Drug Abuse

DRUG USERS VERSUS OUTREACH WORKERS IN COMBATING AIDS

Part I Agency problems in traditional outreach interventions

Robert S. Broadhead, Douglas D. Heckathorn, Jean Paul C. Grund, L. Synn Stern, Denise L. Anthony, Eastern Connecticut Health Outreach Project, University of Connecticut, USA

INTRODUCTION

AIDS prevention efforts for injecting drug users ( IDUs) since 1988 in over 60 inner-city areas within the United States of America have been based on a 'provider-client' model called 'street-based outreach' (National Institute on Drug Abuse, 1992; Wiebel, 1992; Brown and Beschner, 1993). The model involves hiring a small number of community members, usually ex-addicts or people with street credentials, to contact and work with members of their own community as clients. They do this by going into neighbourhoods as 'outreach workers' to distribute AIDS prevention materials and information, and to recruit IDUs to various programmes and services, including research interviews conducted by social scientists.

Research has shown that outreach projects operate under conditions that cause hierarchy and super vision to break down easily. The result is an array of organisational problems that push outreach projects towards inertia, and invite high levels of mal- and non performance by outreach workers as rational adaptations to their work conditions. Nevertheless, research has also documented that IDUs responded very impressively to the outreach services that they received; IDUs adopted many risk-reduction measures at high rates, and they volunteered and substantially augmented the efforts of outreach workers. Put simply, IDUs went well beyond the role of being mere clients; their response to traditional outreach was far more robust and far-reaching than were the efforts of outreach projects themselves.

In light of users' responsiveness, many researchers have called for future AIDS prevention efforts based on an intervention model that relies on IDUs as active collaborators (Wiebel, 1988; Chit wood et al., 1990; Des Jarlais and Friedman, 1990; Carlson and Needle, 1991). Such an intervention would contrast with the traditional model that turns IDUs into clients of, and dependent on, paid staff of outreach workers. The model would also draw upon and strengthen the sharing rituals and norms of reciprocity that already underlie and sustain drug user networks in the first place (Preble and Casey, 1969; Grund, 1993). It would do so by enhancing the mutual opportunities and incentives for IDUs to work with their own peers, and to invest themselves in their own intervention. One such model, called a 'peer-driven intervention' ( PDI ), began operating in March 1994 in eastern Connecticut ( Broadhead and Heckathorn, 1994). The PDI is being systematically compared with a traditional outreach intervention operating in a different community site. This two part article* presents some of the preliminary ethnographic and impact data on the operation of both interventions for the first eight months of a five year study, comparing the relative effectiveness of a PDI with a traditional outreach intervention.

In the first section of this paper, using agency theory, we provide a research overview of the common organisation problems that have been found to hinder traditional outreach projects and blunt their impact. In the second section, we document IDUs' unexpected responsiveness to the outreach efforts that they received across the country, however inefficient and misdirected those efforts were. In the second paper, section one, based on the 'theory of group-mediated social control' (Heckathorn, 1990), will describe the theoretical basis and operational features of the PDI that was implemented in eastern Connecticut. In the second section of the second paper, we will describe the 'provider-client' model and operational features of the traditional outreach intervention that was implemented in a different area in eastern Connecticut, and that is being compared with the PDI. In the final section; we will present preliminary impact data comparing the effectiveness of the PDI with the traditional outreach intervention in terms of recruitment power, educational effectiveness of IDUs in the community, and comparative intervention costs.

AGENCY PROBLEMS IN TRADITIONAL OUTREACH INTERVENTIONS

Ethnographic analyses have accumulated on the inner workings of outreach projects, including how outreach workers perform in the community in reaching IDUs (NIDA, 1991; Longshore, 1992). The discussion below draws on this literature, and from a longer analysis published elsewhere (Broadhead and Heckathorn, 1994). The analysis is based on an ethnographic study of an AIDS outreach project to drug injectors in San Francisco which consisted of a year and a half of participant observation, during which time the first author and two, full-time, associate ethnographers were trained as outreach workers, and were deployed as active members of several outreach teams in targeted areas of the city. It includes many 'natural' interviews with outreach workers that occurred in the field while working together, as well as formal interviews with 24 of the 33 outreach workers employed by the San Francisco project between July 1988 and January 1990. All quotes below that lack citation come from these interviews. Also, in October 1989, the first author and an associate spent 2 weeks on the street s with an AIDS outreach intervention for drug injectors in New York City at two different sites - Brooklyn and Queens. The observations in San Francisco and New York, and the available studies of outreach in other cities, reveal several commonalities in the manner in which outreach projects function across the country, and the organisational problems they experience.

