Episode 16: Also, Michigan?

It has been so long since we last had a new episode! But we're finally here again!

So let's talk Michigan. Here's midwestern populations using current boundaries:

So Michigan's growth is a lot slower until 1830, has a sudden acceleration, then slows down again, until growth picks up again in the 1860s. We describe factors driving that in the episode.

Here's our estimate of annual migration:

So you can see the big bump up in migration in the 1830s. And you can see Michigan's disproportionate commitment to Unionism in the early 1860s. So again, this is about what we talked about in the episode.

Now, we talked a lot about the Mitchell Map. What does it look like?

You can find it here. Here's zoomed in on Michigan:



See the shark? Lansing is the eyeball.

But notice how the southern tip of Michigan is about even with the north shore of Erie.

Here's Google Maps:


So yeah. Biiiiiig difference.

And here's a map of the Toledo Strip that was being disputed:


You may want data about Detroit! We talked about it a lot. Here's what we think Detroit's population looked like:

It's really kind of sad. They keep growing, almost hitting scale, then they shrink. Then grow, the shrink. Then when they finally make it, BAM they are at scale basically immediately.

And we mentioned a cool source on early Detroit's Francophone population.  Oh and here's a simple timeline too. And here's our source on Francophone opposition to democratization.

I think that's about all we promised ya for this episode... and don't worry, Wisconsin will not take as long!






Episodes 14, and 15: Illinois!

Wow, I really got behind on writing this blog post! So let's do episodes 14 and 15.

Episode 14: Land Before Lincoln

First off, I mentioned old French settlements. Here's a map:



As you can see, no shortage of forts and towns and trading posts!

Here's 1850 population of Midwest states by origin-area:

As you can see, older-settled states had a large local-state-born population, as you'd expect. Meanwhile, the southern-born population probably outweighed the Yankee-born population in Indiana, Iowa, Missouri, and Illinois... although when you add in the Pennsylvanians and Ohioans and Hoosiers, probably free-state-settlers outnumbered slave-state-settlers for every state except Missouri. Speaking of Missouri: it had a substantial enslaved population, the only state included here to do so.

We can also see the foreign-born share in 1850: low in Indiana, a bit higher in Ohio, Illinois, Missouri, and Iowa, a bit higher still in Michigan, and wow that's a lot of foreigners in Wisconsin and Minnesota: over 1/3 of the population in each case!

Here's the same data, shown in absolute numbers:

As you can see, the native-born-and-residing population is by far the biggest group. But looking at Yankees specifically, we can see large groups in Ohio, Michigan, Illinois, and Wisconsin. When we look at southerners, we see large groups in Missouri, Indiana, Ohio, and Illinois. And for foreigners, the big groups are in Ohio, Illinois, Wisconsin, and Missouri.

Here's the promised map of the 1822 Illinois gubernatorial election:




Green indicates Coles, yellow is Browne (hardy har har), Red is Phillips, Blue is Moore.

And bam, that's episode 14! Now, on to episode 15.

Episode 15: The Keystone of Union

Let's start with some population estimates for Chicago.

But this has way too long a time period to see the data that interests us. So let's zoom in.

Here you can see that Chicago's growth really was remarkable. I should note as well that the declines you see for Cook County aren't really declines, per se, but are probably cases where my imputation method has failed to capture adequate YoY variation, or where Chicago annexed territory from Cook County.

Next up, Illinois' borders! Here's Illinois' original borders:



And here's a map of the Black Hawk War:



As you can see, overwhelmingly in Wisconsin or the Keystone Region parts of Illinois.

Here's growth rates in Chicago vs. the rest of Illinois:

As you can see, in most years Chicago is well above the rest of the state.

And of course we can look at the foreign-born population share as well:

And next we come to the 1870 origins of Chicagoans! So here we go for that:

Pretty much as I described it. I will note here, however, that about 40% of Chicago's Southern-born population was black as early as 1870, long predating the "Great Migration," and about 50% of the black population was southern.

Next up, we can look at political outcomes. Ignore the scales here, just notice the relative coloration of northern Illinois vs. the rest of the country. These are a variety of minor- or major-parties associated with abolitionism and Yankee-centric politics.


As you can see, in almost every case, northern Illinois, the Keystone Region, turned out to be a heavily anti-slavery region.

And finally, we can get to the Illinois governor's race of 1856. The three maps below show the total votes received by each candidate.

The Know-Nothing, with the most votes in the center, especially around the St. Louis suburbs (St. Louis was as or more heavily immigrant than Chicago).

