Female Sex Worker Projects in the Asia-Pacific Region: Three Case Studies

Home , Prostitution in Cyprus, Sex worker, Sonagachi

UNAIDS Female sex worker HIV prevention projects: Lessons learnt from Lessons learnt Papua New Guinea, India and Bangladesh Female sex worker HIV prevention projects:

Female sex worker HIV prevention projects: Lessons learnt from Papua New Guinea, India and Bangladesh

UNAIDS Case study Joint United Nations Programme on HIV/AIDS (UNAIDS) 20 avenue Appia, 1211 Geneva 27, Switzerland Tel. (+4122) 791 46 51 – Fax (+4122) 791 41 87 e-mail: [email protected] – Internet: http://www.unaids.org November 2000

UNAIDS Best Practice Collection Photographs by: Smarajit Jana, Coordinator, Sonagachi Project Joe Anang, Project Manager,Transex Project Carol Jenkins, Travis Jenkins AKM Mohsin, reprinted with permission from the Daily Star

Illustrations and author: Carol Jenkins Senior Scientist Social and Behavioral HIV Prevention Research DAIDS, National Institutes of Health, USA

Cover photo : Henao Sisters during a vocational training meeting at the project house, where YWCA workers provided sewing lessons. Photo: Joe Anang, Project Manager, Transex Project.

UNAIDS/00.45E (English original, November 2000) ISBN : 92-9173-014-9

© Joint United Nations Programme presentation of the material in this on HIV/AIDS (UNAIDS) 2000. work do not imply the expression of any opinion whatsoever on the part All rights reserved. This document, of UNAIDS concerning the legal which is not a formal publication of status of any country, territory, city or UNAIDS, may be freely reviewed, area or of its authorities, or quoted, reproduced or translated, in concerning the delimitation of its part or in full, provided the source is frontiers and boundaries. The Joint United Nations Programme on HIV/AIDS (UNAIDS) is the leading advocate for global acknowledged. The document may action on HIV/AIDS. It brings together seven UN agencies in a common effort to fight the not be sold or used in conjunction The mention of specific companies epidemic: the United Nations Children’s Fund (UNICEF), the United Nations Development with commercial purposes without or of certain manufacturers’ Programme (UNDP), the United Nations Population Fund (UNFPA), the United Nations prior written approval from UNAIDS products does not imply that they International Drug Control Programme (UNDCP), the United Nations Educational, Scientific and (contact: UNAIDS Information Centre). are endorsed or recommended by Cultural Organization (UNESCO), the World Health Organization (WHO) and the World Bank. UNAIDS in preference to others of a The views expressed in documents similar nature that are not UNAIDS both mobilizes the responses to the epidemic of its seven cosponsoring organizations by named authors are solely the mentioned. Errors and omissions and supplements these efforts with special initiatives. Its purpose is to lead and assist an responsibility of those authors. excepted, the names of proprietary expansion ofthe international response to HIV on all fronts: medical, public health, social, products are distinguished by initial economic, cultural,political and human rights. UNAIDS works with a broad range of partners – The designations employed and the capital letters. governmental and NGO, business, scientific and lay – to share knowledge, skills and best practice across boundaries. UNAIDS, 20 avenue Appia, 1211 Geneva 27, Switzerland Tel. (+4122) 791 46 51 – Fax (+4122) 791 41 87 e-mail: [email protected] – Internet: http://www.unaids.org Produced with environment-friendly materials Female sex worker HIV prevention projects :

Lessons learnt from Papua New Guinea, India and Bangladesh

Geneva, Switzerland 2000 Contents

Contents

Introduction ...... 5 Background ...... 7 Lessons in Best Practices in Sex Worker Interventions ...... 9 The Individual vs. the Social ...... 9 Formative Research ...... 10 Courage and Clarity of Commitment ...... 11 Measuring Effectiveness...... 12 Strategies for Strength and Replication...... 15 Efficiency and Management...... 16 Specific Points ...... 17 Conclusions...... 18

The Transex Project: Sex and Transport Workers, Police and Security Men in Papua New Guinea ...... 19 Introduction...... 19 Situational Background...... 20 Relevance ...... 23 Formative Stages...... 25 Implementation ...... 28 Working strategies...... 28 Outreach and peer education ...... 29 Condoms ...... 30 Health care ...... 32 Empowerment...... 33 Monitoring and Evaluation...... 35 Increasing condom use ...... 36 Reducing line-ups...... 42 Condom use and regular partners ...... 44 Increasing knowledge...... 46

2 Contents

Condom and lubrication distribution ...... 47 Resource materials ...... 49 Efficiency...... 50 Ethical soundness ...... 51 Replicability ...... 51 Sustainability ...... 52 Lessons Learnt ...... 52 Acknowledgements ...... 54 References ...... 55

Sonagachi: A Sex Worker Project in a Red Light District of Calcutta, India...... 57 Introduction...... 57 Situational Background...... 59 Relevance ...... 62 Formative Stages...... 62 Implementation ...... 66 Working strategies...... 66 Health care ...... 67 Peer education ...... 69 Condoms ...... 70 Empowerment...... 71 Clients and lovers ...... 75 Monitoring and Evaluation...... 76 Effectiveness ...... 77 Efficiency...... 83 Ethical soundness ...... 84 Replicability and sustainability ...... 85 Lessons Learnt ...... 86 Further Reading ...... 88

3 Contents

SHAKTI: A Brothel and Street-Based Sex Worker Project in Bangladesh...... 91 Introduction...... 91 Situational Background...... 93 Tangail brothel ...... 93 Street sex work in Dhaka ...... 95 Relevance ...... 97 Formative Stages...... 98 Brothel baseline studies ...... 98 Street-based sex worker baseline studies...... 100 Implementation ...... 101 Working strategies...... 101 Peer education ...... 101 Health care ...... 103 Empowerment: in the brothel...... 105 Empowerment: on the streets of Dhaka ...... 109 A turning point ...... 110 Condoms ...... 111 Clients, non-clients and lovers ...... 114 Monitoring and Evaluation...... 117 Effectiveness ...... 118 Efficiency...... 122 Ethical soundness ...... 122 Replicability and sustainability ...... 123 Lessons Learnt ...... 123 References ...... 126

4 Introduction

Introduction he set of case studies in this collection emerged from a session Tentitled Best Practices in Female Sex Worker Projects held at the Fourth International Congress on AIDS in Asia and the Pacific in Manila, October 1997. Preparation for the session began in May 1997 when UNAIDS sponsored an effort to learn about sex worker projects in the region. Lists of known potential projects were elicited from sex work networks, well-connected individuals and researchers. A call for responses was placed on the SEA-AIDS list server. After intensive communication with 25 sex worker projects in the Asia-Pacific region, five were selected that asserted they could produce data to document their impact and effectiveness. It was decided to omit male and transgender sex worker projects, so as not to sideline them, and to place them in their own category for a future review. It was also decided to omit the national sex worker programme in Thailand as it was being documented on its own as a successful best practice case study.

The selected projects were requested to send a sex worker and a manager to the Congress, able to discuss the nature of their projects, each from her own point of view. The session took place in several lan- guages with translations, a time-consuming exercise, but because of the candid and forthright personalities of the sex workers involved, was nonetheless very lively. Those attending felt the session was a success in demonstrating some of the effective strategies available and showing the strength of sex workers themselves in bringing about the success of their projects. The presenters at the Congress were: • Dr. Smarajit Jana, Coordinator, Sonagachi Project, Calcutta, India • Anima Bannerjee, Mohila Samanwaya Committee, a sex worker organization developed as part of the Sonagachi Project • A.M. Quddus, Field Coordinator, SHAKTI Project, Bangladesh

5 Introduction

• Aklima Begum, Peer Educator, SHAKTI Project • Kim Yel. Outreach Worker, Svay Pak Project, Cambodia • Hong Chanta, brothel owner, Svay Pak Project • Joe Anang, Manager, Transex Project-Port Moresby, Papua New Guinea • Alice Michaels, Peer Educator, Transex Project-Port Moresby • Candelaria Cantillo, Project Coordinator, Talikala, Davao City, Philippines • Michele Valera, Lawig Bubai, a sex worker organization developed as part of the Talikala project Carol Jenkins and Irene Fonacier-Fellizar were the facilitators. The case studies that follow are constructed from information gathered from those projects, both before, and for many months following the session. With one exception, all project sites were visited. After visits and attempts to secure adequate proof of effectiveness, two projects were omitted from the case study collection, Svay Pak and Talikala. While screening projects, we recognized that some sex worker projects might have been as effective as those selected, but because they lacked good documentation, it would be difficult to defend their approaches. Others appeared sound in some respects, but aimed at removing sex workers from prostitution, which was rarely successful and is not the principal aim of most HIV prevention projects. Others were not considered simply because communicating with them was very difficult, either due to the lack of (or poor) electronic communications, or because their managerial personnel were unable to com- municate in English, or another accessible language to the reviewers. These fell out of our network. The evolving nature of UNAIDS best practices criteria also played a part in the eventual selection of these case studies. These guidelines

6 Introduction

emphasize effectiveness and impact, but not entirely on criteria of an epidemiological nature. Other factors, such as sustainability, ethical soundness, relevance, efficiency, were also considered. At first it was difficult to find any project which met all of these criteria. Eventually, UNAIDS came to realize that documenting weaknesses can be just as useful as showing success. None of the projects selected is a perfect example of all criteria. It is doubtful that any such project exists. What these case studies represent is a set of experiences and lessons that might clarify for others the areas of strength and weakness typical in successful female sex worker projects. To the greatest extent possible, we have shown the real difficulties and triumphs of each of the projects. These case studies demonstrate how sincere organizations have tried to deal with the complex problems presented to their projects in different (and sometimes similar) ways. And they exemplify some of the region’s best efforts at preventing HIV among a diversified and highly vulnerable group of women. Background he three projects selected for case studies represent a range Tof situations, geographical locations, problems and solutions. Two of these, Sonagachi and SHAKTI, involve brothel-based sex workers in South Asia. The Transex project works with club or street-based sex workers in the Pacific nation of Papua New Guinea. SHAKTI includes both brothel and street-based sex workers in Bangladesh. In each case, situations differ considerably. Sonagachi is a large, densely populated red light district in the centre of Calcutta, housing 5,000 sex workers, of whom two-thirds are Indian and one-third hail from Bangladesh and Nepal. The project emanates from the All India Institute of Hygiene and Tropical Medicine, funded by WHO in its earlier days and by the United Kingdom Department for International Development (DfID) more recently. Although it began when HIV prevalence was still quite low, prevalence has been increasing in West Bengal. In other regions of India, HIV prevalence has risen very rapidly. SHAKTI is the name of a project implemented by CARE, Bangladesh.

7 Introduction

It began at a brothel in Tangail, a medium-sized city in Bangladesh, and expanded to include street-based women in the capital city of Dhaka. DfID provided funding. The brothel at Tangail houses about 600 sex workers and the street-based intervention targets about 3000 workers. SHAKTI began when HIV was only sporadically found in returning migrants and a few other groups of people in Bangladesh. As of 1998, HIV prevalence remained fairly low in Bangladesh, but appeared to be increasing slowly. The Transex project, implemented by the Papua New Guinea Institute of Medical Research, has two branches in the two major cities of Papua New Guinea, Lae and Port Moresby. In both cities, the project works with independent, street-level sex workers. The project is mainly funded by AusAID and, to a smaller extent, had funding from UNAIDS for a single year. It is named Transex because it focuses on transport workers, i.e. sailors, truckers and dock workers, as well as police and security men, in addition to the women who sell them sex. HIV prevalence was rising rapidly as the project started and continues to do so.

The process by which these projects have been documented requires some mention. All projects were visited and documents reviewed. Taped interviews were conducted with some managers. Mid-term or final evaluations were carefully examined where available. Raw data were accessed for several projects and analysed. Sex workers from three projects were especially helpful, as were outreach or field workers. The first drafts of the case studies were examined by each project manager and then sent for review and comments to persons uncon- nected with the project in the same country as well as to persons outside each country. Each reviewer was a person closely involved with HIV prevention activities, including several known authors of published AIDS-related materials. After their comments were received, project managers were asked to respond, and then a full review of all materials took place in Geneva with UNAIDS representatives. As this process extended over a year, updated information was sought from each project just before the completion of the last draft. Therefore, it is hoped that this document is a fair and honest representation of the realities faced by the selected projects and their participants and that

8 Introduction

the reader will glean useful insights from the strengths and weaknesses these projects demonstrate. Lessons in Best Practices in Sex Worker Interventions he efforts made to document and analyse five female sex Tworker projects selected for the session at the conference in Manila in 1997 have resulted in three best practice case studies. Two have been set aside because of inadequate documentation and lack of sound data to support claims of success. The other three illustrate various aspects of good practice and some not quite so good. There are lessons to be learned from all. In fact, it might be easier for most of us to see why some action fails than how an activity succeeds in bringing about change in complex and difficult situations. This section will address several broad issues, illustrated by discussions of the case studies. Other more specific points will be listed below more briefly. The interested reader can refer to the full case studies for a more complete discussion. The Individual vs. the Social n nearly all settings, female sex workers are a stigmatized group Iof people. Their very existence challenges the standard family and reproduction-oriented sexual morality found in most societies. Yet they exist nearly everywhere, clearly indicating that they fulfil a function for society. Hypocritically, most mainstream societies have relegated them to the margins, abused them, exploited them and restricted their rights as citizens. As women (in contrast to male sex workers), they are doubly powerless. With the advent of the HIV pandemic, they have been the first group in many nations to be targeted as vectors and seen as dangerous to the general population. They are seen as the agents of infection and their clients as unwitting victims. Usually the sex trade or industry itself is left untouched. The contrast- ing perspective is that the sex worker is a person whose livelihood

9 Introduction

places her in a highly vulnerable situation for acquiring HIV. That livelihood, being illegal, is surrounded by layers of uncontrolled and therefore abusive persons. Sex worker projects must grapple with these vari- ant perspectives, both in the society at large surrounding the sex worker and in the views of project personnel. Sometimes sex workers themselves are ambivalent about their position in the epidemic, at times seeing themselves as actors, and at other times as victims. The strategies taken by HIV prevention projects for female sex workers reflect these perspectives and set the tone for the way in which the project is implemented.

Most projects seem to address high-risk behaviour on the level of the individual, with persuasive methods, such as advice, counselling, and peer education; enabling approaches that remove social constraints to safer sex (or conversely, put barriers to unsafe sex in place) demonstrate greater success. These enabling approaches are exemplified by the projects documented here in various ways, which illustrate the way they function. Formative Research irst, one must understand the structure of the sex industry, the Fposition of women in it, who has power over them and who does not, and what the sex worker lacks to be able to live and work in a healthier setting. Baseline quantitative survey research, while usually required, is not able to supply the in-depth information needed to design strategies for change.

The Transex project was based on an optimal use of qualitative situational assessments and formative research. This enabled the project to tailor the activities to local contexts and to be able to adjust to the realities in different sub-populations and cities. Sex workers themselves were involved in the project design and made their priorities clear: they wanted help against police harassment. The Sonagachi project conducted extensive surveys of sex workers, clients, boyfriends, and

10 Introduction

sex workers’ children’s needs, gradually adding each piece of research, as issues become clearer. More qualitative information was gathered somewhat informally, but was reflected repeatedly in the published documents on the project. SHAKTI conducted a formative survey, but it was not designed to uncover the power structure of the brothel. This information was gathered in a less formal and less organized fashion, leaving project implementers somewhat confused about how to pro- ceed. The force of events, however, in Bangladesh, moved the project forward. Courage and Clarity of Commitment ut there is more. It also takes courage to face powerful polit- Bical and social structures once these are delineated in the formative phase. The Sonagachi project is hailed globally by sex worker organizations and AIDS activists alike for its integral involvement of sex workers. This could not have taken place without a palpa- ble demonstration of courage and commitment on the part of project personnel. In this case, several community-based groups had been working in the Sonagachi area prior to the HIV intervention. They had themselves confronted the local gangs, a portion of the power structure, and had developed understanding of the social scene and the strength to stand up to it. As the HIV intervention began, they partic- ipated and helped intervention personnel to develop the strategy of empowerment that has been so successful. Gradually, women’s groups, legal rights organizations, and even government agencies have joined the sex workers’ efforts to reform the social system around prostitution in Calcutta.

The Transex project in Papua New Guinea (PNG) operates in quite a different social setting. There are far fewer layers of control over the sex worker than found in the Indian sub-continent. The one salient power broker is the police force and the project took them on directly, addressing the highly sensitive issues of group rape or line- ups. This has been one of the most important aspects of the project,

11 Introduction

empowering the women through confronting police practices, sexual violence and issues of social marginalization. Gradually, the commitment of the project and its staff has become apparent to sex workers, who appear to be beginning to bond in a self-run organization. Both of these projects created a precedent by addressing social and sexual norms.

The SHAKTI project has attempted to bring about improved condom usage without confronting the power structures and, in some ways, has been less successful than the other projects. Its original model of conflict resolution did not resolve the major conflicts surrounding the sex worker and hence, she has remained relatively isolated in her efforts to induce safer sex in her trade, one that remains highly stigmatized in Bangladesh. Nonetheless, the participation of SHAKTI’s sex workers, particularly those who led in a fight against the closure of brothels in 1999, has been critical in bringing about a court decision citing prostitution as a legal way to make a living. While it is under- stood that all constraints in the social arena are not amenable to rapid change, the SHAKTI and Sonagachi projects demonstrate with clarity that, even in highly repressive and abusive environments, the rights of sex workers can be addressed and sex workers themselves can be enabled to act. Measuring Effectiveness nless a project creates sound modes of measuring its success, Uit is very difficult to defend its practices, even if they are sub- jectively thought to be successful. When screening numerous projects in the Asia-Pacific region, this issue loomed largest. Most projects are quite capable of maintaining process indicators, i.e. the number of condoms given out, the number of meetings held or peer educators trained. Measures of impact or effectiveness, however, are often less well developed. In some cases, data are collected which could be used to demonstrate impact, but there are no project personnel available with the skills to analyse and disseminate the results. In the case of

12 Introduction

SHAKTI, such a skilled person was available during the design phase and good indicators were set up, but for a period in the project’s life, those skills were not replaced. A similar skills gap occurred in the Transex project.

The types of indicators used must fit what the project is trying to accomplish and must accommodate all its major effects. While all of the projects discussed aim at the control of HIV transmission, prevalence levels of HIV do not demonstrate impact very well. Incidence levels would be an improvement but require a cohort study, which becomes a major investment in research and is not feasible for most intervention projects. Levels of other STDs are considerably more useful in demonstrating improved sexual health, but are not specific to project components. It is not possible to separate out the effect of improved recognition of symptoms, improved access to treatment, improved diagnosis and treatment, or improved condom use from such data. Further, some STDs, such as chlamydia, are very often asymptomatic and not a good indicator in women. Results of STD surveys are very sensitive to modes of sample collection (e.g. how well a vaginal or cervical swab is taken), modes of laboratory testing (especially where cultures are required), and modes of sampling. Unless well-trained persons are in place to carry out such surveys, they are frequently subject to failure, or at least, to questionable results.

In the case of Sonagachi, STD indicators produced a problem for several of the above-mentioned reasons. Eventually, the laboratory methods were improved. In the case of SHAKTI, laboratory methods were apparently well controlled, although some confusion resulted over the interpretation of negative controls on the PCR chlamydia results. Unfortunately, those with the clearest results, i.e. PCR-tested gonorrhoea and current syphilis (via VDRL at dilutions ≥1:8), showed no improvement over time. The Transex project was not able to develop its STD indicators at the beginning of the project due to difficulties with availability of personnel as well as with the government. This work has begun but will require several years before an assessment of

13 Introduction

the impact of the project can take place. It is not likely that most sex worker projects will be able to utilize laboratory-based indicators and simpler modes of monitoring STD levels must be developed.

Self-reported condom use is always a difficult measure, though heavily depended upon by most projects. It is well recognized that women are under pressure by these projects to claim improved condom use and are highly likely to exaggerate. The large differences between mid-term survey results and simultaneous monitoring results for condom use at the Tangail brothel illustrates well the intense pressure felt by sex workers when asked about their condom use by their own dedicated peer educators (as opposed to peer educators assigned to educate other cohorts of sex workers). SHAKTI attempted to address this question by collecting used condom covers, but this method did not appear to be adequate.

Transex tried to diminish overreporting by means of improved interview techniques, but had no other modes of verification. Condom use is not a monitoring indicator and data on this aspect of outcome are only available through repeated surveys. Fortunately, the protracted period of initial baseline data collection allowed a trend analysis, and the police component did have a follow-up survey. Its full evaluation was designed from the beginning of the project to diminish bias by moving outreach workers from one city to the other to serve as interviewers in the follow-up survey.

While Sonagachi also simply asked women about their condom use in its earlier repeated surveys, more recently it has moved to asking how many condoms they are buying. This is a far better measure, assuming poor women will not buy what they do not plan to use, and is only possible when free condoms are not being distributed. The very strong social marketing component of Sonagachi has enabled this to come about. Such a measure, however, does not accommodate the number of condoms supplied by clients or boyfriends, but, in the absence of targeted interventions for these groups, may well act as a proxy for all

14 Introduction

condoms available to Sonagachi’s women. A similar move to selling condoms has occurred in the SHAKTI project and, in both cases, the sale of condoms by sex workers serves well to confirm the rising levels of condom use.

Monitoring is an on-going process. The majority of monitoring indicators for all projects focus on process or progress, not outcome. These are of value if project personnel are able to critically review their meaning and adjust activities accordingly. In all the projects discussed, time pressure and overworked personnel have been factors working against the optimal use of monitoring indicators.

