Women’s Land Ownership & Child Diarrhea – Uganda

Public Policy Course: International Poverty and Economic Development

The goal of this rigorous, hands-on course was to help students understand how economists approach problems of development. Topics included various types of impact evaluations related to health, nutrition, HIV/AIDS, education, public infrastructure and economics.  The course was devoted to in-depth analysis of microeconomic models and empirical methods and included two problem sets. The first project was aimed at evaluating a conditional cash transfer program aimed at reducing sex worker risk behavior in Ecuador. For my final project, I was interested in the relationship between women’s land ownership and the impact this has on children’s health outcomes in Uganda. More specifically, I wanted to look at childhood diarrhea since it is the second leading cause of death in children under 5 years of age.

Research Question: Does women’s land ownership in Uganda have an impact on the incidence of diarrhea among children under 5 in their household?

Background:

• Well-defined and secure land tenure and titling is hypothesized to provide many benefits to the owner including increase access to credit, enhanced land security and the ability to use land as collateral. (World Bank, 2007).
• Most analyses have studied asset holdings at the level of the household rather than at the level of the individual. This does not capture distribution of ownership of land or other productive assets by gender. (Alvarado, Deere, et al. 2012)
• Women who own land are significantly more likely to have the final say in household decisions, a measure of empowerment. Similarly, children of mothers who own land are significantly less likely to be severely underweight (Allendorf, 2007).
• Proper land titles provides security for households to invest in water and sanitation improvements (IADB 2004).

Methods:

I used cross sectional data from the Demographic and Health Survey (DHS) from 2011 and ran OLS regressions with fixed effects for urban/rural. The 2011 Uganda Demographic and Health Survey (UDHS) is a nationally representative survey of 10,086 households with 9,247 women age 15-49 and 2,573 men age 15-54. We began by running OLS regressions where Diarrhea among children < 5 was our dependent variable and female land ownership was our independent variable. We made dummies for female land ownership, we lumped together any type of ownership, whether its alone or jointly.

Female Land Ownership, Marriage and Wealth

• First we wanted to see what would happen if we added marriage status, wealth index, and fixed effects for rural/urban.
• We know that land is often acquired from marriage and we know that wealthy households are probably more likely to have a lower incidence of diarrhea (for a variety of reasons). Of these 3 characteristics, female land ownership has the greatest coefficient.
• Without controlling for any variables, women that own land are 2.8 percentage points less likely to have a child with diarrhea in the past 2 weeks.

Control Variables:

• Then we thought about all the household and individual factors that would impact the incidence of diarrhea,
• Some of these factors included household floor material because that could impact contamination, what type of toilet the household had, whether there was soap or water present for hand washing, how many households shared the toilet, as well as individual characteristics like the women’s educational attainment, husbands educational level, whether the child was given baby formula or breastfed…. 

Male Land Ownership vs Female Land Ownership Findings

• A woman owning land alone, jointly or both jointly and alone is 14.6 percentage points less likely to have a child that had diarrhea in the past two weeks
• A man owning land alone, jointly or both jointly and alone is about 7.6 percentage points less likely to have a child that had diarrhea in the past two weeks
• A woman owning land alone was 10.4 percentage points less likely to have a child<5 with diarrhea in the past two weeks

Problems with these Regressions

• The relationship between a man and woman in household and the type of land ownership they have individually or jointly…How do we start to think about controlling for this if there was another person of the opposite sex in a household that owned land?
• We did fixed effects by urban rural, but would have like to have a higher level, like district. We tried by cluster, but there was multicolinearity with urban/rural.
• Are we just capturing more single female headed households with children? What does this mean if we are?

Making the connection with empowerment via decision making questions.

• First we wanted to see what would happen if we added marriage status, wealth index, and fixed effects for rural/urban.
• Women’s land ownership has the greatest coefficient.
• Women who own land are 2.8 percentage points less likely to have a child with diarrhea in the two weeks preceding the survey.
• What factors impact the incidence of diarrhea among children < 5 years old?