The Democrat, with total votes fairly evenly distributed, with some large clusters in the mid-north and in Cook County:

And finally, the Republican, raking in a huge amount of votes in Cook County, and large vote tallies throughout the Keystone region, and then virtually no electoral support in Little Egypt:

The result, highlighted by winner of each county:

As you can see, the Republican did win a few southern counties around St. Louis, thanks to the Know-Nothings splitting the vote, and in the southeast, for reasons unclear to me. But the vast majority of Republican support came in the north, where heavy majorities in the Keystone Region propelled Bissell to victory, and ultimately enabled Lincoln to beat Stephen Douglas in his home state.


Okay, that's all for episodes 14 and 15! See you next time, for Michigan!

Episodes 11 and 12: The Settlement of Ohio and Indiana

I didn't do an Episode 11 post, but this one will cover some of the stuff in that episode as well.

So to start with, let's look at the population of various midwestern states.

So the first thing to notice is that the line is pretty close to zero with fairly little growth up to 1780 or so. If you're curious, here are NW territory growth rates from 1700 to 2015

That's the lagged 10-year compound growth rate, by the way, so it's smoothing out a lot of annual volatility. And you can clearly see the negative growth after the Proclamation of 1763, and then around the American Revolution, and then huge spikes after the area was opened for settlement. Peak rates of Midwestern growth were from 1790 to 1810.

So that's the whole region. Let's look at individual states.

As we discussed in the episode, Ohio was by far the population leader in the midwest before 1860. Towards the end, Illinois began to catch up. By the mid 1880s, Illinois was the largest midwestern state, and Ohio has never quite caught up. But for our purposes, Ohio was really big, and then Indiana was the second most populous state. Wisconsin doesn't even really register until 1840, and Minnesota is irrelevant until 1860.

From this population data, we can then derive our usual migration estimates. I showed these estimates for the whole midwest in the Episode 10 blog post, but now I'll show them disaggregated by state.

As you can see, Ohio is the leader early on, then Indiana reaches parity by the 1820s, but by the 1830s and 18403 states like Illinois, Minnesota, and Michigan have begun to draw migrants.

Okay, so we've got our population and migration data. Cool. Now, I want to note, this is domestic net migration. Students of midwestern history, or midwesterners themselves, will note that we totally ignored a huge factor in midwestern population growth: immigration! Don't worry, we will get to immigration. But that comes a bit later. Large-scale immigration doesn't really get going until the 1830s, and isn't huge until the 1840s.

But now, let's look at those treaty-maps we promised.

First up, here's a map of the treaties that made Indiana:

Source: https://en.wikipedia.org/wiki/File:Indiana_Indian_treaties.svg

Source: https://en.wikipedia.org/wiki/File:Indiana_Indian_treaties.svg

You can also see a map of Ohio. Notice the "Treaty of Greenville" in 1795 that we mentioned in the episode includes that small sliver of eastern Indiana, and a huge chunk of Ohio:

Source: https://en.wikipedia.org/wiki/File:Greenville_Treaty_Line_Map.png

Source: https://en.wikipedia.org/wiki/File:Greenville_Treaty_Line_Map.png

If you look at the dates on the Indiana treaties, you'll see the procession northwards we discuss on the episode, which would necessarily make it easier for southerners to get there than Yankees.

From there, we can use the data supplied by David S. Rose's work to chart out which Indiana counties were dominated by which origin-groups. Again, we want to really laud this work, it supplied a lot of insight, analysis, and raw data that we would have been hard-pressed to get elsewhere. Rose's work, while a bit dated now, is nonetheless interesting and impressive, and still of great relevance.

So, to begin with, here's a color-coding of counties by dominant state-origin-group in 1850:

Now, to get this, I did do some imputations on where Ohioans were from, with some being treated as "native Ohioan," but others being treated as Yankees or Pennsylvanians.  As you can see, Yankees are generally dominant in the north, and in a few of the counties along the southeastern edge, while Southern-origin settlers dominated in the south, and midland-settlers like Pennsylvanians dominated the middle section of the state.

We can get a bit more detailed, however, and look at each county for each origin-group. So here they are:

These charts should make the trends already described even clearer. Yankees were in the northern rim and, to a lesser extent, a few southeastern counties.Pennsylvanians and Ohioans were more widely dispersed, but still with a northern and eastern bias. Southerners dominated the south and center, especially Kentuckians, who made up a huge share of the southerners.

That pretty well wraps up the data pieces for these episodes, but please let us know if you have any questions or comments we can answer!






