Study participants were recruited through a two-stage systematic sampling procedure conducted in selected migrant worker residential colonies (either organized labour colonies or illegally occupied land with small houses without a proper roof) and worksites. Geographical maps drawn for each district were used to list all migrant worker residential areas and worksites and to select cluster areas within these sites for recruitment. Clusters were created by combining smaller sites and dividing larger sites such that each cluster offered an area with approximately 5000 male migrant workers. Three clusters were then selected randomly from within each district, and migrant men within the chosen clusters were systematically sampled to obtain a minimum of 2500 participants per state. The sample size was determined using an estimated proportion of 15% of men having sex outside marriage, an assumed difference of 3% increase in the proportion with every unit increase in degree of mobility, a confidence level of 95% and power of 80%.
A total of 11 635 eligible male migrant workers satisfied the inclusion criteria: age 18 years or older, having migrated to at least two places in the past 2 years for work, and were selected for the cross-sectional survey. Of these, 145 (1.2%) men refused to be interviewed, and 271 (2.3%) men did not complete their interviews and were thus excluded from the analyses, providing our final sample size of 11 219.
Interviews were conducted by multilingual graduates or postgraduates in sociology, anthropology or statistics. Interviewers were trained and experienced in quantitative data collection techniques and field-based public health and HIV/AIDS research. Data were obtained through face-to-face interviews conducted in private locations close to the respondents' residence or workplace. Verbal consent was obtained from all respondents before being interviewed. Data quality and management involved immediate review by field staff after interviews to ensure accuracy and completion, same-day review by the field supervisor and weekly transport of surveys to the data management team. Trained data entry officers then entered the survey data weekly and processed it monthly to verify consistency and accuracy, using SPSS. Ethical approval for the study was obtained from the institutional review boards of the Population Council and the University of Manitoba, Canada.
Data collected included the sociodemographic characteristics of the migrants, including age, highest level of education completed, marital status, current residence with wife, income, and occupation (construction work, market place loading and unloading, daily wage work, industrial work, stone cutting, and fishing). Information was also collected on whether a labourer had worked within a contract system (either the first migratory move for the current work was under contract or because his current work was under contract).
Risky behaviour was measured by collecting information on men's sexual behaviour, frequency of condom use, and alcohol use with paid sex workers and unpaid, non-spousal sex partners in destination areas (past 12 months) and in their places of origin (past 24 months). These data were used to create variables such as alcohol use before non-marital sex, the number of sex worker partners, the number of non-spousal unpaid female sex partners, and the occurrence of sex with a sex worker or non-spousal unpaid female sex partner in place of origin. They were also used to create measures of consistent condom use. Consistent condom use is defined as the use of a condom every time that the respondent had sex. Consistent condom use with a range of partners, sex workers, non-spousal unpaid female sex partners, and in any non-marital sex encounter, was measured. An additional variable measured whether the participant had ever used condoms.
The survey also collected information on HIV risk perceptions and STI symptoms. Participant HIV risk perception was assessed by responses to a question on their perceived risk of getting HIV and if the risk was high, moderate and low. Participants were defined as having recent STI symptoms if they indicated any of the following in the past 12 months: genital ulcers; swelling in groin area; itching in genital area; or frequent painful urination. Alcohol use was assessed based on responses to alcohol consumed in the past 30 days and the type of alcohol (e.g. beer, whisky, country liquor).
Data analysis involved generating descriptive statistics on the prevalence of contracted labour and, for the subsample of contract labourers, sociodemographic and HIV risk profiles. Chi-square analyses were used to detect differences in HIV risk indicators by sociodemographics among contract labourers; the significance for all analyses was set at P < 0.05. Logistic regression models and models adjusted for sociodemographics (age, education, income, occupation and marital status) were also used to assess associations between contract systems and HIV risk indicators. Logistic regression results are presented for the key independent variable, being under contract, which is an easily identifiable characteristic of a migrant labourer and can be used for appropriate targeting of HIV prevention interventions. All statistical analyses were conducted using SPSS 11.0.
