Search

We propose families of measures for vulnerability to both individual and aggregate poverty. We view vulnerability as the burden caused by the threat of poverty. Exposure to states of the world where hardship may strike causes distress ex-ante, even if ex-post the dice are benign and poverty (consumption below a critical norm) does not materialise. We ensure our measures meet a set of desirable axioms-in the case of aggregate vulnerability, this set includes sensitivity to the degree of correlation of outcomes across individuals. We argue that individual vulnerability levels are not enough to compose the aggregate picture, e.g. because society may be wary of the threat of simultaneous, widespread poverty. We illustrate with data from Ethiopia, where rainfall risks are found to raise vulnerability to poverty despite sustained increases in consumption. [ABSTRACT FROM AUTHOR]

This article investigates whether public investments that led to improvements in road quality and increased access to agricultural extension services led to faster consumption growth and lower rates of poverty in rural Ethiopia. Estimating an Instrumental Variables model using Generalized Methods of Moments and controlling for household fixed effects, we find evidence of positive impacts with meaningful magnitudes. Receiving at least one extension visit reduces headcount poverty by 9.8 percentage points and increases consumption growth by 7.1 percentage points. Access to all-weather roads reduces poverty by 6.9 percentage points and increases consumption growth by 16.3 percentage points. These results are robust to changes in model specification and estimation methods. [ABSTRACT FROM AUTHOR]

Much of the literature on consumption smoothing and on risk sharing has focused on the ability of the household as a unit to protect its consumption. Little is known about the ability of individual members of the household to keep consumption smooth over time or relative to other members of the household. We use data on adult nutrition in Ethiopia to investigate whether individuals are able to smooth their consumption over time and within the household. We find that poorer households are not able to do so. Furthermore, poor southern households do not engage in complete risk sharing; women in these households bear the brunt of adverse shocks. This result implies that the collective model of household organization, which imposes Pareto efficiency on allocations, is rejected for these households. Finally, we obtain estimates of the relative Pareto weights in household allocation. We find that a wife's relative position is better if customary laws on settlements at divorce are favorable or if she comes from a relatively wealthy background and that poor southern women have lower Pareto weights in allocation. [ABSTRACT FROM AUTHOR]