The Antebellum in Census Numbers

[MattLul]

For a while I've been gathering the US census records from 1790-1860 and slowly charting them along many different values usually separated by the States that would become the Confederacy and the rest... I figured I'd share what I have so far.

To be clear I started charting data points that sounded interesting and not with a specific goal to prove a significant trend, so some of what I share won't show a significant trend proving a point. With that said some of the trends that don't show a change are significant in and of themselves.

My primary motivation is that often I find myself (and others) referring to believed trends or specific smaller trends of growth in slaves, white populations, free populations, etc... so wanted to chart it all from 1790-1860 to get a big picture along many ways to look at that data.

It's been a side project for a while throwing this all in a spreadsheet, summarizing each stat in each decade, linking them to a single page, charting them etc. Originally I considered importing the data into a database and charting it via that or by creating my own little app (I'm a Software Engineer). I chose to just do a spreadsheet though now in hindsight it probably would have been faster to do the other and could chart things probably easier... I might still do that as my next step.

In any case here are the charts, I think I broke it down pretty much in every way that one could though if anyone has any requests feel free. Additionally I tried graphing some data in different ways, if anyone has any request in that regard just let me know.

I'll also add a disclaimer that it's entirely possible I linked some data wrong and some data point in a chart is wrong, let me know if you are curious about a specific data point. There was a ton of linking, copying and pasting, formulas, etc. So plenty of places for human error. Another reason I might import the data into a DB and map it out like that next.

All data gathered from https://www.nhgis.org/]https://www.nhgis.org/[/url]

First are two charts mapping the percent of slaves in each region. So in 1790 36% of the CSA States population were slaves. We can see the CSA grow from about 35.5% to about 38% in 1860. While the rest of the US shrunk from a slave percent of about 6.8% down to just under 2% by 1860.

The two things I found most interesting is the relatively small range of change within the CSA States. Basically they always had a large percent of their population as slaves (though it grew slightly)... and the relatively large fall in the salve population outside the US.

This also reveals how much of the slave population had always been centered in the slave states, at least proportionally, I have totals below.





Same data in a stacked graph



Here we look at totals. You can see the CSA slave numbers grow at a pretty consistent and large level, while outside the numbers very slowly grew, but far slower than overall population growth.



Same data in a stacked graph



Here we look at the CSA vs US and their percentage of the total slaves in the US. Unsurprisingly we see a trend of the CSA States gaining a larger portion of the slaves. We can also see again that the CSA States always had a massive portion of the Slaves from the forming of this Nation on.



These two pie charts show the proportion change a little better




Now we start to look at the Free Colored population, in this case the US and CSA's percentage of all the Free Colored population. We can see here that the US always had the majority of the free population and it fluctuates with a slow trend in gaining a bit more while remaining mostly the same divide. Keep in mind that this is despite massive slave population growth in the South. The US goes from having about 2/3rds of the Free Colored population to having almost 3/4ths by 1860.

View attachment 225

Now we look at totals and we can see the free colored population grow in both regions, though clearly much larger growth outside of the CSA states. The South has about half of the free colored population the rest of the US has in 1790, and only about 1/4th by 1860.

This one I found interesting since it shows more about the free colored population change than about any specific region. This is the Free colored population as a percentage of each regions total population. We see it grows in both regions and then declines in both regions. The decline starts in the US between 1810 and 1820 while it starts in the CSA in between 1830 and 1840. Keeping ind mind these are proportional numbers, so population growth in each region effects this number. In the end the Free colored population was gaining proportional ground in both regions and started losing it by 1860. As we see in the previous graph the numbers grew in total, just not as fast as the rest of the population.


Full data citation
https://www.nhgis.org/]https://www.nhgis.org/[/url]
Minnesota Population Center. National Historical Geographic Information
System: Version 11.0 [Database]. Minneapolis: University of Minnesota. 2016.
http://doi.org/10.18128/D050.V11.0.

 

[MattLul]

Here is the total White population. We can see population growth in both regions and can see the trend most of us are familiar with, where the White population in the US outpaced the CSA White population, starting at almost a half of the rest of the US White population and ending at almost a quarter of the rest of the US White population.



Finally I tried to do a complete breakdown via a stacked chart. Keep in mind that the 1860 entry comes up short of 100% due to it being the only census to add groups beyond White, Free, and Slave (such as Indian). I chose not to chart those since the other charts didn't include such data points. We can see the US White percentage grow from 55% to 67% of the total population and the CSA White go from 24.5% to 17%.


