The Geographic Concentration of Knowledge : Scale, Agglomeration and Congestion in Innovation Across Us States

The endogenous innovation research program is one of the most important theoretical agendas in modern growth theory. Progress, however, has been slowed due to an apparent empirical anomaly. In their simplest and most appealing forms the models suggest a scale effect that seems inconsistent with the data. We argue that existing empirical tests of the theory do not account for the importance of space. More importantly, they do not use a measure of innovation as a dependent variable. Our formulation directly tests the prediction of scale effects and the relative importance of agglomeration effects versus congestion effects in influencing the rate of innovation. If scale effects are viewed through the agglomeration effects, as suggested by our model, then there is strong evidence that rates of innovation are significantly higher when resource constraints are relaxed through a greater population level