[Intermarriage and Social Distance Among U.S. Immigrants at the Turn of the Century
Author(s): Deanna L. Pagnini and S. Philip Morgan
Source: The American Journal of Sociology, Vol. 96, No. 2 (Sep., 1990), pp. 405-432 Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2781107]

The pattern of assortative mating among European immigrants and native whites is examined by ethnicity and generation using a na*tional sample drawn from the 1910 census manuscripts and a sam*ple of marriages registered in New York City between 1908 and 1912. The pattern of assortative mating is virtually identical in the two data sets. Endogamy was strong for all groups examined, but was castelike for the "new" ethnics from eastern and southern Europe. Marriages between "old" and "new" ethnics were espe*cially rare. The pattern of ethnic intermarriage was nearly identical for men and women. [. . .] While the existence of ethnic and generational endogamy at the turn of the century is not surprising, its strength has not previously been estimated with appropriate statistical tech*niques. Further, these techniques reveal more detailed features of the pattern of assortative mating that are not well known and pro-vide important facts for theories of immigrant assimilation and as*sortative mating.

[. . .] This im*migrant flow included a number of what historians termed "new" immi*grants. These included immigrants from southern and eastern Europe: the Italians, Russians, Poles, and Jews, who had not emigrated in large numbers before the 1880s. These groups were very different from the "old" British, Irish, Scandinavian, and German immigrants in appear*ance, customs, and demographic characteristics (Lieberson 1980; Higham 1963; Abrams 1973; Kraut 1982). [. . .] The groups most inclined to congregate in the cities were the Jews, Irish, Poles, and Italians. [. . .]

While physical distance was a factor affecting mate selection, social distance was important also. The ability to speak English, which allowed for interaction with those of other ethnic groups and with natives, was an important aspect of social distance. The new immigrants were the least likely to speak English (table 2, col. 2), especially the Poles and Italians (see Cheney 1988).4

Further social distance arose from the xenophobia of the native whites. Native whites perceived the new immigrants as a threat to the old Ameri*can order and as a possible contamination of the white American race. While each immigrant group had its own stereotyped reputation, the new immigrants were always perceived as the worst. The new immigrants were seen as inferior in looks, morals, and intellect (see Hall 1915, pp. 69-70). According to a sociologist at the time, that the new immigrants "are morally below the races of Northern Europe is as certain as any social fact" (Ross 1915, p. 74). [. . .]

Xenophobia, coupled with the high birthrate of the new immigrants and the low fertility of the native whites, led to the fear that the American race was being diluted by "unworthy bloods." Many believed that "the peoples of the Mediterranean region were biologically different than those of northern and western Europe and that the differences sprang from an inferiority of blood and could be observed in certain social characteris*tics" (Handlin 1957, p. 77). Ross stated that "it is fair to say that the blood now injected into the veins of our people is sub-common" (1915, p. 73). Those concerned over the intermingling of blood clearly opposed inter-marriages. The fact that some immigrants were seen as providing good genetic stock while others were undesirable provided ideological support for a range of sanctions against intermarriage. Resistance to intermar*riage may have come from within the immigrant groups as well, as the older generation tried to maintain the traditional culture.

[. . .]

Net of the social distance parameter between old and new immigrants, an examination of residuals revealed several additional areas of special affinity. First, as mentioned previously, there was a much greater ten*dency for British and Irish immigrants to marry each other than other-wise implied by model 2 (such a match was over five times more likely than one with other old immigrant groups [. . .]

The next pattern of attraction crosses the new-old ethnic boundary: Germans were much more likely to marry Poles than any other new immigrant group. Furthermore, Germans were the only old immigrant group especially likely to marry Poles. If national data are used, German-Pole marriages were 5.5 times more likely than any other new-old immi*grant match. The greater ethnic detail in the NYC data allows us to say that disproportionate numbers of these German-Pole marriages are with Poles from Austria-Poland, not those from Russia-Poland.

