A0900
Title: Testing linearity of network interaction functions
Authors: Francesca Rossi - University of Verona (Italy) [presenting]
Abhimanyu Gupta - University of Essex (United Kingdom)
Jungyoon Lee - Royal Holloway, University of London (United Kingdom)
Abstract: A computationally straightforward test is proposed and theoretically justified for the linearity of a network interaction function. Such functions commonly arise due to practitioner-imposed specifications or optimizing behavior. The test is nonparametric but based on the Lagrange Multiplier principle. This entails estimation only under the null hypothesis, which imposes a familiar, easy-to-estimate linear spatial autoregressive model. Monte Carlo simulations show excellent size control and power. An empirical study with Finnish data illustrates the test's practical usefulness, shedding light on debates in the public finance literature on the presence of tax competition among neighboring municipalities.
