This paper addresses the question of whether or not a theory of total factor productivity (TFP) is needed in order to explain the documented large per capita income differences across countries. As the argument that it is needed has been reached by calculating TFP empirically, we show that the way the estimates of TFP have been computed is not an innocuous issue. To prove our point, we discuss how two well-known textbooks on growth theory present the arguments and the problems associated with these expositions. We conclude that the tautological nature of the estimates of TFP lies at the heart of an important question that the empirical literature on economic growth has been dealing with during recent years. Hence, our arguments cast doubt on the need for a theory of TFP.