Research Methods, Econometrics, & Statistics

How does heterogeneity over space and time affect econometric estimates and forecasts? Should an econometrician use the least aggregated data possible? The papers in this session answer these questions in the context of crop rotation, land development, and commodity prices.

Misspecified models can produce spurious econometric results. The papers in this session use nonparametric methods to avoid misspecification when modeling yield distributions, estimating willingness to pay, and imputing missing data.

Panel/longitudinal data allow researchers to undertake analysis which is not possible with cross-sectional analysis and the relevance of the research for policy and program development is enhanced with panel data. In an era of tight budgets having relevant panel data for policy and programs is essential. The presentations in this session highlight panel data usefulness and statistical issues, the use of state farm business management program panel data sets, the use of panel data for policy in Canada, and specific research applications using panel data sources.