| Abstract: | What are the implications of the historically observed economic policy instability in Latin American countries (LACs) for macroeconometric testing? Two pressing restrictions on the econometrician using time-series of LACs arise: time-varying parameters and time-varying specifications. Such an instability also has profound impacts on time-series measurements of national accounts at constant prices. This, together with the “second best methodology” used in LACs for computing real GNP, implies that LACs figures on GNP growth reflect growth in gross production rather than in value added. LACs time-series for private consumption are unreliable. Crucial data set constraints in LACs further complicate the task for the econometrician. |
