The objective of this paper is to estimate the impact of regional innovation on employ-ment and population growth in the U.S. Midwest. A measure of effective local innovation is created by fitting a first-order spatial autoregressive model to create a measure of regional innovation using patents. In the second stage, employment and population growth are explained as a function of effective innovation plus other growth related characteristics using a spatial error model. Using a spatial bootstrap routine to infer parameter significance, we find local innovative activity is positively associated with employment growth during the period 1990-2005, although rural and agriculturally dependent counties remain at a disadvantage. Population growth is also positively associated with effective innovation, although to a lesser degree.