The same data but with totals, can really see the US White population grow so fast... as ell as the growing CSA slave population. You can see how outside the CSA the growth shrinks the South's portion.

amweiner said:

Hey @MattL, I only skimmed your OP, but this is AMAZING data!! Thank you so much for this thoughtful work, and for sharing it with us. I look forward to going over this more closely, and really appreciate you taking the time to compile this..it can't be easy.

Respectfully,
Adam

Thanks, it definitely has taken a lot longer than expected, slowly been chipping at it.

For reference here's my source Google Spreadsheet, in case anyone wants to look at the raw data (some columns are hidden though, only for data points I'm not using, not all the census summaries have the same data points) and in case someone sees an anomaly and might find an error I made linking all these data points.

https://docs.google.com/spreadsheets/d/15Hp-CY9wJ5a-8KTufBm5WM6K8jPrjw2cjlAJE4s2BVU/edit?usp=sharing

 

[MattLul]

[url=https://civilwartalk.com/goto/post?id=1605233]mofederal said:[/url]

I see that you broke them down by CS and US. My question is on which side did Mo., Ky., and Maryland fall. on, seeing as they were considered as Cs states also. My interests lie on the actual state's numbers.

[url=https://civilwartalk.com/goto/post?id=1605233]mofederal said:[/url]

I didn't mean my earlier post to sound bad. What you did is amazing work, and I know how hard it is to do that work. You deserve an award for that. I mean that.

Not taken that way, a perfectly valid question... The short answer is used the typically accepted states for US vs CSA, so Mo, Ky, and Maryland count as US... A slightly less short answer (for you at least) is I linked the source spreadsheets in my second post so feel free to look and you can see in there, there's a column where I have "CSA" listed for CSA states, and it's empty for non-CSA states (I'll have to flag them US but was quicker to just put it as nothing for US).

The long answer is that's a good question and point. If you look at the spreadsheet you'll see the census data includes territories before they were States and things get a bit murky. Fortunately what becomes the CSA (from the more accepted classification vs the contemporary wishful thinking) is more or less straightforward...

I will probably post a summary of what States/Territories are flagged as US vs CSA per census decade in some list, or chart so people don't have to dig it out of the spreadsheet.

You question brings to mind a few other comparisons that might be interesting:

* Three divisions... US, CSA, and Border States. I suspect the trends among the three would be quite interesting.
* Trends via counties, most of that data should be accessible though a real trick is how the counties shift and change over this time, which might make a crude rendering a bit less useful, though maybe still of some use. For example one county might divide down to four others, that might cause it to lose a percentage of slaves. That of course might not mean they actually lessened in slave percentage, just county lines moved to shift the data. Trying to correlate all that data would be a ton of manual work and hard to even figure out how to represent that. With that said I suspect a county breakdown would be very illuminating for sub-State trends
* Per state comparisons... I can see a line chart with every slave and it's slave percentage, white percentage, etc. Could see which States trend together and apart, etc.

Before I approach any of these I'll probably import it into a DB and start building a little app (maybe a web app) around it since this would get much more repetitive to break down the data this way.

I've started to add elector information into my sheet to try and track white per elector representation, adjusted house seat representation per white person (that whole 3/5ths rule comes into play here). I'm seeing some interesting trends already, not only in representation skewing more than I thought it would, but in ways that I didn't expect, trying to verify my data is correct and what might explain the trend I'm seeing. I never really looked into how house seat count assignment worked but just looking at that the numbers don't seem to be quite as consistent as I thought they would so wondering if I'm using wrong numbers or if Congress weighed certain states more arbitrarily than I thought they would (towards the North a bit in the beginning, then interestingly towards the South, even beyond the 3/5ths rule).

The beauty of looking at data trends, it helps highlight things that you might just not have known or understood, especially human influence which usually results in inequality.

 

[MattLul]

I've expanded the total slaves data to show growth rate from previous decade, broken down by US/CSA and total... I also include average growth rates for both and total during 1790-1860. Additionally I created a speculative chart expanding the average growth rate per region to 1900 to get an idea how slavery might have grown beyond 1860 if it hadn't stopped.