Monitoring growing levels of self-esteem or empowerment has been a weakness in all these projects, particularly in a measurable form. Qualitative documentation does exist, e.g. anecdotes and newspaper articles. Only at Sonagachi has a quantifiable measure been developed, i.e. the number of sex worker meetings attended. This is possible only because of the advanced nature of the intervention’s approach to sex worker rights and self-directed activities, such as training other sex workers elsewhere. Strategies for Strength and Replication ll these projects have aimed at replication in other areas of Atheir respective countries or regions. The West Bengal Sexual Health Project incorporates the Sonagachi experience directly and the Sonagachi sex workers’ organization, the Durbar Mahila Samanwaya Committee, has trained hundreds of sex workers elsewhere in West Bengal. Without the development of strong sex workers’ organizations, each project has had to rely upon the interest of other NGOs. This has not been successful as yet in Papua New Guinea and is only partially successful in Bangladesh. However, in Bangladesh the strong alliance of SHAKTI with government during its early phases was very effective in demonstrating the possibility of targeted interventions for prostitutes in a highly conservative setting. Its strength grew as it linked with other

15 Introduction

legal and human rights agencies. In Sonagachi as in Papua New Guinea, the implementing agency was a government affiliate. Nonetheless, other government sectors had to be convinced of the value of the project. Partnerships with NGOs were well developed in Sonagachi and have only recently developed in Papua New Guinea. In all cases, it is clear that replicating the project must be done in a way that will allow local groups to adapt the overall strategies to their own situations and access the experience of the older project in the process. Planning for expansion or replication at the early stages of such a project is important for it allows the time for development of understanding and cooperative relationships with government agencies and NGOs. Efficiency and Management hile levels of efficiency in such projects will never approxi- Wmate to those in successful profit-making businesses, they can be tightened and improved. As a rule, it is more efficient to address the top of a hierarchy that is in control of activities and behaviours of those lower down, as the Transex project did with the police, than to spend a great deal of time and money directly reaching the masses at the bottom. Sonagachi also conducted repeated formal and informal advocacy meetings with senior government officials to win their cooperation. To a large extent, this has paid off and allowed major international meetings of sex workers to take place in Calcutta.

Large budgets are not always helpful. They can mislead inexperienced project managers into thinking there is ample room for mistakes to be made. Good budgetary control is required and needs direct coordination with project strategies and specific activities. Even small budgets can thereby be well utilized.

AIDS prevention projects may attract NGOs and individuals who simply understand that there is funding available for this kind of work. When they realize the intensity of the problem, they may or may not be willing to commit themselves and a large part of their lives to working

16 Introduction

with stigmatized people. Leadership in sex worker projects requires honest commitment. It is not a job that suits everyone. The individuals who led the projects discussed were deeply committed and personally involved. Their families became involved and their personal time was sacrificed. While this may lead to burnout, it may be necessary, at least at the beginning of such projects. Others in such projects who are deeply committed also experience burnout and the project must build in methods to diminish feelings of ineffectiveness, monotony, and grief. Persons living with HIV, whenever possible, should be incorporated into project activities. This can help them and help everyone else remember why they are working so hard. Specific Points eer education is one of a number of ways to convey informa- Ption and persuade people to change their behaviour. Its great advantage is that peers can utilize their normal venues and modes of communication. If peer educators are separated from their colleagues with special privileges, the value of the approach may be lost.

• Addressing clients with specific programmes tailored to their needs strengthens sex workers’ abilities to negotiate condom use. • Boyfriends, husbands and other non-commercial sexual partners frequently place sex workers at more risk than do their clients, particularly after they have learned to persuade clients to use condoms. This aspect of the lives of sex workers needs specialized attention. • Personnel well trained in current STD diagnostics and treatment must oversee the establishment and management of clinics for sex worker projects. Sex workers are people with other health problems as well and these also need attention. • STD and other studies undertaken with sex workers must be carried out with their full understanding and assure their right to refuse.

17 Introduction

Well-designed studies conducted properly are of value to sex workers too; poor studies waste their time and are seen as abusive. Results need to be fed back to the sex worker community in an appropriate fashion to ensure their understanding, and timely treatment must be given. Anything less is unethical and will destroy their confidence in the project and its messages. • Project staff and sometimes sex workers too need training in human sexuality in order to be able to speak about sex with ease and convey explicit messages. • Lubrication, female condoms and encouraging non-penetrative sex should be made part of the package of safer sex options. • Documentation is more than periodic surveys and counting monitoring indicators. Writing and/or using tapes and film to document the history of a project is a worthwhile endeavour of its own and has potential value to many others. Conclusions he importance of designing and implementing successful tar- Tgeted interventions for sex workers as part of HIV prevention and control cannot be over-emphasized. In almost every country, sex workers comprise a focal point of the epidemic, because, as women who provide sex for many men, infections can easily accumulate among them. They are the victims of discrimination, often violently intense, trafficking, legal persecution and societal ambivalence — as well as one of the first occupational groups to become heavily infected. From them, the infection passes back to their clients and into the general population of women, men and children. One of the clearest public health lessons emerging from the HIV pandemic is that protecting the human rights of sex workers is one of the best ways to protect the rest of society from HIV.

18 Transex, Papua New Guinea

The Transex Project: Sex and Transport Workers, Police and Security Men in Papua New Guinea

Introduction he Transex project, for transport Tand sex workers, began in 1996. It was built on a foundation of extensive research conducted by the Papua New Guinea Institute of Medical Research (PNGIMR), exploring both behavioural risk factors and STD epidemiology. With funding from the regional AIDSCAP (FHI) project, in 1994 the PNGIMR conducted an ethnographic situational assessment of men in the transport industry (truckers, sailors, dock workers) and the women who sell them sex. In 1995 participatory project design workshops held with sex workers, hotel managers, and maritime union, trucking, and shipping company representatives, led to the development of a proposal and an appli- cation for funding. The Transex project was then included within a larger three- year HIV and Sexual Health Project funded by AusAID in cooperation with the Department of Health (DoH). Specific activities were additionally funded by USAID, WHO, UNFPA, and UNAIDS. The

19 Transex, Papua New Guinea

implementing agency has been gramme, NGOs, and community the PNGIMR, a statutory body of groups. The Transex project was the government. originally conceived to be a project implemented by an NGO, but The project operated, from its as no NGO emerged ready to beginning, in the shadow of a handle its objectives, it has been dramatically increasing HIV epi- brought forward for the next demic, a free-fall economy and a phase of funding as Transex Plus. roller-coaster political scene. The The new project will include urban areas, where the project youth and other groups and be was based, experienced implemented by the National markedly increasing levels of AIDS Council, where its lessons crime in the wake of a structural will be spread more widely. adjustment programme that devalued the currency and led to Situational rapid inflation. General compla- Background cency about AIDS and lack of understanding of behaviour apua New Guinea is a change programmes made for Pcountry of about 4.3 mil- slow progress. Despite these lion people spread out over a obstacles and the short life of land mass approximately the size this project, the project provided of Thailand. Except for a few the first successful example of a areas, population density is quite targeted intervention for HIV in low and people are scattered Papua New Guinea (PNG). Sex over the landscape. About 85% workers developed their own of the people live in rural areas organization, police and trans- and speak over 860 different lan- port workers were trained as guages or dialects. About one- peer educators, condom use third are literate in English or increased in high-risk groups, Melanesian Pidgin, the lingua HIV testing increased, and the franca. Most have little access to project provided the first real functioning health or other serv- home-grown expertise in peer ices. Traditional lifestyles are education and numerous other under pressure as consumer HIV issues to the national pro- demand and a cash economy

20 Transex, Papua New Guinea

expand without real growth in in late 1995. The Transex project income-earning opportunities. was designed to be executed The country is rich in natural through an urban community- resources, with numerous gold based NGO, with PNGIMR acting mines, oil fields and logging as major adviser. However, when camps, but, due to years of poor funds arrived, no NGO was will- management, it has become a ing to implement it because of rich nation of poor people. the stigma of working with sex workers. In mid-1997, a further Risk factors for HIV infection are fruitless attempt was made to high in Papua New Guinea, with find an NGO to help in imple- high levels of STDs in many areas menting the project. Poor coor- around the country and wide- dination between national and spread risky sexual practices provincial health departments as (Jenkins and Passey, 1998). The well as mutual distrust between epidemiological patterns of HIV government and NGOs has ham- infection are influenced by sev- pered a solid national response eral factors, including the to the epidemic. In 1997 legisla- absence of injecting drug use, tion to establish a National AIDS high levels of infertility and STDs, Council was passed and renewed a diffuse urban and rural sex efforts to coordinate the public trade, as well as clear patterns of and private sectors began. spread along transport routes (Malau et al., 1994). Over the In a national study conducted in past decade, most political lead- 1992 in rural and periurban ers chose not to recognize HIV as areas, about half of all women a serious threat to the nation, investigated revealed that they despite numerous studies by UN exchanged sex for money or agencies as well as local researchers gifts (National Sex and that sent clear warnings (Jenkins, Reproduction Research Team 1993a; Mugriditchian and Jenkins, and Jenkins, 1994). While that 1993). Through AusAID, Australia, study was not carried out on a the main foreign donor, funded representative sample, subse- the initiation of the first major quent studies corroborated the programme for HIV prevention high frequency of commercial

21 Transex, Papua New Guinea

sex. Research conducted in three they were frequently made to do urban areas (Goroka, Port the same to obtain protection at Moresby and Lae, with total night from security men. populations of about 315 000) Therefore, a targeted one-year estimated that approximately intervention was designed 15 000 women were working specifically to diminish these more or less independently, at practices. least part-time, as sex workers on the streets, along the Prostitution is illegal in PNG, Highlands Highway, in small though few women are impris- guest houses and at hotels oned for this reason. Sex work in (Jenkins, 1994a, 1994b). Their PNG is not brothel-based and is main clients were reported to be seen as an unfortunate but rea- government workers, truckers, sonable way for women to gain sailors, dock workers, security an income for their families and men and police. The project, themselves. In the urban settle- therefore, aimed at including as ments or slums, theft, perpe- many client groups as possible in trated by “raskol” gangs, and sex addition to the sex workers in work carried out by women are the two main port cities of the important income producers country, Lae and Port Moresby. (Jenkins, 1996). Sharing the fruits Lae is the largest port and the of these activities with families origin of the Highlands Highway, ensures protection and minimal the nation’s main road artery stigmatization. Outside of these through the most populated sec- areas, sex workers are stigma- tion of the country. tized and families often hide the fact from outsiders that their During the formative research, women are sex workers. Most sex workers indicated that their people consider these women as most serious problem was being carriers of infection, while, in picked up on the streets and keeping with a strong sexual taken by police for group rape double standard, their male sessions, known as line-ups, at clients are seen as blameless vic- the police barracks or other tims. Most women themselves locales. They also indicated that accept the notion that men have

22 Transex, Papua New Guinea

uncontrollable sexual desires however, harass them and obtain that must be satisfied, and sex on demand without paying. accept that rape is the natural The greatest abuse of sex work- outcome of this. Sex workers, ers in PNG comes from the police. therefore, represent the classic “necessary evil” and are gener- Relevance ally tolerated. The HIV epidemic in PNG has hardly been documented. While most women work alone, Although blood for transfusion some also utilize boskrus, men or has been screened since the late women who find them clients. 1980s, donors have increasingly Those working at hotels may been drawn from young school rent their own rooms and have a students; hence, few positive regular clientele or may go to cases are found through blood the rooms of clients contacted screening. Sentinel surveillance at through hotel workers. Guest antenatal and STD clinics took houses serve as both contact and place sporadically for several action venues. A great deal of years but was abandoned in sex takes place outdoors, with 1993, due to the lack of confiden- street-based women working tiality and other problems. More during the daytime in busy areas, recently, it has been reinstated. and sometimes at night as well. As this can be quite dangerous, Passive detection at hospital clin- these women often have a was- ics has been almost the only man, a man who guards the area source of information on HIV she is working in, keeps out infection, and even these cases intruders and helps her collect are not always reported. Figure 1 her money if someone tries to shows the latest available figures cheat her. The men or women on on HIV infection in PNG. the periphery of the trade are paid by the sex worker, either in Therefore, the Transex project, as money or in kind, but do not the only behavioural change pro- exert violent power over the gramme working with highly vul- woman as is found in sex work nerable groups in the country, has cultures elsewhere. The police, been extremely relevant. When

23 Transex, Papua New Guinea

Figure 1. Reported new cases of HIV infection in PNG by year

700

600

500

400

300

200

100

0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

the project began to train sex worker peer educators in August 1996, the women were already aware of probable deaths from AIDS in their communities.

The PNGIMR had conducted community- based intervention research in the health field for many years in rural communities. It has a well-developed infrastructure with branches in several parts of the country and a staff including an STD epi- demiologist, STD laboratory technicians and an experienced social science unit. This was the Institute’s first major urban programme and first HIV intervention.

24 Transex, Papua New Guinea

Formative Stages reparatory qualitative research Phad taken place several years before the project began. This included observations conducted by researchers working on trucks and on ships, and private interviews with 85 sex workers, 116 clients and 36 key informants, i.e. men and women on the periphery of the sex trade. The Sex Workers Outreach Project in Sydney provided a visiting consultant, who pointed out that one of the most difficult aspects of the sex worker scene in PNG would be bringing independent, often competing, street-level women workers together, both for purposes of training and for empowerment. Her report also remarked that moralistic attitudes on the part of some staff members and others would be a deterrent to the project.

The formative research found that the situation for sex workers and dock workers differed considerably between the two cities. Sailors, on the other hand, moved so readily around the coast that their lives were little affected by loca- tion. Sex workers in Port Moresby were more often living with their families and had larger incomes, whereas those in Lae were poorer and often homeless, sleeping on the streets or in parks. Dock workers in Lae were recruited as village

25 Transex, Papua New Guinea

groups and came to the city Rapport building with sex work- accompanied by relatives from ers proved to be a long and deli- their home communities. This cate process, as police continued placed a certain degree of con- to harass them, even arresting trol on their sexual behaviour, them for prostitution. The sex whereas in Port Moresby dock workers thought the project was workers were recruited from the contributing to their problems. urban settlements where many After one such arrest of 19 sex sex workers lived as well. Many workers in November 1996, acted as pimps for sex workers, many of whom had become bringing them to sailor clients. involved with the project, they scattered, some leaving the city. In mid-1995, a trial quantitative As no agency other than the baseline was begun with sex Transex project had ever reached workers (n=79), boskru or was- out to these women, fear of man (n=20), and a mixed group exposure was great. It took sev- of clients (n=68). Subsequently, eral months, during which the more complete quantitative project’s outreach workers baseline surveys were completed became quite discouraged, to with sailors (n=251), truckers find them again and make con- (n=405), dock workers (n=202), tact. The project tried to protect and sex workers in both Lae sex workers and demonstrate its (n=181) and Port Moresby sincerity. No publicity in the (n=297). In Port Moresby, police- media for the sex worker aspect men (n=130), policewomen of the project was allowed. Staff (n=53), and security men (n=154) training was intensified to try to were also surveyed. Focus group overcome all expression of the discussions were held with moralistic stance and poor gen- policemen’s wives. Results of der-related attitudes sometimes baseline surveys among dock exhibited by the male staff. Role workers on rates of self-reported plays, single sex and mixed sex STD symptoms as well as rates of group sessions with monthly pri- accessing commercial sex con- vate feedback to individuals firmed the differences by city about their behaviour seemed to found in the earlier study. help. Police were addressed on

26 Transex, Papua New Guinea

numerous occasions, not only organized in Port Moresby and about their own sexual practices, does not run at all after 18:00, a but also in an effort to help them project vehicle was essential. understand the damage done to Personal attacks on project per- HIV prevention when they raided sonnel in Port Moresby, when sex workers’ venues. they carried out work in settlements where many sex workers As there was no one in the coun- lived, continued to occur and the try with experience in training project vehicle was frequently peer educators when the project out of order due to these attacks. began, the South Pacific Commission was requested to For various reasons, trucking send their AIDS educator, who firms in Lae (where the trucker had been training youth peer intervention was based) were educators elsewhere in the less cooperative than other pri- Pacific, to PNG to help the proj- vate sector groups. Conditions ect get started. After preliminary on the Highlands Highway were training for all new staff, they very poor in recent years, with and a police counsellor selected inadequate road maintenance by the Police Commissioner and theft from highway gangs attended a workshop on peer acting as a major constraint on educator training techniques. the industry. Profits were threatened and AIDS prevention was Security issues loomed large and not a high priority for trucking about 10% of the project’s funds firm managers under those cir- had to be designated for guards cumstances. Eventually, after a and a radio security system. Even long period of consultation, they with this degree of protection, decided that truckers could be the Port Moresby office was reached and educated while they attacked by armed thieves were working. In an attempt to (locally called raskols) during the try this approach, outreach first year of the project, who workers rode for short distances stole and destroyed the only with truckers and met them at project vehicle in Port Moresby. the few places they stopped As public transport is not well along the Highway. This arrange-

27 Transex, Papua New Guinea

ment was not very satisfactory. tion of change towards safer Instead, other approaches evolved. sex, and the maintenance of Dispatchers were trained and these new practices, aided by a given condoms to distribute, and wider normative acceptance of individual truckers came for- condom use in the society. This ward, independent of their com- latter was to be effected, in panies, to become peer educa- part, by a social marketing proj- tors. Eventually, the trucking ect that was never imple- firms joined the project’s efforts. mented. The project’s overall Two of the larger companies aim emphasized facilitating a allowed workplace training ses- shift towards safer sex practices sions and offered to build rest through peer education, but- stops along the road for their tressed by several enabling drivers. These would allow out- strategies. One such strategy reach workers and peer educa- was to increase the variety of tors to access the drivers more options, by making both male easily. and female condoms easily accessible and introducing lubri- From the beginning sailors and cation. While most client groups dock workers were more accessi- of adult men managed to ble to the project. Shipping com- obtain STD treatment, sex work- panies, unions and workers’ asso- ers typically did not, as the staff ciations were very cooperative, at government STD clinics often allowing outreach workers to treated them harshly. The proj- enter the wharf, board ships, and ect sought to find ways to interview men whenever they had improve access to good-quality a break in their work schedules. STD treatment for sex workers.

Implementation In order to create an enabling environment for improved con- Working strategies dom use, major client groups The project was designed as a were targeted and the per- behavioural change programme ceived needs of sex workers to move people through early addressed. These needs included levels of awareness, considera- reducing the frequency of rape

28 Transex, Papua New Guinea

and harassment by police and, Outreach and peer education for many street-based sex work- While conducting the baseline ers, provision of a place to surveys, initial outreach began. bathe. A strategy was devel- After each interview, the oped to work on the very sensi- informant was given a short talk tive topic of line-ups with the on STD/HIV prevention with the police and the Police Commissioner aid of a flip chart developed was confronted privately but previously by the PNGIMR in directly with this issue. cooperation with other agencies, given condoms and taught Project personnel repeatedly how to use them. Nearly two reassured concerned groups that thousand persons were reached they were not going to moralize in this way over the early about prostitution or rehabili- months of the project. During tate sex workers, but that they this time, criteria for selection of would work with other NGOs to peer educators were developed provide skills training for sex and potential peer educators workers who wished to give up were identified. Workplace pol- the trade or simply supplement icy workshops were held with their income. Such non-moralis- shipping firms, police, security tic attitudes are not widespread firms employing a total of about in PNG, but the project’s success 3900 men, and eventually truck- has demonstrated their value. As ing firms. Trucking firms and a the project has grown, its per- few large security firms were sonnel have been called upon resistant at first. Security men at extensively by other agencies, one of the cooperative compa- within and outside of govern- nies suggested that their work- ment, to help develop different shop be televised in order to aspects of AIDS prevention. encourage cooperation from Although this reduced the the resistant ones. This was amount of project work staff done and was covered by the members were able to accom- newspapers as well. Eventually plish, it was a necessary and all registered security firms, as valuable act of bridge-building well as trucking companies, in the wider society. became cooperative.

