Immigration and the American Nations

The United States is regions apart when it comes to attitudes about immigration, immigrants and immigration policies

By Colin Woodard

Immigration policy, as you’ve no doubt noticed, has become one of the most animating issues in American politics. Constructing border fortifications and rounding up undocumented migrants has become a central feature of Trumpism, so much so that its leader ordered Congressional Republicans to spike a bipartisan deal that would have addressed many of the movement’s longstanding concerns. (“Unless we get EVERYTHING needed to shut down the INVASION of Millions & Millions of people,” Trump wrote of his opposition on Truth Social, “many from parts unknown, into our once great, but soon to be great again, Country!”) Trump leads an ethnonational authoritarian movement, so it’s not surprising that he and his followers call immigrants – few of whom are white and Protestant —  invaders, rapists, criminals, disease-spreaders and “vermin” who are “poisoning the blood of our country.”

So we know immigration is an explosive political issue, but it’s also polarized on regional lines. The reasons go back to the way the continent was settled in the 17th and 18th centuries – by rival colonization projects and settlement streams – and the dramatic effect they had on the geography of immigration in the late 19th and early 20th centuries, when attitudes about American identity and belonging experienced a dramatic reboot.

The regional divide is significant. The federal government sets immigration policies, but states have a choice in whether or not they want their agencies and state and local police forces to help federal agents detain or deport migrants. Some state legislatures have passed laws prohibiting cooperation – so called “sanctuary” states – while others adopted laws that require local and state officials to cooperate with federal Immigration and Customs Enforcement (ICE), providing them holding cells, jail beds, access to interview detainees and assistance in apprehension and custody transfer. The Immigrant Legal Resource Center monitors such enforcement policies and provides each state with a pro- or anti-immigrant score. In their 2023 index, the states with the most anti-immigrant policies – Florida, Texas, Iowa, Alabama and West Virginia – were controlled by the Deep South, Greater Appalachia, and the Midlands. The pro-immigrant states – Oregon, Illinois, New Jersey, Washington, California, Vermont and Connecticut – were all dominated by Left Coast, Yankeedom, New Netherland or, in California’s case, El Norte. No Yankee, New Netherland or Left Coast state had enacted anti-immigrant policies, while almost most Deep Southern and Greater Appalachian states had.

The regional gap first opened up in the early 21st century after the Deep South and Greater Appalachia began experiencing their first significant foreign in-migration since the United States banned the import of enslaved people in 1817. Coupled with post-9/11 border security anxiety, in the aughts and 2010s, states controlled by these regional cultures enacted a spate of laws to facilitate the deportation of unauthorized immigrants or to make life so difficult and insecure they would “self-deport.” (Ironically, this phrase was first created by Chicano satirist Lal Alcaraz and fellow comedian Esteban Zul to mock anti-immigrant campaigners, but was then popularized by Republican politicians like Pete Wilson and Mitt Romney who adopted it in earnest.) Many unauthorized immigrants fled Alabama in 2011 after legislators passed a 2011 law making it illegal for them to reside there or for people to give them rides or rent them homes and also directed law enforcement to screen the status of all students enrolling in public schools. South Carolina and Georgia passed similar laws before the courts overturned most of their provisions as being unconstitutional intrusions on federal authority over immigration. (Arizona, a state split between El Norte and the Far West, had a “check your papers” law the Supreme Court overturned for the same reason.) When researchers at Queen’s University of Charlotte in North Carolina analyzed 3500 state and local immigration policy measures passed across the U.S. between 2005 and 2017 they found those in Southeastern states – all of which are controlled by the Deep South, Greater Appalachia, Tidewater or a combination thereof – were significantly more restrictive than those passed in other regions.

The situation has accelerated since, with the governors of Texas (controlled by Greater Appalachia and Deep South) and Florida (Deep South) transporting and dumping migrants on the streets of Yankeedom and New Netherland without prior warning. Last year Texas adopted a law empowering local and state police to arrest migrants entering its territory and Louisiana followed suit this spring. Both states argue these intrusions on federal control are justified because the migrants constitute an “invasion.”

