In the early 1980s, an acknowledgement of the economic failure of the African states, which was statistically verifiable through poor growth figures, came alongside the liberal reforms implemented in Africa under the auspices of international financial institutions. The failure appeared all the more obvious given that many African states had moved from strong growth at the end of the colonial period to stagnation in the 1960s and recession in the 1970s. This article aims to reveal the origins of this statistical evidence that emerged in the context of a liberal turn. It traces the history of growth figures in Africa and considers the actual conditions under which they were compiled by statisticians, their successive meanings, and their different uses by contemporaries. It shows that the statistical grand narratives on the failures of the African states involved reusing figures that were considered to be uncertain at early stages of development and changing the meaning of those figures. From the late colonial period, the figures continuously provided conflicting interpretations and several reinterpretations, before formalizing the failure of the African states and the need for liberal reforms in the early 1980s. In the end, statistics appears not only as a technique for government and legitimization, but also as a means by which different eras can project into each other by conferring present meaning to past—or future—figures.