Induced Innovation in U.S. Agriculture: Time-series, Direct Econometric, and Nonparametric Tests

The hypothesis of induced innovation is tested for U.S. agriculture using a high-quality state-level panel data set and three disparate testing techniques – time series, direct econometric, and nonparametric. We find little support for the hypothesis. That conclusion is robust across testing techniques. However, as with all empirical tests of this hypothesis conducted to date, ours focus only on the demand side of the hypothesis. The hypothesis could have been rejected simply because the marginal cost of developing and implementing input-saving technologies for the relatively expensive inputs is greater than for the relatively cheap inputs.