Episode 10: Yankees vs. the Dutch

So this was a pretty fun episode for us, both researching it and actually recording the show. There's just a lot of oddness, not least the Dutch drainpipe issue.

As such, even though it's not the first item chronologically in the show, here's a picture of Dutch drain pipes:

See those spouts hanging off the front of the houses? Well, that's what Elkanah Watson was all up in arms about. Nifty stuff, eh?

Anywho, to the maps and data! Let's start with maps. First of all, here's the Northwest Territory, our topic of interest:

And here's a map of all the competing land claims in the "west" around the 1780s, 90s, and 1800s:

And, finally, here's a map by a modern cultural geographer that more-or-less traces out the areas that were settled by the Yankee Diaspora:

The dark blue "Yankeedom" bit is what we're really talking about, although I would say the Yankee Diaspora extends easily into the lighter blue Midlands, and sometimes down into the red of "Greater Appalachia." But the dark blue makes a nice core.

So we've got some nice maps to visualize our story.

Now, let's look at some data.

I mentioned some gross migration rates. But I want to provide something simpler: the net domestic migration of New York and New England from 1780-1900:

I want to draw your attention especially to the period from 1790 to 1810, the years we are discussing today. During this period, immigration into the US was low for a variety of reasons, even as domestic migration was quite unidirectionally westward. We see that New England's migration gets steeply negative, even as New York's gets positive. This is the Yankee diaspora in data. By the 1830s, New York begins to go negative as the Midwest opens up, and especially as immigration into New York rises sharply in the 1830s and 1840s, and many of those immigrants then head west themselves.

We can also do something really fun: we can look at the total net migration for New York and New England, and then compare it to the total net migration for the future states of the Northwest Territory:

Notice anything interesting? Cuz I do. New York and New England have a very close, negative relationship with the Northwest territory. In other words, a very large amount of the migration into the Northwest Territory can be explained by reference to migration coming from New York and New England. This relationship breaks down in the 1870s and 1880s unfortunately, but it's pretty tight up through the 1860s. All that to say: yes, the Midwest really is settled by the Yankee Diaspora to a large extent.

We've talked a lot of Albany in this episode, so it's only fitting I give you a population chart for Albany. Here you go:

As we discussed in the show: small population until the eve of the American Revolution, then a huge boom.

However, it's worth noting that, in terms of the growth rate, the early periods really were not so slow. In fact, growth from founding in 1614 to population estimates in 1710 suggest that this period had higher growth rates than the period from 1710 to 1790, or from 1810 to 1820.

But the real champion-periods for growth in Albany are, of course, 1790 to 1810, and 1820 to 1830. These periods routinely saw annual population growth of more than 5%, fueled largely by the influx of Yankee migrants, who rapidly outnumbered their Dutch predecessors in Albany.

A final note on Albany: we discussed the urban fires there, but I left out a historically tantalizing tidbit. There were in fact a series of fires in the Dutch parts of the city, and elsewhere. But several of these fires, especially one particularly large one that destroyed a major Dutch district, were alleged to be arson, and "hooded individuals" were spotted fleeing the scene rather suspiciously.

It is possible that these fires that reshaped Albany were not accidental, but in fact set for the purpose of destroying Dutch houses, enabling Yankee Americanization efforts to keep rolling forward. Elkanah Watson news clippings of these fires in his journals, though that's not saying much, because he kept news clippings of basically everything. Still, given how gleefully Watson exaggerated the story of the vigilante waterspout-chopping, and how clearly he felt architecture was a key element of American culture, it is kind of darkly curious to wonder if maybe this Yankee crusader took matters into his own hands, letting flames do what we could not accomplish politically.

Some resource links:

1. Liz Covart's article on Albany.

2. Liz Covart's excellent Podcast about early American history.

3. Michael Barone's book, which is really a must-read if you want to engage in the debate about what historical American migration means today.

Episode 8: Enslaved Migration, the Finale

We're finally done with slavery! yaaaay!

I've already presented extensive data on population and slavery. So today, I'll just offer some tables on the progress of emancipation.

When I assembled this dataset, I had this article in mind

Pretty remarkable I think that we've got a chart of the great American social movements... except for slavery. Whoops.

So what does it look like if we include slavery?

Well, I don't have the underlying data for the above chart so can't duplicate it (I've asked the authors; no response), I can provide a chart showing the rise of emancipation. Now, to do that, there are different methods. We can look at any emancipation or initial emancipation; we can look at full emancipation or complete emancipation; or we can look at partial emancipation.

