47
I began with the census definition of “city”—larger than 2,500 in total population. Then I discovered towns far smaller than 2,500 that posted signs, passed ordinances, spread the word informally, burned houses, or took other steps to keep out African Americans (and sometimes other groups). So I enlarged my definition of “town” to include places from 1,000 to 2,500. When jurisdictions even smaller than 1,000 came to my attention for excluding African Americans, I included them as well, although I did not try to study these hamlets systematically. (In many states, I have not been able to study towns smaller than 2,500 systematically and have merely taken note of information on them when I obtain it in the course of my research.)
Such small towns can be important, partly because when they do expand, usually they remain sundown. Malcolm Ross investigated East Alton, Illinois, for example, for the Fair Employment Practices Commission during World War II. The town’s industrial patriarch, F. W. Olin, told him that East Alton had an ordinance dating back to 1895, when a “Negro boy” committed some crime, and men had gone hunting for him with shotguns. He got away, but his angry pursuers reportedly swore that no Negro would ever again set foot in East Alton. Ross noted that during the next fifty years, East Alton had grown from a few families to a sizable town without any Negro ever having stayed the night. During World War II, Olin’s munitions plant employed more than 13,000 workers—not one of them African American or Native American. In 1940, shortly before Ross wrote, its population had increased to 4,680, with 1 stray African American. By 1960, East Alton had 7,309 residents but only 4 African Americans, probably none of whom lived in an independent household. It finally cracked in the 1990s.
Larger cities tested my operational definition in a different way. A cutoff of ten proved too low to do justice to large cities widely known to keep out African Americans, such as Cicero. In 1951, as we have seen, the governor had to call out the Illinois National Guard to stop a riot against one African American who had tried (and failed) to move into Cicero. “Of primary significance in understanding the violence,” sociologist William Gremley points out, “is the fact that it was widely believed by the residents of the community that no Negroes lived in Cicero.” Actually, the U.S. Census in 1950 showed 31 African Americans in the city, but they were apparently live-in servants, biracial adopted children, or individuals living unobtrusively in rental property. Cicero clearly defined itself as a sundown town, no matter what the census said. Indeed, according to a report issued after the 1951 riot, “It is said that no Negroes now live within the limits of Cicero, although one or two families have done so in the past.” But “fewer than ten African Americans” would have missed Cicero. So for cities larger than 10,000, I changed my definition for “all-white town” to “less than 0.1% black,” decade after decade. For towns smaller than 10,000, “fewer than ten blacks” remained in force.