Agency theory (Jensen and Meckling, 1976; Eisenhardt,1985; White,1985) provides a useful perspective for analysing the inner workings of outreach projects. The theory focuses on the informational ` asymmetries between individuals who contract for a service (principals) and those who hire them (agents). For example, in the relationship between patients (principals) and physicians (agents), the latter's vastly greater access to specialised medical knowledge creates opportunities to control patients through evasion, dissimulation, mystification and many other deceptive practices (Waitzkin, 1991). Similarly, in the relationship between clients (principals) and lawyers (agents), the latter can use their specialised legal knowledge in ways that lead clients to act against their own interests (Bok,1978). Any bureaucracy can be seen as a chain of principal-agent relationships that link principals ('superordinates') to agents ('subordipates') charged with fulfilling their delegated responsibilities. However, subordinates' differential control over information frequently enhances their power and abilities to control their superordinates or to pursue their own interests despite their superiors' wishes to the contrary.

Outreach projects can be analysed at many different levels, including that of funding agencies and principal investigators, principal investigators and outreach supervisors, and supervisors and outreach workers. Here, we limit the analysis to the latter relationship, because it is the closest to the street level at which outreach interactions occur with IDUs, and because analysing higher-level agency problems would exceed the scope of this article. However, some of the problems identified in the performance of outreach workers may derive from agency problems occurring at higher organisational levels (for example, see Broadhead and Margolis, 1993).

Outreach and the problem of adverse selection

According to agency theory, two fundamental types of problems arise when an agent's interests fail to coincide with those of a principal. The first problem occurs before a principal hires an agent's services, and is termed 'adverse selection'. In general, agents with the strongest desire to offer their services to a principal are those who are least qualified or motivated. To begin with, applicants who respond to job ads do not come from a random sample of all people who are qualified, because most such people are already satisfied with their current employment. Instead, most responses come from people who are unemployed or are in the process of losing their current jobs. This group contains a larger proportion of workers with problems in competence or reliability than does the potential applicant pool at large. In addition, identifying the best candidates for a job is especially difficult because the least qualified applicants have the greatest incentive to withhold information revealing their deficiencies.

Adverse selection problems arise to varying degrees in all organisations, because available information on applicants is often fragmentary and unreliable.** Consequently, much organisational energy is devoted to overcoming this problem. For example, well-managed police departments employ psychological tests to identify and eliminate bullies and sadists who are attracted by the opportunity to use force. Similarly, financial institutions employ background checks to eliminate con artists who seek opportunities to steal. Such screening devices are, at best, only partially effective.

Research indicates that outreach projects have experienced severe adverse selection problems in identifying highly motivated and capable candidates for hire. This is because some of the credentials that outreach projects looked for in hiring outreach workers are the same kind that often keep people from getting a good job; e.g. a former drug habit, a prison record, socialisation to lower class culture, street smarts about hustles and con jobs, previous gang membership or a former career in prostitution. Some people hired as outreach workers have conventional credentials, such as college degrees and journeyman skills, but they all must have street credentials (Broadhead and Fox, 1990). Such credentials, however, by their very nature, portend problems. Among street cultures, as Stephens (1991, p. 47) explained, 'high status is conferred on those who are most adept at conning'. Obviously, in actively recruiting people with solid street credentials, outreach projects set themselves up to hire some applicants who are already prepared to pull a con job on the projects themselves. For example, the most frequent and simplest con job occurs during employment interviews when applicants express a heartfelt desire to help drug addicts protect their health. After being hired, it becomes apparent, sometimes quickly, that their heart was never in the job. Thus, 'going into the field' comes to include visiting the mall and shopping, goofing off with clients, chasing around, hanging out in bars, drinking beer, playing pool and getting high. As one outreach supervisor noted, 'All their problems that we deal with on the street we have in our very own agency'. It is inevitable that such agency problems include varying amounts of what Agar ( 1973 ) aptly called 'ripping and running', because many out reach workers come from, and all are asked to work in, drug-using and other hustling scenes.

Adverse selection problems are aggravated by the wealth of hustling opportunities available to outreach workers in the community. Outreach projects find that identifying outreach workers who can work in, but not join, hustling/drug scenes is fundamentally problematic. Stephens ( 1991 ) found that, as interaction with street based drug and hustling scenes increases, the greater becomes the likelihood of people's commitment or attachment to them. Thus, some outreach workers have been found to use their job as a cover for running various street-based hustles. As one outreachworker observed while working a neighbourhood in San Francisco: 'This job would be the perfect cover if you wanted to run a scam.'

Later the outreach worker was discharged by the project after he was discovered fencing merchandise on the job. Another outreach worker was confronted several times by project directors over rumours that he was orchestrating the sale of drugs while distributing bleach and condoms. But such schemes are difficult to prove. In this case, the outreach worker was a team supervisor and, in observance of the strong street ethic to never 'snitch', the outreach workers under him refused to tell what they knew to the project directors.

In encouraging outreach workers to use their street-based experience, outreach projects expect outreach workers to develop trusting relationships with people participating in complex and lucrative black market systems offering goods and services in high demand. Some outreach workers have been found to take advantage of the opportunities that they cultivate, either by dabbling to make a quick return on an investment, or getting more intensely involved.

At the same time, research confirms that many outreach workers are committed to their job and strive to conduct themselves like professionals (Johnson, 1988; Broadhead and Fox, 1990; Margolis, 1990; Rivera Beckman, 1992). For such outreach workers, the most demoralising experience that they face is having to tolerate the shirking and con jobs of their colleagues ( Broadhead and Fox, 1993 ) . Thus, a very sincere outreach worker who was able to stay with the San Francisco project for only a few months explained:

When I worked with a volunteer agency, the volunteers worked harder and longer than we did . . . So here is an agency where everyone is paid, but so little is happening, at least with the team I was with. Eventually I just felt like I was wasting my time. I even started to schedule personal things into my own work time, which I didn't think was right.

In contrast, an outreach worker who constantly had to deal with his partner's large-scale con job eventually quit in desperation:

I hate Sam, man, I just want to kill that dude!

I'm just ready to say 'Screw this job!' . . . And Sam, he's still dealing' on the job! He's got four guys that I know of workin' for him . . . The reason it bothers me is that it makes me look like a fool. I'm out there trying to do something about this epidemic.

Outreach and the problem of moral hazard

A second type of agency problem occurs after an agent's services have been retained. If a principal lacks the effective means to monitor an agent's performance, the latter can act in ways that serve his or her interest at the principal's expense. Problems of postcontractual opportunism is termed 'moral hazard', though it need not entail behaviour that is either immoral or illegal.***

All organisations contend with the problems resulting from moral hazards. Businesses lose far less money to robbery than to embezzlement, because it is impossible to watch all employees all the time, especially those in positions of trust. Similarly, no reliable means for preventing corruption among police or truck drivers have been found, because much of their work occurs out in the field (Manning 1977; Friedman, 1982; Ouellet, 1994). According to Lawlor ( 1990, p. 154), problems of moral hazard are endemic to the administration of social services. But outreach projects have been found to be severely limited in their ability to monitor the behaviour of outreach workers, for two reasons.

First, outreach workers enjoy considerable autonomy in the field for long periods of time, relatively free of supervision or colleague control. Such autonomy is a generic feature of occupations at the 'street level' (Lipsky, 1980), and many outreach workers regard it as a major prerequisite of their job. Once in the field, outreach workers have many opportunities to shirk. Consequently, research has documented that outreach workers frequently organise their days to accommodate personal matters. such as educational programmes, artistic pursuits, avocations, other jobs and con jobs. As they a generally paid very low salaries, many outreach workers keep 'banker's hours'. Working only par; time is a creative way of converting a low full-time salary into good hourly wages (Broadhead and Fox, 1993).

A second factor found to thwart the effective monitoring of outreach workers derives from local norms. In honouring the ethics of the street, as described above, outreach workers are loath to 'snitch on one another, which makes project supervision especially difficult. For example, in joining an Outreach team, one outreach worker described the

advice she was given by her new teammates: 'I was told, whatever happens in the team, stays in the team. Don't bring problems out in the staff meeting that are our business. We keep our own problems to ourselves.'

Given the lack of effective monitoring, outreach workers have extensive opportunities to act in ways that conflict with the official aims of outreach projects. These divergent actions result from ( 1 ) political conflicts, ( 2 ) conflicts between local culture and the goals of outreach, ( 3 ) the status needs of outreach workers and (4) outreach workers' reactions to the occupational risks of outreach.

Political conflicts

Political conflicts between outreach workers and outreach projects adversely affect AIDS prevention efforts. Minority communities in the USA have been for some time disproportionately at risk of contracting HIV (National Commission on AIDS, 1991). In addition, as Quimby and Friedman ( 1989, p. 405 ) noted, African American and Hispanic communities tend to see AIDS in the context of broader problems of poverty, drug addiction, inadequate education and unemployment'. As such, in 1990, at nearly the same time that the Centers for Disease Control were announcing the success of outreach services, members of the Black Leadership Commis sion on AIDS in New York City were holding their own press conference. From their perspective, the national AIDS outreach effort was a cop-out by the federal government, reflecting a failure to deal with the underlying problems that afflict minority communities. As reported in the New York Tim, (1990, p. N14), the Black Leadership Commission 'criticized public health officials in New York City for the bleach distribution, saying they were giving the poor a sop rather than real help'; and, 'bleach distribution amounts to endorsing inexpensive ways to stop AIDS from spreading among users but failing to come up with the millions of dollars needed to help users get off drugs'.

Outreach workers who share this position with their communities have difficulty remaining committed to their work. Thus, for example, a young Latino outreach worker felt compelled to quit after struggling for about 6 months, during which time he resumed a cocaine habit:

I just couldn't handle it anymore. There was something kind a weird about going up to old dope fiends and saying, 'Hey man, want some bleach?' when it's like, 'Well, you won't die from AIDS but, man, you might OD in two weeks.' The project didn't even address that! Its like, 'Oh, sure, we want to help you, so here, take some free bleach,' you know (pointing his finger to his temple indicating that this is crazy) . . . What about the real problems? It doesn't address the problem that all these kids in our community are dying from crack and from violence about drugs.