If we only care about when a place made its first step towards emancipation, i.e. when the political will for emancipation became sufficiently dominant to pass a law, then we care about "initial" emancipation. But if we care about when slavery was actually abolished, i.e. when nobody was enslaved anymore, then we care about when the emancipatory process was completed. On the other hand, maybe we want to give some reward for even partial, hesitant, early emancipations: so we score based on partial emancipation.

The below chart shows the number of admitted states in the union that were free states according to each method, as well as a line showing the total number of admitted states.

As we've talked about, the solid "free state" bloc doesn't really become apparent until the 1840s when many states with gradual, partial, or incomplete emancipation move towards full emancipation. Then in 1863 we get the Emancipation Proclamation covering some states, and then in 1866 the 13th amendment becomes effective in the rest.

But, hold on. Much of the debate about slavery wasn't about the states. It was about the territories. So now, let's ask how many future states are free.

Above, you can see a new story. Even in the 1840s and 1850s as more states were completing their emancipations, the territories were becoming less free. Initial or partial emancipations were being rolled back by legislative and judicial actions. But then of course, in the 1860s, slavery was ended throughout the country.

That's the march of slavery. It took the country 80 years from its founding to abolition for this process to be completed. At the territorial level, there were advances for the cause of freedom, and setbacks, like the case of Utah that we mentioned in the show. Freedom did not advance uniformly or without temporary defeats.

And with that, I conclude my blog for episode 8. Hope you enjoyed the show!

Episodes 6 & 7: Enslaved Migration, Parts 3 & 4

Whoops! I forgot to post a blog for the last episode! So this post will have data for episodes 6 and 7. So much data!

As can be seen, U.S. cotton production began with just 3,000 bales in 1791, but rose to over 4 million bales on the eve of the Civil War. That war decimated cotton production, though it should be added that I think this graph excludes cotton production in areas outside of Union control, so actual cotton production is probably substantially higher.

The above chart shows the arrivals of enslaved people from 1770-1807. As you can see, arrivals were high before the War of Independence, fell during it, then spiked up again afterwards as pent-up demand was met. But arrivals wouldn't surpass the 1786 level until 1794, after the invention of the cotton gin.

I don't have a great explanation for the dip from 1797 to 1800. It's possible it may relate to piracy by the Barbary states, or potentially economic disruptions due to the Napoleonic Wars and their attendant violence.

Good data on domestic migration of enslaved people is hard to find. But here are maps from the Schomburg Center for Research in Black Culture:


The estimates I offer come from scraping lots of different articles and books. But put broadly, we don't have a great estimate of enslaved migration, and there are very significant limitations to estimating how many enslaved migrants migrated as a result of a sale, versus migration accompanying the same master. While it may seem unimportant to some, these were different experiences that I'd like to be able to separate. Unfortunately, it's very hard to do so.

Also, you can find a really cool decadal interactive map of enslaved migrations here.

That's it for Episode 6! Not tons of stuff, but some. Now on to Episode 7!

Talking about "internal expansion," we can see that from 1820-1840, population growth in the south was overwhelmingly concentrated in cotton-belt states (and Missouri, where St. Louis led population growth benefiting from hefty German immigration).

Alabama, Tennessee, Georgia, Mississippi, all have a whole leg up on Kentucky or Louisiana, while the old south in the Carolinas and the Chesapeake states saw much smaller growth.

But new states like Florida, Texas, and Arkansas also look really small. That's just because of low initial population. If we present growth as a percent of initial population, we get:

Holy cow! Texas saw huge growth! Arkanss, Missouri, Mississippi, Alabama, and Florida all more than quadrupled in size, while Louisiana, Georgia, and Tennessee all doubled. Meanwhile, Kentucky, the Carolinas, and the Chesapeake states saw much, much slower growth. Southerners were moving westward, especially to areas where cotton could be grown profitably, like Texas.

And... that's it for Episode 7! These episodes were a lot of narrative and elaboration on the data. If you've got any questions or comments, feel free to leave them below, and I'll try to get back to you!



Episode 5: Enslaved Migration, Part 2

Today's episode wrapped up our bit on early enslaved migration. Many of the charts from last week are still relevant, especially the chart about the composition of immigration. Check that post out for a reminder of the data.

This post will have comparatively little to add. The South Carolina and Louisiana stories are comparatively data-poor on their own. In a future post I will have a fun chart visualizing the development of emancipation throughout the United States, but I'm gonna save that for later. And in the blog post for the next episode, I'll provide data on the growth of the enslaved population and the prominence of slavery by state.

That mean's that for today's blog post there's really not much data to present. The only chart I think is due here is a final chart showing where enslaved people arrived in the U.S., versus where they resided in 1775. So, here goes.

