Title: Women's Property Rights and Gendered Policies: Implications for Women's Long-term Welfare in Rural Tanzania
Abstract: Abstract This paper evaluates effects of community-level women's property and inheritance rights on women's economic outcomes using a 13 year longitudinal panel from rural Tanzania. In the preferred model specification, inverse probability weighting is applied to a woman-level fixed effects model to control for individual-level time invariant heterogeneity and attrition. Results indicate that changes in women's property and inheritance rights are significantly associated with women's employment outside the home, self-employment and earnings. Results are not limited to sub-groups of marginalised women. Findings indicate lack of gender equity in sub-Saharan Africa may inhibit economic development for women and society as a whole. Acknowledgements Support for this paper was provided by the Hewlett Foundation and Population Reference Bureau Dissertation Fellowship in Population, Reproductive Health and Economic Development. The author thanks Sudhanshu Handa, Gustavo Angeles, Kathleen Beegle, Christine Durrance, Joachim De Weerdt, Krista Perreira, Kavita Singh and two anonymous referees for helpful comments. This research has been approved by the UNC Behavioral IRB Board (Study # 07-0753). Notes An Online Appendix is available for this article which can be accessed via the online version of the journal available at www.informaworld.com/fjds 1. Women constitute nearly 60 per cent of the more than 25 million people estimated to be living with HIV/AIDS in sub-Saharan Africa. In many places, HIV prevalence among girls and young women aged 15–19 years is four to seven times higher than among boys of the same age group (Walsh, Citation2005). In addition to low socioeconomic status which fuels transmission, evidence suggests that women are biologically more likely to be infected than men (Loewenson and Whiteside, Citation1997; World Bank, Citation1998). 2. See the 1966 ‘International Covenant on Economic, Social and Cultural Rights,’ the 1979 ‘Convention on the Elimination of All Forms of Discrimination against Women’ and, more recently, the ‘Women's Declaration on Population Policies,’ the 1995 ‘Beijing Platform for Action,’ and the 2003 ‘Commission on Human Rights Resolution 2003/22 on women's equal ownership, access to and control over land and the equal rights to own property and to adequate housing.’ 3. De Soto's solution is to formalise all informal use of land, giving the poor the ability to use property as collateral to obtain loans and to further invest funds. 4. Despite population-level poverty being viewed as a risk factor for HIV/AIDS, there is considerable debate as to the relationship between resources, wealth and HIV within a given resource poor country. See, for instance, Bingenheimer (Citation2007), Mishra et al. (Citation2007) or Piot et al. (Citation2007) for overviews and evidence. 5. However, it should be noted this practice is increasingly being abandoned where the death of the husband is thought to be caused by AIDS (UNAIDS, 1999). 6. For example, AIDS-related losses can reduce African household incomes by up to 80 per cent, food consumption by 15–30 per cent and primary school enrolment by 20–40 per cent (Whiteside, Citation2002). 7. The reforms begin with the appointment of the Presidential Commission on Land Matters in 1991 and resulted in the passage of the Land Act and the Village Land Act in 1999. The commission made recommendations in three broad areas: (1) policy and questions of law; (2) administration and adjudication of disputes; and (3) gender equity. For detailed description of the reform process see Tsikata (Citation2003). 8. The Village Land Act specifies that the Village Council govern land-related issues, of which at least one third of its total of 25 members should be women (Walker, Citation2002). 9. For example, the Law of Marriage Act set a minimum legal age of marriage of 15 years old for women and 18 years old for men. In addition, it required registration of marriage and provided legal terms for divorce for the protection of the family members. See Tenga and Peter (Citation1996) or Mtengeti-Migiro (Citation1997) for further discussion of terms included. 10. The research project is also known as the Economic Impact of Adult Mortality (EIAM) Study. Funding was provided by the World Bank Research Committee, the United States Agency for International Development (USAID) and the Danish Agency for Development Assistance (DANIDA) (World Bank, Citation2004). 11. Community adult mortality rates were taken from the 1988 census and a subsequent enumeration for the survey (WB-DRG, 2004). 12. ‘Sick’ households were those who either had an adult death (aged 15–50 years) due to illness in the last year or an adult too sick to work at the time of the survey or both. ‘Well’ households were all others with no recent deaths or sick members (WB-DRG, 2004). 13. Note that many of the women who moved out of sample communities are re-interviewed in KHDS (approximately half), however since they have no community-level variables including WPIR, they cannot be included in the sample and are treated as attriters. 14. For example, it could be argued that WPIR is a choice variable on the individual women level. This would occur if specific women or groups of women try to influence or change WPIR. If this is the case, a problem would arise if these women with high WPIR would be expected to have better outcomes (time use or labour force participation) on average than women in communities with poor WPIR because of correlated unobservables at the individual level. 