Interviews with this outreach worker's team members found that they agreed with his position but felt that there was nothing they could do, which increased their frustration. For example, one exclaimed:

I could relate to Julio a lot, but at the same time, I've got kids and a wife . . . You know, I hate the fucking system but what could I do, right? I'm stuck. At least Julio is willing to rough it and that's why [he quit]. He said, I don't give a shit, I'll rough it because the contradictions really got to him.

Instead of quitting, such outreach workers put in their time, but adjust their investment of time and effort downward on the job according to their reduced commitment to the work.

Identity conflicts between local culture and outreach

Research has documented that outreach workers' indigenous identities in the community frequently conflict with carrying out their prevention efforts. Two examples from the San Francisco project suggest the extent and complexity of the problem. One outreach worker, who had been a prostitute and hustler years earlier, was recruited by the San Francisco out reach project because of her credentials and experience. She had since become a born-again Christian. Her identity conflicts with the project were twofold. In being asked to work with prostitutes, her job as an outreach worker constantly drew attention to a former identity she wanted to forget. In addition, in being required to be non-judgemental towards her clients, she was prohibited from spreading the religious message closest to her heart, for which she wore a large, glittering pin that said 'THINK JESUS'. Her tenure with the project ended after several weeks of erratic performance. As Stephens (l991, p. 53) explained:

. . . a person is more likely to adopt new role identities (or roles) when these are compatible with existing roles and self concepts . . . When one of these new role identities does not fit. it is 'sloughed off'.

A second outreach worker exemplified the kinds of identity conflicts that 'streetwise' community members commonly experience in becoming outreach workers, even though outreach projects actively seek to hire members with solid street credentials. As the outreach worker emphasised in an interview, 'I'm a homeboy, and I'll be a homeboy `till the day I die'. As evidence of this he emphasised, 'I only talk to people that I've been introduced to'. To be a homeboy is to be 'cool' and 'in the know'. Some homeboys place a premium on being seen as a 'bad ass' (Katz, 1988). Thus, it is hard for some homeboys to don a friendly demeanour and talk to strangers about AIDS or work the streets distributing prevention materials. They worry that handing out bleach and

condoms makes them look uncool or that they sold out and became a 'do-gooder'. Thus, as an outreach supervisor pointed out, 'Homeboys can get the best contacts, but they're guilty of "going native" . . . They have trouble distinguishing between when they are working and just hangin'.' As one outreach worker quipped, 'I can't go native. I am a native.'

In general, outreach workers were found to work best with clients who were most like themselves This meant that for any outreach worker Where are many types of clients with whom they are unprepared to work. Specifically, outreach workers indigenous ingroup/outgroup alignments commonly reflect the same narrow attitudes and prejudice that are held by their peers. Thus, for example a straight Latino outreach worker, perhaps a former heroin addict, may feel confident about accessing and relating to people like himself. But he can be at a loss in having to work with Latina IDUs, or gay Latino or transvestite cocaine injectors, or speed-using male or female prostitutes, or Black crack addicts, or non-drug-using sexual partners of IDUs, or white runaway drug-using youth. Outreach project specifically hire indigenous members of targeted communities to work as outreach workers but research has confirmed that much of what outreach workers bring to the job can-compromise their performance on the job. After hiring people with street credentials, outreach projects can run blind for several months in trying to discern those they can trust.

Status needs of outreach workers

In successfully establishing themselves in specific communities, outreach workers find that they enjoy a kind of popularity and kinship with user populations; being well known and admired is a powerful reward. Yet becoming established has been found to produce inertia. Outreach workers find the prospects of having to break into new drug networks stressful In turn, outreach workers find their work more satisfying if they stay with clients who know and respect them, instead of venturing into new situations as strangers and again have to face IDUs' deep seated distrust. Thus, outreach workers commonly tend to restrict their work to areas in which they are well known and feel comfortable, at the expense of breaking into new territories. Thus, their outreach efforts bog down easily.

Status conflicts also arise between outreach workers and their clients. Outreach projects ask out reach workers to maintain a non judgemental attitude towards their clients' lifestyles and practices. Yet, many outreach workers have been resocialised by drug treatment programmes to attitudes that are highly negative towards drug use and addicts, especially in New York City. As reported by Rivera Beckman ( 1991), almost all the 23 outreach workers who staffed the AIDS outreach project she studied identified themselves as members of Narcotics Anonymous (NA). In turn, the philosophy of NA towards active users is highly disparaging: drug use per se is a repudiated activity that cannot be tolerated; users are denigrated; recovering addicts who 'slipped' are denounced by NA members and stripped of rights and entitlements that can only be earned back through humiliating submission to NA strictures and control. As a result, given their NA membership, most outreach workers in New York City had an aversion to working with IDUs and, as reported by Rivera-Beckman ( 1991 ), they commonly refused even to place bleach in IDUs' hands. Instead, IDUs had to approach outreach workers, whose way of working was to stand behind portable tables set up on the street:

In this process, then, the outreach workers' presentation of self asserts a superior status to the user. As a direct consequence, few users are being reached. Few users will walk up to the table, fewer still will stand and talk to an outreach worker whose demeanour, if not message, denigrates their status. Thus, supplies of bleach and condoms needed by users remain on tables instead of reaching their hands.