The above chart shows, in the red columns, the estimated number of enslaved people who arrived in each area. "Arrival" here means "disappeared from immigration record-keeping systems." It doesn't mean "ultimately resided there." As you can see, Virginia has a way bigger population compared to arrivals than most regions, while South Carolina has way more arrivals than you'd expect given the population.

There are three explanations for this. Any of this data could be wrong. The arrivals data is from a mixture of Census and Transatlantic Slave Trade Database info. They could be allocating regional disembarkations incorrectly, a note they explain in the usage guide for their data. It's possible some of those Charleston disembarkations actually disembarked in Virginia, but the boat from the Caribbean made a first stop in Charleston, and the TSTD incorrectly attributes too many disembarkations to the first port of call (Charleston) rather than a later port of call like Alexandria or Jamestown.

The second possible problem is in the enslaved population data. There was no Census in 1775. Many of the colonies held local Censuses in the 1760s and 1770s, but their methods were not all the same, and the quality of the data is not always very good. For example, in South Carolina and North Carolina specifically, official administrative and Census data from before 1775 give very widely diverging estimates depending on the exact source and year of collection. I have deferred to numbers provided in the US Census' Historical Statistics Section Z where available, but the reality is that margins of error of 10,000 or 20,000 enslaved people in the major slave-holding colonies are certainly possible. Those are big margins.

Finally, there's the explanation I noted in the Podcast itself: many enslaved people arrived in a given port, were sold to a domestic owner, cleared customs/immigration, and then were sold and moved again domestically. Intercoastal shipping of enslaved people as well as land-based trade is a documented fact. Census records show Charleston harbor exporting at least a few hundred enslaved Africans each year, mostly to more northerly colonies. Over 75 years, this could be 5,000 or 10,000 enslaved people sold away from South Carolina, towards other colonies like Virginia or New York. Likewise, enslaved people arriving in Delaware or Maryland were almost certainly sometimes sold into Virginia.

I can't give a good estimate weighting these three potential sources for error. Undoubtedly they all play some role. But exactly how much? I can't really say.

Luckily, beginning in the next episode, we're into topics where I have much more, and more reliable, data. 

Episode 4: Enslaved Migration, Part 1

We've begun our segment on enslaved migration. It's gonna be a rocky road for a little while here at Migration Nation, because this topic is fraught with difficult issues. However, the plus side is that I have lots and lots of data to share, so the blogs should be interesting. Let's dive right in.

Let's start with Hernando de Soto. For those that don't know, he was a Spanish explorer and conquistadore apparently searching for the mythical "Fountain of Youth." More importantly, he's a guy who wandered through the southeastern United States for years with a sizable party, and came away thinking it wasn't worth colonizing seriously. This tells you something about Spanish vs. English vs. French priorities. Spaniards wanted to conquer. French wanted to trade. British wanted to settle. Hernando de Soto found few societies whose conquest could yield great riches. So Spain saw the area as a throw-away. But Britain found fertile areas ripe for settlement.

So here's a map of De Soto's possible route, from Wikipedia:

Next up: some basic western hemisphere slavery statistics. This is from the Transatlantic Slave Trade Database. We mentioned that the Caribbean was the really big destination for the forced migrations of enslaved people. We weren't kidding. Here's some estimates of slave trading from Africa to the Western Hemisphere:

As you can see, the biggest American receiving-region, the Chesapeake, comes in 12th on the list of destinations. In other words, the future United States was really not a heavy-hitter when it came to the transatlantic slave trade. Even if we restrict to just 1650-1775, to exclude periods where British America wasn't settled, the Chesapeake still only reaches 10th place. 

So when Brian and I talk about the Caribbean and Latin American colonies being the really lucrative plantation-slavery economies, we're talking about Brazil, Jamaica, Barbados, Saint-Domingue, Dutch Guiana, Martinique, etc. More than twice as many enslaved people were forced to migrate just to Jamaica than to all of the future United States. Tiny Barbados received almost twice as many forced, enslaved immigrants as the whole U.S. Saint-Domingue was not far behind. Brazil received fully six times as many enslaved people as the United States over the 1650-1775 period. All totaled up, the entire French Caribbean received over twice as many enslaved people as the U.S., despite having a land area about the same size as Virginia alone, with 3/4 of that land in sparsely-populated French Guiana. Martinique, at just 436 square miles, received 183,000 forced, enslaved immigrants. Virginia, at 43,000 squares miles, received no more than 100,000, even when we include domestic inflows.

