Poverty and Weather Shocks: A Panel Data Analysis of Structural and Stochastic Poverty in Zambia

While it is generally accepted that climate change will exacerbate poverty for small and medium sized farmers in Sub-Saharan Africa (SSA) over the coming years, at least due to rising variability and rainfall shocks (Mulenga, Wineman, and Sitko 2017; Hallegatte et al. 2016), a number of questions remain unanswered. Which types of the poor are more exposed to climate risk and how do the impacts of climate and weather shocks vary across stochastically and structurally poor households? Addressing these questions is crucial for improved targeting of interventions intended to build the resilience of smallholder farmers. Smallholder farmers’ reliance almost entirely on rain-fed agriculture and their limited capacity to cope with weather shocks exposes them to climate risks. Weather shocks negatively impact smallholders through their effects on agricultural productivity, which is the mainstay of rural smallholder households. If left unchecked, weather shocks can lead to increased poverty incidence and intensity. In this paper, we utilize data from a nationally representative two-wave panel of recent agricultural household surveys to conduct a high resolution analysis of the spatial distribution of poverty, and how the different types of poverty are impacted by exposure to climate change variability. The data allows us to (a) control for observed and unobserved sources of household heterogeneity, and (b) distinguish between the structurally poor, i.e., those households that have very little assets or savings, and the stochastically poor, i.e., those households that have low savings but enough assets that they could liquidate if necessary to smooth consumption during a climate shock Out of the 14,508 rural households interviewed in Zambia in 2012 and 2015, about 51% were structurally poor (low income and assets) and 5% were stochastically poor (low income and high assets). About 23% of households that were structurally poor in 2012 remained structurally poor in 2015, hence, chronically poor. A third of the structurally not poor in 2012 fell into poverty in 2015, while about 19% of poor households in 2012 managed to escape poverty in 2015. Structurally poor households in Zambia are more exposed to drought risk. Lower than normal rainfall, as measured by a negative precipitation index, significantly increases the probability of being structurally poor by 2.3 percentage points. Three implications follow from our findings. First, there is a need for well-structured and targeted social promotion programs to lift the viable but chronically and structurally poor and stochastically poor households from poverty. This can be achieved within the agricultural sector by using the electronic voucher delivery systems to better target large-scale, anti-poverty programs such as the farmer input support program. Along with improved targeting, the use of the electronic based voucher systems crowds-in private sector investments, which make available diverse inputs for farmers and also help develop the rural nonfarm sector where farmers can earn extra incomes. Smart-subsidies should be flanked by output market linkages and/or market development in order to enhance market participation and help improve incomes from agricultural production. Second, for those not commercially viable, there is a need for a better targeted and sustained social welfare program specifically meant for this group. Thus there is need for sustained social protection (e.g., social cash transfers) in order to prevent the non-poor from falling into poverty. And lastly, the intricate linkages among climate variability, climate risk, and poverty call for more support to enable farmers not only adapt to, but also mitigate climate change and variability. Such support may be v directed towards climate-smart agriculture adoption, autonomous and planned adaptation, improved extension, and climate information services.