Rivera-Beckman (1991, p. 45).

Outreach workers' reactions to the occupational risks of outreach

Outreach workers enter communities and work directly with IDUs, which generally entails walking the streets of blighted, crime-ridden neighbourhoods. Individuals are hired as outreach workers, in part, because of their personal knowledge of the areas in which they will be working. Yet, as several outreach workers noted, in the past they sometimes avoided those very areas because of what they knew about them. As one outreach worker reported, 'When they said I was going to work down there as an outreach worker I said, "Oh no, that is not where I want to be"!'.

Most inner-city areas containing large concentrations of IDUs have high predatory crime rates, and many new outreach workers initially feel anxious about working in them. On a day-to-day basis, it is not uncommon to see people involved in confrontations and shouting matches; or undercover police running people down, hassling and rousting people in various locations and making arrests. In such areas, the threat of physical violence is palpable and ever present. Besides physical assault, the risk of being psychologically and emotionally assaulted is also high. Outreach workers' clients live in extremely deprived circumstances, and the majority are homeless, or addicted, unhealthy and impoverished. Outreach workers speak often, and with considerable emotion, of the psychological and emotional assaults that they experience in witnessing their clients' suffering and deprivation ( Broadhead et al., 1990; Margolis,1990).

However, research has documented that outreach workers' adjust to their work situation in ways that go far beyond merely protecting their physical safety. To reduce their exposure to disturbing situations, outreach workers tend to restrict themselves to open, public spaces. For example, an outreach worker reported in a staff meeting, 'This week we went into the City Hotel and I want to tell you, I've never seen five floors of such absolute filth in my life like we saw there'. The staff agreed that outreach workers are better off staying out of such places and positioning themselves to hand out prevention materials to clients as they come and go. Outreach workers typically position themselves where community members will know where to find them (Johnson et al.,1990). Thus, for outreach workers in New York City, as described by Rivera-Beckman (1991, p.40):

A typical day consists of setting up a table in front of a neighborhood park, by a methadone maintenance clinic . . . or soup kitchen and waiting for interested clients. This sedentary approach may be combined with a limited amount of outreach worker movement around the vicinity of the table.

Outreach workers in San Francisco and Chicago have been found to work the streets much more often, but they follow a similar strategy: maintain daily routines such that IDUs know where to find them if they want to, rather than outreach workers feeling obliged to find their clients (Johnson, 1988; Wiebel, 1988; Rivera-Beckman,1992). Thus, outreach workers' become highly circumscribed and routinised in their efforts. The strategies that outreach workers fashion to minimise the risks of outreach generally lead them to cover only a small part of the territory that they are supposed to serve.

In addition, outreach workers learn that in working to be accepted and trusted by IDUs, it is streetwise to copy IDUs' street demeanour. Members of drug using scenes try to avoid drawing attention to them selves and their activities. As a veteran outreach worker remarked, 'All those guys that seem to be just standing around out there are not just doin' nothin'! They're workin', watchin' what's goin' on, keepin' track of business: sellin', coppin', makin' connections'. Thus, outreach workers strive for a demeanour that is equally low profiled and muted. They dress down in keeping with street styles and work to be seen as cool and part of the scene. As Johnson (1988, p.33) noted of outreach workers in Houston, 'Doing "hanging out" is an attempt to blend in. Clothes are a means of camouflage. Actions are a method of concealment'. Thus, even handing bleach to IDUs followed street etiquette, as one outreach worker explained in San Francisco: 'I just hold it in my hand and pass it to them, just like it was a drug deal.' Being low profile and cool helps outreach workers allay IDUs' suspicions and forge trusting relationships with them (Broadhead and Fox,1990). But the style works steadily against aggressive and widespread distribution of AIDS prevention materials.

IDUS' RESPONSIVENESS TO TRADITIONAL OUTREACH EFFORTS

In 1988 the federal government, through the National Institute on Drug Abuse (NIDA), began funding outreach projects in over 60 inner-city areas throughout the country and, despite their many problems, substantial research has documented that IDUs' response to outreach efforts was very impressive (Brown and Beschner,1993). Researchers early on reported that IDUs began disinfecting their needles with bleach, and reducing their practice of sharing needles at high levels. IDUs also increased their use of condoms, though less successfully.

