This paper develops a gravity model of immigration. Tests of the model using panel data for 16 OECD countries for 1991-2000 confirm the model's high explanatory power, and examples illustrate its usefulness for testing other hypothesized determinants of immigration.

Measures of national product can be misleading because there is nonmarket production. There are also distortions due to transactional activities, which are expenditures to support transactions, not actual output consumed. For 1950-89, this study recalculates output for the United States,...