Everything Old Is New: Migration in the 1850s
Census Data Shows That Some States Have Oddly Persistent Migration Patterns
So, as regular readers know, I’m starting a Podcast about the history of migration. Pending a few more technical details being worked out, I’ll be able to give you more information, like a website and stuff, next week. Exciting! So look forward to that. In the meantime, I’ve been doing some history-of-migration research that I thought would be fun to share.
There is not that much widely-used data on migration from the 19th century. From 1850 on we can use birth-state information from the Censuses, but before that we’re incredibly limited in what we can say about migration. So unfortunately, we really only have two solid data points for internal migration before the civil war: 1850 and 1860.
Luckily, the Censuses of 1850 and 1860 kind of rock. Really, these two censuses are orders of magnitude more detailed and useful than the Censuses that came before, and whoever it was that caused this increase in detail really deserves some kind of recognition.
So, let’s look at some fun factoids we can extract from the 1850 and 1860 Census.
Birthplace and Migration
Powerful, But Limited, Indicators
So the key statistic that the 1850 and 1860 Census gives us is state of birth for the non-enslaved population. It also tells us state of residence. So if birth state doesn’t equal residence state, bingo, that’s migration. You can’t create new New York-born people in Pennsylvania, ergo, net migration and mortality are the sole drivers of changes in these birth/residence mismatched populations.
But crucially, we only get this data for the free (and usually white) population. That’s important. Enslaved migration vanishes from the record. We know large-scale enslaved migration happened, and I believe that data might be available if you use microdata (I’m not, for expediency). But enslaved migration is likely to mirror free migration in southern states, as enslaved people were sold towards new and expanding agricultural areas, or else forced to accompany their masters on westward migrations. The major exception is probably Louisiana, where the extraordinarily lethal conditions of sugar cultivation created high laborer-mortality, leading to constantly high demand for new enslaved labor, leading to persistent enslaved inflows even if free/white inflows were low.
So before I give you fun maps and charts, let’s look at a few factoids.
Take New York from 1850 to 1860, for example. We can see that from 1850 to 1860, a minimum of 319,000 New York-born people migrated out of New York, and the number of native-born non-New Yorkers actually decreased. In other words, in this example, New York experienced huge losses to domestic out-migration from 1850 to 1860: inflows from other states were fairly small and at best break-even, while outflows to other states were very large.
Meanwhile, the foreign-born population increased by 345,000, and even the number of New York-born individuals within the state increased by 451,000. That means that the net increase in New York-born people regardless of residence from 1850 to 1860 was about 770,000, of whom about 40% migrated away from New York. New York State’s population rose by 783,000. For reference, New York City increased from about 700,000 people to almost 1.2 million during this time, meaning that New York City got about 60% of the population growth in New York State, despite having just 22% of the 1850 population. That’s a huge growth rate. And perhaps most strikingly, the foreign-born population of New York City rose by 220,000, or about 65%.
If this all sounds familiar, it’s because this is pretty much the same demographic profile New York has today.
I’m not trying to say that migration patterns are somehow super-stable over 150 years, just noting that we’ve been here before. Notably, the 1850s were also a period of extremely high immigration, and a period with a similar trend in the foreign-born share of the population as we have today.
Another fun state to look at might be Vermont. Between 1850 and 1860, Vermont’s population basically stagnated at 315,000 people; total population increase was under 1,000 people. But that doesn’t mean nothing happened, it just means there were counterbalancing trends. For example, the number of Vermont-born people rose by about 36,000, and 6,000 stayed in Vermont… but 30,000 moved to other states. Meanwhile, the foreign-born population declined by about 1,000, and the native-born in-migrant population declined by about 4,000. In other words, Vermont between 1850 and 1860 saw about 80% of its new Vermont-born residents move away, and experienced a reduced ability to attract or retain outsiders from abroad or the rest of the US. I don’t know what was happening in Vermont in the 1850s, but it sounds unpleasant.