For example, in San Francisco during the winter/spring of 1986, before outreach distribution of bleach and condoms began, only 3 % of the city's estimated 15 000 IDUs reported that they regularly disinfected their syringes. Outreach workers began distributing bleach in the streets in July 1986. One year later, '55.4% interviewed reported using bleach . . . [and analysis ofl needle-sharing partners in the past year showed significant shrinkage in the reported size of needle sharing circles and increased numbers of persons who reported not sharing needles' (Wattersetal.,1990a,pp.592-3).

Similarly, in 1986, 9.8% of IDUs reported not sharing needles, which increased to 21 % one year later. During that same year, only 4.3 % of the respondents reported using condoms at least half the time they had sex. By 1987, 32.7% reported using condoms in general and '18.6% reported using them at least half of the time' (Watters et al.,1990a, p.593). Watters et al. (1990b, pp.3-4) reported that from the baseline measures taken in the winter/spring of 1986, 'there was a near doubling of HIV seroprevalence [to] early 1987, from 7% to 13%. After this point the curve is relatively flat' through late 1989. Watters et al. (1990b, p. 4) emphasised that 'major behavior change occurred immediately following the implementation of outreach and bleach distribution'.

Risk reduction by IDUs in response to outreach efforts in other cities were similarly impressive, as researchers reported in New York City, Miami, Chicago, Denver, Baltimore, Cleveland, Hartford and other sites (Chitwood et al.,1990; Neaigus et al. 1990; Weeks, 1990; NIDA, I991; Stephens et al., 1991; Wiebel and Lampinen, 1991; Booth and Wiebel, 1992; Brown and Beschner, 1993). Such changes occurred so rapidly following the implementation of outreach projects that secular trends, such as growing awareness of how HIV is

transmitted, are not able to account for the changes themselves (see Stephens et al., 1991; Booth and Wiebel,1992). In sum, research indicates that IDUs responded positively to outreach projects in making significant risk-reduction changes, a response far greater than what many experts expected. As Des Jarlais et al. (1991, p.1279) concluded, 'Intravenous drug users have surprised many policy makers and researchers by exhibiting large-scale AIDS risk reduction'.

Indeed, IDUs' responsiveness went beyond r isk reduction changes per se. In projects throughout the country, outreach workers found, and ethnographers documented, that IDUs volunteered and helped out reach workers carry out AIDS prevention efforts in many ways ( Broadhead and Fox,1990; Johnson et al., 1990; Rivera-Beckman, 1992). IDUs frequently introduced outreach workers to other users, and vouched for outreach workers in new communities. IDUs commonly helped outreach workers fill and prepare bleach bottles, and helped outreach workers distribute bleach, condoms and prevention information. IDUs were also commonly found to aid out reach workers in locating users to be interviewed or to find users who needed to return for follow up interviews. As the directors of the San Francisco out reach project reported:

In short, the IV drug users became deeply involved in helping us gather health information regarding AIDS and its means of transmission They generally looked favorably on such efforts to involve them voluntarily and encouraged their friends to cooperate in a similar fashion.

Feldman and Biernacki (1988, pp.31-2)

Similarly, in New York City, the Association for Drug Abuse Prevention and Treatment (n.d.: 3) reported that, 'users will often volunteer to help you set up your table and to bring their friends to it or distribute literature on the street . . .'. Ethnographers have even found operators of high-volume shooting galleries enforcing risk-reduction norms, as Ouellet et al. (1991, p.80) described in Chicago

Although Slim allows syringe sharing, he said, 'I discourage that,' and he makes sure everyone who needs bleach has it . . . To share a syringe in Slim's gallery is unusual; to share without first cleaning it with bleach violates gallery norms.

The Chicago outreach project concluded:

In fact, we have found that, as addicts become aware of the threat that AIDS poses, they are quite capable of assimilating a strong sense of social responsibility which can be readily channeled to include an assumed role of prevention advocacy.

Wiebel, (1988, D. 147)

In sum, although outreach projects were riddled with agency problems, it appears that such projects served as a catalyst for risk-reduction changes, and IDUs and other drug-scene members clearly augmented outreach workers' efforts. In the course of doing so, IDUs further disseminated and reinforced the strength of prevention norms within the larger IDU community. What is now known about the limitations of traditional outreach and the unexpected responsiveness of IDUs suggests the potential for a new approach to AIDS prevention that relies on, and works to strengthen, the capabilities of drug users to promote risk reduction among their peers.

IDUs' responsiveness suggests that future prevention efforts be based on collaborations between drug users and health professionals, and that IDUs are capable of assuming a role that is far more active than being mere clients. One model of such a collaboration, termed a peer-driven intervention (PDI), began operations in March 1994 and is now being analysed and compared with a traditional outreach intervention in a different site. In Part II of this paper, in the next issue, we describe the theoretical basis and design of the PDI.

Robert S. Broadhead, Douglas D. Heckathorn, Jean-Paul C. Grund, L. Synn Stern, Denise L. Anthony, Eastern Connecticut Health Outreach Project, Department of Sociology, University of Connecticut, Storrs, CT 06269-2068, USA.