Prevalence of contract labour and HIV risk among contract labourers
One-third of migrant male workers surveyed (35%, n/N = 3880/11 219) were contracted labourers, and the proportion of men reporting contract labour varied by state: 47% in Andhra Pradesh; 42% in Tamil Nadu; 31% in Karnataka and 19% in Maharashtra (see Table 1). Among the contract labourers, 70% were under the age of 30 years. Over half (55%) were married; 34% resided away from their wife as a result of migrant work. Seventeen per cent (17%) were illiterate, having received no formal education; an additional 16% received only primary education. The median income of contract labourers was 3500 rupees (approximately US$90, at 1US$ = 39 rupees in the year 2008) per month, and were primarily occupied in the construction (29%), industrial (24%), daily wage (12%), and loading and unloading industries (7%).
Alcohol use and HIV risks among contract labourers
More than three-quarters of contract labourers (77%) reported drinking of any alcohol type in the past month, with 19% reporting the use of a variety of alcohol (i.e. beer, whisky, country liquor) in this same timeframe (see Table 1). Thirty-seven per cent of the contracted labourers reported that at the last sexual encounter with a non-marital sex partner they had imbibed alcoholic drinks before sex. More than one in six men (17%) reported sex with a sex worker in the past year; 12% reported sex with multiple sex workers in the past year. Almost one-quarter of men (24%) reported sex with a non-spousal, unpaid female partner in the past year; 9% reported sex with multiple non-spousal, unpaid female partners in the past year. The majority of contracted labourers (65%) had never used condoms with any partner. Among men reporting sex with a sex worker in the past year, however, 58% reporting consistent condom use. Among men reporting sex with a non-spousal unpaid female sex partner in the past year, only 19% of men reported consistent condom use. Almost two-thirds of men (64%) reported at least one of the four STI-associated symptoms in the 12 months before the survey. Despite such high sexual risk behaviours, only 3.2% of the contracted labour sample reported moderate or high HIV risk perceptions.
Associations between sociodemographic characteristics and HIV risks among contract labourers
The proportion of contracted labourers visiting sex workers was found to be high among men aged 20–34 years (19.0%) compared with those younger (<20 years of age: 6%) or older (35+ years of age: 14%); those who are single (18%) or married and reside away from their wife (22%) compared with those living with their wife (13%); those who are middle (23%) or lower (20%) income rather than upper (14%) income; those working as fishermen (32%) or stone cutters (23%) than men involved in any other occupation (see Table 2). Considerably higher proportions of men working as loading and unloading labourers (31%), fishermen (35%) and industrial labourers (30%) reported having sex with non-spousal unpaid female partners than men working in other occupations. Inconsistent condom use was more common among those who were married and residing with their wives (76%) rather than single (60%) or married and not residing with their wife (54%); lower (80%) rather than moderate (61%) or higher (63%) income; loading and unloading labourers (79%), stone cutting men (81%), fishermen (75%) than men in other occupations.
Differences in HIV risks between contracted and non-contracted male migrant labourers
Regression analyses adjusted for sociodemographic variables document that contracted migrant labourers were significantly more likely than non-contracted migrant labourers to drinking alcohol in the past 30 days [adjusted odds ratio (AOR) 1.5, 95% confidence interval (CI) 1.4–1.6], diverse alcohol use (AOR 1.5, 95% CI 1.4–1.7), sex with a sex worker (AOR 1.4, 95% CI 1.3–1.6), sex with multiple sex workers (AOR 1.3, 95% CI 1.2–1.5), sex with a non-spousal unpaid woman (AOR 1.8, 95% CI 1.6–2.0), sex with multiple non-spousal unpaid women (AOR 1.4, 95% CI 1.2–1.6), sex with a non-marital partner in place of origin (AOR 1.3, 95% CI 1.2–1.4), and STI symptoms (AOR 1.7, 95% CI 1.6–1.9) (see Table 3). Contracted labourers were significantly less likely than non-contract labourers to report never having used condoms (AOR 0.8, 95% CI 0.7–0.9), but were also less likely to report consistent condom use with sex workers (AOR 0.7, 95% CI 0.6–0.9) and moderate or high HIV risk perceptions (AOR 0.7, 95% CI 0.6–0.9).
