Some of the statewide retreats indicated in
Table 1
are dramatic. For example, African Americans lived in every Indiana county but one in 1890. By 1930, six counties had none and another fourteen had fewer than ten African American residents, even though many more African Americans now lived in the state. I have confirmed eighteen Indiana counties as sundown throughout or in substantial part. Moreover, even when
Table 1
does not show a dramatic decline, looking at the actual number of African Americans in each county does. For example, in 1890, every county in the state of Maine had at least eighteen African Americans, except one with just two and another with nine. By 1930, Maine looked very different. Now five counties had eight or fewer African Americans. Several showed striking drops in their black populations: Lincoln County from 26 to 5, for example, and Piscataquis from 19 to just 1. Hancock County dropped from 56 in 1890 to just 3, yet Hancock had more than 30,000 people in 1930. Geography does not seem to account for these declines; the counties with fewer than eight African Americans were sprinkled about, not concentrated in Maine’s isolated rural north.
The Great Retreat and the Great Migration
These decreases to no or only a few African Americans by 1930 came in the teeth of huge increases in the black population nationally and in many northern states. Nationally, the number of African Americans went up by nearly 60%, from 7,388,000 in 1890 to 11,759,000 in 1930. Moreover, beginning about 1915, African Americans from Dixie started moving north in large numbers, a movement now known as the “Great Migration,” in response to the impact of World War I, which simultaneously increased the demand for American products abroad and interfered with European migration to northern cities.
26
More than 1,000,000 African Americans moved north between 1915 and 1930. Thus the absolute declines in black population by 1930 in many northern counties are all the more staggering. Without a retreat to the cities, these increases in overall black populations would have caused the number of counties with zero or few blacks to plummet.
Coming in the middle of the deepening racism of the Nadir, this Great Migration prompted even more white northerners to view African Americans as a threat. A 1916 editorial from Beloit, Wisconsin, exemplifies the “Negro as problem” rhetoric:
The Negro problem has moved north. Rather, the Negro problem has spread from south to north.... Within a few years, experts predict the Negro population of the North will be tripled. It’s your problem, or it will be when the Negro moves next door.... With the black tide setting north, the southern Negro, formerly a docile tool, is demanding better pay, better food, and better treatment. . . . It’s a national problem now, instead of a sectional problem. And it has got to be solved.
27
Historians and sociologists took note of the growing urban concentration of African Americans between 1890 and 1930, continuing to about 1960. One of the foremost writers on race relations of the era, T. J. Woofter Jr., put it this way: “It is remarkable that Negro city population should have increased by a million and a half between 1900 and 1920; but it is astounding that a million of this increase should have been concentrated in the metropolitan centers of the East and the Middle West.” More than half of this increase was absorbed by just 24 cities, each having black populations of more than 25,000, he observed. “This emphasizes the astonishing degree of concentration that has taken place.”
28
But neither Woofter nor other commentators noted the
decreases
in black populations—often to zero or to a single household—in smaller cities and towns across the North. The Great Migration seems to have masked the Great Retreat. Scores of books discuss the Great Migration; none tells of the Great Retreat (by this or any other name). The increased black population in, say, Chicago got ascribed to migration from Mississippi, which was largely true; hence the internal migration of African Americans from small towns in Illinois to Chicago went unnoticed. Not grasping the extent of anti-black sentiment in smaller northern towns during the Nadir, social scientists somehow found it “natural” for people from tiny Glen Allan, Mississippi, to wind up in Chicago; for those from Brownsville, Tennessee, to move to Decatur, Illinois; and for inhabitants of Ninety Six, South Carolina, to move to Washington, D.C. This is not how other migrations to the North worked. People from small villages in Italy often wound up in places such as Barre, Vermont, or West Frankfort, Illinois, as well as St. Louis. Norwegians went to Mount Horeb, Wisconsin, not just Minneapolis. But this was not true of African Americans—not after the 1890s, anyway.
Indeed, historians and social scientists have used the Great Migration to “explain” the increased racism in the North. That is, they used documents such as the Beloit editorial to explain the increased segregation African Americans experienced: the masses of newcomers strained the system, threatened whites’ jobs, upset existing equilibriums, and the like. But the Great Migration did not cause the Great Retreat. Whites were already driving African Americans from small towns across the Midwest
before
those towns experienced any substantial migration from the South. They continued to drive out blacks from towns that never saw any sizable influx after 1915. The Great Retreat started in 1890, a product of the increasing white racism of the Nadir. It cannot be understood as a reaction to a migration that started in 1915.
