...there is a 'striking and easily observable' clustering of immigrants in high-benefit states.
In Immigration and Welfare Magnets (NBER Working Paper No. 6813) , NBER Research Associate George Borjas concludes that "the generous welfare benefits offered by some states have magnetic effects and alter the geographic sorting of immigrants in the United States." He finds that there is a "striking and easily observable" clustering of immigrants in high-benefit states, such as California. This is especially true for immigrants receiving welfare as opposed to those who are not. Natives do not cluster in the same way, perhaps because they find it expensive or costly in other ways to move from one state to another. Immigrants have already decided to make the costly move to the United States, and then must only decide which state is most advantageous to them. The extra cost of reaching that state may be small.
To reach his conclusion, Borjas uses data from the 1980 and 1990 censuses, looking at where immigrant households (those whose heads are resident aliens or naturalized citizens born outside the United States) have settled and making comparisons with native households (those with heads born in the United States). In classifying households, Borjas determines if a household received Aid to Families with Dependent Children (AFDC), Supplemental Security Income, and general assistance in the year prior to the census. The census data does not include information on Medicaid and Food Stamps. His calculations take account of the general trends in welfare.
California, Borjas notes, has become relatively more generous in its welfare provisions. In 1970, California benefits were at the median, with as many states giving more as those giving less. By 1990, California's AFDC benefit package was almost the most generous in the nation. It was 20 percent larger than that provided by New York; 89 percent larger than the one in Illinois, and almost 280 percent greater than that offered by Texas. The increasing relative generosity of California's welfare system appears to have had an impact.
In 1990, California was home to 9.6 percent of U.S. natives who did not receive welfare and 11.5 percent of U.S. natives who did. It was also home to 27.6 percent of the nation's immigrant households that did not receive welfare and 37.6 percent of immigrant households that did.
If only those immigrant households whose head has arrived in the United States five years prior to the census are included, the clustering becomes even clearer. Some 45.4 percent of recent immigrants receiving welfare live in California, as compared to only 28.9 percent of those recent immigrants who do not receive welfare. Much of this clustering is because less-educated immigrants are more likely to live in California than less-educated natives. This is true even within groups of immigrants from a specific nation.
The same clustering is shown in the numbers for recent female-headed immigrant households with children. Borjas also determines that the clustering in California holds if immigrants from countries sending large numbers of refugees to the United States are excluded and if immigrants of Mexican origin are excluded. So the clustering in California can't be said to be entirely the result of California being adjacent to Mexico, or merely a favorite location for refugees. Further, his analysis indicates that clustering is not attributable to ethnic enclaves in California that often help new immigrants get settled.
-- David R. Francis