Here at Nationhood Lab we were curious if the gaps seen at the policy level reflected regional differences in public opinion on these issues. Most national surveys aren’t nearly large enough to power the American Nations model, which requires county- or at least congressional district-level results. A wonderful exception is the Democracy Fund + UCLA Nationscape project, a massive public data set of weekly polls conducted from summer 2019 through January 2021 that asked more than half a million Americans about a broad range of issues and attitudes, including a number of questions related to immigration and immigrants.

The Nationscape data reveals Americans in every region actually agree on a number of key issues: large majorities everywhere disapprove of Trump’s policy of separating children from their migrant parents (so as to deter would-be asylum seekers via terror); they oppose the building of the border wall and Trump’s Muslim ban; they want the “dreamers” – undocumented people brought to the U.S. as small children – to be allowed to become citizens; and they think there should be a path to citizenship for law-abiding migrants who entered illegally. They also think the immigration system needs a fundamental overhaul. As with abortion and gun control, the American people are much closer together on the issues than their elected representatives are.

There are significant regional differences in several salient issues, however. Asked if they supported deporting all undocumented immigrants – some 11 million people – slight pluralities in Greater Appalachia (40-39) and New France (41-38) agreed, while the policy was rejected by huge margins in New Netherland (30-51), El Norte (28-53), Left Coast (19-54), and Florida’s Spanish Caribbean section (30-50.) Residents of Tidewater, the Midlands, and Yankeedom disagreed with the policy by closer margins (see chart below.)

There was a similar pattern in support for charging people who enter the country illegally with a federal crime. (It is, in fact, illegal to do so, with the first offense being a misdemeanor and subsequent ones a felony, but most migrants are deported instead because of the human and financial costs of trying and incarcerating them and caring for minors while they are imprisoned.)  Pluralities again agree in Greater Appalachia (44-34) and New France (42-35) and were joined by respondents in the Deep South (42-37) and the Midlands (41-38.) Respondents in Far West and Tidewater were statistically tied on the issue, while pluralities opposed the measure in Yankeedom (38-40), Spanish Caribbean (36-43), El Norte (34-45), New Netherland (35-46), and Left Coast (21-47).

The following pattern held throughout the data, including in the size of the dissenting blocks on the wall, Muslim ban, and other issues. Greater Appalachia, New France, and Deep South are the most conservative regions on immigration. The Left Coast, El Norte, New Netherland and Spanish Caribbean are the most liberal. Yankeedom and Tidewater lean toward immigrants; the Midlands and Far West lean against. Which leads us to ask why this is the case.

The “American Nations” have always had distinct stances about immigration and identity, and those attitudes had an enormous influence on where immigrants did and did not go in both the colonial era and the great 1880 to 1924 immigration wave that reshaped assumptions about American identity and the criteria for “in group” belonging. Several of these “nations” had narratives that aspired to define the United States as a whole, including distinct ideas of who could fully belong and who could not. In brief, they are as follows.

Yankeedom’s Puritan founders had strict religious and moral requirements for citizenship, with the effect that 17th and 18th century Yankeedom was overwhelmingly English and Calvinist. After the Civil War, however, it welcomed immigrants, but with requirement that they assimilate into Yankee norms. The region’s sophisticated network of public schools were tasked with making good Protestant Yankees of immigrant children while adults were expected to learn Yankee ways and leave their own identities behind. Hector St. Jean de Crevecoeur, a French mapmaker and writer who lived in the Yankee upstate of New York from 1759 to 1779, recognized this even then. “He is an American who leaving behind him all his ancient prejudices and manners receives new ones from the new mode of life he has embraced, the new government he obeys, and the new rank he holds,” he famously wrote in 1782. “Here individuals of all nations are melted into a new race of men.” In Yankee Michigan two centuries later, Henry Ford set up special schools for his immigrant workers to learn “American” conduct, dress, and customs; at their graduation ceremonies they strolled onto a stage and behind a giant “melting pot” wearing their Hungarian, Slovak, or Italian folk dress and emerged changed into identical “American” suits. Newcomers were meant to “melt” into the existing culture.