In India the use of contract labour in employment is increasingly common, particularly in the industries of construction, industrial production, mining, fishing and market labour , which have traditionally employed migrants. Furthermore, migrant workers are a recognized HIV bridge population. Evidence in the present study affirms these previous findings by documenting that more than one-third of male migrant workers are contracted labourers. In addition, findings document a hierarchical structure within these contract systems that support the social welfare and health of these workers and thus could be used to support HIV prevention efforts.
A notable proportion of mobile contracted labour in the study reported alcohol use in conjunction with risky sex, sex with multiple sex workers, sex with multiple non-spousal unpaid women, and sex with a non-marital partner in the place of origin. Despite multiple partnering, these men reported low condom use and HIV risk perceptions; corresponding to these risky sex behaviours, STI symptoms in the past 6 months were reported by the majority of these men. These findings are consistent with the results noted in other national and international settings regarding the behaviours of migrant male workers [2,11,14,24,31]. Although age and singlehood have effects on HIV risk behaviours irrespective of a labourer's contract status, they are not easily identifiable characteristics for targeted programme interventions. On the other hand, contract status is an easily identifiable characteristic and the contract system itself provides a potential mechanism for delivering health promotion interventions. There is no easy mechanism to group individuals by their demographic or social factors for designing or implementing preventive programmes.
Notably, the current findings further document that contracted labourers are significantly more likely to report alcohol use and HIV risk behaviours (e.g. sex with sex workers and other non-marital partners, unprotected sex) than non-contracted labourers. This may be attributable to contract labourers often living separate from their wives and migrating within male groups who may socialize through the consumption of country and cheap liquor, watching adult (pornographic) movies in theatres, and visiting beer bars or sex workers. More than two-thirds of contracted labour who had sex with a non-spousal unpaid female partner reported that the sexual partner was a workmate. This suggests that HIV prevention programmes are required for both male migrant workers and women at worksites.
This analysis provides evidence and guidance to efforts by the Indian government for HIV intervention programmes for male migrant workers currently underway in the third National AIDS Control Programme . The use of contract systems as structures to implement HIV prevention interventions among male migrant workers is a potentially feasible mechanism. This recommendation for using contract systems builds on the argument that individual behaviours associated with HIV risk may be difficult to regulate directly . The challenge, however, is to have a macrosocial view of structures that adds to a workable response to the pandemic  in India.
The response for HIV prevention among migrant men in India can be greatly enhanced via delivery through mapped contractors and contract systems. For example, the findings indicate that the process of recruiting migrant men in real estate, bridge, dam and road construction industries includes multiple layers of the contract system: the construction agency authorities issue a subcontract to licensed contractors (also known as ‘big contractors’) at district or state levels to recruit and hire men (Table 4). These ‘big contractors’ have connections with various skilled subcontractors who in turn have small contractors/agents with more localized connections to recruit and hire workers. The small contractors/agents are the lowest denominators in the chain of the contract system. The small contractors are often skilled men from the local village who are already connected to the subcontractor(s) and upon receiving a specific ‘contract’ or ‘work order’ are able to recruit men for that particular job from their village of origin. The small contractors often accompany men in their migration to the workplace and stay as small workgroup overseers until the work is completed and provide information on lives at the places of destination. This chain of the contract system varies and is based on the type of work, volume of work and number of skills required for completing the work, and could become a potential for implementing HIV prevention programmes.
The contractors would also be interested in investing in HIV prevention interventions as it protects and promotes the health of their workers, which in turn increases productivity. Structural interventions, especially when bolstered with political, organizational support and government policy and regulation, are feasible, effective, and cost-effective . This may mean that it requires greater participation of the contractors at all levels and such participation should be encouraged by the organizational, political and governmental authority. One intervention model could be the education of the contractors within prevention interventions on how to self-assess HIV risk and promote behavioural change using oneself as an example and personally endorsing the benefits of change. The trained contractors could become advocates for further communication with the contracted labourer on behavioural change and HIV prevention. This requires an organizational policy in each sector of employment, including construction industries, fishing, mining, loading and unloading work.