 As we'll discuss more later on, the map of major slave-receiving areas is basically a map of where sugar was grown. But there will be a lot more detail on sugar and slavery in the next episode.

I'm also from here on out gonna try to link to Podcasts we like. So, here's the site for Revolutions. It's good. The series on the Haitian Revolution is particularly relevant to the current discussion we're having at Migration Nation.

Now we're on to indentured servants. The data is really fuzzy here. I've got pretty good data on enslaved immigration. I've got good total estimates of prisoner movements, which I've just distributed evenly. And I've got reasonable estimates of the total number of indentured immigrants.

But allocating them by year is a bit trickier. To do it, I just come up with a share of non-prisoner, non-enslaved migrants for each year who seem likely to be be indentured. That share is higher during peaceful times and higher before Bacon's Rebellion in 1676. The chart below shows my annual estimates for each kind of immigration from 1630-1825. 

As you can fairly clearly see, indentured migrants were the dominant group until the 1680s. Free migrants briefly took the lead until 1700, when I begin allocating prisoner migrants. Then enslaved migrants take the lead, until a brief free migrant dominance in the 1710s. Free and enslaved migration remain close until 1740, but after the 7 Year's War until the Revolutionary War, enslaved migration completely dominates free migration. 

But then, after the revolution, things change. Free migration heats up, while enslaved inflows are lower, except in the odd year of 1807, which will be discussed more later. Prisoner inflows certainly cease. I have also opted to show indentured inflows ending in 1803. This is not technically correct; some indentured immigration continues until 1917. But legal changes in the UK in 1803 make British indenturehood vastly less prominent, and most immigrants in this period were from the U.K. U.S. courts and laws were also increasingly hostile to the enforcement of indenturehood contracts against white servants, and in 1833, debtors' prisons were closed down. This resulted in easier ways for indentured servants to escape their bonds. The issue is I don't have even ballpark-estimates available on post-Revolutionary indentured migration. It does seem that it became much less frequent, especially after 1803. But I don't have data on how much less frequent. So I feather it out after the Revolution, but understand this is probably inaccurate.

We can also combine all the unfree immigrants together.

As should be clear, the period 1630-1775 was the Age of Unfree Immigration. Of the about 865,000 immigrants who came to the future United States during this period, 610,000 were either enslaved, indentured, or prisoners. The prototypical pre-Revolutionary immigrant, then, is not a free Puritan settling for religious freedom, or a British craftsman looking for new opportunities. No. When you think about pre-Revolutionary immigration, think about bonded laborers, slaves, and prisoners. Most of our nations' founders were the dregs of society, the poor, the downtrodden, the indebted, the criminal, the enslaved, and the kidnapped. The stories we tell of planters and Puritans are stories of the elite minority of immigrants.

So where did the enslaved immigrants go? Well, I'm actually not going to give you data on that today. Because it would be a spoiler! The next episode will finish addressing the distribution of enslaved people before 1775, and that is when I'll publish my estimates of inflows for each region.

I guess that's all for this post. Until next time!

Episode 2: The Birth of Bourbon, Part 2, History Notes

So for the new Episode, I just have a few data notes, mostly concentrated early in the episode. The whole James Wilkinson bit has relatively few data points that spring to mind, and I presented Kentucky migration data in the previous post. So check that out if you're interested in the Kentucky data.

So first off, let me offer some information about western Pennsylvania.

The above chart shows the Census-reported population for each county in western Pennsylvania. I've chopped Pittsburgh out of Allegheny County, then summed up the westsylvanian counties, minus Pittsburgh; i.e. the potential rural population in the area. Census data collected in 1790 and 1800 obviously didn't capture 13,000 militiamen showing up in the mid-1790s. So the high red line is the total westsylvanian rural population. For comparison, I've put the population of Philadelphia.

As you can see, Westsylvanians did have something to complain about. There were more of them than there were Philadelphians, yet it should be transparently obvious that Philadelphia has played a larger role in early American politics than Westsylvanians. That's because of them mountains there. Tellingly, this imbalance between Philadelphia and Pennsylvania's growing central and western populations led to a shift of the state capital in 1812 from Philadelphia to Harrisburg, a smaller city out on the eastern fringes of the Appalachian mountains. Still separated from the west, but at least symbolically closer. By that time, of course, there were better roads through the mountains too.

So how about that demographically significant army? 13,000 militiamen (presumably plus support personnel?) in an area with just 66,000 or so residents. That's pretty crazy. This is an issue I haven't thought about very much in the past, but I really wonder about the demographic footprint of armies now.

