SHOCK METRICS FOR RESILIENCE ANALYSIS: AN INVESTIGATION OF THE RELATIONSHIP BETWEEN SUBJECTIVE AND OBJECTIVE INDICATORS

Abstract

As shocks become more prevalent, virulent, and widespread, their adverse consequences are expected to disproportionally affect those who are most vulnerable. The recent interest in resilience highlights the need for more precise measurements of shocks. This study investigates the extent to which subjective and objective measures of drought shocks are correlated. It also attempts to determine the factors, beyond “objective” measured rainfall, that determine households’ self-report of drought. First, I investigate whether type of employment, irrigation sources, and socioeconomic status explain the heterogeneity in drought reports. Second, I estimate the effects of the timing of interview on the probability of reporting a drought and the extent to which shock reports are path dependent. Using a panel of 1,298 households from India, the analysis reveals a negative but imperfect correlation between self-reported drought and measured rainfall. I find that wealthier households are more likely to report drought at low levels of rainfall. While rain dependent occupations are positively correlated with drought reports, I find no evidence of correlation with type of irrigation. The timing of survey interviews proves to be important: households interviewed during the rainy season are less likely to report a drought. I also find evidence for the path dependency of self-reports.