Now let us tour the country, seeing the profusion of sundown towns almost everywhere, beginning in the Midwest. In the process, we shall visit towns that excluded not only African Americans, but also Chinese, Jewish, Native, and Mexican Americans—and in a few cases Catholics, labor union members, homosexuals, and some others. We shall see that prime real estate—elite suburbs, beach resorts, mountain vacation spots, and islands—has typically been off-limits. And we shall encounter whole subregions where African Americans are generally not allowed, even in unincorporated rural areas.
The Great Retreat in the Heartland
I did more research in Illinois than in any other single state.
Table 1
shows that African Americans lived in every Illinois county in 1890. By 1930, six counties had none, while another eleven had fewer than ten African American residents. Without a doubt, exclusion underlies these numbers. In Illinois and elsewhere, entire counties developed and enforced the policy of keeping out African Americans. Many of the towns confirmed as sundown towns in my research are county seats, and when they went sundown, often—not always—the rest of the county followed suit. I have confirmed that ten of these seventeen counties had gone sundown by 1930 and suspect all seventeen
29
did.
Various written and oral sources tell of Illinois counties that kept out African Americans as a matter of county policy. Malcolm Ross of the Fair Employment Practices Commission wrote about Calhoun County, for example, “Calhoun County is recorded in the 1940 census as ‘8,207 whites; no Negroes; no other races.’ This is not by accident. Calhoun people see to it that no Negroes settle there.” According to an 83-year-old lifelong resident of Mason County, north of Springfield, the sheriff “would meet [blacks] at the county line and tell them not to come in.”
30
Mason County has remained all white for many decades, despite its location between Springfield and Peoria, both with large African American populations, and on the Illinois River, an important trade route.
31
Table 1
is a useful way to summarize the entire northern United States, but county data can only hint at the extent of the problem, because county is such a broad unit of analysis. Illinois may have had seventeen sundown counties in 1930, but it had far more sundown towns than that. Several entire counties in Illinois allowed no African Americans except in one or two isolated locations, for example, but that one place sufficed to remove such a county from
Table 1
. Town is a more useful jurisdiction to examine. Most sundown towns in Illinois lie in counties that never appear in
Table 1
. In 1970, when sundown towns were probably at their maximum, Illinois had 621 towns larger than 1,000 people, ranging from Wyanet, with 1,005, to Chicago.
32
Of these, 424 or almost 70% were “all-white” (as defined in Chapter 1) in census after census. In addition, my universe of towns must include 50 hamlets smaller than 1,000 that came to my attention because of evidence confirming them as sundown towns. Therefore my list of Illinois towns totaled 671, ranging from tiny hamlets to Chicago. Of these 671 towns, 474 or 71% were all-white,
33
while 197 had African Americans.
Of course, the mere fact that they were all-white does not confirm the 474 as sundown towns. That requires information as to their racial policies in the past. I was able to get such material on 146 of the 424 all-white towns larger than 1,000. Of these 146, I have confirmed 145 as sundown towns or suburbs, or 99.5%.
34
In addition, the 50 hamlets smaller than 1,000 in population were confirmed as sundown towns. Confirmed Illinois sundown towns range in size from communities of just a few hundred people to Cicero, which in 1970
35
had 67,058 residents, and Pekin, which in 1970 had 31,375 and another 3,500 in its suburbs.
If 145 of the 146 suspected sundown towns larger than 1,000 on which we have information indeed turned out to be confirmed, what can we predict about the remaining 278 towns, on which we have no historical information beyond census data? Our best estimate would be that 99.5%—the same proportion as among the towns we have checked out—or about 277 of them would be sundown towns. There is no good reason to suppose the next towns will be different from those we know.
36
We add to that estimate of 277 the 145 towns that I have confirmed, plus the 50 hamlets, and our best single estimate is that 472 of 474 all-white towns and hamlets were all white on purpose. Therefore
our best single estimate is that 472 of the 474 all-white towns and hamlets were all-white on purpose.
Of course, we would not be surprised if “only” 465 (98%) of the 474 towns turned out to be sundown, or if 473 were sundown. Applying the principles of inferential statistics, we can calculate a range within which we can be confident the true number of sundown towns will fall. Statisticians call this the “confidence limits” for our best estimate of 472 or 99.5%.