A whopping nearly 11.1 million slaves by 1900 if it kept growing at the same average rates, with an average of 30% growth in slaves each decade in the South.

 

[MattLul]

Another interesting data point, here are the seceding states, their date of secession, percent of families owning slaves, and percent of slaves of the population It's actually pretty impressive how closely the order of secession roughly matches their portions of slaves and slave owners.



For thoroughness here are the US state sorted by slave percent as a population, as is not surprising all the top States seceded and the border States come in next where support was most split elsewhere

 

[MattLul]

So I started to map population representation data according to the Constitutional 3/5ths rule. Was curious how that trended with the growing slavery in the South and the decline of growth outside of it.

This first chart is basically White + Free Colored + (Slaves * 0.6). 0.6 being 3/5ths.

I then compare it to the White population of the region. Considering the White population basically controlled things, how the representation numbers compared to them seemed the most interesting.




You can see how the population representation grew larger than the white population in the South as slavery grew, basically giving those White people an outsized representation, that relatively speaking they got over presented more than those in the rest of the US. Of course the US population growth overall was fast enough to eventually outpace that.

Here is the population rep to white population ratio. Roughly this is the percentage of population representation each white person had in their region due to slave 3/5ths representation.



Again as the slave population increased in the South you can see it grow from 137% to over 141%. While the rest of the US decreased as slaves became an ever shrining portion of the population.


I've started to map the other interesting representation factors and will post some charts regarding elector representation which includes the above representation combined with smaller states having extra representation due to the two senate seats as well as seemingly an arbitrary level of assignment each time they updated it and made the numbers fit, some States just got overrepresented and others under.

 

[MattLul]

Some more representations of Population Representation vs White Population. Wanted to compare the percentage of all White Population for each region compare to their population representation taken into account slaves (and a small amount of free colored people).



Interestingly this graph shows the inverse result. The US White percent was higher than their share of population representation. While the CSA was the opposite, their White percent was lower than their share of population representation.

More specifically here are the numbers for 1790 and 1860










These pie charts really show how outsized the White people in the South had representation due to the 3/5ths rule, though it also shows the growing population outside of the South eventually growing beyond that.

 

[MattLul]

wausaubob said:

It is all due to the fact that European immigration was almost entirely northern. When immigration picked up, the slave states began to lose representation.
The demographic affect also reflects that legal and illegal importation of slaves continued after 1787, and was accelerating after 1850.
An anti-slavery president already had the legal tools to smash the illegal slave trade. If he had an adequate budget he could also hire the military and law enforcement tools he would need.
15,000 slaves per year, is only about 3o ships at 500@ and it adds up fast if it continues for 50 years. If more slaves are acquired through Louisiana and Florida, it adds up quickly.
There was a unbelievable amount at steak with respect to political control and the future of free immigration.

I agree European immigration North, I absolutely do not with illegal slave importing South. Do you have a source for that number? The estimates I have seen vary wildly because it's simply hard to track. Some are higher and plenty are far lower than that.. The expansion of the slave population likely had for less to do with slave importation but the known natural expansion of the slave population itself by births, either willingly or through some sort of forced breeding or other gray area. Certainly illegal importation played a role but not the primary method of slave population expansion.

Also the reality is people have tried to stop the importing of something all throughout history, it never is fully effective, often fairly ineffective.

 

[MattLul]

Somehow I missed charting this... but this is the Black and US total percent... Basically slaves + Free Colored. Obviously neither region had large numbers of Free colored so it doesn't change the number greatly, but it does bump the total Black sin the South up to 40% in 1860.



The decline of Blacks in the US matches the decline of slavery since the Free Colored population stayed mostly the same percentage (meaning it grew in total number to keep up with population growth).

 

[MattLul]

damYankee said:

I appreciate the hard work but there are a few things I'd like to point out,
A. The was no CSA in 1790.
B. Europeans immigrated to southern states. ( duh ) how do you think white people got there? )
The first place Sicilian immigrants settled in the US was New Orleans. French settlers also. Every coastal city in the South was established by European settlers.
German and British immigrants in Texas were instrumental in the creation of the cattle industry. Until WWII there were several newspapers printed in German throughout Kansas, Iowa, Nebraska and Texas.
European immigration to the US, Canada, Central and South American, Australia, India, So. Africa ebbed and flowed in direct relationship to war, famine and epidemics in Europe, contrary to our myopia, it isn't always about us.