REFERENCES

Agar M (1973). Ripping and Running: A Formal Ethnography of Urban Heroin Addicts. New York: Seminar Press.

Association for Drug Abuse Prevention and Treatment. Outreach Tactics. The do's and the don'ts. Unpublished paper, New York City.

Bok S (1978) . Lying: Moral Choice in Public and PritWate Life. New York: Random House.

Booth R, Wiebel WW (1992). Effectiveness of reducing needle-related risks for HIV through indigenous outreach to injection drug users. American Journal on Addictions 1: 277-87.

Broadhead RS, Fox KJ (1990). Takin' it to the streets: AIDS outreach as ethnography. Journal of Contemporary Ethnography 19:322-48.

Broadhead RS, Heckathom DD (1994). AIDS preventionoutreach among injection drug users: agency problems and new approaches. Social Problems 4 1: 473-95.

Broadhead RS, Margolis E (1993). Drug policy in the time of AIDS: the development ofoutreach inSan Francisco. SociokeicalQuanerly 34: 497-522

Broadhead RS, Fox KJ, Espada F (1990). AIDS outreach work ers. Society 27: 66-70.

Brown B, Beschner GM (1993). Handbook on Risk ofAIDS: Injection Drug Users and Sexual Partners. Westport, CT: Greenwood Press.

Carlson G, Needle R (1991). Sponsoring addict self-organization (addicts againstAIDS): Acase study. InNational Institute on Drug Abuse (Ed. ), Community-Based AIDS Pre1)ention: Studies of Intravenous Drug Users and Their Sexual Partners:Proceedings of the First AnnualNADR National Meeting, pp. 342-9 (DHHS Pub. no. 80M 91-1752). Washington DC: US Government Printing Office.

Chitwood DD, McCoy CB, Inciardi JA, McBride DC, Comerford M, Trapido E, McCoy HV, Page JB, Griffin J, Fletcher MA, Ashman MA (1990). HIV seropositivity of needles from shooting galleries in south Florida. AmericanJournal of PublicHealth80: 150-2.

Des Jarlais DC, Friedman SR (1990). Shooting galleries and AIDS: Infectionprobabilities and 'tough' policies. American Journal of Public Health 80: 142-5.

Des Jarlais DC, Abdul-Quader A, Tross S (1991). The next problem: maintenance of AIDS risk reduction among intravenous drug users. The InternationalJournal of the Addictions 26: 1279-92.

Eisenhardt K (1985). Control: organizational andeconomicapproaches. Management Science 3 1: 134-49.

Feldman HW, Biemacki P (1988) . The ethnography of needle sharing among intravenous drug users and implications for public policies and intervention strategies. In RJ Battjes, RW Pickens (Eds) , Needle Sharing Among Intravenous Drug Abuse: National and International Perspectives, pp. 28-40. National Institute on Dnug Abuse Research MonographNo. 80, Washington DC: US Govemment Printing Office.

Friedman SR (1982). TeamsterRank and File: Power, Bureaucracy, andRebellionat Workandina Union. New York: Columbia University Press.

GrundJ-PC (1993). Drug UseAsA Sodai Ritual: Functionality, Symbolism and Determinants of Self Regulation. Rotterdam: ErasmusUniversiteits Drukkerij.

Heckathom DD (1990). Collective sanctions an(l compliance norms: A formal theory of group-mediated social control. American Sodological Review 55: 366-84.

JensenMC, MecklingWH (1976). Theoryofthe firm: manage rial behavior, agency costs, and ownership structure.Journal of Financial Economics 3: 305-60.

JohnsonJ (1988). Community health outreach workers and AIDS intervention: An ethnographic analysis. Masœr's Thesis, Department of Sociology, University of Houston.

Johnson J, Williams ML, Kotarba JA (1990). Proactive and reactive strategies for delivering community based HIV preventionservices: Anethnographic analysis. AIDS Education and Prevention 2: 191-200.

Katz J (1988) . Seductions of Crime: Moral and Sensual Attractions in Doing Evil. New York: Basic Books.

LawlorEJ (1990).Whenapossible jobbecomes impossible: Politics, public health, and the management of the AIDS epi demic. In: EC Hargrove, JCGlidewell (Eds), ImpossibleJobsin Public Management, pp. 152-76. Lawrence: University PressofKansas.

LipskyM (1980). Street-LevelBureaucracy: Dilemmas of the Individual in Public Services. New York: Russell Sage.

Longshore D (1992). AIDS education for drug users: Existing research and new directions.Journal of Drug Issues 22: 1-16.

Manning PK (1977). Police Work: The Social Organization of Polidng. Cambridge, MA: MIT Press. ,

Margolis E (1990). Visual ethnography: lbols for mapping the AIDS epidemic. Journal of Contemporary Ethnography 19: 370-91.

National Commission on AIDS (1,991) . The Twin Epidemics of Substance Use and HIV. Report no. 4. Washington DC: US Govemment Printing Office.

