The Inherent Bias of Bad Models

Gross domestic product, an official measure of a country’s economic activity, is a useful metric. Governments can get a sense of likely revenue from taxation, for example, and investors can evaluate promising markets. It’s also helpful for planning: a growing GDP suggests an increasing demand for energy, since GDP tends to correlate with energy consumption.

But in sub-Saharan Africa, write Michael Dioha, PhD and Rose M. Mutiso, Ph.D. , “the usual methods of calculating GDP frequently underestimate true economic activity, especially where informal employment is common.” When GDP is recalculated to account for some of these informal economic sectors, the results can be dramatic. In the case of Nigeria, revising the GDP revealed that the true number was 60% higher. That meant that previous forecasts of energy demand for the country were likely too low.

These and other faulty models for understanding sub-Saharan Africa, argue Dioha and Mutiso, “perpetuate bias in scenarios that greatly undervalue future energy demand and emissions, resulting in lower investments.” They propose bringing more African analysts and institutions to the task of developing the energy models that domestic and international organizations use to advance electrification and development goals on the continent.

Read more about the importance of bringing expertise, data, and model development “home” to African countries.