29 Transex, Papua New Guinea

Peer educator training methods 14 police stations), 16 police- and materials were developed women (from 10 stations), 14 and adopted for each different wives of policemen, 20 security target group. Comic books were men, 163 sailors (on at least 24 designed and tested, with the different ships), 65 dock workers, help of a well-known cartoon and 14 hotel workers were artist, Biliso Osake. An excellent trained as peer educators. In Lae, flip chart by this artist developed by March 1997, a total of 125 earlier and printed in a large for- peer educators had completed mat was reprinted in a smaller, the full course (13 dock workers, hand-held format for use in the 12 sailors, 21 truckers, 63 sex project as well as by others. A flip workers and 16 from other chart specifically for teaching the groups, e.g. prisoners, artists, use of the female condom was musicians, and slum dwellers). also developed. Modules cover- Outreach workers were assigned ing a series of topics were a cohort of trained peer educa- designed for each peer educator tors to follow up. Given the num- training scheme. For sex workers, ber of peer educators, it was not special emphasis was placed on likely that the existing number of the use of condoms with outreach staff could monitor the boyfriends and husbands, and, frequency and quality of peer for condom-resistant clients, on education interactions taking how to learn to ignore the con- place. In 1997 outreach workers dom and concentrate instead on recorded discussions with and the woman’s pleasure. Sex-posi- distributed materials to about tive attitudes were always main- 14,000 persons. tained in these messages. Project personnel were frequently By mid-1998 in Port Moresby, 403 called upon to include others volunteer peer educators had within their training sessions. completed training. A total of While this was positive in the sense 357 sex workers had attended of overall AIDS prevention and 130 training sessions, with 75 acceptance of the project’s completing all learning modules. approaches by a wider portion of In addition, 36 policemen (from the community, it diluted the pro-

30 Transex, Papua New Guinea

ject’s focus on targeting specific project had to distribute poor- groups, took a great deal of time quality condoms as these were and caused problems for project the only ones available. A bet- personnel. For example, the ter-quality condom was pur- Correctional Institutional Services chased with the help of in Lae became involved and put AusAID, but the project’s four men through the peer educa- demands could not be met. The tion training and two through HIV project provided a regular sup- counsellor training. This led to the ply to all police stations in the provision of condoms for the project area (including those prison, an unexpected and very along the Highlands Highway) worthwhile action. Organizations, and the demand from sex and such as UNFPA, the Salvation Army, transport workers continued to ADRA (Adventist Development grow. Supplying many groups and Relief Agency) and others and individuals beyond those in came regularly to project person- the designated target groups nel for help, resources and training caused more shortages. Others, and to offer various kinds of sup- such as students and youth port in return. The Salvation Army, groups and even various health for example, developed a pro- service facilities, turned to the gramme of counselling and care to project for condoms. Their only which the project could refer HIV- other access to free condoms positive sex workers. The larger was family planning and STD AusAID HIV project repeatedly clinics, where staff expressed called upon project personnel to negative attitudes about con- speak at meetings and to train oth- doms or refused to give any at ers. There was a great need to all on grounds of morality, even spread expertise in PNG during this to STD patients. While project period, and the project has played personnel felt they should pro- a major role in this regard. tect the supply for their target groups, they also felt they could Condoms not refuse to give condoms to all Condoms were given free and those who requested them. It obtained from the DoH’s sup- was obvious that this situation ply. During the first year, the could not continue and could

31 Transex, Papua New Guinea

only be corrected by ensuring Health care an adequate condom supply, The AusAID HIV and Sexual either free or through social Health project employed a spe- marketing. cific officer whose task was to improve the government STD During the project’s first year, clinics. Although new clinics USAID donated 2,000 female con- were constructed and training doms and during the second year, sessions held, most STD clinic UNFPA supplied 15,000 female staff did not rapidly alter their condoms. These were a major attitudes and sex workers contin- boon to the project. The latter ued to complain about the way supply arrived after months dur- they were treated. The training ing which no free male condoms of peer educators emphasized had been available. Previous stud- rapid recognition of symptoms ies had shown the acceptability of and treatment for STDs. Project female condoms to sex workers outreach workers met with STD and to urban women in general clinic staff and arranged specific (Jenkins, 1995). The use of female times when they could accom- condoms added a major innova- pany sex workers to the clinic. tive aspect to the project and Although this approach filled a gap when male condoms appeared to secure more friendly were not available. treatment, sex workers were still stigmatized by having to come at In general, sex workers were a special time. Gradually, they enthusiastic female condom were designated as ’project users, as were many men who clients’ by clinic staff and even came to the project office to further stigmatized, especially by gain access to female condoms. doctors, who tried to pressure Female condoms were not avail- them to undergo HIV testing. able in shops in PNG, but Private clinics were too costly for recently have been made avail- most (but not all) sex workers able at government family plan- reached by the project. After ning clinics. considerable negotiation with both AusAID and the DoH, a clinic was established at the proj-

32 Transex, Papua New Guinea

ect site in Port Moresby, provi- lived. A theatre group in Lae sionally for the execution of an called SEEDS and a troupe of STD study. Two female project musicians became associated outreach workers, with previous with the project, and helped training as nurses, received HIV with the dissemination of AIDS counsellor training and special- prevention messages in the set- ized STD training as well. tlements and with World AIDS Periodic visits by physicians, labo- Day activities. In Port Moresby, ratory back-up through the outreach workers and peer edu- health services, and the process- cators themselves staged cam- ing of samples at the PNGIMR paign days in the settlements. headquarters in Goroka were begun. Sex workers rapidly By October 1997, two sex work- began using the clinic and partic- ers had joined the staff as part- ipated in the study. The DoH sup- time outreach workers, in an plied all medicine and basic con- effort to bring about greater sumables free of charge. In Lae, involvement in decision-making sex workers could access HIV test- by sex workers in the project’s ing and professional counselling management. One sex worker at the project site, but still had to was invited to speak at the be referred to the local hospital’s Fourth International Conference STD clinic for treatment. These on AIDS in Asia and the Pacific, efforts encouraged sex workers and to attend the satellite work- to be tested for HIV, an impor- shop on networking among sex tant step in prevention. workers. Her appearance in Manila was the first time a PNG Empowerment sex worker ever publicly admit- When the project began, 5-8 ted her trade. Both Papua New sites where sex workers could be Guineans and others in the audi- found in each city were desig- ence were touched by her hon- nated as target sites, but later esty. Following that, several expanded by another 6-8 in each other sex workers addressed city. In particular, sex workers groups of church-related NGOs requested that the project work in Port Moresby, and the in the settlements in which they response was very positive, lead-

33 Transex, Papua New Guinea

ing to improved attitudes of one of the outreach workers. acceptance and willingness to Discussions with the city authori- promote condom use, even in ties to build public showers and church youth groups. toilets did take place, but nothing happened. In Port Moresby, It was recognized that having a an open-sided house where sex few sex workers on the staff workers could gather and sleep would not improve conditions was built on the property owned for the women as a whole, and by the project. On many nights, efforts continued to develop sex workers slept at the Port associations run by sex workers. Moresby house and used the It became increasingly clear, from shower facilities. A more suit- the rising numbers of women able drop-in centre with tempo- who dropped in to the project rary sleeping and shower houses to request information arrangements would be an and supplies, that they were talk- improvement. ing to each other. In Port Moresby, a small group of During the formative research, women established a group sex workers expressed their called the Henao Sisters (the interest in learning about make- project house is on Henao Drive). up and other ways to improve This sex worker-run group their appearance, as well as received funding in its own bank acquiring cooking and banking account from an anonymous for- skills. In Port Moresby, a profes- eign client, a good example of sional hairdresser held sessions the power of enlightenment in at which personal hygiene, hair- the sex industry. dressing and the use of make-up were demonstrated. By early Developing the means to meet 1998, the YWCA joined as a the other needs of sex workers partner to provide literacy train- was slow. Building a shower for ing for 53 women twice a week them to use at the Lae house, for and vocational training for example, was not possible as the another 50. Women who were house was rented. Instead, they already literate received com- used the shower in the room of puter training.

34 Transex, Papua New Guinea

Numerous sex workers associated Monitoring and with the project took on the edu- Evaluation cation of women and men in their own communities. They On-going monitoring of process borrowed videos and other indicators was accomplished materials on a regular basis from through weekly submission of the project offices to address forms from outreach workers, church groups, youth groups and showing how many peer educa- other gatherings. Some of these tors visited for follow-up, how women clearly showed evidence many condoms and educational of improved self-esteem. materials were distributed, and However, for the majority, heavy comments on difficulties and alcohol consumption, violent signs of success. These were to personal relationships as well as be summarized monthly by the ambient violence in their com- two project managers who munities, continued to counter were also responsible for con- efforts at personal development. ducting and monitoring staff and peer educator training. Weekly meetings were to be held by managers with all staff

Henao Sisters during a vocational training meeting at the project house, where YWCA workers provided sewing lessons.

35 Transex, Papua New Guinea

at which progress and problems documents and interviewed all were discussed. levels of stakeholders and staff. They produced a formal evalua- Impact on behaviour was meas- tion in preparation for the ured by baseline and follow-up extension of the project into its surveys. A follow-up survey was second three-year funding cycle completed with police in Port (Lepani and Stephens, 1998). Moresby after only 9 months of Although the survey data were intervention, as the police com- not ready when they conducted ponent was funded only for one their work, all of these sources year. Among sex workers, quan- of information are reviewed titative evidence for progres- here. sively increasing condom use was provided by the data The project design had also obtained from these surveys, called for a baseline STD survey which were completed over a among sex workers, but for vari- long period of time, during ous reasons this study did not which the peer educator train- take place until the second half ing had begun. Shortage of staff of 1998. At that time, the preva- made separate research compo- lence of HIV in Port Moresby sex nents difficult to execute. workers was already 16.8% and Baseline surveys included infor- in Lae 3%. Levels of syphilis, mation on sex acts with non- chlamydia, gonorrhoea and tri- commercial and commercial chomaniasis were reported partners, risk perceptions and respectively, as 31.3%, 32.8%, behaviours. In order to minimize 39.4% and 21.2% in Port bias, the project design called Moresby, and 33.7%, 30.2%,33% for the outreach team from one and 44.1% in Lae (Mgone, city to move to the other in Passey and Russell, 1999). order to conduct the final end- of-project survey. A mid-term Increasing condom use evaluation was conducted in While it is commonly recognized 1997 and an external evaluation that altering condom use habits was conducted in 1998. is generally not easy or rapid, Evaluators examined project the earliest phases of interac-

36 Transex, Papua New Guinea

tion with sex workers saw a these sites on many days, talking rapid rise in condom use. The to women and handing out con- women were able, with free doms, before interviews were condoms, to increase the pro- possible. portion of acts covered, both with clients and regular part- Analysis of the trends in condom ners. Full or consistent use has use among sex workers is shown been more difficult to increase. in Figure 2. Work began in Port The earliest baseline survey Moresby before it did in Lae. The among sex workers was carried surveys were always conducted out between August and ahead of the identification and October 1995 in Port Moresby. selection of women for peer As the project had no presence educator training, but it seems among sex workers at the time, clear that the large amount of it was very difficult to find condoms and considerable infor- women who would admit to mation disseminated through- being sex workers when out the baseline survey period approached on the streets. Only was making an impact, even as one interviewer tested a short the data were being collected. A survey instrument, mainly by review of the figures suggests going to the settlement where that condom use rose very sex workers lived and using, quickly in both cities in the ear- essentially, a snowball sampling lier phase of the project. technique. When project funds Although sample sizes of sex arrived and outreach workers workers varied between 74 for had been trained, the baseline the earliest survey in Port survey instrument was Moresby and 442 in Lae in 1997, expanded to include informa- an upward trend is clear, espe- tion on knowledge, risk percep- cially in Port Moresby. tion and STDs, as well as con- Proportions of acts of inter- dom use. The main areas where course covered by condoms, the women find their clients with their upper and lower 95% were mapped and six selected confidence intervals, are shown where the work could begin. in Figure 2. Interviewers went in pairs to

37 Transex, Papua New Guinea

Figure 2. Trends in condom use among sex workers in Port Moresby and Lae

100

90 POM 80 Lae 70

0 Sep-95 Mar-96 Jun-96 Oct-96 Feb-97 Oct-98

In Port Moresby the trend is statistically significant, but in Lae an early upward trend slowed down. Survey results show that economic factors may explain some of the differences seen. Figure 3 shows the mean number of clients reported for the previous week by sex workers in both cities compared to the average income per client. When the project began, Lae sex workers took more clients but made less money per client than did workers in Port Moresby. Consistent condom use the previous week, i.e. 100%, did not change in either city.

38 Transex, Papua New Guinea

Figure 3. Numbers of clients and income per client reported for the previous week between 1996 and 1998 among Port Moresby and Lae sex workers

40 Mean number of clients 35 Kina per client 30 (1 Kina = US$ 1.05)

0 POM POM Lea Lea 1996 1998 1996 1998

As time passed, and economic conditions worsened, sex workers in both cities earned significantly less money per client. However, in Port Moresby the women were able to compensate by taking more clients (mean of 2.9 in 1996 and 4.9 in 1998), but this did not occur in Lae (mean of 3.2 in 1996 and 2.6 in 1998). It may be that the competition for clients in Lae became more intense, incomes declined and so did insistence on condom use. Were the project’s activities in Lae working to diminish the number of men who went to sex workers?

39 Transex, Papua New Guinea

Changes in the number of commercial and casual partners reported by the project’s male client groups over the previous week, indicate that this may be the case. Figure 4 shows a significant decline in commercial or casual sex among dock workers, most of whom were in Lae, and among truckers, all of whom were in Lae. Sailors were not separated by city as they travelled readily from port to port. The project only worked among police and security men in Port Moresby. Neither sailors nor police significantly reduced their commercial sexual activities, and

Figure 4. Changes in levels of commercial or casual sex among male client groups, 1996 to 1998

50 1996 1998 40

% 30

(s)= p≤ 0.05

(s) (ns)= p> 0.05 (s) Truckers Police (ns) Sailors (ns) Security (s) Dock workers

40 Transex, Papua New Guinea

security men significantly increased their commercial sex partners.

These groups of men seem to have adopted different strategies after exposure to HIV prevention education. Some had less commercial sex as well as increasing their condom use, while others increased condom use but did not decrease their purchases of sex. Figure 5 shows the changes in condom use with commercial or casual partners among these men. In all cases, condom use rose, whether the men decreased, increased or

Figure 5. Changes in condom use with commercial or casual partners between 1996 and 1998 among male client groups

50 1996 40 1998 % 30

0 Truckers Dock workers Sailors Police Security (s) (s) (s) (s) (s) (s)= p≤ 0.05

41 Transex, Papua New Guinea

made no change in the level of commercial or casual sex. These changes were least significant among truckers and dock workers, the groups that decreased their commercial sexual activity most significantly. Consistent condom use did not change among any of the men.

Among police in Port Moresby, a survey was carried out after 9 months of active intervention, because funding for this component lasted only one year. Condom use had risen more steeply than it did in later stages of the project, and the proportion of men seeking sex with “rot meri” (women of the streets) had already decreased from 25% to 18%. One of the issues that arose in discussions with policemen’s wives was that they would be insulted and angry if they saw condoms carried home by their husbands, which may have discouraged greater condom use among these men.

Reducing line-ups The strategy for reducing line-ups called for pressure from the top through the hierarchy of police, i.e. from the Police Commissioner downwards; and pressure from the side, i.e. from the policemen’s wives, who were informed about the practice and the risk that it involved for them. It appears that the efforts to target line-ups began to be effective. At baseline, 10% of the men stated they

42 Transex, Papua New Guinea

had been in a line-up during the previous week, which reduced to 4.2% in the following 9 months (p=.01). The average number of men involved in each act did not greatly change, 4.22 vs. 3.58 at that time. In 1998, unfortunately, the question was changed on the survey instrument to number of line-ups in the past month and results could not be directly compared. In 1998, 8% of men reported involvement in a line-up during the previous month. As a month, the period queried, was 4 times as long as a week, it is likely that the proportion of men engaging in line-ups continued to decrease, but results are inconclusive. Anecdotally, as of mid-1998, sex workers reported that sexual harassment continues to be diminished but, without monitoring indicators, this can only be confirmed by repeated surveys.

While it has always been difficult for a sex worker to gain redress for rape or theft, it has been nearly impossible for her to prosecute a policeman. In the course of the project, there were some indications of change in this regard. In August 1997, six policemen, who took advantage of the situation to rape two sex workers, raided one of the guest houses included in the project’s outreach. The sex workers were arrested, but they managed to lay complaints against the police officers. All six policemen were

43 Transex, Papua New Guinea

jailed, pending trial, and the sex workers were set free. The newspapers covered this event. A few months before, another woman had made a complaint of rape against the police. When women are willing to stand up for their right not to be sexually harassed by police, both human rights and HIV prevention are well served. Police peer educators were an important link into the police sub-culture, a valuable lesson in this project.

Condom use and regular partners During peer educator training, strong emphasis was placed on using condoms with regular, non-commercial partners. Between 1996 and 1998, sex workers reported an increase in regular partners (64% up to 80% - Port Moresby and 57%

Police peer educators contributed significantly to discussions on reducing harassment of sex workers and reducing risky sex among police.

44 Transex, Papua New Guinea

to 68% - Lae) and levels of condom use (proportion of last week’s acts covered) with them (36% to 48% - Port Moresby and 34% to 67% - Lae). Even consistent condom use increased significantly, as shown in Figure 6. Figure 6. Consistent (100%) condom use reported for the previous week among sex workers with regular partners

40 1996 % 30 1998

0 Port Moresby (s) Lae (s) (s)= p≤ 0.05

Among the men, truckers, dock workers and security men reported significant increases in use with regular partners, as shown in Figure 7. The consistency of use did not improve, however. This pattern is typical. Although many people can learn to

45 Transex, Papua New Guinea

Figure 7. Condom use with regular partners reported for the previous week among male target groups

50 1996 40 1998 % 30

10 (s)= p≤ 0.05 (ns)= p> 0.05 0 Truckers Dock workers Sailors Police Security (s) (s) (ns) (ns) (s)

increase their use of condoms, it requires very strong commitment to personal safety to use condoms every time.

Increasing knowledge In all groups, levels of knowledge significantly increased over the course of the project, regarding both modes of transmission and prevention. Sources of information also shifted, with the majority in every group mentioning the project’s outreach workers or peer educators. Workmates were a significant source of information, indicating that the diffu- sion model of peer education was effective. Most other sources of information,

46 Transex, Papua New Guinea

e.g. television, radio, health workers, actually decreased between 1996 and 1998.

Condom and lubrication distribution The project tried to maintain data on condom distribution, showing the numbers distributed to each target group by month. This did not work well as an indicator because condoms were given without payment, were often given to encourage use (as opposed to on- demand), and were given in large amounts to non-target groups as well. Evidence of a major increase in demand does exist, however. For example, the police called for increasing numbers of deliveries of condoms at the police stations in Port Moresby on repeated occasions. Industry management was very supportive of condom distribution and some companies began putting condoms in pay packets. Five condom depots were set up at the wharves where sailor and dock worker peer educators could collect and sign for their supplies. Others, for example, fire fighters, also made regular use of the depots. In total, between 1996 and 1998, it is believed the project distributed between 1.5 and 2 million male condoms and about 8,000 female condoms. Even without good data on condom use, there is little doubt that demand increased enormously, causing numerous stock-outs.

47 Transex, Papua New Guinea

During periods of condom short- Female condoms were well age, the sex workers were encour- received by the sex workers and aged to buy condoms at shops, and reportedly used with considerable prices in the city were monitored success. Many told their friends so that suitable advice could be about them and many new given. In focus group discussions women became associated with during the final evaluation some the project through a desire for the people expressed concern about female condom. The statement in the quality of condoms available the box below from a sailor whose from the government, reporting ship was berthed in Lae illustrates breakage and the need to double- the impact of female condoms on up. Some men wanted more vari- sexual risk-taking. ety to choose from. Overall, supply problems had not been solved. The Earlier research on the female planned social marketing compo- condom demonstrated that many nent, if implemented, could help Papua New Guineans appreci- solve these problems as could more ated well-lubricated sex. With sensitization and training of DoH encouragement from profes- personnel. sional sex worker organizations

When we got on shore the first thing that came into my mind was to find a woman. I went to Club 96 where there was a dance going on. While there I managed to get a girl out, around in the bush she started negotiating about the male and the female condom.“Which condom do you prefer, the male or the female condom?” I paused for a while and thought, if I refuse, then no sex for me.Maybe I should accept the female one. I knew this must be one of the trained sex workers, so I pretended and insisted on having sex without a condom. The lady refused and replied,“I was taught about AIDS and STD.Who would know if either of us has a disease? It’s safe with condoms, so which one do you prefer?” I had no choice and accepted the female condom.After sex I had all my sense back and thanked the woman for what she had done and paid her extra. Otherwise, I would have had many sleepless nights. She was so polite. I learnt something from her.

48 Transex, Papua New Guinea

elsewhere, extra lubrication was to avoid placing blame on sex introduced to PNG sex workers workers and to illustrate the pos- for protection from abrasion. A sibility of acquiring HIV from the sachet with the Pidgin label Swit semen of previous men in the Gris (Sweet Grease) was line-up. The flip chart given to designed for a water-based each peer educator also had a lubricant and 50,000 were pur- specific drawing illustrating this chased. The peer educators dis- sexual practice as did the comic tributed these to sex workers, books for sailors and for youth, and transport workers also in order to bring group sex to began requesting lubricant. immediate attention when dis- cussing risk behaviours in the Resource materials PNG context. The project developed five targeted comic books in Melanesian Demand for the comic books has Pidgin, an STD picture album, a been very high and over 30,000 poster and several flip charts have been distributed. The proj- used by the peer educators. One ect repeatedly experienced comic book was created about a shortage of funds for reprinting young sex worker and a trucker these very popular materials. The on the Highlands Highway; one designed for sailors was another about a party held with adopted by the South Pacific sex workers on a ship; another Commission for use in other one about STD treatment for sex areas of the Pacific. Other comic workers; and one specifically for books for youth, not directly police. The police comic book related to the project’s target targeted line-ups and is called groups, were also developed, ’Hit ‘n Run’, a term used by the and over 20,000 of these have police themselves. In the story, a also been distributed through sex worker who states she is HIV- the project offices. A poster negative because she just had an specifically for women and suit- HIV test, is forced into sex with able for family planning clinics numerous police, several of and other women-oriented serv- whom are later shown to be HIV- ices, was also designed, printed, positive. The story was designed and 5,000 have been distributed.

49 Transex, Papua New Guinea

An AIDS educational video film agencies. A few papers pre- distributed by the British Red sented at meetings were pro- Cross, using electronic anima- duced from the formative tion, was professionally dubbed research (Jenkins, 1992, 1993b, into Pidgin and several hundred 1994c, 1994d) and from the ini- of these videos were distributed tial findings on the police com- to sailor peer educators for use ponent (Anang and Jenkins, on their ships. Other language 1997). versions, in Tagalog, for example, were made available for use with Efficiency foreign sailors at PNG ports. The project was funded for the first year by both AusAID and In addition, project personnel UNAIDS, with the entire fund- worked with staff at the local ing for the second and third Maritime College to develop years from AusAID. The total training modules on STD/AIDS to annual budget of about be used within the first aid US$ 215 000 supported salaries course. The local police training for two managers, 14 outreach college also instituted a unit for workers, a hotline, rental of their training programme with one and purchase of another help from the project. house, maintenance of these houses, office equipment (fax, In March 1997 a hotline was phone, computer, VCR, over- installed in the Port Moresby head projector), the purchase office and three outreach work- and maintenance of two cars, ers from Port Moresby and two the development and printing from Lae were sent to an HIV of educational materials, the counsellor training course. Two purchase and packaging of sex workers were trained as water-based lubricant, security counsellors as well and helped to systems, and training costs. manage the hotline. Networking with other HIV prevention projects has helped To date, no documents have meet the needs of target been written on the project, groups that the Transex project other than reports to funding could not provide.