The third pattern of attraction is between the central and eastern Euro*pean Jews. This is the strongest pattern of intermarriage by far. In the national data, eastern-central European Jewish intermarriages are close to 500 times more likely than any other Jewish intermarriage. In the NYC data the effect is roughly two times larger-Jews from central and eastern Europe were over 900 times more likely to marry among themselves than to marry other groups. Thus, while there was social distance between these two groups of Jews, marriages with other Jews were clearly more common than marriages with non-Jews. Attitudinal data presented by Bogardus (1928, table 4) from native-born Jews are consistent with the actual marital data we present. He demonstrated that Jews clearly pre*ferred to marry other Jews.

The final two zones of attraction also involve Jews but are different in two respects. First, they are only fitted to the NYC data; there were too few Jewish intermarriages in the national data to detect these patterns. Second, these are the only asymmetric effects fitted in either table. To explain, a Jewish husband (from central or eastern Europe) was less likely to marry an Italian or Polish woman than a woman from the old immi*grant stock (i.e., a British, Scandinavian, or German woman). Net of other effects, Jewish men were only about .3 times as likely to marry these non-Jewish new ethnics as those from the old ethnic groups. The result is that the old-new social distance effect is partly neutralized for Jewish men. Again, such preferences for old stock groups, following the prefer*ence for Jewish intermarriage, can be seen in Bogardus's (1928, table 4) attitudinal data. For "Americans," new immigrants, and especially Jews, were the least-desirable spouses. For Jews, the old immigrants were the first preference as intermarriage partners after other Jews. The NYC data suggest these differential social distances are canceling one another out. We are unable to explain why this effect is asymmetric, holding only for Jewish men. But perhaps successful Jewish men were able to marry "up" in status while Jewish women were less able to do so. The final effect operates in the opposite direction for eastern European Jewish women. They were three times more likely to marry Italian men than Polish men (or, net of other effects, men from old immigrant groups). Perhaps Jewish women, unable to marry "up" as Jewish men did, "settled" for Italian men who far outnumbered the available Italian women. Detailed tabula*tions (using the NYC data) not shown here reveal that these Jewish-Italian marriages most often involved men from southern Italy and Jew*ish women from Hungary and Russia. Both of these asymmetric effects operate regardless of spouses' generations.

[. . .] the greater endogamy for new (compared to old) immigrants that we document still persists, although endogamy is much less extreme for all white ethnics now. In the contemporary period, Lieberson and Waters (1988) partly attribute this greater endogamy of the new immigrants to less intermarriage in the past, giving these groups less highly mixed an*cestral categories. Our paper shows which groups intermarried at the turn of the century, yielding more mixed groups. An important difference in the historical and the contemporary pattern lies in the strength of the old-new distinction. Marriages across this old-new distinction were quite rare in 1910. Historians of the turn of the century, and writers at the time, described the discrimination, differences, and social distance between the old and new immigrant groups. In large measure, our analysis of data on assortative mating fits well these characterizations of intergroup rela*tions. This old-new distinction is much less visible now (see Alba and Golden 1986), a change that reflects the socioeconomic integration of these groups (see Neidert and Farley 1985; Lieberson and Waters 1988).

[. . .] Lieberson and Waters show that intermarriage rates between two groups are higher when the proportion of Catholics in the two groups is similar. This result implies that ethnic intermarriages are frequently religiously endogamous. We found this to be true at the turn of the century for Jews, and possibly for Poles and Italians. But we found no heightened tendency for Irish to marry Poles and Italians-a straightfor*ward expectation if religion played a dominant role in assortative mating. Other work (see Peach 1980a, 1981) has challenged the claim that there was a Catholic melting pot including Irish, Poles, and Italians. The ab*sence of any special affinity between these groups illustrates the over*whelming power of nonreligious factors reflected in the old versus new distinction. These factors included (1) high levels of residential and occu*pational segregation, (2) period of immigration and the selectivity operat*ing in different immigrant streams, which produced ethnic socioeconomic differentials, and (3) great social distance between groups supported by an ideology heavily laced with xenophobia.