15. For consistency between cross-sectional and panel specifications, all binary models are estimated using LPMs. Parallel analysis using probit models do not change results. 16. For example, a minimum of three experts representing the most knowledgeable people in the following areas were called: births, deaths and migration (village secretary/chairman was suggested), economic activities (community development officer), education (chairman of education committee), health (chairman of health committee), agriculture and livestock, cultural practices (World Bank, Citation2004). 17. In actuality 49 survey instruments were filled at the community level, two of which represented two clusters each for a total of 51 community clusters (World Bank, Citation2004). 18. The answers to WPIR questions are asked to the group of experts as described above. When there was disagreement among the experts, the majority opinion was honoured (World Bank, Citation2004). 19. Passages 1 and 2 in the baseline did not ask this question in the community questionnaire. Therefore indicators of exemption from widow inheritance for women in these passages were replaced with the value for Passage 3. In either case, this represents a value collected approximately 6–12 months after individual level outcomes. Although it is unlikely for the indicator to have changed for a significant portion of the sample, it would result in an underestimation of the change over the panel period and would tend to bias results toward zero or no significant effect of WPIR as a whole. 20. The pharmaceutical in name is paracetamol, an aspirin substitute. Other pharmaceuticals were tested in the model but were dropped because of their low correlation with re-interview. In addition, following Beegle et al. Citation2006a, alternative relationships to the household head and indicator of child living elsewhere were also tested as potential predictors of attrition and found to have no correlation with re-interview. Note that Beegle and colleagues are only instrumenting mobility and therefore these mentioned variables have increased validity for their analysis versus this analysis. Additionally, mortality related enumeration rates connected with sample selection, as well as shock variables at the community level (drought, flood and so forth) were tested and found to lack statistical power with the inclusion of other Z variables and controls and, therefore, were not utilised. 21. Fixed effects models are used as an alternative to random effects models. The Hausman test rejects the random effects model in favour of fixed effects in all cases but self-employment. 22. As an alternative, conditional logit models are run on panel specifications to estimate working outside the home, working self-employed and indicator of savings account. However, the sample size drops to 220 and thus is not utilised. 23. The variables used in all waves are: (1) log of dwelling value, (2) log of land value, (3) log of livestock value, (4) toilet, (5) tv, (6) car, (7) bicycle, (8) refrigerator, (9) stove, (10) radio, (11) dwelling has cement floors, (12) per capita number of rooms in dwelling, (13) main lighting source is electricity. Mean values and proportion contribution to the factor index are available from the author upon request. 24. Although many of these variables are subject to recall bias, documentation indicates that individual members were interviewed separately in private for individual level outcomes such as mentioned above, improving accuracy and confidentiality of responses. 25. All expenditure and savings indicators were deflated to the baseline (1991) value of the Tanzanian shilling on a yearly basis, using the CPI obtained from the Tanzanian Bureau of Statistics on 18 July 2007. 26. Note other community-level indicators of high women's status were tested and also found to be insignificant including proportion of women with higher education and proportion of women with higher education in comparison to men in the same community. 27. Although these women are distinctly different, they share common hardships. While widowed women may be subject to abuse by in-laws, divorced or separated women may be denied property by her husband and forced to return to her natal family. In many cases returning home results in forced re-payment of bride price by the woman to her family, and it is not necessarily accompanied by welfare improvements (Manji, Citation2000). To test differences empirically, the groups were separated and coefficients from the cross-sectional models were tested against each other. Results indicate that the groups do not differ significantly with respect to the outcomes of interest and are therefore retained as one indicator. 28. The alpha scale reliability associated with the indices are 0.70 for the baseline and 0.59 for the end-line. 29. Countries with equal WPIR include Burkina Faso, Malawi, Mozambique, Niger, Rwanda, South Africa, Tanzania, Uganda and Zimbabwe. Ethiopia and Eritrea have gender-neutral laws and Botswana, Kenya and Namibia are in the process of reform (UN-HABITAT, 2006).
Publication Year: 2011
Publication Date: 2011-01-01
Language: en
Type: article
Indexed In: ['crossref', 'pubmed']
Access and Citation
Cited By Count: 97
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot