Presentation is loading. Please wait.

Presentation is loading. Please wait.

Rwanda: The impact of conflict on fertility Kati Schindler & Tilman Brück Gender and Conflict Research Workshop 10/06/2010.

Similar presentations


Presentation on theme: "Rwanda: The impact of conflict on fertility Kati Schindler & Tilman Brück Gender and Conflict Research Workshop 10/06/2010."— Presentation transcript:

1 Rwanda: The impact of conflict on fertility Kati Schindler & Tilman Brück Gender and Conflict Research Workshop 10/06/2010

2 2 Outline of presentation 1.Genocide in Rwanda 2.Research question 3.Hypotheses 4.Data and conflict measures 5.Estimation approach 6.Results descriptive regressions 7.Further steps

3 3 Genocide in Rwanda -100 days in 1994 -targeted violence against Tutsi and (to a lesser extent) against moderate Hutu -800.000 deaths -highest mortality among Tutsi men -severely unbalanced sex ratios (88 for prime age adults in 2002) missing men high proportion of widow-headed households marriage market: women have lower probability to marry -return of 700,000 old caseload refugees to Rwanda -internal migration -former rebels take over power; new legislation and policies towards gender

4 4 Research Question What determines fertility in post-war Rwanda? What is impact of conflict exposure?  individual-level effect  age-group effect  locational effect Relevance of topic understanding household vulnerability to poverty in post-conflict understanding group-dynamics in fertility for future conflict onset

5 5 Hypotheses: impact of conflict on fertility (1)Victims may have a perceived need to replace the lost ones. Preferences on fertility may be driven by either individual or group characteristics, including a heightened sense of group identity. (2)More men and boys died than women and girls during the genocide. This in turn reduces the number of unions and fertility. (3)The genocide may have regional effects, e.g. by impacting on health infrastructure or employment opportunities as a result of regional economic divergences induced by the genocide.

6 6 Data Rwanda Demographic & Health Surveys pooling three cross-sectional waves: 1992, 2000, 2005 sample: prime age women (15-49 years) representative for Rwanda and 12 provinces information about: birth histories, maternal & child health, domestic violence, sibling mortality, socio-economic characteristics Limitations few information on partner & household members no consumption or income data construct asset index through ordinal PCA (Kolenikov/Angeles 2009) no community questionnaire no conflict intensity measure (in post-genocide waves)

7 7 Conflict measure 1: Sibling deaths in 1994 captures individual-level (or family-level) exposure to violence based on sibling mortality module (DHS 2000 & 2005) every respondent is asked for characteristics of all siblings includes both siblings sharing residence and living in separate hh but: issue of comparability of indicator across waves  DHS 1992: ethnicity (being Tutsi)  DHS 2000 & 2005: sibling(s) died in 1994 Fig. 2: Distribution of sibling mortality over time

8 8 Conflict measure 2: Widowhood captures individual-level exposure to violence but: no information on cause for husband’s death  we cannot distinguish genocide / HIV / accident widows assumption: widowhood is not a choice variable Fig. 3: Distribution of marital status by age group

9 9 Conflict measure 3: Sex ratios captures the aggregate effect of conflict-related male mortality calculated from census data: census 1991 (matched with DHS 1992) census 2002 (matched with DHS 2000 & DHS 2005) per 5-year cohort and province Fig. 4: Sex ratio by age group

10 10 Conflict measure 4: Proportion of Tutsi population captures province-level conflict intensity calculated from DHS 1992 in regressions: provincial dummies used Fig. 5: Proportion of Tutsi population by province

11 11 Estimation approach Determinants of fertility dependent var: number of children ever born to a woman poisson regression model potential endogeneity problem: education, child mortality sub-sample of women of age 44-49 years in every survey wave reference group in peace conflict-exposed group 1 conflict-exposed group2 age 38-43 DHS 1992 DHS 2005 DHS 2000 1994 genocide age 33-38 time

12 12 Descriptive evidence: Trends in fertility

13 13 Determinants of fertility (poisson) (I)

14 14 Determinants of fertility (poisson) (II)

15 15 Further steps So far, analysis compared drivers of fertility within the oldest age group of women Next steps: Expand the sample to all prime age women to compare the impact of conflict across time and across women of different age during the genocide but: interpretation of coefficients becomes more deliberate, while it is hard to disentangle the pure impact of genocide Include proxies for conflict-related health problems to differentiate between trauma-effect and replacement Explore the determinants of age at first birth and spacing between birth

16 16 Thank you!


Download ppt "Rwanda: The impact of conflict on fertility Kati Schindler & Tilman Brück Gender and Conflict Research Workshop 10/06/2010."

Similar presentations


Ads by Google