Consider the Civil War siege of Vicksburg. Vicksburg's county had 20,000-25,000 residents. The siege, lasting almost 2 months, involved 100,000 combatants. How insane is that? The battle of Gettysburg, fought over 3 days, involved almost 180,000 men in a county with just 30,000 people. Now, these battles occurred in a larger context of war, and in a period with better infrastructure, which could limit the demographic impact.


But 13,000 guys camping out in western Pennsylvania for a few weeks or months, at a time when probably few of them would have even considered that journey otherwise, seems likely to have some kind of impact. Aside from any local dalliances, it seems plausible that military procurement would alter demand for goods and services, altering urbanization patterns and industrial mix and labor demand. And it would also seem almost inevitable that some of these guys would say, "Well hey, I made it this far: why not just stick around, or go farther west?" Militiamen are likely to be able-modied, male, young adults: ideal candidates for migration. Add in draftee discontent, likely antifederalist sentiment among many militiamen, and the transaction cost concerns discussed in the episode, and I'd be shocked if the follow-on migration from Washington's army is under 500 or 1,000 people.

I meant to give you a route map in my last post, but I forgot. So here's a belated sketch map of the major westward migration routes, starting points given as New York and Philadelphia, the two biggest ports of entry as of 1780-1800. Alexandria, Virginia was also a big entrepôt at that time, and its routes basically differed by having some roads across northern and central Virginia, but by all accounts those routes were much less traveled, and many migrants arriving by Alexandria still headed up to Philadelphia. Immigrants arriving via Boston would eventually link up to the Mohawk Trail in the north (sleigh-riders!). Immigrants via southern cities like Savannah and Charleston were not doing much westward migration in the 1780s, or honestly even up to the early 1800s. Most migration of southerners in the early period headed north via Kentucky and Tennessee, or around the coastal areas of the Gulf. Not until the War of 1812 did mass migration across the southern states really get going in earnest.

As you can see, my "just four routes" was a bit of a simplification. But the remarkable thing is I found little to no evidence of migration in routes through northern Pennsylvania/Southern New York, or directly through West Virginia, or directly through the TN/NC borderlands, etc. There are obvious topographical reasons why (i.e. big reason' mountains), but still, as a 1790s migrant, I would be pretty frustrated that to go west, I had to spend half the journey doing cross/cutting north/south moves.

So aside from the route map above, I figured it'd be good to give a map of north america in 1790, to give a sense of what's going on. So, here it is:

From wikipedia. Notice independent Vermont. That's what some Kentuckians wanted.

From wikipedia. Notice independent Vermont. That's what some Kentuckians wanted.

So this is something I've become really interested in since starting research for this podcast: the weird absence of Spain from the classic American history textbook narrative. Sure, it crops up here or there, but American history as told by my high school and college education pretty much cast the US as existing in an international millieu of France, Britain, and essentially nobody else.

But look at that map. Spain is everywhere. Spain is all the way up in Canada. Spain is claiming Alaska. Spain has forts in Florida. Spain is climbin' in yo window. This is a big deal. So this Spanish secession issue is really interesting, because it represents a kind of geopolitical what-might-have-been that's really interesting. The War of 1812 is often considered a "2nd war of independence." Because obviously, if the US was going to be re-colonized, it would have been by Britain.

But maybe that view is wrong. What if we've forgotten that a plausible post-1783 world was one of a fragmentary United States being gobbled up piecemeal by Spain? Attempts were made in Westsylvania and Kentucky. Later on, (former) Spanish holdings will be gobbled up by an expansionist United States. Indeed, if any readers know a really good history of U.S.-Spanish relations including this early period, I'd be really interested. Let me know.



Episode 1: The Birth of Bourbon History Notes

Accompanying each episode, we'll have a post (or posts, if Brian blogs too) offering some more information and visuals about our topic. So this one will cover Episode 1. These posts probably won't be the most poetic writing, but hopefully they'll provide a little extra material for those interested.

So, here's some more information about the episode.

First off, we mentioned some changes in gross migration rates and the population-redistribution effect of migration. So here's some charts on that. First, a chart showing gross migration rates in the US, and a gross migration rate estimate for China. The source for the Chinese data is here. The average net migration rate estimate is based on a state-level migration database I've built going back to 1780, based on a plethora of sources. I regard it as very reliable. The gross migration rate after 1850 I regard as very reliable too. Before 1850, I do not regard it as being nearly as reliable. So treat those very high gross rates with some caution; they're based on very limited evidence.

