This paper identifies worsening R&D allocative efficiency as a potential driver of declining US economic growth. Within a simple endogenous growth framework, I develop a closed-form solution of the growth rate that can be decomposed into a frontier growth-rate, only achievable with the growth-maximizing resource allocation, and an allocative efficiency measure, measuring the gap between realized and frontier growth. Combining model with data on the innovation activity of US firms I estimate that allocative efficiency declined significantly from 1975 to 2014. Comparing the 1975- 94 period to the 2005-14 period, I find that declining allocative efficiency predicts 40% slower economic growth in the latter period, which can explain the entire concurrent decline in economic growth as documented in the literature. I discuss potential drivers of declining allocative efficiency including waning federal support of R&D, institutional and technological change, and increasing labor market power over inventors.