We investigate the strength of the Penn effect in the most recent version of the Penn World Tables (PWTs). We find that the earlier findings of a Penn effect are confirmed, but that there is some evidence for nonlinearity. Developed and developing countries display different types of nonlinear behaviors. The nonlinear behaviors are likely attributable to differences across countries and do not change when additional control variables are added. We confirm earlier findings of large RMB misalignment in the mid-2000's, but find that by 2011, the RMB seems near equilibrium. While the Penn effect is quite robust across datasets, estimated misalignment can noticeably change from a linear to a nonlinear specification, and from dataset to dataset.
↧