COVID-19 water and electricity subsidies in Ghana: How do the poor benefit?

Abstract

This paper investigates the extent of welfare distribution amongst households in Ghana due to the government's subsidies on electricity and water as part of its relief package to households on the fallout of the COVID-19 pandemic. The study seeks to induce policy discussions on effective targeting and the consequential welfare distribution of subsidy interventions in Ghana. Subsidising the price of basic household necessities is expected to immensely benefit the poor whose expenditure is largely occupied by these necessity goods, but, as pertains in developing countries, schemes for effective targeting of the poor are nearly non-existent. This mostly leads to misalignments in income redistribution that is eventually anti-poor. Data from Ghana Living Standard Survey (seventh wave) which captures household final expenditure on goods and services in 2017, including electricity and water, is used for the analysis. First, we simulate household consumption changes as a result of the pandemic on the 2017 data, using on real-time information in the course of the pandemic on household income and expenditure changes. Next, we estimate a complete demand system for all households, and their elasticities of demand for electricity, water and food using the Quadratic Almost Ideal Demand System model. Using the computed elasticities, Taylor's approximations are then used to simulate consumer welfare in the form of compensating variation (CV) from price changes due to government subsidies. Our results indicate that the imposed subsidies promote consumer welfare of all households, especially, households living in rural areas. The blanket subsidy on water better identified the poor and the lifeline households than the subsidy on electricity which had a yardstick for targeting. However, we find that the subsidies on the household utilities are regressive in the sense that lifeline consumer households and the poorest households in both urban and rural areas are rather less advantaged in welfare redistribution in both subsidies.