Whereas the current research provides important insights to support better HIV prevention efforts for migrant men and possible ways for the implementation of interventions, findings must be interpreted with consideration of some limitations. As is true of a number of investigations, the information collected is based on self-reports, which is subject to recall and social desirability biases. To the extent possible, these biases have been reduced by using a shorter-term recall period and anonymous interviews. Contracted labour was defined based on simple questions on a given participant's actual association with contract on at least one of two occasions. As this variable does not involve any reference period, the estimate of the proportion of contracted labour obtained in this study could be an underestimate of ever contracted labour and an overestimate of current contracted labour. STI symptoms were also based solely on self-report and lack information on health provider diagnosis. Nonetheless, despite these study limitations, this large-scale study offers for the first time an analysis of contracted labour in high HIV prevalence states within India, and documents the utility of contract systems to address HIV risk in a major HIV bridge population within the country.
Finally, the large infrastructure of contract systems available in India can offer several strategic advantages for implementing HIV prevention policy to migrant men. They include opportunities for peer-based counselling, treatment for STI and condom distribution besides behavioural change communication from contractors to men recruited for work at a large scale in India.
The authors would like to acknowledge the endless efforts of other study team members Dr. S. Ramachandran and Dr. A.K. Ravi Shankar from Annamalai University, Ms. Raluca Buzdugan from the University of Manitoba, and Dr. Kanchan Mukherjee from Tata Institute of Social Sciences for their participation throughout the study. They also thank Dr. Anita Raj from Boston University and the anonymous reviewers for their valuable comments on the earlier version of the paper.
Sponsorship: This research was funded by the Bill and Melinda Gates Foundation.
The views expressed herein are those of the authors and do not necessarily reflect the official policy or position of the Bill and Melinda Gates Foundation.
Conflicts of interest: None.
2. Jochelson K, Mothibeli M, Leger JP. Human immunodeficiency virus and migrant labour in South Africa. Int J Health Serv 1991; 21:157–173.
3. Anarfi JK. Sexuality, migration and AIDS in Ghana – a socio-behavioral study. Health Transit Rev 1993; 3(Suppl.):45–67.
4. Wolffers I, Fernandez I, Verghis S, Vink M. Sexual behavior and vulnerability of migrant workers for HIV infection. Culture, Health Sexuality 2002; 4:459–473.
5. Ford NJ, Kittisuksathit S. Sexual hazards for migrant workers. World Health Forum 1996; 17:283–285.
6. Campbell C. Migrancy, masculine identities and AIDS: the psychological context of HIV transmission on the South African gold mines. Soc Sci Med 1997; 45:273–281.
7. Chirwa WC. Migrant labour, sexual networking and multipartnered sex in Malawi. Health Transit Rev 1997; 7(Suppl. 3):5–15.
8. Lurie M, Harrison A, Wilkinson D, Abdool Karim SS. Circular migration and sexual networking in rural KwaZulu/Natal: implications for the spread of HIV and other sexually transmitted diseases. Health Transit Rev 1997; 7(Suppl. 3):15–24.
9. Brockerhoff M, Biddlecom AF. Migration, sexual behavior and the risk of HIV infection in Kenya. Int Migr Rev 1999; 33:833–856.
10. Wolffers I, Fernandez I. Migration and AIDS. Lancet 1995; 346:1303–1304.
11. Poudel KC, Jimba M, Okumura J, Sharma M, Poudel-Tandukar K, Wakai S. Migration in far western Nepal: a time bomb for a future HIV/AIDS epidemic? Trop Doctor 2004; 34:30–31.
12. Pison G, Guenno BL, Lagarde E, Enel C, Seck C. Seasonal migration: a risk factor for HIV infection in Rural Senegal. J Acquir Immune Defic Syndr 1993; 6:196–200.
13. Gangakhedkar RR, Bentley M, Divekar A, Gadkari D, Mehendale SM, Shepherd ME, et al
. Spread of HIV infection in married monogamous women in India. JAMA 1997; 278:2090–2092.