50 Transex, Papua New Guinea

Optimal efficiency was compro- mation were attached to survey mised by several factors. The sen- data. The project was approved ior adviser left the project in mid- by PNG’s national medical ethi- 1997 and skills were lost in behav- cal review board. Problems ioural data collection, analysis arose, however, with govern- and documentation. The man- ment personnel to whom sex agement structure changed sev- workers were referred for HIV eral times. Due to the general testing. Confidentiality was complacency about AIDS, recruit- breached and project personnel ment of outreach workers was were told the names of HIV-pos- very slow and, when found, they itive sex workers because health needed a great deal of training. officials wanted to find them Eventually several outreach work- again, in a mistaken notion that ers also left the project, which they could follow up on their required the training of new per- partners or even stop them from sons. Budgetary issues delayed selling sex. Project personnel the project’s start-up dates, espe- were placed in a very awkward cially in Lae. Most of the forma- position and tried to make tive research was conducted prior health workers attend to issues to the arrival of project funding. of confidentiality. The develop- The set-up phase required 6 ment of experience with HIV- months, although baseline sur- positive sex workers began veys continued for longer due to largely through the Henao staffing limitations. In addition, Sisters, the organization run by the project experienced problems sex workers. The Salvation Army with staff members, their spouses, has also begun to provide some lack of condoms, political unrest care for people living with HIV on the streets, temporary lack of and AIDS. transportation, and the intensive use of project personnel by other Replicability agencies. As a pilot project based at a research institution, it is not Ethical soundness likely that future projects, mod- During all formative research, no elled on this one and imple- names or other identifying infor- mented by other types of agen-

51 Transex, Papua New Guinea

cies, could maintain the same Council, through which future level of data collection and funding will be channelled, will analysis without special invest- be responsible for developing ment. Replicability of key com- greater sustainability in the components, such as the peer educa- ing years. tor approach, is certainly likely to be successful. Important issues Lessons Learnt revolve around payment of peer educators, their training and The success of the Transex proj- supervision. During the final ect, in spite of numerous obsta- evaluation, stigma, shame and cles, is due to several factors. It negative attitudes from health was based on extensive sound workers continued to be identi- qualitative research on the fied by sex workers as barriers to real contexts of risk-taking in vul- improved use of STD services. Sex nerable groups. Hence, it has workers prefer to go to the pro- been able to develop materials ject’s own clinics. Without and peer education modules tai- improvement of the attitudes of lored to the specific life situa- service providers, and a reductions of those participating. tion in moralistic attitudes over- These materials and modules all, future HIV prevention proj- have been successful. Providing ects in PNG cannot expect the options, i.e. female as well as same level of success. male condoms, and introducing lubricant to sex workers, has Sustainability been innovative and highly use- The project is not at present sus- ful. Training to diminish tainable without donor funding. moralistic and judgmental Although project personnel have attitudes among staff proved tried to facilitate the develop- to be successful and a valuable ment of an AIDS-dedicated NGO lesson to all observers. The proj- and a sex worker self-help organ- ect showed that the develop- ization, neither of these organi- ment of meaningful relation- zations could function without ships with target groups is a considerable investment from key issue, requiring time and donors. The new National AIDS empathy.

52 Transex, Papua New Guinea

The project has targeted several tized or marginalized groups, is important groups of clients of a major challenge everywhere. sex workers at the same time Communication, job role defini- as the sex workers, who were tions and other issues were not relegated to a later stage or problems throughout the life of lesser position of importance. the project. Numerous recom- This has certainly enabled sex mendations made by the mid- workers to negotiate condom term evaluation were not acted use more easily. Addressing upon. Continued training of harassment of sex workers by staff was not well implemented. police has not only benefited It is clear that some components, the police and their families, but e.g. the hotline, were not very has demonstrated that non-judg- useful and could have been mental approaches even to dropped. Workplace policy highly sensitive sexual practices workshops were highly use- such as group rape, can be effec- ful, but could have been more tive. Evaluation of such changes productive if staff with appropri- would be better served by hav- ate skills were available. Some ing sex workers themselves mon- issues were known but not itor events. planned for, such as care and counselling for HIV-positive sex The project has demonstrated workers, and these emerged that, in Papua New Guinea, sex during the course of the project. workers can become a major Issues dependent on govern- part of the solution, despite ment, such as condom supply their illegal status. Their self-run and the improvement of STD organization has the potential to services, remained unsatisfactory carry out important work in both at the project’s end. As the first prevention and care and it seems targeted intervention using peer that government and donors education as a strategy, the proj- have come to recognize this. ect has attracted a great deal of attention, and many extra Management of large commu- demands were made of project nity-based HIV prevention proj- personnel. Skills in manage- ects, particularly with stigma- ment, advocacy, data analysis,

53 Transex, Papua New Guinea

and documentation were not Acknowledgements well developed among project personnel and would have been Many thanks to the staff of the valuable. However, as in many PNGIMR, the DoH and the developing nations, the pool of AusAID Sexual Health Project for people available with such skills access to documents and data is small. Continued learning with which to construct this and capacity-building must be review. taken seriously by donors and country-based planners alike for the better implementation of targeted HIV preventions.

54 Transex, Papua New Guinea

References Guinea, Prevention Research Unit/GPA, WHO, Sept, 23 pp. 1997 Anang, J. and Jenkins, C. Encouraging safer sex among 1994b Jenkins, C. Behavioural police, security men and sex Risk Assessment for HIV/AIDS workers in Port Moresby, Papua among Workers in the Transport New Guinea. Paper presented at Industry, Papua New Guinea, the 4th International Congress AIDSCAP (FHI), Bangkok, Sept, on AIDS in Asia and the Pacific, 19 pp. Manila, Oct 1997 (Abst A (O) 1994c Jenkins, C. Knowledge vs. 045, p. 45). Behaviour Change: Preventing 1992 Jenkins, C. The Risk of AIDS AIDS Transmission in PNG. 30th in Papua New Guinea. Paper Annual Symposium of the presented at Second Conference Medical Society of Papua New on AIDS in Asia and the Pacific, Guinea, Sept 7-10, Mt. Hagen New Delhi, Nov 8-12 (Abstr). (Abstr). 1993a Jenkins, C. Fear of AIDS - 1994d Jenkins, C. Sex as Work in the second outbreak in Papua Papua New Guinea (Poster), Xth New Guinea (June 2); AIDS and International Conference on the economy of Papua New AIDS, Yokohama, Aug 6-12 Guinea (June 9); A national AIDS (Abstract Book 2, p. 325). prevention programme for 1995 Jenkins, C. A Study of the Papua New Guinea (June 6). Acceptability of the Female The Post Courier: 11. Condom in Urban Papua New 1993b Jenkins, C. Sex and Guinea, UNFPA, Port Moresby, Society in Papua New Guinea, May, 22 pp. paper presented at IXth 1996 Jenkins, C. Final Report: International Conference on Poverty, Nutrition and Health AIDS, Berlin, June 7-11 (Abstr). Care: A Case Study in Four 1994a Jenkins, C. Situational Communities in Papua New Assessment of Commercial Sex Guinea, 53 pp., report submitted Workers in Urban Papua New to The World Bank, June.

55 Transex, Papua New Guinea

1998 Jenkins, C. and Passey, M. STD/AIDS Situation in Papua New Papua New Guinea. In: Sexually Guinea and Proposed AIDSCAP Transmitted Diseases in Asia and Activities. Nov 14-Dec 4. AIDSCAP the Pacific, eds. Brown, T., Chan, Report. R., Mugrditchian, D., Mulhall, B., Plummer, D., Sarda, R. and 1994 The National Sex and Sittitrai, W. East-West Reproduction Research Team Center/Thai Red Cross and C. Jenkins. Sexual and Society/Venereology. Armidale: Reproductive Knowledge and Venereology Publ. Co, Behaviour in Papua New pp. 230-255. Guinea. Papua New Guinea Institute of Medical Research 1998 Lepani, K. and Stephens, D. Monograph No. 10. Goroka: Evaluation of the IMR Transex PNG Institute of Medical Project: Port Moresby and Lae. Research. Unpublished report for AusAID, 28 pp. 1994 Malau, C., O’Leary, M., Jenkins, C. and Faraclas, N. HIV/AIDS prevention and control in Papua New Guinea. In: Current Opinion in Infectious Diseases. AIDS in Asia and the Pacific, ed. Kaldor, JM. 7: S117- S124. 1999 Mgone, C., Passey, M. and Russell, D. Prevalence of HIV and STDs among commercial sex workers in Port Moresby and Lae. Unpublished PNGIMR Report. 1993 Mugriditchian, D. and Jenkins, C. Assessment of the

56 Sonagachi, India

Sonagachi: A Sex Worker Project in a Red Light District of Calcutta, India

Introduction

he Sonagachi project is a well- Tknown project, having had wide media coverage in recognition of its impact on the lives of sex workers and the social change approach it has added to standard public health strategies. As an HIV intervention it has had several evaluations, both external and internal. It was initiated by the All India Institute of Hygiene and Public Health (AIIH&PH) in 1992 as the STD/HIV Intervention Programme (SHIP), in consultation with the National AIDS Control Organization (NACO) of India, the Ministry of Health and Family Welfare of West Bengal, and WHO. Later donors included NORAD, DfID, and HORIZONS/USAID. It also includes two NGO partners, the Health and Eco-Defence Society and the Human Development and Research Institute. As a relatively mature project, it has broad- ened its support base in the area since 1994 by enlisting a women’s organization (Sramajibi Mahila Sangha) and a legal aid agency (Socio-legal Aid Research and Training Centre). Most

57 Sonagachi, India

importantly, the Sonagachi sex ceive. There are examples, how- workers formed their own ever, of government policies in organization in 1995, the Durbar Thailand and Senegal that incor- Mahila Samanwaya Committee porate some of the same aims (DMSC), which has become the and mechanisms. The essential major mover of the programme. difference between the process that has taken place in Gradually, what started as a nar- Sonagachi and elsewhere in the rowly conceived HIV prevention developing world is that the sex project, has become a social workers themselves, including movement in West Bengal of no their families and friends, have mean importance, and is draw- taken the lead and carry out the ing in sex workers and women’s work of the intervention. This rights organizations from all includes support of the behav- over India and beyond. As about ioural, clinical, financial, politi- one-quarter of the sex workers cal and legal aspects of the proj- in the Sonagachi area are from ect. As such it provides a Nepal and Bangladesh, linkages remarkable example of power- to sex worker projects in those less women moving into the 21st countries have also been devel- century with an altered image oped. The potential to move sex of what they can accomplish for work out of its feudal social and themselves, and a lesson for economic bases to emerging those involved in the women’s alternative forms of service pro- movement, poverty reduction, vision underlies the project’s and public health in general. It growing vision, articulated by requires that thorough consid- the sex workers themselves. In eration is given to society’s so doing, sex work could stance on prostitution and that become an industry-regulated sex workers can have a voice in business with mechanisms to the ensuing debate. sustain safer sex practices. Such a vision is far beyond what most The project is complex with HIV prevention projects with sex numerous components and is workers, particularly in develop- well documented in both English ing counties, are ready to con- and Bengali. Its original specific

58 Sonagachi, India

aims were simple: to reduce lev- able, to date, to deal with all the els of STDs, increase condom use, factors that place the Sonagachi and to develop an effective strat- sex workers at risk. These imper- egy that could be replicated else- fections, however, have not where. In this the Sonagachi destroyed its obvious effective- project has been successful. In ness or credibility. There is little terms of HIV and sexual health, doubt that the Sonagachi project current syphilis infections and is one of the best examples of an clinically detected genital ulcers HIV prevention project that have diminished. HIV positivity addresses the short-term needs rates in eastern India are gener- to control the epidemic as well as ally lower than elsewhere in the longer-term requirements to India. Among the sex workers at reduce vulnerability of women in Sonagachi, they also remain rela- the sex industry. tively low, at approximately 6% (as of 1998). The project is now Situational expanded to many other areas in Background Calcutta and West Bengal. Of greatest significance is the Calcutta, sometimes called ’the emphasis on meeting the felt city of joy’ is also a city of needs of sex workers and extreme poverty, of Mother enabling them to take charge of Theresa’s domain, and the histor- solving their own problems. ical seat of the British Raj. It is a Ultimately, this is the final goal, crowded and underserved city, one appreciated by sex workers struggling to clean itself up and themselves from around the modernize. It is also the tradi- world. tional centre of Bengali culture and society, a culture that takes The project has had contentious great pride in its poetry, arts, moments dealing with a large music and political thought and array of actors, sensitivities, and action. The relentless oppression conflicting interests. It has also of women in India is powerful had some difficulties in tech- precisely because it is deeply niques and modes of monitoring integrated into class, caste, and and evaluation. It has not been ethnic hierarchies. Resistance to

59 Sonagachi, India

class exploitation and social den- donment, running away from igration has fuelled numerous the cruelty of husbands and in- movements in modern Bengali laws, and becoming a family out- history. It is within this cultural cast due to personal attempts to and historical context that the break through the intense Sonagachi project has grown. restrictions on the lives of young Surrounding West Bengal are the women. To this list today we states of Bihar and Uttar Pradesh must add poverty from birth, in India and the nearby countries fake marriages and sale, situa- of Nepal and Bangladesh, all of tional poverty due to family which are substantially poor and losses, trickery and trafficking, disadvantaged, both ecologically abduction and sale. In addition, and socially. From these areas some family traditions of dancers hail a significant minority of and entertainers, e.g. Agrawalis, Sonagachi’s resident sex workers. have gradually become family lines of sex workers. Sonagachi is the derived pronun- ciation of a lane called Sanah The pattern of recruitment into Ghazi, named after a dacoit or the sex trade includes cross-bor- hoodlum, recast as a religious der movements and a significant teacher. The Sonagachi area was number of young girls, approxi- first identified as a significant mately 15% under the age of 15 red light district of Calcutta at the time the project began, when the British enacted the entered in that way. The sex Indian Contagious Diseases Act industry is controlled by police, in 1868. This was done in an politicians, local hoodlum gangs, effort to control STDs reaching brothel owners, madams and England via British soldiers serv- pimps. While some sex workers ing in India. At that time, about are independent, others, particu- 30,000 sex workers were esti- larly those newer to the trade, mated to work in the area. A sur- have madams and/or pimps. vey conducted in the late 1800s Some madam-controlled sex noted three major reasons why workers known as chhukris are women entered prostitution. completely bonded and each one These were seduction and aban- is obliged to earn enough to pay

60 Sonagachi, India

off an advance given to her or, clients and then form longer- more often, to the persons who term relationships. While some ‘sold’ her into the trade. Another of these men actually care for madam-controlled arrangement their women lovers and con- is called the adhia system, in tribute to their expenses, many which the sex worker turns over are abusive and exploitative. 50% of her income to the Nonetheless, many women need madam. Pimps take 25% of the these companions for emotional sex worker’s income and are and practical reasons, and are organized in a hierarchical and tied to these relationships. regulated system headed by mukia, or political headmen. The Sonagachi project area Rooms are rented at high rates. includes four main colonies of The sale of sex is legal in India if brothels, located relatively close it takes place “within a room”. In to each other, known as order to avoid arrest, freelance Sonagachi, Rambagan and or ‘flying’ sex workers must also Sethbagan, Jorabagan, and rent rooms at high hourly rates. Rabindra Sarani. Collectively, The Prevention of Immoral this is the largest red light area Traffic Act, aimed at regulating in Calcutta, with about 370 sep- the flow of new women into the arate brothels housing about trade, in fact simply makes it eas- 4000 sex workers. The total ier for sex workers to be number swells to about 5000 in harassed, extorted, beaten, September and October during raped, and otherwise abused, the Hindu Puja season, a time of when the Act is used to raid revelry and greater earnings for brothels on the pretext of sex workers. In addition, on the removing younger women. The crowded streets and tiny lanes attitude of the surrounding soci- below the multi-storey tene- ety is generally to despise and ment brothel buildings, about stigmatize sex workers on moral 1500 flying sex workers operate. and social grounds. The whole neighbourhood is served by innumerable small In addition, many sex workers shops, itinerant traders, veg- have babus, men who begin as etable hawkers, youth clubs,

61 Sonagachi, India

wine shops, and medical clinics, income, which is largely based some run by quacks. Prices at on age and attractiveness. these establishments are higher Higher-income women have than in other neighbourhoods pleasant rooms, sometimes with in Calcutta of similar class. air conditioning, while the Hordes of people can be seen at lower-income group may share any time of day and night trav- a single room among up to five elling on bicycles, scooters, rick- women, with beds separated by shaws and on foot. It is a curtains. Ages range from 13 to crowded, dirty, lively, and dan- 45 with incomes from Rs. 10 to gerous area where daily about 2,000 per night (39 Rs.=1US$). 20,000 men come to drink and Despite crowded quarters, the find sex. Not only the sex work- women live their lives in consid- ers, but their clients as well, are erable isolation from each other preyed upon by hoodlums. As in and from the larger society. most such areas, the police Madams and pimps impose rules either ignore or are involved in that keep sex workers from visit- the ambient violence. ing each other. Older sex workers, unless they become madams, are unable to secure many clients and many work at housekeeping chores or babysit- ting to make a living. Eventually, many become destitute.

Relevance The AIDS epidemic in India is a rapidly growing one. As one of At a street corner with Sonagachi pimps learning about HIV/AIDS. the world’s most populated countries, its absolute numbers of HIV-infected and affected Each building has several broth- people are likely to become els in it with 5 to 25 rooms. Sex staggering in the next decade. workers fall into different cate- With high proportions of very gories according to their poor people and cultural pat-

62 Sonagachi, India

terns dictating very low status ogy, socio-demographics and for women, Indian women are behaviour. Flying sex workers rapidly becoming the bearers of were excluded. After an initial the HIV burden. Sex workers count of 362 brothels housing represent those with the lowest 3,664 sex workers, the sampling status and the highest risk, methods were designed. Sex among all women. The All India workers were grouped into Institute of Hygiene and Public those with charges higher than Health (AIIH&PH) has been Rs. 100 per client (Category A), known for its various health those with charges between Rs. interventions and hence, has 50 and Rs. 100 (Category B) and been able to lead in the devel- those with charges less than Rs. opment of a broad-based sup- 50 (Category C). Sex workers port system for the project. This were sampled randomly in pro- support mechanism is known as portion to the number in each the Conglomerate, and includes category. Sixty-five brothels NGOs, community-based organi- were randomly selected out of zations (CBOs), and government the possible 362 and about 6-7 institutions. At the time the sex workers interviewed in each project began, the prevalence of (with non-responder replace- HIV was very low in West ment), for a total sample size of Bengal, providing an opportu- 450 (12% of the possible total). nity for a creative and flexible In addition, 360 women from intervention. this sample were willing to undergo an STD examination Formative Stages conducted at a two-room clinic set up in a volunteer local club The initial baseline survey, con- in the heart of the red light dis- ducted by personnel from the trict. STD specialists from the AIIH&PH with the aid of an NGO Calcutta Medical College that had been working in the attended the clinic, and labora- area, the Society for Community tory work was carried out at Development, was carried out the Bacteriology Department of from April to June 1992. The the Calcutta School of Tropical survey covered STD epidemiol- Medicine.

