Social Mobility and Immigrants or Immigrants and Social Mobility
回应Raftery的评论，澄清了关于四格表数据错误和移民变量遗漏的争议，指出移民并非影响社会流动的既定因素，而是被不同国家吸引的人口。
Raftery raises three issues in his comment. First, we got some of our four-fold tables wrong-including farm fathers when we said we didn't. He seems to be right-and for two populations, the Philippines and Puerto Rico, our errors are consequential. Our Canadian error turns out not to be so. The difference between his and our collapsing of the Hungarian data can well be seen as a matter of honest disagreement-one on which we are willing to yield if he sees the matter as important. For reasons he does not indicate, he also gets a different value for mobility in Yugoslavia. In general our measures and his are in agreement and, as he reports in his equations 1 and 2, one gets rather the same results with one set as with the other. The second issue Raftery raises has to do with an additional variable: we did not talk about immigration. He is not entirely correct here. We devoted two columns of text to two asides noting that the four countries with the most circulatory mobility have histories of providing permanent homes to unusually large numbers of immigrants relative to their sizes. We presented zero-order correlations between structural variables and percent foreign born and speculated about their interpretation. We did not pursue the matter in the article because we could not plausibly include immigration or percent foreign born in a causal logic promoting differences in circulatory mobility. We saw immigrants not as a given that influences how a social order functions, but as people attracted differentially to countries. In this context the dependent variable of our original article, circulatory mobility, is inde-