The Deep South and Tidewater were ruled by slaveholding oligarchs and aristocrats respectively who saw no need for free immigrants and created an economic and social environment that offered few reasons for any to come. During the Antebellum period they argued that the United States was a collection of ethnostates belonging to the superior “Anglo-Saxon people,” and that others were not entitled to full citizenship or – in the case of African-Americans – humanity. In the Civil War the in-group was redefined as the aristocratic “Anglo-Norman race” (which conveniently embraced the French Huguenot elite) who imagined they were fighting a reprise of the Norman Invasion in 1066 against the swamp-dwelling “Anglo-Saxons” of the northern regions. These regions were tightly bound to narrow ethnoracial and religious criteria for belonging and remained so right into living memory, though the Tidewater has rapidly transformed since for reasons outlined in American Nations.

From the outset the Midlands and New Netherland each embraced different strains of pluralism and multiculturalism, in which immigrants were not only welcomed but encouraged to retain their cultural practices, identities, and languages. With the creation of the U.S., Midlander intellectuals would argue individuals could only achieve liberty and the pursuit of happiness within the context of their own value systems and, thus, their cultural distinctiveness had to be respected and even nurtured. “The process of Americanization…is not one of assimilation or conformation to any particularly ethnic type,” argued Marion Dexter Learned, a Delaware-born Midlander who headed the Germanic department at the University of Pennsylvania in the early 20th century. “Nor is it simply a process of amalgamation, a mixing of elements… (but) is rather an organic union, an intergrowth of race elements.” Learned said Americans should be a “composite people” composed of overlapping but still distinct ethnic cultures. America, in this tradition, is a mosaic not a melting pot.

 Greater Appalachia’s mythic narrative was developed in response to the 1880 to 1924 immigration wave, whose decidedly un-Protestant character had panicked many “old stock” Anglo Protestants. In the midst of this “invasion,” the intellectual elite of Appalachia – many of them transplants from Yankeedom or natives who’d been educated there – asserted their region was a repository of unadulterated Anglo-Saxon Protestant settlers, a time capsule where millions of people were living, speaking, and worshiping just as their pioneering 18th century ancestors had, uncorrupted by unsavory aliens and degenerate cosmopolitans. “In these isolated communities…we find the purest Anglo-Saxon stock in all the United States,” asserted Ellen Churchill Semple, Louisville native and president of the Association of American Geographers, in 1901. “The stock has been kept free from the tide of foreign immigrants which has been pouring in recent years into the States.” Thus was born the notion that there were “real Americans” who were members of an American ethnicity that was British, Evangelical Protestant, English-speaking, and white and to whom the country was supposed to belong.

Two other regions – Far West and Left Coast — had narratives of regional purpose that are not particularly relevant to today’s immigration debate. The Far West’s, keyed off Frederick Jackson Turner’s Frontier Thesis, held that Euro-American settlers would be molded into a single “American” people by the region’s challenging conditions. The Left Coast’s, a fusion of the Yankee and Greater Appalachian myths, initially combined white supremacy and assimilative ideology, but has since pivoted to the cultural pluralism of the Midlands and New Netherland. The rest — El Norte, New France, First Nation, Greater Polynesia and Spanish Caribbean – have stories emphasizing having been distinct societies before having been conquered and incorporated into the United States. Their stories aspire to conserve and strengthen local culture and historical memory rather than to try to set the agenda for the continent as a whole.

With this in mind, check out this map of where the 1880-1924 immigrants did and did not go. It comes from the 1900 census, when census takers asked each person if they were foreign born or not but, unlike in 1910 and 1920, didn’t exclude people of color from their tabulations. It shows the percentage of the population in each county (and region) that was an immigrant.

