Swedish-born return immigrants by parents country of birth
Andreas Raneke, Statistics Sweden
The immigration and emigration in Statistics Sweden’s national population projection are projected for seven birth-groups, of which Swedish-born constitutes one. To estimate the future re-immigration of Swedish-born, a population living abroad are created by adding emigrated persons and subtracting re-immigrants. This population are then used in a linear regression model applied to re-immigration rates and information about emigration three years earlier to project the re-immigration. The model does not consider the background of the Swedish-born living abroad, i.e. their parents country of birth. In the present model for the future re-immigration, irrespective of their background, all Swedish-born have the same propensity to re-immigrate. Since the estimations of re-immigrants have been biased in the latest projection is it useful to see if there are any diversities in re-immigration depending on the parents background. The aim of this study is to further develop the model that predicts the re-immigration of Swedish-born. Parents country of birth are added to the population living abroad and re-immigration rates are estimated for Swedish born with; two Swedish-born parents, two foreign born parents and one Swedish-born and one foreign born parent. The results shows that Swedish-born with two Swedish-born parents have the lowest rate to emigrate and the highest rate for return immigration. This leads to a slowly increasing number of emigrants and also an increasing re-immigration. Swedish-born with two foreign born parents have the highest propensity to emigrate but doesn’t return to Sweden in the same extend as those with two Swedish-born parents. This leads to an increasing number of persons with two foreign born parents living abroad. However, compared to the latest population projection, including parents background doesn't give any considerable differences in re-immigration during the projected period. Therefore it might not be necessary to include this information in the population projection.
Presented in Poster Session 3