ABSTRACTMaurice Scott has argued that the neoclassical production function and growth accounting are fundamentally flawed as tools for understanding the growth process. If the role of capital were correctly evaluated, then (he argues) the famous ‘residual’ of growth accounting would disappear. Contrary to these claims, this paper seeks to show that growth accounting gives correct answers to interesting questions, even when all technical progress is embodied in new capital goods and even when depreciation is entirely due to obsolesce