The notion of “economic development” dominates aid policy. A nation’s per capita Gross Domestic Product (gdp) determines whether that country is considered developed or less developed, and the standard measure of any developmental progress is gdp growth. This article investigates the evolution of the concept of economic development, as it emerged from a specific combination of modeling, statistics and political circumstances. In this story, Arthur Lewis played a decisive role, but only by building upon Colin Clark’s first global national income statistics, an indispensible foundation for Lewis’s seminal model of economic development. This model was embraced by policy makers longing for a theoretical framework to clarify and operationalize the hitherto vague concept of development. More importantly, however, the statistical indicator on which Lewis based his theory had already been universally accepted. In other words: statistics came before theory. This holds important lessons for alternative development ideas. It explains why the idea of economic development remains so firmly entrenched and suggests the conditions that might be necessary for an alternative theory to take hold.