We can also look at states making big gains. For example, Illinois saw its accumulated net balance oif migration improve by 135,000 people, and gained 212,000 foreigners on top of that. Illinois gained more foreigners than any state besides New York, and gained the second-most native in-migrants, behind Iowa. So we can say that these were good times for Illinois. And, unsurprisingly, these were good times for Chicago, as it grew from 30,000 people to 115,000 people during the 1850s. Abraham Lincoln’s Illinois was the big migration-winning state of his day, the cutting-edge state for politics and demographics. And just across the Mississippi, freshly-minted Iowa gained 250,000 people from the rest of the nation, as well as 85,000 foreigners. For perspective, Iowa’s 1850 population was under 200,000 people.
The slave-state frontiers were growing too, but not as much. Texas drew a solid net of 130,000 domestic migrants at least, but just a measly 25,000 increase in the foreign-born population. Arkansas eked out just 2,000 net increase in foreigners, while gaining 100,000 domestic migrants. For perspective, Arkansas had about 210,000 people in 1850, of whom 55,000 were enslaved. Texas had about the same population.
This lackluster performance partly reflects the bias of the data, however. Texas and Arkansas were slave states, and the nativity-migration data neglects the enslaved population. In Texas, the enslaved population grew by 124,000, almost certainly much of that increase came from migration. In Arkansas, the enslaved population grew by 63,000. If we assume 75% of these increases were through migration, then Arkansas and Texas have a stronger record, but still can’t beat Illinois or Iowa.
The slave states simply did not see the same volume of westward migration as the free states, not least because the free states managed to attract huge flows of immigrants.
Mapping 1850s Migration
But What’s the Best Statistics?
So I promised you maps. Now, the statistics I’ve given are fairly easy to digest verbally. But trying to figure out the best way to map them is trickier. I’ve decided to go with a component-level approach. So I’ll present change in population from 1850–1860 as a percent of 1850 total resident population for: the foreign-born, enslaved people, people born-and-residing in a given state, out-of-staters residing in a given state, and people born in a given state residing elsewhere. I will also present a map of the change in out-of-staters minus the change in locals-moved-to-other-states, which can be seen as a ballpark estimate of net migration for a state. It’s really far from perfect. but it’s better than nothing.
Finally, I’ll present each state’s “adjusted growth anomaly.” This is equal to each state’s growth rate for the non-foreign population, minus the aggregate non-foreign growth rate for the nation. I’ve found for more recent periods that this number is a surprisingly good approximation of net migration. All graphs are on the same scale, from component-level changes equal to -50% of 1850 population, to component-level changes equal to 50% of 1850 population. When changes exceed 50%, they’re simply colored in the darkest color on their end of the scale.
So, here we go:
Change in the Foreign-Born
The foreign-born population grew almost everywhere in the 1850s (Vermont is the sole exception). But in some places, it grew with astounding speed. In California, the gold rush brought immigrants from the eastern states, while growing labor demand led to an influx of Chinese workers. Outside of California, the fastest foreign growth was in the upper midwest, were German immigrants settled in huge numbers, alongside smaller numbers of other northern Europeans, like Norwegians and the Dutch. Absolute growth in New York was substantial, but the good soils and cheap lands of the midwest were the real appeal for immigrants. Many of these immigrants probably stayed in New York or other eastern cities for a few years, however, so the foreign-born inflows to Wisconsin, Minnesota, and Iowa probably include as much domestic migration of foreigners as true immigration.
Change in the Enslaved Population
As you can see, Texas, Mississippi, and Florida dominate this chart. They’re the states that saw the fastest growth in the enslaved population, which means they’re probably the states that saw the highest rate of enslaved net migration. So we can see slavery’s march before the civil war, moving out towards new lands on the former fringes of the south. Notably, the enslaved population shrunk slightly in Delaware and Maryland.
Change in Locals
Retention of locals, by which I mean the growth rate of the born-in-state-of-residence population, is an interesting variable because it reveals a telling trend. Essentially, areas with high rates should either have higher natural increase due to fertility and mortality factors, or else should have much lower out-migration (or much higher in-migration by young families). Frontier areas could have higher birth rates, but could also have higher death rates in some cases. Meanwhile, they may have attracted young families. But it also seems likely that we’re seeing a migration effect. People born on the frontier didn’t, at least as of 1850–1860, tend to move away a whole lot (perhaps because many were quite young). Meanwhile, growth of the local-born population was quite low in most of the longer-established states.
