Child stunting in growth currently affects 164 million children globally, and has dire consequences for the future well-being of the affected children. Wars disproportionately affect children and is believed to raise levels of stunting due to malnutrition and diseases. Using the 2002 Eritrean Demographic and Health Survey, this thesis adopts a differences-in-differences methodology and finds that the 1998-2000 border war between Eritrea and Ethiopia raised levels of stunting in affected regions by 12 %, which is more than a quarter of the non-conflict level. A second investigation studies idiosyncratic war-related shocks, and does not find any significant and negative effects on child stunting from mobilisation, war-related deaths among male family members, and displacement still ongoing in 2002. Thirdly, this thesis finds that the levels of stunting among children in the Debub region were significantly more severely affected by conflict if living closer to a main road, indicating that dependency on trade with Ethiopia was an important risk factor.
The analysis of coping strategies in this thesis further develops the idea that the conflict-stunting relationship is asymmetric and highly influenced by the options available to households in mitigating the effects of negative shocks. An analysis is provided on how households adapted production patterns, asset holdings and fertility to cope with the constraints created by the conflict, and how they made use of supportive networks. Of particular interest is the role of Eritrean women in mitigating the effects of war. Eritrean women have to a significant degree substituted for the labour of men in general, and also for mobilised men. The level of fertility during the 1998-2002 time period was lower in families that were not able to secure the basic needs of their last-born child still alive, of which stunting is a manifestation. Mothers of stunted children have furthermore wanted to postpone further childbearing which implies that a significant part of the reduction has been a desired adjustment for the mothers, yet they have not necessarily reduced their overall fertility desires. Consistent with economic theory, community assistance or mutual insurance systems do a much better job in mitigating crises with repeated risk exposure and when facing idiosyncratic risk than in an extraordinary conflict situation. Stunting due to war is significantly more dependent upon the ability to convert own wealth into emergency consumption and on growing own crops than stunting due to other causes. These findings point to a breakdown in social contracts during war, which is particularly problematic for the most vulnerable.