Project CEMAPRE internal
| Title | Nonlinear models with spatial dependence |
| Participants | Luís Silveira Santos (Principal Investigator) |
| Summary | The present project aims to contribute to the spatial econometrics literature, regarding the specification and estimation of a parametric model for fractional dependent variables on the interval [0,1]. In recent years, several authors discussed the importance of modelling cross-sectional dependence in spatial data, due to the interactions established between different units (regions, firms, …) inside the same country or a well-defined area. Ignoring such dependence structure may lead misleading economic interpretation and biased estimates. With regard to the complexity of introducing these spatial interactions in the estimation, innovative methods and algorithms will be introduced. The finite sample properties of a new estimator will be discussed based on an extensive Monte Carlo simulation study. Finally, these methods will be applied to study the determinants of the proportion of regional expenditure on high-tech R&D in the United States Metropolitan Statistical Areas. Unit-specific and spillover effects with be accounted for, in terms of social, economic and technological growth over the neighboring areas. |