Change in Domestic In-Migrants
Once again, we see that inflows of people born in other states caused the populations of the frontier states to explode: Oregon, California, Texas, Minnesota, and Iowa more than double due to this migration of natives from other states alone. Meanwhile, states like Tennessee, Vermont, Georgia, and Maine actually see meaningful decreases in their out-of-state population, suggesting inflows to those states had really dropped off by the 1850s.
Change in Locals Moving Out-of-State
Aside from Minnesota, where the big increase in locals moving elsewhere is due to a statistical fluke involving Minnesota’s territorial status, the biggest increases in “state diasporas” can be found in the states that were settled in the early 1800s: Ohio and Indiana. Near-frontier areas like Illinois and Mississippi also turn out to be large “sending states” during the 1850s. The northeast also had fairly high outflows. We can also see that frontier states in the north became donors fairly rapidly (Illinois, Iowa) while southern frontier states don’t seem to as rapidly become sending-states (Louisiana, Arkansas, Texas). This further suggests that northern migration into new territories was, by the 1850s, substantially outpacing southerner expansion towards the west, even aside from direct benefits from immigration. That is to say, in the westward race between Yankees and Rebels, the Yankees were evidently the more fleet-of-foot. I should say though, this map excludes slavery, which could slightly bolster the southern states. Even so though, southern whites were less likely to move west than northern whites during the 1850s.
Change in Net Stocks of Domestic Migrants
The above chart shows the change in out-of-staters residing in a given state, minus the change in people born in that state residing elsewhere. In other words, it shows the change in the net stocks of lifetime accumulated migration per state. Basically, it shows which states were seeing migration boost their population, versus shrink it. The results are clear: domestic migration was a huge boost to the frontier, while mostly pulling people from the middle of the country: Ohio, Kentucky, Tennessee, etc. Another locus of migration existed in the northeast. But crucially, some of the eastern states actually gained domestic migrants, including Connecticut, New Jersey, Virginia, and North Carolina. The margins are small, and for North Carolina and Virginia could be heavily impacted by any variation in enslaved migration. But it does appear that the old colonies were seeing a lessening in out-flows during the 1850s.
Growth Anomaly During the 1850s
The above map charts the “growth anomaly” for each state, defined, as I explained earlier, by the difference between the growth rate for that state’s non-foreign population, and the national non-foreign population. This value tends to correlate pretty closely with net migration in more recent years. And, in general, it confirms the estimates I adopted as well. The correlation between the growth anomalies and the change in net stocks is about 99%. When I drop the 5 most extreme high and low values, it’s still 85%. In other words, these two estimates probably give us a pretty reliable guess at the size of net migration in each state. Notably, most of the states here have a similar record, but there are a few differences: those eastern states with gaining records under the net balance method are now net sending states, except New Jersey, which remains a net gaining state.
Conclusion
There’s No Reason We Can’t Assess Migration Since the 1850s
Existing Census records can allow us to use multiple different methods to estimate net migration at the state level. These different methods tend to confirm each other fairly closely, and match anecdotal evidence. In other words, we have multiple, corroborating sources on migration since the 1850s. There’s no reason for this record to be ignored in discussions of migration: Ohio and New York, for example, have been losing domestic migrants for over 100 years. I’ve raised this point before: the long run history of migration challenges many of the vogue stories we tell about regional competition. When we assess migration in the 1850s, we can see the massive scale of westward migration: it routinely led to double-digit growth rates for many states. Such enormous flows are rare today. However, this is not because Americans are less mobile; in fact American migration rates are similar to or higher than in the 1800s; they just happen to also be less unidirectional. Finally, the strong correlation between the two methods used suggests that the growth-anomaly based method is useful for assessing general migration trends in the antebellum period, possibly even for periods before 1850, for which data is extremely sparse.
For my post analyzing the libertarian invasion of New Hampshire, click here.
For my post discussing the urban success of Des Moines, click here.
If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!
Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!
I’m a graduate of the George Washington University’s Elliott School with an MA in International Trade and Investment Policy, and an economist at USDA’s Foreign Agricultural Service. I like to learn about migration, the cotton industry, airplanes, trade policy, space, Africa, and faith. I’m married to a kickass Kentucky woman named Ruth.
My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research. More’s the pity.