National Institute on Drug Abuse (1991) . Community-based AlDS prevention: studies ofintravenous drug users and their sexual partners. Proceedings of the First Annual NADR National Meeting. Washington DC: US Govemment Print ingOffice.

National Institute on Drug Abuse (1992). TheNIDA Standard Intervention Modelfor Injection Drug Users not in Treatment: Intervention Manual. Division of Applied Research, Community Research Branch, Rockville, MD.

NeaigusA, SuhanM, FriedmanSR, Goldsmith DS, Stepherson B, Mota P, Pascal J, Des Jarlais DC (1990). Effects of out reach intervention on risk reduction among intravenous drug users. AIDS EducationandPrevention2: 253-71.

New York Times (1990) . Black Group Attacks U sing Bleach to SlowSpread of AIDS. June 17, p. N14.

Ouellet LJ (1994) . Pedal to the Metal: The WorkLitaesof Truckers . Philadelphia:Temple University Press.

Ouellet LJ, JimenezAD, JohnsonWA, WiebelWW(l991). Shooting galleries and HIV disease: Variations in places for

unjecting illicit drugs. Crime anl Delinquency 3 7: 64-85 .

Preble E, CaseyJJ (1969). Taking care of business - the heroin user's life on the street. InternatiotlalJournal of Addictunts 4 1-24.

Quimby E, Friedman SR (1989). Dynamics of black mobiliz l- f tion against AIDS in New York City. Social Problems 4:

403-15. :

Rivera-Beckman J (1991) . Process ethnography report on tb. AlDSoutreachand prevention program: a report tothe New l York Division of Substance Abuse Services. New York Cit s

National Development Research Institutes. l

Rivera-BeckmanJ ( 1992). Voices from an underground neeclle i exchange. Paper presented at the Third Intemational Con- t ference on Drug Related Harm Reduction, March, Me l- t boume, Australia.

Stephens RC (1991). The StreetAddict Role: ATheory of Herl ym Addicuon. New York: State University of New York Press

Stephens RC, Feucht TE, Roman SW (1991). Effects of ln intervention program on AlDS-related drug and nee(l l. t behavior among intravenous drug users. AmericanJourtlal (~t Public Health 81: 568-71.

Waitzken H (1991). The Politics of Medical Encounters. New Haven: Yale University Press.

Watters JK, Downing M, Case P, LorvickJ, Yu-Teh Cheng, Fer gusson B (1990a) . AIDS prevention for intravenous dmg users in the community: street based education and risk behavior. American Journal of Community Psychology 18: 587-96.

Watters JL, Yu Teh Cheng, Segal M, Lorvick J, Case P, Taylor F, Carlson JR (1990b). Epidemiolgy and prevention of HIV in heterosexual IV drug users inSan Francisco,1986-1989. Paper presented at the Sixth International Conference on AIDS, June, SanFrancisco.

,,

WeeksMR(1990). ProjectCOPE: PreventingAIDSamonglnjec aonDrugUsersandtheirSexualPartners-Comprehensive Data Report. Hartford, CT: Project COPE.

Weibel WW (1988) . Combining ethnographic and epidemio logic methods in targeted AIDS interventions: the Chicago model. In: RF Battjes, RW Pickens (Eds), Needle Sharing among Intravenous DrugAbusers: National and International Perspecaves, pp.137-50. National Institute on Drug Abuse Research, Monograph no. 80. Washington DC: US Gov emment Printing Office.

Weibel WW (1992). The Indigenous Leader Outreach Model: Intervennon Manual. Rockville, MD: National Institute on Drug Abuse, Division of Applied Research, Community Research Branch.

Weibel WW, Lampinen TM (1991) . Primary prevention of HIV/1 among intravenous drug users. Journal of Primary l're vention 12:3.

White H (1985). Agency as control. In: J Pratt, R Zeckhause (Eds), Principals and Agents: The Structure of Business, pp. 187-214. Boston: Harvard Business School Press.

* The second part will appear in issue 4 of IJDPvolume 6, 1995

¡ Some of the outreach workers were interviewed during and after their tenure with the San Fransisco project. The names of project staff members appearing in the text are pseudonyms.

** It should be noted that adverse selection problems are symmetrical. Job candidates do not have a monopoly on deception or incomplete disclosure. For example, to attract highly skilled candidates, employers sometimes make promises regarding opportunities for advancement that are not subsequently honoured.

*** The problems of adverse selection and moral hazard are conceptually distinct. However, they are sometimes difficult to discriminate in particular cases because they become intertwined. For example, inadequate monitoring of agent performance can lead to postcontractual opportunism (moral hazards) such as running con jobs on the project. Subsequently, the prospect for con jobs can serve to attract recruits who are on the lookout for opportunities to exploit, thereby creating a problem of precontractual opportunism (adverse selection). In either case, agency theory locates the source of the problem in informational asymmetries.