14. Morris M, Podhisita C, Wawer MJ, Handcock MS. Bridge populations in the spread of HIV/AIDS in Thailand. AIDS 1996; 10:1265–1271.
15. Entz AT, Ruffolo VP, Chinveschakitvanich V, Soskolne V, van Griensven GJP. HIV-1 prevalence, HIV-1 subtypes and risk factors among fishermen in the gulf of Thailand and the Andaman sea. AIDS 2000; 14:1027–1034.
16. Chandrasekaran P, Dallabetta G, Loo V, Rao S, Gayle H, Alexander A. Containing HIV/AIDS in India: the unfinished agenda. Lancet Infect Dis 2005; 6:508–521.
17. Decosas J, Kane F, Anarfi JK, Sodji KD, Wagner HU. Migration and AIDS. Lancet 1995; 346:826–828.
19. Verma RK, Saggurti N, Das M, Rama Rao S, Jain A. Patterns and implications of male migration for HIV prevention strategies in Andhra Pradesh
. Technical brief from Population Council India, No. 1, 2007. March. Also available at: www.popcouncil.org/pdfs/India_TechBrief.pdf
. Accessed: September 2008.
20. Nepal B. Population mobility and spread of HIV across the Indo-Nepal border. J Health, Popul Nutr 2007; 25:267–277.
21. Gurung HN. Social demography and expressions. Kathmandu: New Era; 1998, 17–39.
22. Poudel KC, Okumura J, Sherchand JB, Jimba M, Murakami I, Wakai S. Mumbai disease in far western Nepal: HIV infection and syphilis among male migrant-returnees and non-migrants. Trop Med Intl Health 2003; 8:933–939.
23. Halli SS, Blanchard J, Satihal DG, Moses S. Migration and HIV transmission in rural south India: an ethnographic study. Culture, Health Sexuality 2007; 9:85–94.
24. Mishra A. Risk of sexually transmitted infections among migrant men: findings from a survey in Delhi. Asian Pacific Migr J 2004; 13:89–106.
25. Gupta I, Mitra A. Knowledge of HIV/AIDS among migrants in Delhi slums. J Health Popul Dev Countries 1999; 2:26–32.
26. Rao A, Nag M, Mishra K, Dey A. Sexual behavior pattern of truck drivers and their helpers in relation to female sex workers. Ind J Soc Work 1994; 55:603–615.
27. Rao KS, Pilli RD, Rao AS, Chalam PS. Sexual lifestyle of long distance lorry drivers in India: questionnaire survey. BMJ 1999; 318:162–163.
28. McGreevey W, Alkenbrack S, Stover J. Construction workplace interventions for prevention, care, support and treatment of HIV/AIDS
. Washington, DC: Futures Group International, POLICY Project; 23 May 2003. 16 pp.
30. Registrar General of India. Census of India: D Series (Migration tables): 2001
. New Delhi, India: Registrar General of India; 2001.
31. Tuan NA, Fylkesnes K, Thang BD, Hien NT, Long NT, Kinh NV, et al
. Human immunodeficiency virus (HIV) infection patterns and risk behaviors in different population groups and provinces in Viet Nam. Bull WHO 2007; 85:35–41.
32. NACO. Targeted interventions under NACP III: operational guidelines – Vol. II (migrants and truckers).
New Delhi, India: National AIDS Control Organisation, Ministry of Health and Family Welfare, Government of India; 2007.
33. Blankenship KM, Bray SJ, Merson MH. Structural interventions in public health. AIDS 2000; 14(Suppl. 1):S11–S21.
34. Sumartojo E. Structural factors in HIV prevention: concepts, examples, and implications for research. AIDS 2000; 14(Suppl. 1):S3–S10.
35. Sweat M, Kerrigan D, Moreno L, Rosario S, Gomez B, Jerez H. Cost-effectiveness of environmental–structural communication interventions for HIV prevention in the female sex industry in the Dominican Republic. J Health Commun 2006; 11:123–142.
Keywords:© 2008 Lippincott Williams & Wilkins, Inc.
contract system; HIV; India; migration; mobility; prevention; risk behaviours