A. Absolutely, clearly there was no CSA in any of these time periods... The states flagged as CSA are simply States that would become part of the CSA. Obviously this can pose problems, such as Tennessee not being a state in 1790 at all! In some census entries the future states are simply Territory (like Mississippi) and of course the territory often had somewhat different borders than the future State. Many of those issues mostly go away when looking at the whole region though, State specific looks for those States during those times could be problematic though.

I link the source spreadsheet and you are welcome to take a look and identify any flaw in my methodology.

I think looking at the numbers we have for how the regions progressed up until the Civil War is very useful. Maybe you disagree.


B. Agreed... I'm not sure what your point is, could you please clarify? European immigrant into the South declined... that is probably something I can chart and would be useful in looking at the slave numbers, etc.

 

[MattLul]

wausaubob said:

I don't think the African American population in the southwest states was due to natural increase. The conditions were difficult. The slaves were being imported from other US states and from abroad.
As you note, smugglers don't keep statistics.
However, the white population was growing based on continued immigration and the slave population was growing apace.

A good point... I'll have to see what sort of chartable numbers I can find on this, I'm certainly curious what sort of data can give us an impression. I know there are studies on the age ranks of slaves during this era and this obviously contributes to our understanding of infant and child mortality for slaves (which was significantly higher than Whites). Of course I believe (without digging up the data right now) that Southern mortality rates were worse than the North in general too.

 

[MattLul]

damYankee said:

The problem is mapping, if we overlay a map of the southern states in 1790 with the southern states in 1860, more than half of what was the actual CSA (1860) was the territory of either Spain or France in 1790.

True, such considerations are important, maybe I'll throw up map pictures for these time periods just for reference, though right now I'm assuming anyone who reads this knows roughly US geography during this time. I will say there is a continuity between the original US States and the States that made up US vs CSA. It's not perfect, like anything, but certainly the people of the time considered a Southern heritage relevant to the secession and formation of the CSA so I think it's useful to see the trend over time, obviously considering other factors as well, such as geographical expansion.

You Concluded the reason that the white population of the northern states outpaced that of the south was because of the lack of white immigration from Europe.
It was more complex then that. The stars crossed when the cotton gin became available about the same time we acquired Florida, Mississippi, Alabama, Louisiana, Arkansas and later, Texas.
My question is how many young whites born in the states that later became the CSA, migrated away to pursue opportunities in the west, or the north?
What was the birth rates?
As for European immigration it is true more ended up in the north.

Well I didn't fully expand upon that, but my opinion is that mostly or at least a good portion of Northern vs Southern population expansion is through immigration, at least the difference in population expansion (both experiencing natural expansion.

You do race good questions... 10 of my ancestors fought in the Civil war, 5 on each the US and CSA. Much of my Union ancestors were actually Southern descendants who migrated West as you say, though some were Northern (one from PA I believe and another from NJ that stopped by in Georgia though). Most of my ancestor goes back to the South though likewise some lines do lead North in a similar expansion West pattern.

Before 1850 finding data of Westerners or Northerners born in the South would be difficult though the 1850 and 1860 census does include such data... I'm not sure such data has been aggregated however, at least in the source I'm using (nhgis). I'll have to take a look. If it does I can chart such things easily, otherwise I might chart some things if I look into the more raw datasets out there that might include it.

As far as birth rates, I agree this is an additional data point to consider. Some of such data has been aggregated from census records (one can look at 1860 vs 1850 and find native born children 10 or under for example) though again I'll have to see if it's been aggregated in an easy format already.

Eventually I want to get the raw County (and State like I am now) data into my own database and run some more complex queries against it all. Maybe even correlate it to a map etc. My Software Engineering mind runs wild with the possibilities of correlations, mapping, etc. Though all about time

As far as immigration I'm putting some charts together for 1850-1860... I think they really reveal how impactful that was.Click to expand...

 

[MattLul]

Ok some charts on 1850 & 1860 charts regarding foreign born people in the US. Unfortunately this data is not available on census records before 1850 (though some other sources have been used by some to get estimates), though it really gives an idea where things were right before the Civil War.

First is a comparison of the percentage of the CSA and US States Foreign Born Whites as a percent of the total of all Foreign Born Whites. We can see the CSA dropped a slight bit from 6.58% to 5.67%. The US increased a bit from 93.42% to 94.33%.



