Poverty-targeted cash transfers are reducing some HIV risk behaviours

Cash transfers to orphans and vulnerable children in Kenya are associated with later onset of sexual activity and fewer sexual partners, but no increase in condom use among those already sexually active, according to long-term follow-up from Kenya’s national cash transfer programme presented last week at the Nineteenth International AIDS Conference in Washington DC.

What are cash transfers?

While there has been a small dip globally in new HIV infections, rates are still soaring among adolescents and women in sub-Saharan Africa. This is due in large part to their marginalized social position, and to poverty, gender inequality, and violence against women.

In order to address socio-economic concerns fueling the epidemic globally amongst women and girls in sub-Saharan, governments and researchers have considered the effect that cash transfers may have on behaviour change, and therefore risk of HIV infection.

Cash transfers generally come in two forms:

• conditional cash transfers, in which cash is given in reward for positive behavior or service access;
• more general economic development and empowerment programmes aimed at reducing poverty, but with the either intended or unintended side-effect of reducing HIV risk.

The evidence of the HIV preventative effect of cash transfers is promising. Speaking at a session entitled "Is Money Alone Enough: The 'Value' of Cash Transfers" during the conference, Charlotte Watts of the London School of Hygiene and Tropical Medicine presented findings on several studies that have been conducted in the past several years.

As an example of potential efficacy, Watts noted that the Zomba Cash Transfer Programme - which took place in Zomba, Malawi, and gave money to girls attending schools and their families, as well as payment of school fees - showed 64% lower HIV prevalence in the intervention arm when compared to the control arm. Importantly, these results did not differ significantly if money was given on a conditional or unconditional basis.

Another Malawi-based study has shown that even small economic incentives can double the number of people returning to collect HIV results. A study conducted in Tanzania showed that a conditional cash transfer given to participants who remained free of STIs resulted in a 25% lower prevalence when considering the intervention versus the control arm, with the impact greater among poorer households in rural areas.

Studies presented by Lorraine Sherr of the University College London and Morten Skovdal of the University of Bergen showed that community buy-in to programmes was important, in order to de-stigmatise the link between cash transfer and HIV; to allow for broader discussions of social and behavioral change; and to ensure that programmes were tailored to local needs.

Watts said that sustaining results outside of these programmes is difficult, however, especially for populations considered most at risk, such as youth and women. The Tanzanian STI study showed that behaviour changes were only sustained among men, and not among women; the South African micro-finance study showed that there was no long-term impact on HIV incidence on adolescents in the community.

Results do vary. A study in rural Malawi considering economic rewards for remaining HIV negative over a year period showed no effect on HIV status. Watts said that while cash transfers had a clear impact on uptake of services, the long-term impact on HIV risk was less clear.

Michelle Reme of the London School of Hygiene and Tropical Medicine said that structural interventions should be considered not only for their value in reducing HIV-risk behavior, but also other health and development objectives. Pointing again to the Zomba study, Reme noted that beyond the 64% reduction in HIV risk, the study resulted in a 35% reduction in school drop-out rate; 40% reduction in early marriages; 76% reduction in genital herpes risk; 58% reduction in depression risk; and 30% reduction in teen pregnancies. While the cost per HIV infection averted ranged from $5,000 to $12,500 - much higher than the cost per HIV infection averted by other interventions, such as VCT, PMTCT, and male circumcision - these costs do not take into account the programme's positive impact on other aspects of participants' lives, and society as a whole. 

Recognising that HIV risk stems from myriad factors, and not poverty alone, Watts pointed to a programme in South Africa which considered the coupling effect of micro-finance loans enhanced with community mobilisation and participatory training in gender, violence and HIV. The results were promising, with younger women especially more likely to improve HIV testing and safe sex behaviors, with 64% testing for HIV, and a 25% increasing safe sexual practice. Additionally, participants experienced 55% less violence in the past year; researchers noted that this was due to a multi-pronged approach, with micro-finance alone not enough to reduce violence or empower women.

Just as cash transfer programmes may impact more than just HIV, programmes not specifically tailored to HIV reduction may have the benefit of reducing risky behaviour. Presenting new research at a late-breaker session on Friday, Sudhanshu Handa of the University of North Carolina at Chapel Hill said that the Kenyan government's grant programme for orphans and vulnerable children impacted positively on age at first sexual debut, despite the programme not being specifically intended to address HIV risk.

The programme - run by the Ministry of Gender, Children and Social Development -provides a $20 monthly transfer to "ultra-poor" orphans and vulnerable children under the age of 17. It is the largest social protection program in Kenya, serving 120,000 households, and 350,000 orphans and vulnerable children.

By considering baseline and follow-up interviews on sexual behavior with two residents aged 15-20 of nearly 2,000 households, Handa said that those receiving the grant were 30% less likely to sexually debut within the study period of 2007-2011. Additionally, for those who were sexually active, the program had a "protective effect on having more than three unprotected sexual acts in the last months," explained Handa, and females were less likely to have fewer than two partners in the last twelve months if they received the cash transfer. But other indicators amongst those who did begin having sex during the study period were not as positive, with no significant increase in use of condoms during first sexual act, and no significant decrease in number of participants having sex with partners 10 or more years older.

Considering why the cash transfer has the effect it does on delayed sexual debut, Handa noted that staying in school and having friends who had not debuted was protective, but not a mediator. Change in mental health did not have a protective effect on sexual debut. The study did not consider HIV prevalence or HIV incidence among participants.

Handa noted that other countries in Africa - namely Zambia, Zimbabwe, Malawi, South Africa, and Mozambique - all had similar large-scale unconditional cash transfer programmes for vulnerable groups, and that the Kenyan results "indicate that poverty-targeted transfer programmes may help to reduce HIV risk."

Community-led cash transfers in Zimbabwe: pathways for buy-in and improved child health and development. Nineteenth International AIDS Conference, Washington DC, abstract TUPDE0104, 2012.

View the abstract on the conference website.

Campbell C et al.Social capital and AIDS competent communities: evidence from eastern Zimbabwe. Nineteenth International AIDS Conference, Washington DC, abstract TUPDE0105, 2012.

View the abstract on the conference website.

Handa S et al. Effect of a national social cash transfer program on HIV risk behavior in Kenya. Nineteenth International AIDS Conference, Washington DC, abstract FRLBD01, 2012.