63 Sonagachi, India

Briefly, the social survey found used condoms and only 31% that 85% of the sex workers had heard of AIDS. More, 69%, were between 15 and 29 years knew something about other old. Most stated they entered STDs. the trade due to poverty (49.1%), family disputes (21.6%) Endocervical smears detected or being misguided (15.6%). 10.5% positive for gonorrhoea, Also, 85% were illiterate and with 4.9% confirmed by culture, many were heavily dependent although, for technical reasons, on alcohol, and 9.3% were new- it later appeared that the meth- comers, having been in the ods used were insensitive and trade for less than one year. In underestimated the true rate of this sample 43% of the women gonorrhea. Trichomonas vagi- visited and worked in other nalis infection was detected in areas of the city as well, staying 11.1%. Genital ulcers were seen there for some weeks or months in 6.2 % of the women. before returning to Sonagachi. Serological testing for syphilis The majority had 3 to 4 clients found 25.4% positive by VDRL per day for weekly incomes that (dilution=> 1:8). The HIV positiv- varied between less than Rs. 300 ity rate was 1.1%. to over Rs. 5,000. Many women (27%) reported that men hired Later in 1993 a client survey was them in a group, a situation that conducted by 193 sex workers, seriously diminishes the as they had the easiest access to woman’s control. About 40% of these men. Sex workers from the women had children, but each category of income were the proportion with babus was selected to conduct a brief ques- not determined; 45% of the tionnaire-based interview with women used some form of con- one client each. As these women traception but only 27% did so were mostly illiterate, they had consistently. Condoms were to memorize the questions and always used by only 1.1% and answers. Social workers visited often used by another 1.6% of them in the mornings and filled women. At that time in 1992, the form in with their help. 90.6% of sex workers had never Literate clients filled in their

64 Sonagachi, India

own questionnaires. The sample bars. While vaginal intercourse was biased towards men they was most common, 47% also knew, men who were spending enjoyed oral sex. Anal sex was the night, and fixed clients and, reported as rare, and 4% paid to while useful, cannot be consid- be masturbated. ered representative of all clients. Those clients who Another survey of 200 babus was entered and left the brothel conducted, showing that 26% rather quickly were underrepre- were illiterate, but 9% were grad- sented. Thirty-seven clients of uates of universities. Similarly, sex workers of category A, the 46% had low incomes, less than highest income group, were Rs. 1,000 per month, and 4% had interviewed, as were 75 and 81 incomes over Rs. 2,500 per month. of categories B and C, respec- Although half had heard of AIDS, tively. Results revealed an age 73% had never used condoms. range of 15 to over 40, with a Only 4.5% reported using con- tendency of younger clients to doms regularly. visit less expensive sex workers and more educated clients to In addition, in 1993 the project visit the more expensive undertook another survey of sex women. The mean ages of workers to learn about their sit- clients by sex worker category uations and aspirations regard- were: A-42, B-39, and C-34. Of ing their children. Some 2,338 the clients, 87% were literate women were interviewed, 47% and 44% were businessmen. of whom had children under 14 Others were wage-earners, pro- years old. Despite the consider- fessionals, drivers, students and able felt need for children, police. The proportion of stu- largely for emotional and social dents (8.8%) was as high as that reasons, pregnancy was a serious of drivers (8.3%). Slightly over risk for these women, with 14% half (52.3%) visited the brothels stating they had deliberately ter- once or twice a weekend and minated their pregnancies, many 9.3% visited daily. Most were more than once. A total of 768 alcohol drinkers and spent time abortions had been experienced entertaining themselves in wine by 330 women, or an average of

65 Sonagachi, India

2.3 abortions per woman. One Respect and Recognition, i.e. woman had had 15 abortions giving due respect to the and one live birth. An even women, relying on their knowl- greater problem is the future of edge and recognizing their pro- those born. About half were fession. The project team mem- brought up in the brothel with bers had to make it clear to all little in the way of decent care. that they would not disturb the Some sex workers paid older sex existing power structure or seek workers to raise their children, to remove women from prosti- but the majority kept their chil- tution. Throughout the pro- dren with them, arranging ject’s life, all those who influ- makeshift crèches when work- ence the sex trade, even if in a ing. The others were brought up negative way, have been drawn outside the brothel, either with into its activities. At the same family members or in orphanage time, abuse by thugs, power homes, with 90% of the financ- brokers, unethical researchers, ing supplied by the mother. Less and others has been resisted by than 5% were raised in govern- the sex workers and their sup- ment or NGO-run homes. About porters. The project itself has half of the children had received become a sentinel of resistance some form of education, but through organization and affili- drop-out rates from primary to ation, in order to create a plat- secondary school were very high. form upon which sex workers None managed to go beyond could personally and collec- secondary school. The mothers tively grow. Therefore, allies stressed their need for hostel or among government agencies, boarding facilities with proper NGOs, CBOs, academic institu- schooling and supportive envi- tions, and numerous other ronments for their children. organizations have been incorporated in one way or another. Implementation Publicity in newspapers, dozens of public appearances, confer- Working strategies ences, workshops, and visible The project began with a philos- campaigning have been used to ophy based on Reliance, draw attention to the issues

66 Sonagachi, India

requiring reform. These issues personnel. By April 1996, 180 are those most relevant to the police officers had undergone safety and well-being of sex this training. workers, as defined by themselves. They include their rights Health care as workers, the rights of their A clinic, called the Health children, the safety of their Service Centre, was established neighbourhoods, their concern in 1992 on the premises of a over the trafficking of children local youth club, Palatak, in the in the sex trade, and their legal middle of the Sonagachi area. and constitutional rights. The Health Service Centre is staffed by one of two rotating For project staff, misgivings had STD specialist doctors and is to be overcome. Some felt open 6 days a week from 11:00 endangered by being close to to 14:00 to sex workers and sex workers, and feared catch- members of their families. ing diseases. Others felt embar- Treatment is free and serum rassed having to deal so directly samples are collected oppor- with sex. Hoodlums were a very tunistically for syphilis screen- real threat. Gradually, they ing. Sex workers are encour- overcame their fears and have aged by their peers to attend become true supporters of the the clinic for screening even women. when asymptomatic. Follow-up on treatment is carried out Police raids have been a major through the work of peer edu- problem in this project, keep- cators. On average, in 1998, 730 ing the sex workers insecure women came to the morning and frightened. After every clinic per month. In 1993 raid, staff believed that con- another clinic was opened for dom use declined and STDs rose evening sessions 5 days a week among clinic attendees. nearby in another club, specifi- Therefore, the AIIH&PH in cally for clients. Demand has cooperation with the Calcutta grown greatly and the clinic is Police Department arranged a now crowded; other clinic days training programme for police are needed.

67 Sonagachi, India

In 1995, the peer educators undertook a rapid assessment of the situation of sex workers in the rest of West Bengal. They found 254 red light districts and spots throughout the state. After prioritizing 30, they began organizing and, as of late 1999, 14 new projects had started. New The morning clinic staff lined up for clinics were erected, peer educa- photos with one of many guests who visit tion begun and condom sales regularly. carried out. This effort is funded by the project and managed by Gradually, throughout the life of the DMSC. the project, due to the demand of sex workers, additional clinics of a similar kind have been opened. In 1994, one opened in Sethbagan and another outside the project area in Bowbazar, another red light district in Calcutta. These are operated, on behalf of the AIIH&PH, by the project with an NGO conglomerate. Between 1994 and 1998, 9 Newly constructed DMSC clinic at Durgapur, West Bengal. more clinics were opened in red light districts in Calcutta specifically for sex workers and their Technical aspects of STD care, families. As the years have such as resistance to antibiotics passed, the women have seen and verification of cure, remain the value of good-quality health to be handled. Aggressive services and have noted the screening and more referral and poorer quality and disrespectful follow-up of babus and other treatment in other facilities in clients are required in the next the city. phase of growth.

68 Sonagachi, India

Peer education perform plays about STD/HIV Peer educators were selected prevention as well as the issues and trained, starting with an ini- surrounding prostitution. This tial group of 12. By 1997 there dance/theatre troupe has were 65 trained peer educators become an important medium who were paid the equivalent of by which the sex workers can US$1 per day. These women express themselves, forge a pub- were given 6 weeks of training lic identity and reach a large and utilized specially designed number of persons of all walks of flip charts, requiring little liter- life. It won first prize at a acy, to explain the basics of sex- National Cultural Competition ual health and safer sex to their for Sex Workers, held in Benares peers. The Sonagachi area was in mid-1997 and, in 1998, it per- divided into 11 zones that were formed at the Twelfth World covered by the 65 peer educators AIDS Conference in Geneva. with 7 supervisors. From 10:00 until 13:00 every day, each group contacts 40 to 50 sex workers and 10 to 15 madams. They encourage the women to go for regular check-ups as well as seeking care when they have symptoms. Leaflets are given to those who are literate. In addition video shows, slide presentations, small Komol Gandhar, performing in Geneva at group meetings and ventrilo- the World AIDS Conference 1998. quist performances have been used to reach larger groups of After expanding to other broth- sex workers and clients. els in Calcutta, the number of trained peer educators has They have also developed a per- reached 200 with a coverage of forming group, Komol Gandhar, about 20,000 sex workers and that includes male and transgen- babus. As the project has der sex workers, and perform at expanded, it has become difficult different cultural events. They to continue to pay all the peer

69 Sonagachi, India

educators needed in order to positive people. As the number maintain a ratio of one peer edu- of HIV-positive people grows in cator to 50 sex workers reached. India, the aim is to develop a Now the ratio is 1:160. In order model of community-based care to deal with this problem, the for West Bengal. project has developed an additional cadre of volunteers to help Condoms the peer educators reach In the beginning, the peer edu- women. Monitoring the quality cators distributed free condoms of work of these volunteers rep- and enquired about the extent resents a challenge in scaling-up, of condom usage. Demand for typical of most trainer-of-trainer- the high-quality, imported con- like projects. doms distributed by the project grew faster than the supply. In 1998 the Positive Hotline was During the first month of the initiated and maintained by the project, 3592 condoms were dis- sex workers in response to hear- tributed; by 1994, during the ing about the sad situation of an month of December, 79 420 con- HIV-positive woman who was doms were distributed. This rose not a sex worker. It has started to 110 328 in December 1996. with a single phone at one of the Because of budgetary con- field offices and is staffed at straints, the project was unable night by the children of sex to meet demands. workers. This service to the larger community emerged from In July 1995, the sex workers of recognition of the need for Sonagachi formed a registered counselling and support for HIV- society called the Usha positive people, whether sex Multipurpose Cooperative Society workers or not. A core team has Ltd., a consumers’ cooperative been formed of volunteers, a which would help them save doctor, a nurse and a counsellor. money and avoid the exorbitant In addition to medical care, the interest charged on small loans service offers psycho-social and by money lenders in the area. legal aid support and the team Registering as sex workers, can be sent to the homes of HIV- instead of ‘housewives’, required

70 Sonagachi, India

a long struggle with authorities, chased a piece of land outside of but with a growing capacity for Calcutta at which they have been advocacy and publicity, DMSC able to construct a training cen- was successful. Starting with only tre where they conduct most of 13 members, it now has over their training. The older sex 1000 and regularly gives loans at workers produce handicrafts low-interest rates to sex workers. there when specific market Small investment schemes utiliz- opportunities emerge. ing bank deposits have also been developed. Its funds provide a Empowerment crèche for sex workers’ children Organization has been the key during working hours and it to empowering the sex workers gives employment to older, out- of Sonagachi. Several influences of-work sex workers. In April converged to encourage the sex 1997, it developed its first for- workers of Sonagachi to organ- profit business, the marketing of ize themselves. First, a previous condoms. Bought in bulk, the organization of sex workers, condoms were sold to any group called Mahila Sangha, had wishing to buy. Recorded sales formed in the Sethbagan area in increased from a total of 213 056 the 1970s, to fight against the in 1997 to 443 805 in 1998 and regular and violent extortion of 730 656 in 1999. The project still money from sex workers by a distributes free condoms as well local mafia. After gaining vic- for newcomers to the brothel, to tory, they moved on to literacy men who attend the STD clinics and health programmes. and to very poor sex workers. In Eventually, one of the leaders the meantime, condoms manu- moved over to the Sonagachi factured in India have improved project and led the Sonagachi and importing is no longer con- sex workers in their first appear- sidered necessary. An activist- ance at the Calcutta Book Fair in cum-sales arm of the Usha 1992. Second, regular meetings Cooperative, Besanti Sena, now of the peer educators began to markets condoms in 40 red light knit them together and, to districts of West Bengal. With the resolve conflicts, they formed an profits made, they have pur- Advisory Board and a committee

71 Sonagachi, India

to handle grievances. In 1993, emerged which is participatory the peer educators and supervi- and has its own primer. A small sors held a candlelight proces- newsletter with an attractive sion against AIDS which made glossy format, Namaskar, was the headlines. Step by step, new started, in which they can tell of elements were added. The peers their experiences. These are sold were given a green uniform by the peer educators at fairs jacket with a printed red cross and meetings, particularly the on it and an identity card. With annual prestigious Calcutta Book this they could feel like real Fair. Four editions have been pro- health workers. Women began duced so far. In addition, daugh- to take their clients into the ters of sex workers were clinic to have their STDs treated recruited to be teachers of liter- and were sought after for advice acy and there are 14 such teach- and information. It was found ers so far. This strategy has had that, in the beginning, women the advantage of giving pride to with children were the most the mothers and to the daugh- enthusiastic about becoming ters, and helping them gain sta- peer educators as it gave them a tus in society. Indirectly, their more respectable status in the new identities may make it easier eyes of their children. to negotiate better positions with regard to dowry and mar- Running their own project and riage at a later age. organizations required that some of the women, at least, DMSC began to develop cultural, were literate. The women sports and artistic activities for wanted a literacy project for sex workers’ children and, with themselves. One of the supervi- the help of the project, placed at sors agreed to teach and they first 50, and now 100 children in began, but soon it became mainstream boarding schools apparent that adult learners yearly. A crèche now operates for with special life experiences brothel children as well as sev- could do better if the literacy les- eral other programmes spon- sons grew from their own dis- sored by donors and operated by courses. A new mode of learning collaborating NGOs/CBOs.

72 Sonagachi, India

In 1994, researchers, with the of sex workers in other parts of help of an NGO and police, col- Calcutta and demonstrated lected blood samples from 50 against the unethical and illegal Sonagachi women without their testing of an unqualified AIDS consent. This created panic in the vaccine. They collected money to area and the Institute registered give to the State for flood victims its protest. Eventually, the situa- and, by 1996, began touring tion calmed down but left the other parts of West Bengal to women with determination to examine the situation of sex form an organization that could workers, which, in many cases, deal with local issues, such as they found shocking. They visited that one, which directly affected sex worker projects in their lives. Bangladesh and Nepal. Not long afterwards, representatives of In 1995, the DMSC Committee led the Committee attended interna- a group of 1,000 sex workers to tional meetings of sex workers, demonstrate against police raids meetings on the trafficking of in the Sonagachi area. They children in the sex trade, and joined rallies against the eviction national and international con-

Sex workers’ children dancing and playing in front of the main clinic.

73 Sonagachi, India

gresses on AIDS. In late 1997, the the brothel. This is both against Committee brought together sev- the law, causing the police to eral thousand sex workers, make repeated raids, and a prob- including males and transvestites, lem to older sex workers. The for the First National Conference younger women have less capac- of Sex Workers in Calcutta. They ity to insist on safer sex, are bio- have taken on the role of the logically more vulnerable to HIV Secretariat for the Asia/Pacific Sex infection and draw clients away Workers’ Network. In March from the older ones. The mem- 1998, they held another confer- bers of DMSC have taken the ence, the Follow-up Phase of the stand that they can better regu- First National Conference of Sex late the entry of young women Workers. This conference was into the brothels themselves attended by 2,000 sex workers than can the police or other from 48 red light areas all over agencies. They set up a system India. At this conference they through which they could moni- were able to attain several of tor new arrivals and counsel their objectives, particularly them before they actually enter those related to networking. the brothel trade. At first they They now have a vision and a were attempting to send them platform of goals, aimed at inte- back to their homes, but the grating themselves into a majority simply refused to go, worker’s world of rights and having run away from abusive or responsibilities. The Committee difficult circumstances. With the now has over 250 male and trans- help of the Social Welfare gender members. Department, they now have an arrangement whereby most of Three hundred sex workers have the girls are sent to boarding received legal training. As some schools sponsored by the Social wanted more than the short Welfare Department. Over the course originally given, another years this process appears to 60 had additional legal training. have reduced the proportion of bonded sex workers, usually the One vexing issue has been the youngest, entering the brothel, presence of young sex workers in although indicators have not

74 Sonagachi, India

been recorded for this. With the ber of pimps and madams aim of spreading this type of self- appears to be dropping. At regulatory action, DMSC set up Sonagachi, as elsewhere, the in 1999 three more Self-regula- AIDS epidemic has led to the tory Boards at Sethbagan, realization by many sex workers Tollygunge and Rambagan in that they can avoid the repres- Calcutta. These boards are com- sive and unsafe conditions in ille- prised of members from the state gal brothels by diversifying their government, women’s’ rights work modes. Many have moved groups, other local people and into residential or hotel-based sex workers. Members of DMSC sex work, a situation having frequently take part in meetings implications for future HIV pre- and debates on trafficking, an vention as well as the evolving issue of intense concern in the nature of the relations between South Asian region. the state and the sex trade.

Helping sex workers leave the Clients and lovers sex trade has never been an aim For clients, a clinic has been of the project but, inevitably, established with free treatment, growth in self-esteem and a free condoms and counselling, sense of control over one’s own but no clear mode of establish- life has led some women to ing client-targeted programmes choose to leave the trade. About has emerged. Police, youth one-third of the peer educators groups and the general public who have been active for at least have received several awareness several years have moved out of programmes. Some clients have the brothel. They return daily to been reached at wine shops and conduct their work; about 60 through forms of ‘edutainment’, women have left sex work such as ventriloquist shows. entirely. The small salaries they Babus, however, represent a spe- receive from the project must be cial problem and, as fixed clients, supplemented by other income have received special attention. and many of these women now A meeting in the local Botanical sell saris and other small mer- Gardens was held in June 1997, chandise. In addition, the num- attended by about 150 babus,

75 Sonagachi, India

which discussed the problems of should develop different indica- the Sonagachi area and the role tors. Instead of focusing nar- of babus in the establishment of rowly on effectiveness in STD safer sex norms. This was fol- control, it should find a way to lowed by four more such meet- document the emerging strength ings. Subsequently, a babu com- and self-esteem of sex workers. mittee and a collective, known as By 1998 several new indicators Sathi Sangathan (Companion’s were added. The attendance of Collective), were formed. The clients at the project’s clinics had aim is to work with DMSC to been difficult to monitor as men fight against all forms of harass- sought treatment who may not ment and violence against sex have been clients. A referral card workers and their clients. One of given by the sex worker to the their services has been to escort client (and later handed in at the clients past the bands of gundas, clinic) now makes it possible to or hoodlums that operate on the monitor how many clinic atten- streets of Sonagachi. dees are clients referred by sex workers. The duration of self-rec- Monitoring and ognized symptoms when the sex Evaluation worker presents to the clinic is now recorded. Instead of contin- The project collects data on sev- uing to ask how many sex acts eral different monitoring indica- were covered by condoms, the tors both at the clinics and project now monitors how many through the peer educators. It condoms were sold, on the also relies on periodic cross-sec- assumption that women will buy tional surveys for feedback on what they will actually use. effectiveness, with one com- Levels of empowerment and pleted at the end of 1993, engagement in the community another in mid-year 1995 and movement are now measured by the latest in 1999. The first exter- asking how many times the sex nal evaluation took place in 1996 worker manages to negotiate and another in 1999. One impor- with landlords, police and other tant recommendation made by power brokers, as well as how these teams was that the project many sex worker organization

76 Sonagachi, India

meetings or conferences she from 30.7% to 85.8%. The per- attends. Changing the indicators centage of clients using condoms for monitoring the success of an always or often increased from evolving intervention is essential, 2.7% at baseline to 69%, a con- but must be done with care so siderable increase. By 1998, this that continuity with the past is percentage rose to 90.3. minimally affected. Although the methods used to Effectiveness detect gonorrhoea were too At the end of 1993, 14 months insensitive to detect all cases, after the intervention was imple- these same methods were used mented, the first follow-up sur- for the baseline and first follow- vey was conducted. Sampling up surveys. This showed a methods were essentially the decrease in gonorrhoea from same as the baseline survey and 13.2% to 3.9%. Neither HIV nor 612 women were interviewed. recent syphilis infections Age distribution was similar to showed any significant change the first survey and the literacy in prevalence. rate remained very low. The proportion of new sex workers (in In an attempt to improve the sur- the trade less than one year) was vey methods, decisions were considerably higher (22.6%) than made for the second follow-up at baseline (9.3%). The median survey conducted in July-August number of clients the previous 1995 to alter the sampling and day was 3. Group sex appeared STD detection techniques. This to increase from 27.3% at base- decision meant sacrificing some line to 48.7%. The proportion of of the past data in order to have women reporting oral sex improved measurement capacity appeared to decrease from in the future. However, the 74.4% to 38.1%. It is not unlikely hoped-for improved random that the questions were asked in sampling methods did not work. different ways in each survey, True randomization is very diffi- causing this puzzling result. The cult without identifying the proportion of women reporting women in some way, especially some knowledge of AIDS rose because the sex workers were

77 Sonagachi, India

interviewed at the brothels and had to go voluntarily to the clinic for examination and testing. In fact, more went to the clinic than were randomly selected and it was hard to know who had really been interviewed. In addition, although the laboratory methods to detect gonorrhoea were improved, laboratory capacity was not increased, and only 10 samples could be handled per day. In November 1995, when an external team of experts evaluated the project, it was recommended that in the future the laboratory work should be handled by a commercial laboratory with greater capacity and that a reference laboratory be utilized for quality control.

Nonetheless, clear evidence exists of a reduction in levels of recent syphilis and the presence of genital ulcers. Measures of current syphilis are higher for those who have more clients than others, are illiterate and have not attended any meetings organized by sex workers.

78 Sonagachi, India

Figure 1. Change in proportions of sex workers with genital ulcers, recent syphilis and HIV over time

20 VDRL=>1:8

% 15 Genital ulcer HIV+

0 1992 1993 1995 1998

Although the rates of HIV among sex workers in most parts of India had, by then, shown dramatic increases, by 1998, among 503 randomly sampled women at Sonagachi, only 5.5% were found to be HIV positive. While condom use has increased greatly, it has not reached total or near-total coverage. Only about 50% of the sex workers use condoms all the time with clients and, as is typical in sex worker projects, the greatest proportion of this level was reached early in the project. The project has not placed a great deal of emphasis to date on specific negotiation skills or materials. Options that would enhance

79 Sonagachi, India

usage rates, such as female condoms and lubrication, are not yet incorporated.

Measures of condom use with babus are collected within cross-sectional babu surveys and show a reported rise in use at last sex from 4% to 30% between 1995 and 1999. Babus continue to exercise considerable control over their relationships with the women as well as denying risk for themselves and their own wives. Despite enormous changes in levels of self-esteem for many of the women at Sonagachi, these effects have not reached all. Passivity, fear, and desperate need for money militate against the assertiveness necessary to make demands on a man who may be drunk and poten- tially violent. One important way in which babus can exert power is the giving of their name to sex workers’ children so the children can be enrolled in school. As they can repudiate this assign- ment at a later date with the school system, this gives them continuing leverage and power over the women. In addition, madams and pimps can demand that their workers take clients without condoms and many young women share single rooms with other women, making assertiveness embarrassing and difficult. There are some victories as well. For example, one of the most resistant groups, the Nepalese women, has begun to ask for condoms. Table 1 shows the

80 Sonagachi, India

percentage of sex workers using condoms always or often with clients.