Notice almost nobody emigrated to the Deep South, Tidewater or Greater Appalachia – regional cultures that offered few opportunities and embraced a very restrictive definition of who could fully claim to be an American. The vast majority of immigrants went to Yankeedom (for the jobs), the Midlands and New Netherland (where they were welcomed), El Norte (proximate to Mexico) and the Left Coast (which had received many Asian immigrants prior to the adoption of the racist Chinese Exclusion Act of 1882).

The great 1880-1924 immigration wave profoundly changed ideas about American identity, forcing legacy Protestant America to accept that henceforth Catholics, Orthodox Christians and Jews – at least – would be full-fledged Americans too…except in the southern regions the migrants had almost completely avoided. While the rest of the country became ethnically and religiously diverse, Deep South, Greater Appalachia, Tidewater and New France did not, with profound effects today. This map based on data from the Association of Statisticians of American Religious Bodies’ 2020 US Religion Census shows the dominant religion in each county today, blue being Catholic (the largest cohort of that 1880-1924 immigrant wave) and red being Southern Baptist. Notice the remarkable degree of affinity with the American Nations borders, with the Baptist zone in effect the area where legacy Protestant America faced no competition until very recently.

Now, for comparison, here is the situation today based on the Census Bureau’s American Community Survey data from 2020:

To ease the comparison, here’s 1900 and 2020 side-by-side by regional culture. Notice the share of immigrants remains roughly the same in the regions with the most liberal immigration attitudes today, Left Coast, New Netherland, Spanish Caribbean and El Norte. But the Deep South, Tidewater, and Greater Appalachia have seen a substantial increase – roughly ten-fold in the case of the two lowland Southern regions and a more than doubling in Appalachia. And Yankeedom and the Midlands have seen sharp declines, a drop of about a quarter in the Midlands and more than half in Yankeedom. Today the Deep South has a higher proportion of foreign-born people than either of those regions, and nearly all of that growth has happened since the late 1990s. (Again, this shift in immigrant destinations almost certainly reflects shifting opportunities for entry-level work, with the Yankee-Midland manufacturing towns of the early 20th century having turned into the Rust Belt of the early 21st as corporations moved production to lower wage, lower regulation sites in the southern regions and abroad.)

The implication is that people in regions that have always been prominent immigrant destinations have positive feelings about immigrants, but people in regions that have recently become such after having few if any foreign-born residents are more likely to see them as invaders. Regions that were once major immigrant destinations but no longer are today are in the middle.

This analysis definitely tracks for Deep South and Greater Appalachia. Academic researchers have documented how these regions – traditionally places people emigrated from – suddenly became destinations for Latino migrants starting in the late 1990s, attracted to fill labor shortages in factories and meat and seafood processing plants. Starting from an extremely low baseline, Latino populations grew by hundreds or even thousands of percent in many Southern states during the 1990s and aughts, sprouting new Little Mexico neighborhoods in many cities and spurning migration researchers to start calling the region the Nuevo New South.

“Around 2001, Latino children began to enter southern schools in noticeable numbers and Latino families began to buy homes in rural Alabama and Arkansas,” Jamie Winders, a geographer at Syracuse University’s Maxwell School who studied the issue, wrote in 2007. “In the wake of this collision between the South’s emergence as an immigrant-receiving region and the nation’s post-9/11 eruption into border hysteria, southern states have scrambled to contain the social, political, and cultural challenges that Latino migrants bring to southern communities from an even deeper ‘south.’”

