small for a single encounter. Further, in situations of passion or craving, a one-in-one-thousand chance of transmission may not seem as consequential as it might be when viewed from a coldly rational perspective. Finally, the' 'rewards" of sex and drug taking are obviously powerful inducements to behave in ways that are clearly not in a person's long-term best interests. One has only to look at the recent history of efforts to control venereal disease, teenage pregnancy, or overall drug abuse to appreciate the stubbornness of these behaviors despite public interventions. The problem of controlling the transmission of HIV among intravenous drug users is especially difficult because there are significant ethical and legal objections to preventing transmission without addressing drug abuse and because the behaviors associated with needle sharing are elusive and ritualistic.
Examining the collective-action aspects of the AIDS epidemic provides three important insights about the role and performance of public health officials. First, the underlying dynamics of the transmission of HIV infection suggest that no aggressive and rapid group response-especially among the IV drug-using population-would emerge early in the epidemic. Second, analysis of the collective-action properties of the epidemic suggests what kinds of public health strategies are indicated and what others are less defensible. Third, and more important for the public health official, an understanding of the collective-action dimension of this problem provides insight into the special political dynamics that will exist during the initial stage of the epidemic. In particular, public health officials should not expect the immediate formation of a constituency for a public-health-based "reasonable AIDS policy"; they should instead anticipate disproportionate influence from small existing groups with peripheral agendas. In many states a rapid and politically powerful response to AIDS was generated by groups with a conservative moral agenda not specific to AIDS. This phenomenon can be explained in part as ' 'by-product" behavior; because such groups are fairly small, narrowly directed, and extant, it is relatively easy for them to shift their effort to an issue that overlaps many of their members' interests.
This theoretical discussion suggests that the choice whether to intervene or not is a choice between two equally ineffective alternatives. First, in order to intervene effectively, the commissioner must affect the risk/reward perceptions of people so as to alter their sexual and drug-taking behavior in very specific ways that violate strongly held social norms and values in the high-risk group. Such behavior change is unlikely, and the disease control, ineffective. Second, to refrain from intervention and let nature limit the disease will probably mean that the prevalence of AIDS will reach very high levels before it flattens out. There is some evidence that, without intervention, prevalence can reach levels above 50 percent in some communities, a disastrous outcome. When both these ineffective outcomes are considered soberly, the prospects for rapid and effective control of AIDS look extremely disheartening to a public health commissioner.
The theory that responds to problems of collective action suggests three possible resolutions to the AIDS quandary. Coercion, the exercise of the police powers of state, is a direct and obvious response. Public health officials are empowered both by statute and by tradition to employ coercive measures in the face of such a compelling threat to health as AIDS. For reasons that will be discussed below, however, the use of aggressive public-health police powers for this epidemic met with overwhelming political and logistical obstacles. A second response, based on the so-called by-product theory of organization, relies on the existence of an already-mobilized group to promote or take action for interests that are secondary to their membership. The already-organized gay community in San Francisco, for example, might have been expected to transfer their considerable organizational and political skills (as by-product behavior) to the fight against AIDS, but their response was delayed because fighting AIDS directly conflicted with many of the deeply held values of the gay population: sexual freedom, civil rights, and assimilation into the larger political community. The logic of collective action suggests something about the pattern of interest-group response during the early period of the epidemic: Groups that exist for some other purpose will be first and most effective in mobilizing a response. A public health official who perceives his or her role as a dispassionate broker of interests in AIDS policy should understand that there is no reason to expect those interests to be the same as his or her own or as those of other groups. The third theoretical resolution to this problem requires the emergence of political entrepreneurship. Although this attribute was evident in the leadership provided within groups, such as the Shanti Project in San Francisco and the Gay Men's Health Crisis in New York, and in the scientific research community, there are few striking examples of political entrepreneurship in state and municipal public health departments during the early experience with AIDS. The meaning of all this for the job of the public health executive is primarily that mobilization to control the spread of AIDS should not have been expected to develop indigenously among the affected high-risk groups during the initial phase of the epidemic. People who engage in "consensual" risky behavior apparently had little to gain and much to lose by acting in the collective interest. The public health executive is, in effect, the steward of that collective interest. This role carries perhaps the most difficult public policy responsibility: altering individual behavior.
