The Impact of Women's Land Ownership On Sexual Risk Taking in Kenya
METHODS. This study is based on the Kenya Demographic and Health Survey, a nationally-representative household survey. Data for this study are from the 1998, 2003, and 2008-09 surveys. Women 15-49 years old in sampled households were eligible to be interviewed. Sexual risk was measured as: (1) number of sexual partners other than husband/co-habiting partner in the last 12 months and (2) giving or receiving money, gifts or favors in return for sex in the last 12 months. Land ownership was measured as reporting working on own land, family land, rented land, or someone else’s land. OLS and logit regression analyses were employed to examine whether land ownership predicted sexual risk behavior controlling for potential confounding variables (i.e., age, marital status, wealth, urban/rural residence, and education level of women and their partners) as well as province and year fixed effects.
RESULTS. Findings corroborate with our hypotheses. Firstly, women who worked on land they did not own land had 0.018 more sexual partners other than their husbands/cohabiting partners (p < .01), corresponding to an 18% increase from the average of 0.10. Specifically, among women working on family, rented, or someone else’s land, those working on someone else’s land had the highest number of partners, with 0.028 more partners than women working on their own land (p<.01). Secondly, compared to women working on their own land, women who worked on land they did not own had 1.44 times the odds of engaging in transactional sex (p < .05). Those who worked on family land had the highest likelihood (OR = 1.49; p < .05) of reporting past year transactional sex.
CONCLUSIONS and IMPLICATIONS. Land ownership could empower women to make decisions and take actions regarding their sexual behaviour. These findings suggest that policies enabling or encouraging women to own land might reduce their sexual risk and that establishment and reinforcement of land rights for women may be a viable structural approach to HIV/STI prevention in Kenya. Further research that utilizes a more rigorous design (e.g., longitudinal measures for sexual risk and land ownership) to minimize the threats of selection bias is recommended to verify these findings.