Accelerating pace of climate change has made natural hazards more frequent across the globe, with developing countries bearing the brunt of the burden. In low-income settings, where social protection is limited and disaster-resilience systems are weak, households’ coping strategies often fail during climate shocks, resulting in substantial contractions of household income and consumption (Morduch, 1995, Townsend, 1995, Dercon, 2004, Fafchamps and Lund, 2003, Kazianga and Udry, 2006, Maccini and Yang, 2009, Asfaw and Maggio, 2018, Janzen and Carter, 2019). It is often assumed that everyone in a household perceives the shock similarly, but imperfect risk-sharing within household, which is commonly observed in poor households (Doss, 1996, 2001, Dercon and Krishnan, 2000, Duflo and Udry, 2004, Dubois and Ligon, 2011), may induce unequal welfare impacts among its members. Hence, this project attempts to address this question by exploring the impact of weather shocks on the allocation of resources within household, and consequently, on individual consumption and poverty in the context of Malawi.
This study mobilizes recent extensions of collective household models (Bargain and Donni, 2012, Dunbar et al., 2013, Bargain et al., 2022) to recover individual shares of household consumption accruing to children, women and men, and to test whether resource sharing within household is affected by climate shocks. I combine four waves of Malawi Integrated Household Survey with geocoded rainfall data from Vicente-Serrano et al. (2010) to identify households exposed to a rainfall anomaly (flood or drought) during agricultural growing seasons of 2009/10, 2012/13, 2015/16 and 2018/19. In the estimation model, this indicator of rainfall shock with spatial-temporal variation is originally introduced as a determinant in resource sharing functions, along with fixed effect terms to control for spatial and time-related differences among households. Initial estimation results show that after an exposure to a rainfall shock during recent growing season, women tend to get lower share of household resources, with no effect on children’s shares. This pattern of within-household redistribution makes men’s consumption and poverty relatively more incompressible in times of a rainfall shock, while exacerbating the negative welfare impact of climate hazards for women. Heterogeneity analyses with respect to gender gap in employment suggest that the redistribution of resources from women to men after a climate hazard is likely driven by ‘life-boat’ ethics, that is, nourishing the members with higher marginal productivity and potential to bring money to the household in times of hardship (Pitt et al., 1990). On the other hand, cultural practices favoring women’s roles such as matrilineality or matrilocality seem to attenuate, to a certain extent, the adverse intra-household impacts of climate shocks on women.
Findings of this study will contribute to improving policies aimed at increasing disaster resilience of those who are the most vulnerable to climate shocks. The key policy implication rests on identifying and targeting the most vulnerable or ‘newly’ poor individuals by taking into account both inter-household and intra-household effects of natural hazards (Skoufias, 2003). In addition, integrating a set of observable indicators such as gender patterns of employment or cultural norms in proxy-mean tests will likely refine policy targeting to reach the most affected individuals.