Table 1. Patterns of condom use with clients over time

Sex Workers’ Condom 1992 1993 1995 1998 Use

%%%%

Always uses condoms 1.1 47.2 50.1 50.4

Often uses condoms 1.6 22.1 31.6 40.1

Total 2.7 69.3 81.7 90.5

Table 2 (and Figure 2) show the proportion and the respective confidence interval of clients who used condoms in the previous 24 hours, a question asked in all surveys after the baseline in 1992. Although the change in proportions is not very large, the trend is statistically significant.

Table 2. Proportion of yesterday’s clients who used condoms at all

Client’s Condom Use 1993 1995 1998 No. of clients 2 139 1 362 1 174 No. using condoms 1 528 1 037 921 % using condoms 71.4 76.1 78.5

81 Sonagachi, India

Figure 2. Trend in condom use among clients

82 80 78 76 74

% 72 70 68 66 64 62 1993 1995 1998

One important facet of this project has been the way in which new entrants to the sex trade are reached, with 48% of sex acts covered by condoms on the previous day by women who entered the brothel only 2 months ago or less. This compares to 63% of acts covered by those working for 6 months at the brothel. This success is attributed to peer pressure and improved self-control among sex workers. In the 14 clinic catch- ment areas now run by DMSC in West Bengal, survey data between 1996 and 1998 show a rise in condom use from 28.8% to 52.2%.

82 Sonagachi, India

The project has relied so far on the ability of sex workers to convince each other of the value of using condoms and obtaining STD treatment. The main tool of communication has been the original flip chart developed early in the project. Although some sex workers have learned how to present a slide show on STDs, the majority of peer educators have not been equipped with materials that could expand their messages and help them to present them to others, including other sex workers and clients. Investment in materials for behaviour change communication has been quite scant compared to many other projects.

Police violence, a major barrier to the sex workers’ ability to control their own lives, has diminished somewhat, but few objective indicators exist to measure this. It is easier to demonstrate that sex workers are more able to confront the police than before the intervention.

Efficiency Initially the project was funded by WHO, followed by NORAD for two years. In October, 1994, DfID began its financing. The number of staff working for the Sonagachi project is 49 full-time, 216 part-time and 25 volunteers. The project currently costs about US$ 90 000 per year to run. The salary of the project director has, until recently, been paid by the

83 Sonagachi, India

Government of India. There is a prospective study of question- central office and two field able ethics. Instead a cost-effec- offices in Rambagan and Chetla, tiveness evaluation was com- as well as 3 rented cars. Given pleted in 1999 that calculated the high level of activity, every- the cost per service. It found that one seems overworked and the cost per condom sold was Rs. space is at a premium and often 9.2 (US$ 0.22), per interpersonal shared by different project activ- contact for behaviour change ities. A few university students communications was Rs. 35.3 are conducting their own (US$ 0.84) and per STD patient research on the project, which treated was Rs. 560 (US$ 13.3). aids in documentation. Overall The latter cost in the community- direction and management is based STD programme in conducted through a large Mwanza, United Republic of group of partners called the Tanzania was US$ 10.8. Overall, Conglomerate. Although this the project invests 12% of its might add some burdens to deci- expenditures in advocacy, a much sion-making, it ensures greater higher level than most HIV inter- spread of responsibility and vention projects. Cost recovery strategic thinking. has not really begun, although a fee for service has been insti- The World Bank reviewed the tuted in the extension clinics out- cost of the project, though a full- side of Sonagachi. scale economic study has not taken place. A rough estimate, Ethical soundness expressing the reductions in STDs Ethical concerns permeate as DALYs (disability adjusted life throughout the project’s activi- years), showed the cost per DALY ties. Standard procedures for saved to be within the range of confidentiality and de-linking of Rs. 30 to Rs. 50. An impact-ori- HIV specimens were used during ented cost-effectiveness study the survey work. While sex work- would require knowing how ers are gradually gaining more many HIV infections were pride in themselves, the sur- averted in the intervention and rounding society still greatly stig- control groups, an expensive, matizes them and their children.

84 Sonagachi, India

Many do not wish to have their department store. As most con- names attached to printed state- sumer goods are overpriced in the ments or in any way admit to Sonagachi area, this would their families that they are sex enable the sex workers to make workers. This concern is basic commodities available to respected by the project. Special themselves at a lower cost. A reg- attention is paid to protecting istered organization, the Society the identity of STD- and HIV- for Human Development and infected women. In general, the Social Action, has been formed, project takes a very strong stance composed of the project partners, on ethics with regard to treat- including NGOs and sex worker ment of sex workers. organizations. The new project director is the son of a sex worker Replicability and who has been active in the proj- sustainability ect since its inception. Such an The project has already begun organization has the potential to replicating itself in various ways, provide the basis for the mainte- with clinics, outreach pro- nance and growth of future grammes and peer education efforts to provide sexual health training for other brothel areas and many other basic services, as and flying and street-based sex well as a platform for advocacy workers in and around Calcutta. where the sex workers can play Fourteen clinics are now run by the driving role. Skills, such as the DMSC. As of July 1997, 34 basic accountancy and literacy, peer educators were recruited are needed and gradually these and trained from among approxi- needs are being addressed. mately 1,500 flying sex workers in English language skills are also the vicinity. The intervention has generally lacking, requiring the reached out to cover about 8,000 use of interpreters when interact- of an estimated12,000 street- ing with most outsiders. The based sex workers in Calcutta as capacity of Sonagachi’s women well. The Usha Multipurpose to provide the basis of a regional Cooperative Society has initiated network of sex workers has been the marketing of condoms and recognized and such a network has plans for developing a was launched in 1998. Six sex

85 Sonagachi, India

workers are receiving computer accomplished to a greater training which should help them degree in Sonagachi than else- maintain networks via email, where in the Asia-Pacific region. though language barriers still This project has had a remark- remain. The peer educators able evolution, even if sex themselves have trained 600 workers are not yet ready to other sex workers, 200 each at 3 staff and completely manage workshops for brothel-based sex their own clinics. A great deal workers. of investment has been made in building capacity of staff Lessons Learnt and sex workers. The project, at times, has had detractors and The Sonagachi project has critics, but its basic soundness is demonstrated the great value of reflected, not merely in impact gradually placing the control indicators or the opinions of of a community-based health experts, but in the commitment intervention into the hands and active involvement of sex of the community. It has been workers themselves. This could possible to meet the costs of only be accomplished by treat- expansion, incorporating new ing the sex worker as a components, and staffing them whole person, encouraging due to enthusiastic voluntarism her to recognize and express on the part of the sex workers. her needs and treating her in a Whether an HIV or malaria fully ethical manner. Meeting intervention, this approach is the felt needs of the sex one of the most sustainable, if workers has encouraged them not biomedically perfect, in the to commit themselves to HIV reality of an imperfect world. prevention. Impact and effectiveness in such behavioural change interven- It was unfortunate that the labora- tions are difficult to achieve and tory STD work and sampling has, at to measure, but are far more times, been unreliable. Such proj- likely to be sustained, once ects have considerable scope to achieved, if the target group recruit more expertise. A scien- itself takes over. This has been tific research and training centre

86 Sonagachi, India

is envisaged for the future of The fuller involvement in HIV Sonagachi, which is one way of prevention is needed of assuring greater involvement of important men and women scientists. Nonetheless, biomed- who strongly influence the ical and self-reported behav- safety of the sex trade, e.g. ioural data from repeated sur- police, madams, pimps, babus. veys ensured adequate repli- Repeatedly, this project has cation to demonstrate effec- spawned various organiza- tiveness. In the future, monitor- tions that provide a demo- ing and evaluation in Sonagachi cratic basis for sharing deci- should be able to correct past sion-making power. Reforming mistakes and incorporate the sex trade without involving measures and documentation police, madams, pimps and hood- of personal growth and social lums will be very difficult, and empowerment among the they need to be included. women. Recently, as the number of pimps and madams declines, landlords

The process of struggle that we, the members of Durbar Mahila Samanwaya Committee, are currently engaged in has only just begun. We think our movement has two principal aspects. The first one is to debate, define and re-define the whole host of issues about gender, poverty, sexuality that are being thrown up in the process of the struggle itself… Secondly, the daily oppression that is practised on us with the support of the dominant ideologies, has to be urgently and consistently confronted and resisted. We have to struggle to improve the conditions of our work and material quality of our lives and that can happen through our efforts towards us, sex workers gaining control over the sex industry itself.

(From the Sex Workers’ Manifesto, Theme Paper of the First National Conference of Sex Workers, organized by Durbar Mahila Samanwaya Committee, Calcutta, November, 1997)

87 Sonagachi, India

have begun to harass sex workers Further Reading more directly. The future direction and management of the Chakraborty, AK, and Jana, S. project through the HARD Trust Socio-behavioural aspects of has the potential to include all clients of commercial sex workers players and stakeholders. of Sonagachi, Calcutta. Sonagachi Research Series Report India and the rest of the world No. 2. Dept. of Epidemiology, All will have to reckon with the India Institute of Hygiene and awakened women of Sonagachi Public Health, Calcutta, 1992. as they move towards greater outreach to other sex workers in Chakraborty, AK, Jana, S, Das, A, their region and around the Khodakevich, L, Chakraborty, MS world. The main remaining and Pal, NK. Community based obstacle is the attitude of society survey of STD/HIV infection to sex work and sex workers. among commercial sex workers in Calcutta (India). Part I. Some HIV infection, while a serious and social features of commercial sex sad disease problem for the workers. J. Communicable world, is also an opportunity to Diseases 26 (3):161-167, 1994. set right many of the abuses of the past. The Sonagachi project Chakraborty, AK, Jana, S, Das, A, has begun to wed prevention Khodakevich, L, Chakraborty, MS and cure in the best traditions and Pal, NK. Community based of public health activism. survey of STD/HIV infection among commercial sex workers in Calcutta (India). Part II. Sexual behaviour, knowledge and attitude towards STD. J. Communicable Diseases 26 (3):168-171, 1994.

Dept. of Epidemiology, All India Institute of Hygiene and Public Health. Report of the

88 Sonagachi, India

Community Based Survey of Jana, S., Bandyopadhyay, N., Sexually Transmitted Diseases, Mukherjee, S., Dutta, N., Basu, I., HIV Infection and Sexual and Saha, A. STD/HIV interven- Behaviour among Sex Workers in tion with sex workers in West Calcutta, India (April-June, 1992). Bengal, India. AIDS 12 suppl. B, Dept of Epidemiology, All India S101-S108, 1998. Institute of Hygiene and Public Health,1992. Jana, S, Chakraborty, AK, Chatterjee, BD, Chakraborty, MS, Durbar Mahila Samanwaya van Dam, CJ, and Mehret, M. Committee. First National Knowledge, attitude of CSWs Conference of Sex Workers, Nov towards STD/HIV and prevalence 14-16, 1997, Calcutta. of STD/HIV among CSWs. Presented at the IXth Fox, R. Red light on Calcutta: International Conference on health via schools in Andhra AIDS, Berlin, 1993. Pradesh [Commentary]. Lancet 344:1038 (Oct 25), 1994. Jana, S, Banerjee, B. (eds), A Dream, A Pledge, A Fulfilment: Gonzalez, V., Grosskurth, H., Five Years’ Stint of STD/HIV Guiness, L., Pangare, V., and Intervention Programme at Sethi, G. Evaluation of Four Sonagachi. All India Institute of Partner Projects of the West Hygiene and Public Health, Bengal Sexual Health Project. Calcutta, 1997. Final Report. International Family Health, September 1999. Jana, S, Choudhury, R, Bhaskar, B, Mukherjee, M, and Basu, I. Jana, S, Bailey M. All part of the Putting barricade against AIDS service. AIDS Action 26 (Sept- through mobilisation of female Nov), 1994. sex workers in Calcutta. Proceedings of the 4th Jana, S. Without shame or International Congress on AIDS in stigma. Health Action 10:11 Asia and the Pacific, Manila, Oct., (Sept-Nov), 1994. 1997. Abst # C(P) 118, p. 349.

89 Sonagachi, India

Mukherjee, S, Jana, S, Mitra, M and Choudhury, R. Situational analysis of sex trade with the help of FSWs (female sex workers) and their networking. Proceedings of the 4th International Congress on AIDS in Asia and the Pacific, Manila, Oct., 1997. Abst # C(P) 117, p. 349.

Nayanita, S, Singh, S, Moulik, R, Brahma, S, Jana, S, and Basu, I. Enabling female sex workers to create a milieu of negotiation. Proceedings of the XI International Conference on AIDS, Vancouver, Pub C 1166, p. 464.

Singh, Sujata. Three Year Stint at Sonagachi: An Exposition. The STD/HIV Intervention Programme at Sonagachi, 1992-1995. Aug 1995.

Jana, S. and Banerjee, B, eds. Learning to Change. Seven Years’ Stint of STD/HIV Intervention Programme at Sonagachi. Calcutta: Society for Development and Social Action. August, 1999.

90 SHAKTI, Bangladesh

SHAKTI: A Brothel and Street- Based Sex Worker Project in Bangladesh

Introduction HAKTI (“power” in Bangla), or S“Stopping HIV/AIDS through Knowledge and Training Initiatives”, is the name of a project implemented by CARE, Bangladesh, with several components. The larger SHAKTI project has been working with three groups: brothel-based sex workers, street-based sex workers and injecting drug users (IDUs). The first component began in mid-1995 in a brothel located in a town named Tangail, a few hours drive away from the capital city, Dhaka. The second component started in 1996 among street-based sex workers in Dhaka. The project aims at improving condom use and making treatment of sexually transmitted diseases (STDs) available among about 600 brothel-based, and about 3,000 to 5,000 street-based, female and transgender (hijra) sex workers in a highly conservative and repressive setting. In addition to increased condom use, measurement of changes in STD prevalence was included as an indicator and the project baselines were estab-

91 SHAKTI, Bangladesh

lished by clinical and laboratory other brothels remain open, but research. This project, inspired by over the past several decades, the Sonagachi Project in prostitution in Bangladesh has Calcutta, was conceived as one been undergoing change as that could be expanded to other more women work in hotels, res- parts of the country through a idences, other types of informal strategy of NGO and govern- brothels, through beauty par- ment partnership. This remains lours and other less obvious ven- its aim. ues. Street-based sex work remains the most visible, poorest Bangladesh has low HIV preva- paid, with the lowest status and lence. In 1998 the first well- has the roughest clientele. defined HIV sentinel surveillance Government authorities view took place among high-risk prostitution as a shameful blem- groups, revealing the highest ish on Bangladesh’s social record rate of 2.5% among injecting and seek to abolish it, primarily drug users in Dhaka. A sample of by abolishing brothels and relo- 400 Dhaka street-based female cating slum dwellers into rural sex workers revealed no HIV, areas. It is within this context while similar samples at two that SHAKTI’s success can best be brothels, Tangail and another, seen. showed none at Tangail and 1.5% at the other. Levels of As it was the first major project syphilis infection (TPHA and RPR with sex workers in the nation, positivity) were high, with 57% CARE personnel had many les- among the street women and sons to learn. Objectives of the 46% at Tangail (Azim et al., project have shifted several 1999). Following the completion times, with numerous manage- of surveillance, the brothel at ment changes, additional fund- Tan Bazar, the largest in the ing being required, and different country, experienced intense and partnering strategies being violent repression and was developed in order to meet closed. Gradually over the past changing perceptions of need on years, brothel sex work has been the part of sex workers. The sex diminishing. Tangail and 14 workers themselves have

92 SHAKTI, Bangladesh

changed, not only in the sense of going monitoring by its peer group composition, but in terms educators. of their own understanding of the issues surrounding HIV. While Situational the strictly health-related issues Background have not been easy to solve, the human rights issues have posed a Tangail brothel far more serious barrier to Bangladesh has had brothels for improved sexual health and safer centuries that have gradually sex practice. Against all odds, a become part of the local culture. strong street-based self-run sex Some are residential, relatively worker organization has uncrowded, with small rooms emerged, made up of over 900 rented to individual women. dues-paying women who are Others are crowded, with many determined to gain recognition small rooms housing three or of their rights as citizens of more women together. Although Bangladesh. Hijras have joined as not technically legal, brothel well, recognized by the women prostitution is regulated by the as colleagues. A similar move- local authorities in Bangladesh. ment has taken place, but more Sex work is considered a black- haltingly, in the Tangail brothel. listed trade, and sex workers are Hence, empowerment has highly denigrated by most mem- evolved to hold a primary place bers of society. Police exercise a among the project’s objectives. degree of regulatory power, and for a fee ranging from 5,000 to Throughout its life, the project 15 000 taka (US$ 1 = 50 taka), has been funded by the United place an affidavit with the court Kingdom Department for when a woman enters a brothel. International Development This enables her to avoid arrest (DfID, formerly ODA). SHAKTI when working in the brothel. has undergone an external mid- This system is supposed to ensure term review and an additional that only willing women who are near-final evaluation. Also, quar- at least 18 years old enter the terly monitoring takes place brothels. The system is ambigu- through surveys as well as on- ous and recognizes prostitution,

93 SHAKTI, Bangladesh

but does not make it legal. At present, their women do not Tangail, the police, landowners, take clients. Some women have house owners and/or bariwalli permanent relationships and (residential landladies), madams, give up taking clients altogether, sardanis (senior madams), and but continue to live at the mastans or hoodlums associated brothel. Some women are the with local political parties, really children of sex workers and were run the establishment, which has raised in brothels, and in turn, a regulating body called the many raise their own children at Shamaj. Although most women the brothel. Many send money to are “independent” operators, their mothers and other rela- about 60% of their income goes tives. Tangail sex workers were to the aforementioned authori- not allowed to leave the brothel ties. A minority of newcomers, at will, and when out, could not called chhukris (19%), were wear any shoes or slippers; if totally under the control of sen- they did, they would be fined by ior sex workers or sardanis, and the police. This imposition on made to work for at least 6 their rights was supposed to months to pay them off, i.e. ensure they would not sell sex 100% of their income was paid outside of the brothel. Their chil- to their controllers. dren were unable to register for school because they could not At the time of the baseline sur- name their fathers. vey (Sarkar et al., 1997a), the average age of the women was 24 and 86% were illiterate. Almost all (97%) were Muslim and had been working in the sex trade for an average of 7 years. Most (62%) had long-term clients or lovers, whom they call bhalobashai lokh (literally, good home men), with whom they formed a marriage-like relation- ship. At the time the lokh are At one entrance to Tangail brothel, 1998.

94 SHAKTI, Bangladesh

Although violence has not been a Street sex work in Dhaka major problem at Tangail, sex Street sex workers are the most workers are heavily stigmatized visible of the various types of sex by the surrounding community. workers. In Dhaka, they are ubiq- Police exercise a great deal of uitous, found in most areas of control. Sex workers are made to the city, parks, railway stations, pay extra money to the police if bus and truck stops, cinemas, they take a client for the whole markets and shrines. Many other night. Consequently, few women sex workers operate in Dhaka as have all-night clients. Most of the well, mostly out of hotels and time at the brothel there is a rea- residences that pose as normal sonably peaceful and sociable family homes. It is well known atmosphere, with open court- that many university students, yards where women, children, married women from middle- lovers and friends gather. This class families and others supple- isolated and controlled environment their incomes through the ment has afforded certain advan- sex trade. These women do not tages for the SHAKTI project. stand in public places, however, However, it is not typical of all to acquire their clients, but Bangladeshi brothels, is not sta- instead have agents to bring ble, and it certainly is not an envi- clients to them or make appoint- ronment in which sex workers ments for them to go to specific have adequate control of their sites. Public soliciting to sell sex is own lives and bodies. Pressure to strictly against the law. Street- earn for themselves, their fami- based sex workers comprise the lies and their controllers, while lowest class of sex workers and, facing eviction and forced reha- as such, are far more subject to bilitation at any time, makes violence perpetrated by clients, safer sex a difficult goal to mastans and police. The police achieve. Their lives and liveli- are generally paid off to allow hoods are literally at the mercy of them their right to work, with political forces well beyond their sex and/or money. Territoriality is control. also evident, with some mastans or pimps fighting for specific spaces in which the women asso-

95 SHAKTI, Bangladesh

ciated with them can work. Most policemen who look after them street sex workers are independ- in exchange for sex. ent, in the sense that they decide to work or not, as they need. A Street sex work must be seen as minority is bonded to a dalal or one of the more lucrative street- pimp, but many make use of based occupations in Dhaka, a local men on the streets, e.g. crowded city of 10 million peo- rickshaw pullers, hotel boys, ple. Male and transgender (hijra) guards, tea-shop owners, and sex workers also operate in many others, to find them clients and of the same contact venues as do then pay them with cash or sex the female sex workers. for doing so. In addition, some Generally, it appears the females women have men who serve as have the greatest client num- their protectors; they may find bers, hijras next, and males fol- them clients and, when needed, lowing. In addition to thousands help them avoid the police or of paper pickers, beggars, sellers stand as their husbands to seek of flowers, tea, candies, fruits, or their release from police custody. magazines, the streets are the Some sex workers even have home to many thousands of “station husbands”, particular homeless individuals, both single

Family life on the streets of Dhaka.