We get a much more contrasting picture when we look at the White Foreign Born as a percent of their own regions. The CSA had 3.4% foreign born whites in 1850 and a slight increase to 4.28% in 1860. While in the rest of the US we have a much bigger increase of 13.76% of the US population to 18.19% in 1860. This was a bit larger than I expected, that nearly 1/5th of the US (outside of the future CSA states) was foreign born with the future CSA states being only about 1/20th.



Here are the totals of CSA and US White Foreign Born compared to the total populations in general. You can see how they all relate to each-other according to population growth.



The final chart I could think of is comparing the foreign born white population growth numbers to the white population growth numbers. Basically how much of the 1850-1860 White population growth was due to foreign born Whites. Here the contrast is most apparent. Almost 8% percent of the CSA White population growth was from foreign born Whites while nearly 30% of the rest of the US White population growth was from foreign born Whites.

 

[MattLul]

Breaking down the Black population on it's own

 

[MattLul]

wausaubob said:

The trend up until 1820 was slow, but in the right direction. Probably too slow. But the slave population started to grow and the coerced labor system became entrenched.

A good observation. It really reveals itself when looking at the data.

 

[MattLul]

Here's a chart with three numbers taken from the 1860 census
* Percentage of slave owning households
* Percentage of slave owners out of white males 20+
* Percentage of slaves of total population
The States have their secession dates listed and are in order of first to secede

 

[MattLul]

MattL said:

I've expanded the total slaves data to show growth rate from previous decade, broken down by US/CSA and total... I also include average growth rates for both and total during 1790-1860. Additionally I created a speculative chart expanding the average growth rate per region to 1900 to get an idea how slavery might have grown beyond 1860 if it hadn't stopped.

View attachment 151336

View attachment 151335




A whopping nearly 11.1 million slaves by 1900 if it kept growing at the same average rates, with an average of 30% growth in slaves each decade in the South.

This 1830's blip was bothering me so I redid these with a couple changes

1) 1830 is an average of 1820 and 1840, this also changes the overall average applied to predicted trends 1870+
2) I changed the charted growth from 120% to 20% to better show the changes between years and trend, I also moved it to a scatter plot with a trend line

 

[MattLul]

A look from a per State perspective

 

[MattLul]

wausaubob said:

Six of the seven largest states in terms of population, prohibited coerced labor. The 3/5ths clause also operates to limit representation of the significant coerced labor states in Congress and the Electoral College. The stuff is about to hit the wind generating device.

Yup, pretty much. I've posted some generalized charts regarding how slanted population representation was due to the 3/5ths rule, here are some of the same ones on a per state level. I also don't think I ever posted my elector charts in a general level of CSA vs US so I'll post those and new per State ones

Some terms

Pop Rep = The total population used for House of Representatives Calculations (which then gets used for Electors added to Senate seats as well)

The formula is basically Slaves = 3/5ths (or times 0.6) + all other population

Rep % White = The Representation each White person has. Basically Pop Rep divided by the White population. The higher the more representation you have based on Slaves padding your numbers.

Elect % White = The portion of an elector each White person has. Basically Electors divided by the White population. The higher the more elector representation you have based on Slaves padding your numbers along with smaller States getting the boost due to 2 electors for the Senate seats each State has.









This might be the most important Elector stat and graph. How much more Elector representation do the future CSA States have for each White person compared to the US States. That first number is 1790, so basically when this system was set up, what the balance was believed to be. You can see how slavery really shot that up.








It's harder to see that trend on a per State level due to the Smaller States getting a boost, though the aggregate is above. The individual State dynamics were related but a bit different than the sectional interest dynamics. This boost to the smaller States via the two Senate seats no matter how small your State was, was intended upon the creation of this system, to balance out the bigger States and the fact the North had multiple smaller States vs the South having exaggerated slave interest.





Despite the future CSA States getting an increasingly exaggerated influence among most of these metrics, the US White population growth was just so much greater to gain ground even despite this.Six of the seven largest states in terms of population, prohibited coerced labor. The 3/5ths clause also operates to limit representation of the significant coerced labor states in Congress and the Electoral College. The stuff is about to hit the wind generating device.

 

[Emilemut]

It always appears to me that the bottom six or so of the ten leagues are about even in terms of the player quality level. Are they all on amateur contracts too?