Significantly, all of the examples of anti-immigrant feeling and policies Winders and other academics tracked in the aughts and 2010s were in the Deep South and Greater Appalachia, not Tidewater. This is because this region is no longer really “the South,” the result of phenomena I wrote about in American Nations 13 years ago: the fact the Tidewater wasn’t able to expand Westward in the 18th and 19th centuries (due to the presence of the Greater Appalachians), leaving almost the entirety of this relatively compact region within the “federal halos” that formed in the second half of the 20th century around the District of Columbia and Hampton Roads/Norfolk area, site of the world’s largest naval base. For generations, the presence of the federal government and trillions of dollars in federal spending have allowed millions of people to live social, cultural, and economic lives without reference to the legacy Tidewater culture around them, thus radically transforming the region. Change came gradually (primarily from the 1960s onward) and then all at once. Our polls and data analysis have repeatedly shown a dramatic liberalization of Tidewater attitudes on everything from abortion to gun control to elections in the past two decades. It’s now one of the most liberal places on the continent by many measures – and its attitudes on immigration are now in the middle of the pack, similar to those in Yankeedom rather than Greater Appalachia. I would expect this trend to continue, particularly as northern Virginia suburbs around Washington remain among the most popular immigration destinations in the country.

The Nationscape polls asked respondents about their attitudes to various demographic groups including “undocumented immigrants.” Among Hispanic respondents, pluralities in all the regional cultures save New France had a favorable opinion of this group. The picture was very different among non-Hispanics: undocumented migrants were viewed negatively in every region. And the regional pattern did not match the outlines we saw with other immigration-related metrics. Undocumented people were the least popular in (immigrant-friendly) New Netherland (net -12.9% favorability), followed by (immigrant-unfriendy) Deep South (-9.8) and pro-immigrant Spanish Caribbean (-9.3). They were most popular in Yankeedom (but still underwater at -2.6), hyper-conservative New France (-3.1), and Greater Appalachia (-3.7.) This pattern confounds easy analysis, but we present it here to emphasize a point made many times elsewhere: people’s opinions about legal and extralegal immigration are often widely divergent.

Political scientist Margaret Commins and sociologist Jeremiah Willis, the Queens University of Charlotte researchers mentioned earlier, analyzed the thousand’s of state level immigration laws passed between 2005 and 2017. Those from states in the conventionally defined “Southeast” – almost all of them controlled by some combination of Greater Appalachia and Deep South – were far more restrictive than those in the other four conventionally defined regions (the Northeast, Midwest, Southwest and West.) In fact, their scoring method found the average immigration law passed by states in three of these conventional regions of the country was intended to benefit immigrants; in these three regions only one state (Indiana) is controlled by Greater Appalachia and none by the Deep South. “Our findings suggest that the restrictive tone of Southeastern states is likely related to the region’s fast growth rate in Hispanics and a political ideology that is more conservative than other regions,” they concluded in their findings, published in mid-2020.

An even more precise way of understanding the situation would be to say that 21st century migrants – pursuing unmet labor market demand – have been coming to regions that haven’t ever seen significant “out group” immigration and where the criteria for in-group  membership as “Americans” is more restrictive than in other parts of the federation. Meanwhile the Midlands – a region that has championed and celebrated the cultural rights of European immigrants for more than three centuries – has, for converse labor market reasons, ceased to be a major migrant destination, while almost all those foreigners who have moved there are of Asian and Latin American origin and have not been embraced as their European predecessors were.  These trends have spurred anti-immigrant reaction in these regions and support for openly xenophobic politicians like former (western Iowa) U.S. Rep. Steve King, Texas Governor Greg Abbot, Governor Ron DeSantis of Florida and, on the national stage, former and possibly future President Donald Trump. That said, the more extreme anti-immigrant policies – including Trump’s border wall, Muslim ban, and migrant child seizures – have little support in these regions and are radioactive in the most immigrant-friendly regions, New Netherland, Left Coast, and El Norte.

Americans, by and large, continue to see the country as a nation of immigrants and reject radical anti-immigrant policies, just as they oppose total abortion bans and support tightening gun regulations. But as with these other issues, in our federated and triple-branched system, public support alone is no guarantee of the political outcome.

— Colin Woodard is the director of Nationhood Lab.