96 SHAKTI, Bangladesh

adults and children as well as focused on the danger of going whole families. Some have plas- to prostitutes. The number of tic-covered “humpies” to sleep sex workers in brothels has var- and cook in, others do not. Slums iously been estimated as 10 000 are packed with people and reg- to 100 000. The National ulated by local politicians and Behavioural Surveillance of their mastan gangs. Periodically, 1998, however, surveyed all the police are instructed to clean brothels and found a total of up the streets and hundreds of only 6584 (Jenkins, 1999a). women, sex workers and others, Floating sex workers, male, are arrested and placed in female and transgender, are vagrancy homes, where their now thought to total up to treatment is often abusive. Less 100 000 in the country as a often, but more dramatically, whole. One survey has provided instructions are given to the a national estimate of 12 000 army to clear out slums, either by hijras with 2000 in Dhaka, most fire or bulldozer. When this hap- of whom also sell sex (Jenkins, pens, or when a severe flood 1999b). Several early efforts at drives thousands of people to research conducted with Dhaka from hopeless rural areas, female sex workers demon- street sex work increases. About strated poor knowledge of half of the people in Dhaka live STDs and inadequate utilization on little over US$ 1 per day. The of health services. These studies situation is very dynamic and documented their social isola- unstable, creating an extremely tion and extreme marginaliza- challenging environment in tion (Ahsan, et al., 1995; Khan which to carry out a sex worker and Arefeen, 1995). CARE is an HIV intervention. organization dedicated to working with the poorest and Relevance most marginalized members of Many Bangladeshis consider society. Therefore, the project their society to be sexually con- was seen as relevant to both servative. Perhaps for that rea- the needs of poor, highly vul- son, among others, most AIDS nerable women and the princi- prevention messages have ples of CARE.

97 SHAKTI, Bangladesh

Formative Stages vey were given a standardized clinical examination (n=296) and Brothel baseline studies tested for HIV, syphilis, gonor- Tangail sex workers were dis- rhoea and chlamydia. PCR tech- trustful when first approached niques were available for gonor- by CARE, as other NGOs had rhoea and chlamydia tests come, promised various services, through collaborators in but were never seen again. CARE England, while the HIV and personnel, on the other hand, syphilis tests were conducted at a had never handled a sex worker government facility and later project and were uncertain and repeated at the national surveil- nervous in dealing with highly lance laboratories. For blood socially stigmatized women. sampling, the survey recruited After an initial phase of site 466 sex workers out of a possible selection during the first half of 593. The samples for HIV tests 1996, the project undertook, were unlinked to identifying with the aid of external consult- information. Verbal consent was ants, various types of formative required and cooperative sex research, including a qualitative workers helped secure the coop- assessment, a quantitative base- eration of others by showing line study on behaviour and them how the examination knowledge, and an STD survey. would be performed, using a The entire set-up phase covered speculum and a medical school one year, during which advocacy anatomical model. Women with with those in the power struc- symptoms were treated. tures and rapport building with the sex workers were able to pro- Important issues were revealed ceed. A two-room clinic, with as a result of the formative nurse and physician, was set up research, some of which the proj- at the brothel on land donated ect has been able to handle by the Shamaj. Here STDs were while others remain unsolved. treated, as were other sicknesses. First, the levels of clinical signs of Confidentiality was maintained reproductive tract infections on all records. Those who volun- were high, with 63% of the teered for the baseline STD sur- examined women having cervical

98 SHAKTI, Bangladesh

discharge and 32% having vagi- One aim of the project has been nal discharge. A further 2% had to reduce the time gap between ulcers, 6% warts, and 26% had when symptoms are recognized lower abdominal pain. and when care is sought. Laboratory tests confirmed Generally, the women did not somewhat lower rates, with 19% wish to acknowledge current positive by PCR for chlamydia, STD symptoms for fear of losing 20% for gonorrhoea, and 18% clients. At baseline, women with for a current syphilis infection symptoms waited an average of (RPRD 1:8 and TPHA+). No 7 days before seeking treatment woman had a positive HIV test beyond their own home reme- (Sarkar, et al., 1998; Jenkins, dies. When they did seek help, it 1999c). Tests were not conducted was from paramedics, traditional for other possible causes of dis- healers, and medicine sellers, charge or abdominal pain. The some of whom do not have safe disparity between syndromic injecting practices. clinical diagnosis/treatment and the laboratory confirmatory tests The project aimed at improving revealed that the standard syn- levels of condom use. At base- dromic management strategy line, only 3% of women were would only correctly identify using condoms for all instances about half (45%) of women with of vaginal intercourse in the last chlamydia or gonorrhoea as the 24 hours. Both oral sex and anal others were asymptomatic. In intercourse were reported as addition, one-third would be rare, but were probably underre- overtreated because they would ported due to their shameful receive treatment for pathogens connotation. In the baseline sur- they did not have. Nonetheless, vey, the women had an average without better treatment or of 3.5 clients per day; in addition, screening strategies, the brothel they had intercourse, on aver- clinic has had no option but to age, once a day with their lovers. follow standard syndromic man- Numerous barriers to increased agement guidelines. This issue is condom use existed. First, the a problem for sex worker proj- women saw condoms as a family ects in most countries. planning device, and 45% had

99 SHAKTI, Bangladesh

already adopted a mode of con- a two-week period (Sarkar et al., traception. Others either wanted 1997c). Many lived in slums a pregnancy or relied on men- (37%) while others were home- strual regulation (legal in less (30%). Still others came into Bangladesh) or illegal induced Dhaka to work from nearby abortion. At Tangail, 29.7% of areas, often during the day only. the women had never been More than half (60%) were pregnant and, to those who divorced or separated, while wanted children, infertility was a 21% were currently married. problem. Second, 62% of the Most of these women (80%) had women had lovers. While most worked in other low-paying did not live under the same roof occupations, such as in garment with their private partners all the factories, as domestic servants, in time, the position of the lover hotels or restaurants; 20% had was like a husband and they saw been in brothels. The average no need to use condoms in their time they had spent in sex work private relationships. Third, their was 3.5 years (Abdul-Quader, clients did not like condoms. The 1997). sex workers feared that if they asked these men to use condoms, The STD study among these the clients would think that the women revealed much higher sex workers were infected with a levels of prevalence than among disease (Sarkar et al., 1997a). those in the brothel. In a sample of 247 tested in the same way as Street-based sex worker the brothel sample, 48% had baseline studies chlamydia, 53% gonorrhoea, and In late 1996 work began among 31% had current syphilis infec- street-based sex workers in tions. None had HIV. Almost all Dhaka and by mid-1997, a base- (90%) reported STD symptoms line study was conducted, with a and on examination, 69% had capture-recapture component to vaginal discharge, 62% cervical estimate the population size. The discharge and 11% had genital capture-recapture method yielded ulcers. In the previous three an estimate of 4366 street-based months only 54% of those with female sex workers found within symptoms had sought any treat-

100 SHAKTI, Bangladesh

ment at all, mostly from medi- ❏ to create an enabling cine sellers. Most believed that environment STDs were avoided by staying clean. This approach required mobiliz- ing support from the power Few of the women were using structures surrounding the condoms, mostly for prevention brothel, i.e. the local administra- of pregnancy. They reported an tion, the police, local opinion average of 3.5 clients in the past leaders, clients that could be 24 hours and only 14% used con- reached. The project design did doms for more than half of all not include training in negotia- intercourse in the previous 24 tion of condom use, creative hours. Only 10% had used con- modes of safer sex or intensive doms for all intercourse during work with client groups. It never the past 24 hours, but overall anticipated the need for a strong coverage was 18% of all vaginal advocacy component, and skills and anal sexual intercourse with in this regard were not devel- any partner the previous day. Ten oped. As the project proceeded per cent of the women were and events unfolded, new strate- pregnant while working. gies were formed. Implementation Peer education Peer education has been the Working strategies main approach utilized by the The SHAKTI project has had four project. This began in August major working strategies. These 1996 in the brothel and in 1997 are: among the street women. After ❏ to raise awareness the mid-term review in 1998, the manner of handling peer educa- ❏ to have repeated contact of tion was altered in the direction high quality with sex workers of greater sex worker management of the project. During the ❏ to provide the means of first few years, a total of 50 peer behaviour change (condoms educators were trained in the and STD treatment) brothel, with 28 actively working

101 SHAKTI, Bangladesh

(a ratio of 1 to 20 sex workers). at first, then more, and by the They were each assigned zones in end of 1999 there were 8 drop-in the brothel. A zone was made up centres at which women could of 10-14 apartments. Peer educa- rest, bathe and access STD treat- tors received 50 taka per day and ment, as well as simple general worked a half-day, once per health care. Peer educators were week. Their tasks were to visit the paid when they were at the women in their assigned zones, brothel and assigned spots discuss safe sex, HIV knowledge around the city at which about and STD treatment, and ask 35-40 sex workers could be about condom use in the previ- reached per peer educator. ous 24 hours. They also collected the condom package covers. The trainers of peer educators, Rubbish bins for condom disposal called field trainers, wore white were first placed in strategic loca- laboratory coats and a badge tions, but later in each room. around the brothel, while the Used condoms in the bins were peer educators themselves wore counted, but somewhat irregu- a badge and blue apron when larly. These data were recorded working. This applied to the with colour codes on a monitor- street-based peer educators as ing map, according to zone and well and was copied from the room, by the peer educators. Sonagachi project in Calcutta. Then, the results were summa- The field trainers felt they rized on a monitoring board in needed these uniforms to distin- the brothel. The process of peer guish themselves from sex work- education reached about 70% of ers, so that clients would not the women. The remaining 30% approach them for sex. While the were resistant. As these sex work- peer educators have enjoyed the ers were not monitored, they status these uniforms give them, were not included in the number the uniform marked them as who were using condoms at the somehow different and distinct first follow-up survey. from the other sex workers. The peer educators in both arms of On the streets of Dhaka, drop-in the intervention were privileged centres were established, a few with more attention, small

102 SHAKTI, Bangladesh

amounts of cash and other larly. At the brothel, 21 outreach rewards, such as lessons in liter- workers, 102 peer educators as acy. Several made repeated trips well as 29 peer lover educators overseas and were often asked were trained and continue to to represent the project at vari- work under the new scheme. The ous AIDS meetings. Jealousy number of CARE staff at the among the non-peer educator brothel was drastically cut from sex workers began to grow. The 14 to 5 with the aim of decreas- first step taken to address this ing reliance on CARE staff and was to drop the use of the blue increasing sex worker manage- uniform at the brothel. On the ment of the project. streets, the women were able to use the status symbolized by the uniform to protect themselves from police arrest and did not wish to give it up. Gradually these markers have been dropped, and well-trained peer educators were elevated to outreach workers and given small daily salaries, while a new cadre Hijra peer educators at World AIDS Day of volunteer, unpaid peer educa- ceremonies, 1998. tors was developed. Health care By mid-1999, 55 outreach work- The brothel clinic was staffed by ers and 171 volunteer peer edu- a female doctor and two nurses, cators were working on the one male and one female. streets of Dhaka. These women Although the doctor was dedi- reached about 3200 sex workers cated to helping the sex workers, on a regular basis for both group she had no specific training in and individual discussions. In STD management. In early 1999 addition, 6 hijra peer educators she undertook short training and 3 outreach workers were with a clinic group in Dhaka and trained and through them at later in the year went abroad least 100 hijras are reached regu- for a clinical training course.

103 SHAKTI, Bangladesh

The clinic is open 5 days a week HIV and STD recognition and in the morning from 9:00 to treatment, but require consider- 13:00. The women were free to ably more attention. They are bring their family members for known to give unnecessary intra- any illness. STD drugs were given venous injections of saline or free, but drugs for other illnesses other fluids. were bought elsewhere, with a prescription provided by the In Dhaka, CARE’s partner, the clinic. Immunizations were pro- Marie Stopes Clinic Society, sup- vided for children by the govern- plied services for the drop-in cen- ment health service and special tres from the earliest days of the immunization campaign days intervention. Problems arose were held. Clients did not have repeatedly regarding the way in access to the clinic, but the proj- which the clinic doctors treated ect opened the clinic in the the sex workers. Some covered evenings to the sex workers’ their mouths when they lovers. Eventually, technical sup- addressed the women; others port for clinic supervision was were curt and obviously unwill- accessed through a partner clini- ing to touch them. The sex work- cal NGO. Sustainability of the ers were very sensitive to these brothel clinic is a major concern slights and often complained. In for the sex workers. Through a addition, the clinic hoped to workshop, the district health regain some of the cost of medi- administration assured them cine, but it soon became clear that, if required, they would that many women were too poor assign a doctor to the brothel on to be able to afford the fees. a routine basis. Plans have been Repeated meetings with the made to equip the sex workers Marie Stopes directors gradually with skills necessary to manage led to a deeper understanding of their own clinic, including an the needs of these women. The insurance scheme to cover costs. fee schedule was altered so that women now pay only 10 taka to Medicine sellers, whom the gain a clinic membership card women often consult, were and all other treatment costs are given four training sessions on free. As the street-based sex

104 SHAKTI, Bangladesh

worker organization gained 1998 Tangail sex workers claimed strength, it has been able to that the number of clients had address the doctors more directly dropped. It was suspected that and this has helped improve doc- the high visibility SHAKTI tor-patient interactions. Now brought the brothel, and the many of the clinic activities are run large number of visitors (nearly by the sex workers themselves. 200 in 1997) who wished to see the intervention, may have con- Empowerment : in the brothel tributed to a decreased number The baseline survey identified a of clients. Project management number of concerns of the stopped all unnecessary visiting, brothel sex workers and the proj- and from mid-1998 to mid-1999 ect has attempted to address the brothel was visited only 17 these. They were: times by outsiders.

❏ economic stability and Skills training in alternative savings modes of supplementing income began in February 1998. Several ❏ education, literacy associated local NGOs began to ❏ medical services provide training in embroidery and sewing to 20 women, with a ❏ freedom of movement view to the sale of goods pro- outside the brothel duced by the women. Such activities as sewing, however, cannot ❏ welfare of children be expected to replace or even ❏ social acceptance, including supplement a sex worker’s income burial in Muslim cemeteries and most of the women are little interested in this approach. Economic stability has not been addressed directly. Outreach Many women had hired private workers are paid a small amount tutors prior to the coming of and have voluntarily opted not SHAKTI. Literacy and education to take clients for the few hours are highly appreciated. Early in per day that they work as educa- the project SHAKTI initiated a tors. For a variety of reasons, in programme of literacy with the

105 SHAKTI, Bangladesh

peer educators. Forty women peer educators (who are not attended but 12 dropped out. chhukris) gained more freedom Literate peer educators began to wear shoes out of the brothel. teaching literacy to 60 other sex In the early days of the project, workers, but the women lost regular meetings were held with interest due to the inappropri- the gatekeepers of the brothel ate nature of the curriculum. A and attempts were made to new curriculum was developed encourage them to alter some of through consultation with 8 lit- their practices. In a few instances, erate sex workers and several for example on World AIDS Day, project staff, and now includes many women appeared in public issues such as women’s rights. A wearing shoes. sub-contract was signed with a local NGO to run the education The project began with a conflict programme as of October 1999. resolution model (as opposed to Most importantly, literacy educa- one that focuses on the rights of tion was made available to all sex workers) in order to convince the sex workers and not just the those in the surrounding com- peer educators. munity (e.g. religious and political leaders) that it was not The sardanis, bariwallis, and encouraging illicit sex, only police exercise considerable con- attempting to increase its safety. trol over the sex workers, espe- Numerous formal as well as cially over the chhukris. informal meetings were held Independent sex workers are with these groups. This process supposed to ask permission to was also monitored with indica- leave the brothel and not to tors and a quarterly assessment. wear shoes or sandals. Chhukris For the first few years, the proj- are never allowed to leave until ect carefully avoided taking a their period of indenture is com- strong stand in favour of the sex pleted. Chhukris are also not workers. In 1996, sex workers allowed to have lovers as such an were blocked by the local attachment would diminish the administration from participat- number of clients they would be ing in World AIDS Day activities. willing to take. Gradually, the Protests were not lodged, but

106 SHAKTI, Bangladesh

the next year, without con- done so. Now they state they frontation, they simply appeared never realized that the proscrip- in the activities and nothing was tion against shoes was not a said. Because considerable fear real law, but was only in their still exists that the police, local minds. Staff members had politicians or fundamentalist known there was no legal groups will close the brothel if aspect to the taboo, but had the women offend local society, never considered it wise to tell progress has been slow in chang- the women, lest they offend the ing the social climate around the local society. brothel. In India and Bangladesh, chil- After the mid-term review dren are a major issue in broth- revealed continuing disagree- els. Women want their own or ments about condom use adopted children, partly to among the women, meetings ensure an income when aged as were held at which the women well as to satisfy their own emo- were asked what they could tional needs for nurturing and agree upon. They all stated they attachment. Some sex workers wanted to wear shoes out of have been able to educate their the brothel, but the sardanis children by sending them away did not like this. Although it from the brothel, but most are altered the non-confrontational unable to do this. SHAKTI has stance SHAKTI had taken ear- initiated coordination with lier, the women were advised to another NGO to provide a non- go out in groups of five with formal elementary school for shoes but without make-up, to brothel children. Currently report what happened to the about 45 children are enrolled. sardanis, and to wait a few days The partner NGO has purchased to see if the police came to land on which it hopes to build a threaten the sardanis or them. training centre for the brothel Within a few months, over 200 children. women had gone out of the brothel with shoes without any Numerous events have taken mishap and eventually all have place in which selected peer

107 SHAKTI, Bangladesh

educators have made public could feel equal to any other presentations about the proj- woman. ect, both in Bangladesh and in other nations. Several sex workers attended the Fourth International Asia-Pacific AIDS Conference in Manila in 1997 and the Twelfth World AIDS Conference in Geneva in 1998. Both sex workers and brothel keepers (owners and sardanis) attended the First National Sex 1998 World AIDS Day play performed by Workers Conference of India in Tangail brothel sex workers, with their November, 1997 and the First children. Meeting of the Asia-Pacific Network of Sex Workers in A small committee of sex workers, March 1998. On other occa- known as Mukti Shangho sions, sex workers from Tangail (Freedom Committee), was have attended meetings with formed. The spirit of full participa- sex worker organizations in tion, however, was compromised Calcutta, and Calcutta women by the appointment of its leader have visited Tangail. Picnics, by CARE personnel. Eventually, it parties and other gatherings became apparent that the com- held by CARE for sex workers mittee was stagnant and unsup- and CARE staff are other ported by the sex workers. It was instances in which sex workers modified, renamed Nari Mukti are given a small opportunity Shangho (Women’s Freedom to feel less stigmatized. These Committee), the sex workers are, however, inadequate to elected the woman they wanted meet the needs of all the as leader and the organization women in the brothel and began the process of becoming much larger structural changes formally registered as a social wel- to community attitudes and fare organization. The committee control will be required before collects 50 taka as a monthly sub- a sex worker in Bangladesh

108 SHAKTI, Bangladesh

scription from its growing mem- repeated legal problems, moving bership, receives 10 taka from in and out of the trade, a large each woman who registers at the influx of new women, significant clinic (donated by the partner levels of violence, as well as con- NGO which supervises the clinic) cerns for their children, health and makes a small profit on con- care, income stability and safety. dom sales. Meetings are held reg- Their social status is at least as ularly at which sex workers can low as the brothel women, if not participate in making decisions lower, and their self-esteem about matters important to their reflects their stigmatization. lives. Leadership, however, contin- However, those who survive ues to be problematic as the beyond a few years on the woman elected was later arrested streets are tough and smart. and sent to jail. As the brothels of They are not threatened by sarda- Bangladesh are nodes of political nis and all-encompassing power contention and local politics can structures, as many are in the easily turn violent, growth of the brothels. These women took the self-run sex worker organization lead and in early 1998 formed the continues to be erratic. As of first street sex worker organiza- November 1999, membership tion, Durjoy Nari Shangho stood at 230. (“Undefeatable Women’s Club”). They began to mobilize street Empowerment: on the workers, particularly in a local streets of Dhaka fight to re-open a previously Many of the problems of street- closed brothel in Dhaka City. based sex workers were quite From the beginning they saw different from those of the their purpose as one of creating brothel women. Many had no a collectivity strong enough to place to bathe, wash their demand their rights. These clothes or even to sleep; others include legal rights, burial rights, expressed that their greatest rights to shelter, education for problem was police harassment. their children, rights to safety The barriers to becoming skilled from violence, particularly at the in practising safer sex included hands of the state, and the right high levels of mobility, serious to make a living. Gradually

109 SHAKTI, Bangladesh

Durjoy formed 6 area commit- also shifted the stance of the tees with representation to a SHAKTI project. A young sex central one. They began register- worker named Jesmin at Tan ing their organization as a social Bazar brothel was beheaded by a welfare group. They opened a fake client, an act facilitated by bank account into which they political forces that were work- deposit the 50 taka collected ing to close the brothel. The monthly as membership dues. police and Social Welfare They took on the distribution of Department were mobilized to food, water and medicine during help close the brothel in the Dhaka’s severe flood of 1998, to name of a rehabilitation pro- sex workers and their families, as gramme. Violence and gross dis- these women found it very diffi- regard of human rights ensued. cult accessing relief goods at the This brothel and another nearby regular sites. With the selling of named Nimtoli, also similarly condoms, the organization also affected by these events, repre- accrues some earnings. By sented nearly 30% of all brothel- November 1999, membership based sex workers in Bangladesh. had risen to 950 women. Sex Brothel sex workers throughout workers manage many of the the nation were frightened. drop-in centres that can now Womens’ rights groups and legal provide space for some women rights NGOs gathered their to sleep, rest and bathe, as well forces to defend the brothels. as to undertake specific types of Both Durjoy and Nari Mukti training. Both street and brothel Shangho joined, leading the sex sex workers have received some worker contingent, and staged basic training in their legal numerous public demonstrations rights, and several dozen street and made press releases. Within workers have received training in a short time, about 80 different the martial arts as well. human rights and women’s organizations joined to form a A turning point group called Solidarity to stand A crucial event occurred in July with sex workers against the 1999 that galvanized street and government. They met with the brothel sex workers alike, and Prime Minister, but she would

110 SHAKTI, Bangladesh

not alter the government’s plans. lems in Bangladesh and rallied She stated that she believed the support. only option for sex workers was rehabilitation, even though the In March 2000, all who worked women stated clearly that they on advocacy and networking did not want to be rehabilitated. among the sex workers and The case went to court. other concerned groups experienced an astounding victory. CARE’s stance was simple and The High Court of Bangladesh direct. It supported the sex work- declared that sex work was not ers in their fight for basic human illegal and that the women had rights, and began a stronger a right to make a living. It fur- approach to working with gov- ther declared that state agen- ernment, UN agencies and other cies responsible for evicting NGOs to bring this about. In addi- women from the brothels had tion to numerous local groups, the themselves acted illegally. The International Network of Sex women were instructed that Work Projects, Amnesty and they could petition the lower other international agencies courts to re-open their brothels. were alerted to the serious prob- While it is not known how these events will impact on the future of the project, the court decision clearly gives a signal that human rights, and particularly the constitutional right to work, will be protected for sex workers. This decision strengthens the HIV prevention options for all concerned.

