For a while jokers like Noel Ignatiev have been promulgating “whiteness studies” claiming that groups like the Irish were not initially considered “white”. My comment no longer appears at Reason, but I tried pointing out to Ron Bailey that even turn-of-the-century racialists who embraced the concepts of “Nordic”, “Alpine” and “Mediterranean” still considered European immigrants to be white. Lothrop Stoddard’s notorious book “The Rising Tide of Color Against White World Supremacy” (which you can read for free online) is quite explicit on that point. At the same time, what we consider salient depends on the context. So if the only people who can effectively vote are white men (an important legal distinction inclusive to these supposedly non-white immigrants), intra-white distinctions are going to be salient in politics. In other situations where people are homogenous in race and language they might divide over religion, as in Northern Ireland.

Recently I came across one of the rarest of things, an anonymous comment at the iSteve blog which is actually worth reading. It links to the American Journal of Sociology paper Defining America’s Racial Boundaries: Blacks, Mexicans, and European Immigrants, 1890–1945 by Cybelle Fox and Thomas A. Guglielmo. The paper is unfortunately gated, but anonymous provides an excerpt:

In stark contrast [to the black-white boundary], there was essentially no SEE-white boundary [SEE=Southern and Eastern Europeans]. Contrary to the arguments of many whiteness studies historians and the social scientists who have drawn on their work, we contend that wherever white was a meaningful category, SEEs were almost always included within it, even if they were simultaneously positioned below NWEs [=Northern and Western Europeans]. Some individuals and an occasional institution questioned—or appeared to question—the whiteness of SEEs and other Europeans, blurring the boundary in limited contexts. But the categorization of SEEs as nonwhite was neither widely recognized nor institutionalized. In fact, quite the opposite. Federal agencies including the census, the military, the immigration service, the Civilian Conservation Corps, and others all counted by race and placed SEEs firmly within the white category. No court ever denied Europeans the right to naturalize as free white persons at least in part because race scientists and the “common man” placed SEEs within the boundaries of whiteness. Furthermore, when SEEs saw Whites Only signs in movie theaters, restaurants, swimming pools, playgrounds, buses and streetcars, and at places of employment, they could—with near certainty—be confident that those signs were not meant to exclude them. Similarly, when housing covenants restricted the sale of homes to whites, when unions declared that their membership was restricted to white workers, when schools declared that their doors were open to white children only, and where marriage laws prohibited miscegenation, SEEs quickly learned that the category “white” included them, too.

So yes, Jim Crow really existed, and it did apply to blacks and not European immigrants. This also came up when a bunch of people got irritated at Fabio Rojas for writing about our “post-racist society“. It would be understandable if he said something extreme like “racism doesn’t exist”, but nobody says that (although if race didn’t exist, that’s what you’d expect). He was quite clear in his original post that he was saying that the end of legal sanction for explicit racism was a significant change, which I’d think would be hard to dispute. But perhaps for reasons of “mood affiliation” (and a better example than most of the time Tyler Cowen uses the phrase now), people got upset for him saying positive things about what improvements had happened rather than focusing what bad things exist now. Further back Mencius Moldbug and his acolytes tried to claim that after the civil war slaves were still sharecroppers, so their lot was not meaningfully different. Economic historians actually gather data on what folks earned back then, rather than relying on mere assertion, so I was able to point out that was false.