Investment-saving regression misspecification leads to biased saving-retention coefficient.
The article tests for spatial autocorrelation in panel data models using different statistics. These tests help identify if there is a spatial pattern in the data that needs to be considered in the analysis. The researchers suggest five different tests, with the first one detecting any type of spatial autocorrelation. The other tests help determine the specific type of spatial autocorrelation present. By applying these tests to real data, the researchers found that omitting spatial autocorrelation in the analysis can lead to biased results, particularly in the context of investment-saving relationships.