Condoms For weeks in mid-1999 newspapers carried stories about Solidarity, sex Male condoms were purchased worker groups and NGOs collectively from a single source by the proj- seeking protection of sex workers’ human rights. (Photo: AKM Mohsin, reprinted ect and given out free by peer with permission from the Daily Star) educators. At the brothel, not all

111 SHAKTI, Bangladesh

the women accepted condoms, sure to ask questions and the 30% reportedly because they did not of sex workers who had been like being asked how many clients resistant began taking condoms. they had the previous day and Records were kept at the depots how many condoms were used. of how many condoms were Those who did were supposed to taken, but this took the monitor- receive as many as they ing out of the hands of the peer requested. However, shortages educators. By December 1998, and an attempt to control scarce free condoms ceased to be avail- resources led to complaints by sex able at the brothel. By a decision workers that they were not of the small sex worker-run receiving the number they organization, Nari Mukti needed. In mid-1998 the street- Shangho, a system was set up to based intervention only gave out purchase the same condoms at 3 condoms per day to the women, trade price from commercial i.e. the average number of clients depots in Tangail and sell them reported during the baseline sur- through the peer educators with vey, and the brothel intervention a small profit divided between gave out 4 per day. Problems had the sales women and the organi- arisen when the condom distribu- zation. Meanwhile, government tor had a labour strike and no condoms began appearing at the condoms were available. The gov- shops around the brothel at prices ernment would not supply family lower than those bought by the planning condoms to these inter- sex worker organization. As they ventions at the amounts needed. also were better lubricated, sex The future sustainability of the workers began purchasing them condom supply was in question. directly and to sell to others. Sales of condoms increased, but with Eventually, condoms were avail- the variety of possible modes of able again from the main source access, it became more difficult to and by October 1998, three con- monitor. Figure 1 shows all sex dom depots were set up in the worker sales monitored. brothel at which women could go freely to take as many as they Among the street sex workers, wanted. This alleviated the pres- sales began and free condoms

112 SHAKTI, Bangladesh

stopped being distributed in During 1998, female condoms February 1999. The street-based were introduced to both groups sex worker organization, Durjoy, of sex workers on a trial basis was, by that time well organized and found generally useful, and sales increased monthly, ris- especially among the street ing from 23 385 in Febuary to women. Due to their high cost 62 900 in August. A small profit and encouraging reports of safe was made by the peer educators re-use with washing, re-use with or outreach workers who sold washing was also tried, but fears the condoms and also by Durjoy. of misuse prevented project With this money, among other managers from proceeding fur- activities, they set up a trust ther with this option. Larger tri- fund for the poorest group of als are planned after further sex workers. research and training takes

Figure 1. Growth in condom sales by sex workers

70 Brothel T 60 h Street o 50 u s 40 a n 30 d s 20

0 Jul-99 Apr-99 Oct-99 Jan-99 Jun-99 Feb-99 Mar-99 Sep-99 Nov-99 Aug-99 Dec-98 May-99

113 SHAKTI, Bangladesh

place. In addition, a general brothel keepers, as well as demand for water-based lubrica- selected other stakeholders, i.e. tion has led to the development religious leaders and municipal of a locally-produced product personnel, have continued. To which is also under trial. date, only awareness workshops have been held with some client Clients, non-clients and groups. lovers From the beginning, little Street sex worker clients were emphasis was placed in the enumerated in a study carried SHAKTI project on the clients of out in November-December 1998 sex workers. At the brothel these by sex workers themselves. At 22 include traders and other busi- locations around Dhaka on one nessmen, students, truckers, con- different day each week (during struction workers and others. In 7 successive weeks), thirty mem- Tangail two AIDS information bers of the Durjoy Nari Shangho centres were set up at a bus sta- interviewed five sex workers per tion and a court in the town of day (every second one they met). Tangail, but no targeted pro- Responses were recorded on a grammes for clients were devel- pictorial sheet. In total, for 923 oped. Working with the clients sex workers, data were collected while they are in the brothel is about 4654 different clients. The difficult as they do not wish to be proportions of men in different seen or recognized when enter- occupations are shown in Figure ing the brothel. For this reason 2. Rickshaw pullers were the most the brothel has several frequent, followed by a category entrances. The brothel sub-cul- called “service holders” which, in ture is set up for profit and the Bangladesh, means men who protection of clients and, hold wage jobs, such as govern- although brothel keepers have ment and private sector posi- stated they are concerned about tions, and students. In this study, the loss to business that a rise in police were listed as clients but HIV would bring, concrete solu- either paid nothing or very little, tions to these issues have not as they were often simply given emerged. Discussions with the sex in exchange for protection.

114 SHAKTI, Bangladesh

The study provided the project with information on types of clients, as well as other data, and gave the sex workers insight into the practice of research. It clearly demonstrated that illiterate women could conduct valuable research themselves, with minimal technical supervision.

Figure 2. Clients of street-based sex workers in Dhaka, 1998

20 18 16 14 12 % 10 8 6 4 2 0 Police Students Baby taxi Rickshaw Mixed other Businessmen Service holders

The sex workers’ long-term lovers are important men in the brothel. Although some are abusive and exploitative, most of the women value their relationships with these men. Many of the lovers pay the rent, buy them food and clothing, and

115 SHAKTI, Bangladesh

help with their children. In among others, demand sex from February 1998, a study was carried these women. According to the out, showing that, among 233 of survey in October 1999, 36% of these men, 14% were permanent women gave sex in exchange for residents of the brothel and 56% protection (or some other favour) visited for only one or two days at without pay the previous week, a time. About 30% of women in while 61% had had sex with their the brothel had claimed they did own private partners. Condom not need to use condoms as they use differed accordingly, with took no clients and remained 32% of sexual acts covered faithful to their lovers, but project among the casual or coercive staff disputed this. The study, partners and only 18% covered however, seems to corroborate with longer-term, private part- this, with 28% of the men stating ners. Overall, 23% of sex acts the their women took no clients. In previous week had been covered the three months prior to the sur- by condoms with men who were vey, 16% of lovers reported hav- not paying clients. ing had sex with another sex worker in the same brothel and Another survey among these 16% with a sex worker outside women in 1998 revealed that the brothel. In addition, 60% had 13% had at some time injected wives. If it is true that about 30% drugs, as had 24% of their main of sex workers were really taking partners, far higher levels than no other clients than their long- among brothel women and their term lovers, it is also true that men (Jenkins, 1999a). In addition these sex workers, as well as the these women more often sold sex others, were at risk from their to men in groups, either for serial lovers’ multiple sex partners. or simultaneous group sex, than did the brothel women. In most Street sex workers have long- cases, group sex is riskier because term boyfriends or husbands too the woman’s capacity to control but they also are often coerced her safety is reduced and the men into giving sex in exchange for may be exposed to the semen of protection. Pimps, guards, rick- other men as well. In both 1998 shaw pullers, police and mastans, and 1999, about one-quarter of

116 SHAKTI, Bangladesh

the women sampled had hired sales began, the number of con- themselves out to a group (aver- doms sold was monitored. In mid- aging about 4 men) the previous 1998, new indicators on violence week, and over half had done so among street-based sex workers the previous month. were developed. Data on the previous week’s harassment were col- Monitoring and lected at 444 group education ses- Evaluation sions among 4,750 women in the drop-in centres between Sept During the first few years of 1998 and October 1999. With con- SHAKTI, peer educators continuing collection of these data, ducted most of the monitoring of shown in Figure 3, such an indica- changes in knowledge, intent, tor could demonstrate a reduc- trial and use of condoms. After tion (or increase) in violence over the mid-term review, other modes time and any impact the project of data collection were added. might have on securing greater Biannual surveys conducted by safety for sex workers through independent interviewers were advocacy and better communica- put in place, and, after condom tion with street mastans. Figure 3. Harassment of street sex workers, Dhaka (average proportion harassed weekly, 1998-1999)

16 14 12 Police 10 Mastan Client % 8 6 4 2 0 BeatenRaped Robbed Violence

117 SHAKTI, Bangladesh

The SHAKTI project has always 28% to 64% and the trial of a devoted much effort to monitor- condom within the last 24 hours ing its own activities. Although rose from 12% to 59% (Sarkar et at times changes were made that al., 1997b). Self-reported ‘consis- make comparisons difficult, over- tent’ condom use rose from all the sampling and attention to about 14% to 28% (consistency accuracy, data management and was interpreted as 50% or more analysis were adequate to proof all intercourse with clients duce a clear picture of changing during the previous 24 hours). behaviour, major barriers to greater success, and the impact Surveys conducted in 1998 and of attempts to break those barri- 1999 separated questions about ers. One area in which monitor- sex with clients and private part- ing was not fully adequate was ners more precisely than had the changing composition of the been done earlier, using both the target groups, a lesson for the previous day and the previous future. Nonetheless, the atten- week as recall periods. These tion to monitoring and research showed a clearly rising trend in has made it possible to docu- condom use. Figure 4 illustrates ment the project very well, pro- the trend in condom use among ducing numerous presentations the women at Tangail. at international meetings and several useful publications. Changes in composition of the brothel population continued over Effectiveness time. For several years the number The peer educators collected of young women under 18 monitoring data that showed, increased, until 1999, when a after 14 months (as of October severe crackdown took place in a 1997), knowledge that STDs can nearby brothel and panic spread. be prevented through the use of Towards the end of 1999, the gov- condoms rose, as did intent, trial ernment announced that all and reported condom use. brothels would be closed and the Knowledge of the value of con- women rehabilitated. While it may doms rose from 36% to 87%. The be doubtful that this goal will be intent to use condoms rose from accomplished, the combination of

118 SHAKTI, Bangladesh

Figure 4. Trends in condom use (proportion of acts) in Tangail brothel, 1996-1999

70 with clients last 24 hrs 60 with private 50 partners last week % 40 30 20 10 0 Jun 96Nov 98 Jun 99

shifts in the resident population as brothel. Violence at nearby broth- well as reliance on syndromic man- els as well as at Tangail evoked agement in a group of women police raids and some arrests. with high STD prevalence, pro- Results of the survey showed that duced no significant reduction in the number of bonded women STD levels at the time of the mid- had been reduced, young sex term evaluation in April 1998. workers were fewer and, accord- Figure 5 shows some changes in ingly, the number of clients on the the composition of the brothel previous day was seriously dimin- through time. The number of ished. Nonetheless, clients still pre- bonded sex workers (chhukris) ferred the bonded young women. continued to rise until 1999, as did At that time, 52% of independent the number of those under 18. A sex workers had been without survey started in July 1999 was sus- clients the previous day, while only pended for a month and finished 11% of chhukris had had no in September due to trouble in the clients.

119 SHAKTI, Bangladesh

Figure 5. Changing composition of residents at Tangail brothel

45 40

35 <18 years old 30 bonded no clients 25 yesterday % 20 15 10 5 0 Jun-96 Apr-98 Nov-98 Jun-99

Among the Dhaka street sex workers, the baseline survey did not question use with non-paying partners in a clear manner, but between 1998 and 1999 condom use with private and other non-paying partners improved. Results of surveys over time show a clear improvement in use with all types of partners. Figure 6 illustrates these trends.

While the average age of those sampled remained about 22 to 23, the proportion under 18 dropped from 25% to 14%, and the average number of clients per week rose between 1997 and 1999 from 13 to 18. The sex workers reported many new women appearing on the streets during the 1998 floods and again after the Tan Bazar brothel was closed. No method of

120 SHAKTI, Bangladesh

monitoring the changing composition of the street-based sex worker population ever evolved. The prevalence of current syphilis (TPHA+ and RPR=>8) did not shift between surveys in 1997 and 1998, with 32% and 34% respectively. It is difficult to understand if this was because the clinic work and condom use really failed to reduce the prevalence of syphilis or if the women accessing the clinics were constantly changing.

The earliest survey estimated about 4,300 street-based sex workers in Dhaka but continued estimates have not been

Figure 6. Trends in condom use, 1997 to 1999 (proportion of acts) among street-based sex workers in Dhaka

with clients last 50 24hrs

40 with private partners last week

0 Jan-97 Nov-98 Oct-99

121 SHAKTI, Bangladesh

made. The situation is very fluid, well-conceived behavioural with many women shifting from change model specifically for sex hotels to streets, or from city to workers in Bangladesh. Training city, and in and out of the trade. of staff improved through the The actual number visible at any years with the addition of a one time appears to be highly newsletter, a journals club and influenced by police activity, but increased access to reading it is clear that the women either materials translated into Bangla. find new venues or remain inac- Nonetheless, repeated distur- tive for a while when police bances created by local politics activity is high. Poverty, illiteracy, and shifts in management were lack of remunerative options and major factors that limited effi- high rates of abandonment by ciency. Scaling up to cover all the husbands place many women at nation’s brothels through part- risk of turning to the street sex nerships with other NGOs had trade for sustenance. Despite been planned but lack of fund- their many levels of vulnerability, ing, expertise, and, in the end, street sex workers associated government’s threats to close all with the SHAKTI project have brothels, have worked against demonstrated remarkable capac- this plan. ity to take control of their sexual health. Ethical soundness Sound ethical procedures were Efficiency put in place during the extended The excellent infrastructure of period of formative research. CARE, Bangladesh, CARE’s largest Confidentiality, non-coercive and mission, has certainly played an informed consent processes and important role in facilitating the de-linking of HIV samples were operation of the project. practised. All treatable infections Nonetheless, the project suffered diagnosed during STD surveys in efficiency largely due to the were treated. However, in the lack of experienced personnel at case of some STDs, results were all levels and the continued need slow to reach the brothel due to for training. SHAKTI was the first sending the samples overseas for targeted intervention with a processing. Better collaboration

122 SHAKTI, Bangladesh

with local laboratories has solved at building capacity within the this problem. self-run organizations of sex workers for leadership, manage- Planning for the needs of HIV- ment, and monitoring. Continued positive sex workers has also aid to these organizations will arisen as an issue that the project certainly be required, but never clearly considered in its increasing numbers of functions early phases. A small support have been taken over by sex group for positive people has workers. evolved within the framework of the SHAKTI project, and plans to Lessons Learnt support its growth have been made. One of the most important lessons learnt in this project has Replicability and been that, even in a very con- sustainability servative society, sex workers Throughout the life of the proj- can be reached with a sexual ect, the AIDS programme of the health intervention in a non- Government of Bangladesh has judgmental way with previously been associated with SHAKTI in inexperienced staff. A step-by- numerous ways. The Tangail step approach to behavioural brothel and the street sex worker change, with condoms sold clinics have been included as sen- by the sex workers them- tinel sites in the National HIV selves, has worked to improve Serosurveillance. Replicability condom use. Monitoring of con- through other NGOs will, how- dom use and sales by the sex ever, require investment on the workers has contributed to their part of donors or government. involvement in the outcome of Sustainability specifically of the project. SHAKTI was a well- SHAKTI’s sex worker interven- designed project, informed tions is considered achievable by theory and epidemiology, through the enhancement of a factor which has helped hold it skills among the sex workers together through several changes themselves. Investment for the of management and personnel. last year of the project is aimed Considerable investment in

123 SHAKTI, Bangladesh

monitoring and documenta- and the value of an indicator tion has paid off in numerous that will measure the frequency ways for the project. of contact with the intervention. It may also reflect the lack of The basic approach of the adequate training for service SHAKTI project is clearly replica- providers and the inadequacy ble, with adjustments. There is of syndromic management now a vision of sustainability alone in a group with high lev- through a sex worker-run els of STDs. It surely also reflects project that has evolved due to the continued lower level of the efforts of the sex workers condom use with regular part- themselves to band together for ners, a serious issue requiring strength and safety. Protecting sensitive approaches and better the human rights of sex interactions with the men workers has galvanized staff associated with these and sex workers alike and women. helped them to understand the conditions needed to ensure The problems faced by the long-term HIV prevention. Their SHAKTI project emphasize the empowerment has become a need for developing expert- principal aim of the project ise among staff through con- because it is clear that sustain- tinued training and practice. able operation of clinical services As environmental, social and and an outreach programme will political conditions are con- be best motivated by the women stantly changing, the project has themselves. Skills in advocacy had to be flexible to survive. have emerged as a much-needed Meeting the needs of sex work- capacity when working with ers could not be accomplished by highly marginalized groups. a single project, but has required the building of alliances with The lack of evidence for a drop partners that have won victo- in STD prevalence illustrates the ries for sex workers far need for regular monitoring beyond the immediate of the changing composition expectations of the project. of a dynamic target group

124 SHAKTI, Bangladesh

The SHAKTI project illustrates many of the problems faced by sex worker interventions in most nations and a flexible, sex worker-focused approach to solving them.

125 SHAKTI, Bangladesh

References International Conference on AIDS in the Asia Pacific Region, 1997 Abdul-Quader, Abu. Kuala Lumpur, Oct. Women in Need: Street-based Female Sex Workers in Dhaka 1999b Jenkins, C. Hijras and HIV City. CARE, SHAKTI Project Risk: Transgender Communities Report, July 1997. in Bangladesh. Abstract of paper presented at the 5th 1995 Ahsan, RM, Ahmad, International Conference on N, Eusuf, AZ and Roy, J. AIDS in the Asia Pacific Region, Prostitutes and their Kuala Lumpur, Oct. environment in Bangladesh: A geographical perspective. 1999c Jenkins, C. Resistance to Seminar proceedings: condom use in a Bangladesh Commercial Activity, Women brothel. In: Caldwell, J, Caldwell, and Ecology. International P, Anarfi, J, Awusabo-Asare, Geographical Union/Bangladesh K, Ntozi, J, Orubuloye, IO, Geographical Society. Nov 1994, Marcke, J, Cosford, W, Colombo, Rajendrapur, Centre for R, and Hollings, E. (eds.) Development Management, Resistances to Behavioural Dhaka: Grafosman. Change to Reduce HIV/AIDS Infection in Predominantly 2000 Azim, T, Bogaerts, J, Mian, Heterosexual Epidemics in Third M, Islam, M, Sarker, M, Fatteh, World Countries. Health K, Simmonds, P, Jenkins, C, Transition Centre, Australian Chowdhury, MR, and Mathan, National University, Canberra, VI. Prevalence of HIV and 1999, pp. 211-222. syphilis among high risk groups in Bangladesh. AIDS 14(2) 210- 1995 Khan, ZR and Arefeen, HK. 211. The Health Status of Prostitutes in Bangladesh. Study Report. 1999a Jenkins,C. Sex workers in Dhaka Centre for Social Studies, Bangladesh - Results of the University of Dhaka. National Behavioural Surveillance 1998. 1997a Sarkar, S, Durnadin, F, Paper presented at the 5th Jana, S, Hassan, R, Hoque, E,

126 SHAKTI, Bangladesh

Quddus, MA, and Islam, N. A city of Dhaka in Bangladesh. community-based survey of Proceedings of the 4th commercial sex workers (CSWs) International Congress on AIDS in a brothel of Bangladesh on in Asia and the Pacific, Manila, knowledge, intent, trial, and Oct, Abst. # A(P) 071, p. 184. practice for use of condom. ASCON VI (Sixth Annual 1998 Sarkar, S, Islam, N, Scientific Conference, Durandin, F, Siddiqui, N, Panda, International Centre for S, Jana, S, Corbitt, G, Klpper, P Diarrhoeal Disease Research), and Mandal, D. Low HIV and Bangladesh. Programme and high STD among commercial sex Abstracts, Special Publication No workers in a brothel in 57, p. 37. Bangladesh: scope for prevention of a larger epidemic. 1997b Sarkar, S, Durnadin, F, International Journal of STD and Quddus, MA, Chowdury, FK, AIDS 9:45-47. Islam, N, and Mandal, D. Preliminary results on effectiveness of peer outreach intervention on condom use among commercial sex workers (CSWs) in a brothel of Bangladesh. Proceedings of the 4th International Congress on AIDS in Asia and the Pacific, Manila, Oct, Abst. # C(O) 060, p. 118. 1997c Sarkar, S, Durnadin, F, Quader, AA, Hoque, E, Mandal, D, Ahmed, YH, and Islam, N. Estimation of number of street based female commercial sex workers (CSWs) and their HIV situation in the

127 UNAIDS Female sex worker HIV prevention projects: Lessons learnt from Lessons learnt Papua New Guinea, India and Bangladesh Female sex worker HIV prevention projects:

Female sex worker HIV prevention projects: Lessons learnt from Papua New Guinea, India and Bangladesh

UNAIDS Best Practice Collection