This paper estimates the effects of federal research funding on research outcomes at 68 research universities. We provide a new interpretation of the instrumental variable estimate of the coefficient in a regression of the output of an institution on an input. Absent parameter heterogeneity, it captures the total change in output when an institution obtains an additional unit of the input that may be correlated with the other inputs that affect the output measure. Our instrument for research funding is alumni representation on U.S. Congressional appropriations committees. Our results suggest an increase of $1 million in federal research funding (1996$) to a university results in 10 more articles and 0.2 more patents. The change in citations per article is negative but very small and imprecisely measured. As a first approximation, increasing federal research funding on the margin results in more, but not necessarily higher quality, research output.