Small Area Estimation of poverty using remote sensing data

Solving the problem of poverty begins with providing accurate data to describe the population in poverty. The use of small area estimation (SAE) to estimate poverty in a small area is increasingly needed to get better poverty monitoring and policy making. The success of SAE depends on auxiliary variables used in a model to produce appropriate estimates with smaller variances. Numerous studies have shown that remote sensing data, with its various advantages, can serve as covariates in small area models. However, research on using several remote sensing data to estimate poverty, particularly for very poor households whose per capita monthly expenditure is less than 0.8 of the poverty line, is still limited. This research aims to investigate the potential of remote sensing data as auxiliary variables in a small area model to estimate expenditure per capita of very poor households in West Java, Indonesia. The method used in this research is Small Area Estimation using Fay Herriot Model with Transformation. Here we show the comparison between small area models using administrative data and remote sensing data. Both models produced estimation for unsampled areas and more accurate estimation than direct estimates for sampled areas. Although the relative standard error is slightly higher than that of the administrative data model, remote sensing data is preferable due to its lower cost, comprehensive coverage, and frequent updates. We found that nighttime light data as a useful auxiliary variable for expenditure per capita estimation of very poor households. Overall, we show that there is a potential of using remote sensing data as auxiliary variables in SAE of poverty. These findings can provide a detailed view of poverty, helping the government to implement policies aimed at reducing poverty and improving conditions for very poor households.