37
They find these limits by computing the statistical formula known as the standard error of the difference of two percentages. Here this standard error equals .0205 or 2.05%.
38
The more rigorous confidence band used by statisticians is the 99% limit, the range that is large enough that we can be 99% sure that it includes the true proportion of sundown towns among the unknown towns. Here that range is 5.3%.
39
Accordingly, our estimate for the correct proportion of sundown towns among the unexamined towns would be .995 ± .053 or 94.2% to 104.8%. Of course, numbers above 100% are impossible; we can be 99% confident that the number of sundown towns among the unknown towns is roughly 94% to 100%, or 261 to 278 of the 278 towns.
40
Adding the 195 known sundown towns yields an overall estimate that the number of sundown towns among all 474 overwhelmingly white towns in Illinois lies between 456 (96%) and 473 (99.8%). We can say with a 99% level of confidence that between 96% and 99.8% of all the all-white towns in Illinois were sundown towns.
41
Our best single estimate remains 472, or 99.5%.
Even this total, 472, is not the full number of sundown towns in Illinois. I included communities smaller than 1,000 inhabitants only when informants or written sources brought them to my attention. These 50 confirmed sundown hamlets persuaded me to be suspicious of even very small all-white communities; many other hamlets no doubt kept out blacks.
42
Also, various sundown towns larger than 1,000 in population missed getting on my radar in the first place,
43
owing to nonhousehold African Americans such as prisoners.
These sundown towns are spread out throughout the state. Southern Illinois had many more even than Map 1 shows. Central Illinois has just as many: oral history confirmed some three dozen communities as sundown towns just within a 60-mile radius of Decatur, and written documentation confirmed another dozen. Northern Illinois has even more, owing to the sundown suburbs ringing Chicago. As a correspondent suggested about Ohio, instead of studying sundown towns, perhaps I should have researched the exceptions—towns that never excluded blacks—since that would be a more manageable number.
Similar maps could be drawn, showing most towns in boldface, in most other states in the Midwest, the Ozarks, the Cumberlands, the suburbs of any city from Boston to Los Angeles, and many other areas of the United States. But before we leave Illinois, this statistic of 472 probable sundown towns might come alive if I supply examples. I have chosen three, one from each section of the state.
LaSalle and Peru in northern Illinois are separate towns, each with its own library, city hall, etc., but they share a high school and a common boundary, and most people consider them really one entity. I don’t know when they first became sundown towns. Not one African American lived in the towns on the eve of the Civil War, when their combined population was 8,279. Even back then, the absence of blacks was surprising, since both towns lie on the Illinois River, a major artery, and on the Illinois-Michigan Canal, connecting Lake Michigan to the river at Peru, which opened up a water route from New Orleans to the Great Lakes. By 1860, when railroads became dominant, LaSalle-Peru found itself equally favored, being on a main line of the Illinois Central as well as the Rock Island Line, a major east-west railroad from Chicago. These trade routes surely would have brought African Americans to LaSalle-Peru had they been allowed. In 1864, seven African Americans from nearby Mendota signed up for the army and traveled with their recruiting officer to LaSalle to go up the canal to Joliet to be mustered in. In LaSalle a gang of “Copperheads” attacked them and drove them out of the city.
44
Census takers in 1870 found only one African American in Peru, none in LaSalle. Yet the war had caused many African Americans to wind up in Cairo, whence they diffused through the Midwest, and the Illinois Central directly connects Mississippi, Cairo, and LaSalle-Peru. In 1880, LaSalle was the only city in Illinois (defined as larger than 4,000 in that year) with no African Americans, and Peru was one of only two cities that had just one. An 1889 article in the
Chicago Tribune
noted that this was no accident: “The miners of LaSalle, Peru, and Spring Valley do not allow a Negro in their city limits.” Around this time, the towns apparently posted sundown signs, which stayed up until after World War II. The cities clearly still refused to let African Americans spend the night in 1952, for in that year its high school band director had to skirt the policy to host an integrated college band. By 1970, their populations had grown to 22,508, of whom just five were African American. Again, these numbers are shockingly low, since the cities were now also served by U. S. 6, a major east-west highway from Atlantic to Pacific, and U.S. 51, which runs all the way to New Orleans and was the most important single highway in Mississippi before the advent of the interstate system. An undergraduate at the University of Illinois-Chicago who grew up in LaSalle-Peru in the 1980s and 1990s reported that LaSalle-Peru High School stayed all-white until 1998.
45