Association for Public Policy Analysis & Management

This paper provides estimated impacts of the line extension activity using data from over 10,000 households in 360 communities. We use a quasi-experimental difference-in-differences design with a matched comparison group. The intervention group households were randomly sampled from 178 communities that were in turn randomly selected from the over 300 communities targeted to receive new lines. We identified the comparison group of households using propensity score matching in three stages. The first stage used census and global positioning system data on thousands of communities to identify 546 that then participated in a baseline community survey in 2011 along with the 182 intervention communities. The second stage used the baseline data to narrow the 546 potential comparison communities down to a set of 182. The third stage used baseline household and community survey data to match the households in our 182 comparison communities to those in the 178 intervention communities. Outcome data were obtained from a follow-up household survey administered 20 to 34 months after the electricity lines were built.

We estimate impacts of extending electricity lines on household outcomes in seven domains: connection rates; energy use; educational and child time use; pollution, health, and safety; business activity and adult time use; economic wellbeing; and household mobility and composition. In our analyses we control for a rich set of baseline characteristics including both the baseline value of the outcome and variables designed to account for the process used to identify the high-priority communities targeted to receive new lines funded by MCC. In addition, we carry out an exploratory analysis of the impacts of actually connecting to grid electricity using methods similar to those used to estimate impacts of the line extensions, and testing the robustness of our results using a variety of instrumental variables. These methods cannot fully address all concerns about self-selection, but the analysis still provides valuable information for assessing how connecting to the electric grid influences household outcomes 1.5 to 3 years after connecting.

The paper adds to a small but growing literature providing rigorous evidence of impacts of electricity on household outcomes. With the 2013 announcement of the Power Africa initiative by the U.S. government, under which the U.S. and African governments, multilateral and bilateral development partners, and private businesses are expected to invest about $32 billion in energy sector projects in a number of African countries, our findings provide critical information for policy discussions.