As we discussed in the show, 1850s migration rates are at or below present-day rates of migration. However, the population impact of modern migration is much, much lower. In fact, recent migration has had the least population impact of any migration flows in American history. There is no time in our history when migration has had less impact on aggregate population. So, fun times to be a migration researcher.

It's possible that gross migration rates before 1850 were higher. Growing transportation networks, opening frontiers, and relatively low immigration during much of the period may have all contributed to forces pushing, pulling, and easing travel westward. As you can see, the population impact varied, but typically remained between 1.5% and 2% a year for most states.

The Chinese data is flawed, but the best I have on hand. As you can see, gross migration was far lower than in the US during the late 1990s. In the mid-2000s, Chinese domestic migration has been substantially greater than US domestic migration, but still below U.S. peaks in the 1950s, or potentially higher peaks in the 1810s and 1820s.

The chart above shows the "Migration Effectiveness Index," MEI, for the US and China in given periods. MEI measures the lopsidedness or unidirectionality of migration. It's calculated by dividing the absolute value of all regional net flows by the total gross migration flows. So if all regions have perfectly balanced migration (net flows in each region=0), then MEI is 0. But if all flows are unidirectional, i.e. everyone moves from Virginia to Alabama, and nobody moves from Alabama to Virginia, and that's true for every state-migration-pair, then MEI will be 200, because every migrant is counted twice, once leaving, once arriving.

Generally speaking, MEIs of over 50 are pretty large. If sustained for years on end, it's considered to be an extremely strong migration dynamic, reflecting an enduring imbalance in terms of push-pull factors or demographics. A value over 100 is considered quite extreme; that implies that "recipient regions" receive net inflows equal to more than 50% of gross inflows. That's extremely rare, especially for long periods of time.

As you can see, China's migration, despite being below or near US levels, is more unidirectional than US migration has ever been. This may be partly due to measurement error, as the Chinese government restricts internal migration. But this shows the key point made in the show.

Next topic: historic GDP and income. There are plenty of sources on this. You can find two specific papers I read here and here. The figure given of $50-$100 in GDP per capita I then multiply by some household size best guesses, for our ballpark of "a few hundred bucks a year" in average American income.

Conestoga wagon prices are here; I take this source at face value. If somebody knows a better estimate, I'm interested, but it's not essential. It's also worth noting here: in the show, I quoted my erroneous historical understanding that Conestoga wagons were used to settle the west. They probably were not. Most sources I've found suggest Conestoga-style wagons were no longer being produced after the 1840s, and 1850s at the latest. The Oregon trail wagons were of a different design, made much, much lighter to enable the long journey.

I've got some estimates of Kentucky historic population, 1760-1820. They're below. Now at some point, I'll do a big methodology post on my annual population and migration estimates. Not today. But below you can see my estimates of population, net migration, pop growth rates, and net migration rates.

I know the migration line is small, but you can see it rises a lot after 1790. That's almost certainly due to statehood, better roads, Wilderness Road improvements, the end of Native American violence in the state, and possibly some frontiersmen seeking escape from stronger government interference in the east. By 1820, however, Kentucky migration actually turns negative, as many Kentuckians move west or south. Consider two of the Commonwealth's most famous sons in this period: Jefferson Davis and Abraham Lincoln. Davis' father moves to Kentucky in 1793 and farms there. Davis is born in 1808. Then his family moved to Louisiana, then an even more newly-opened state, in 1811, then Mississippi to run a plantation in 1812.

Lincoln's grandfather moved to Kentucky in the 1780s, a very early frontiersman, but his father moved frequently. He settled down in 1808, and Lincoln is born in 1809, in Kentucky, less than 100 miles from Davis. But by 1814, Lincoln's father has lost a major land suit, forfeiting most of their farm, and so they move to Indiana.  The land of opportunity in the 1780s, 1790s, and 1800s, by the 1810s and 1820s, Kentucky's position in the migratory life cycle had changed.

The above chart shows population growth, and that growth broken down into net migration and natural increase (births - deaths). Immigration, likely small, would also be included in natural increase here. My methodology creates some wonky years around the Censuses, but aside from those you can see that natural increase was fairly stable.

Migration, on the other hand, changed radically. Net migration was just 1 or 2% in the 1780s, but by the 1790s, Kentucky was adding 8% or 9% a year to its population through migration. That number declined from the 1800s and 1810s. No longer the frontier, by 1820 Kentucky was losing people, in no small part due to the sale of enslaved people into the south, a topic we'll discuss more in future episodes.

That's all the material I've got. The comments section is open. Leave me questions and criticisms and I'll respond as needed, including updating this post with more information.