We illustrate by means of a dynamic research and development race that, while at some points in the race, social incentives and private incentives may coincide, at other points they may diverge-- too many researchers remain that race.

Based on a model of perpetual growth through optimal search for better technologies, calibrated on time series data form the Israeli economy, it is shown that capital subsidies produce a definite gain in expected growth rates, but those gain are invariant to the particular restrictions...