Mapping the American Coastal Frontier ca. 1800

This slideshow requires JavaScript.

The last post explained how to use QGIS and historic census data to map population density. This post gets to that tricky “so what?” question.

So why map population density in 1800? My current project examines the social, cultural, and physical transformation of the American coast (my shorthand for coast of the United States of America) over the course of the long nineteenth century. I’m primarily interested in the oceanfront between ports, harbors, and the huge natural bays and sounds, that many a European explorer thought led to China, and how they became such a central—arguably essential—part of the American experience.

I’m currently slogging through the first chapter, a survey of the American coast on the eve of its transformation. It’s partly inspired by Marcus Rediker’s magisterial first chapter in Between the Devil and the Deep Blue Sea and similarly favors fruitful generalization over burdensome qualification. An important subargument of the chapter is to establish the 1800 American coast as a frontier. Well…maybe I do a bit of hair-splitting in the chapter. In any case, one of the many ways to define frontier is through population density. Hence, our maps.

Maps, like chapters, articles, books, and lectures, are arguments. I need to see if my map can make the argument that the early American coast was a frontier. So how does one quantify frontier? Fortunately, the U.S. Census bureau did just that in 1874 when it defined frontier as an area with less than six persons per square mile. Clearly, 0-6 people/mi2 should be mapped. But how to divide the remaining population density figures, which range from less than 1 to 2624 people/mi2? QGIS allows you to group your data in different ways: Equal Area; Equal Count; Natural Breaks; Standard Deviation; Pretty Breaks, and, of course, manual (read the QGIS manual for the ins and outs of each). The slide show above shows several of these renderings.

1800 popdens modcomp.pngFor the map above, I thought it would be interesting to compare population density in 1800 with density figures today. Here’s what you need to know. In 2012 (last available figures), the national average was 89 people/mi2. The average of shore-adjacent coastal counties was 182 people/mi2 – take out Alaska and that figure soars to more than 440 people/mi2. Unsurprisingly, only the largest cities in 1800 match average population densities today. Look a bit closer and a few things stand out about this map – first, there is not much “frontier” on the immediate coast. A few counties on the extreme northern and southern edge, the outer banks region of North Carolina, and two counties in South Carolina. In all, 8 counties. We’ll get back to that in a minute. But worry not, all is not lost. The second issue is that, well, there are some inaccuracies. Most notably, Suffolk County, in Massachusetts is wildly out of place. As NHGIS folks warned: “…the 2008-based boundaries also include occasional gross inaccuracies.”[1] Check.

So, not much U.S. Census “official” frontier in the spot I’m arguing was a frontier. On one hand that’s not very surprising considering the data is county level and does not give any indication of where folks actually lived. An abundance of other sources material demonstrate that the vast, vast majority did not live anywhere near the immediate shoreline. On the other, shucks.

Still, there’s more to discover and QGIS provides a bunch of features to further analyze the data we have. The first step is to pull out all coastal counties. Since my projects is primarily focused on the oceanfront coastline, I identified the 37 ocean-fronting counties [see map below]. As with all maps this required making some judgement calls. Counties facing bays, sounds, and other bodies of water that connected to the ocean were not included. I nixed Suffolk County, for example, because I know from other source material that its few oceanfront environs were sparsely populated if not uninhabited in 1800. Further, including Suffolk would have seriously skewed the coastal county figures; it was more than six standard deviations from its 37 counterparts.

chapter1_image1_coastalpopdensity

Run a basic statistical analysis of the 37 oceanfront counties and here’s what we get:

  • 623,842 people (12% of total population) lived in ocean-facing counties which covered 49,790 mi2 (6% of total land area).
  • The population density in ocean-facing counties was 12.5 people/mi2 as compared to the national average of 6.1 people/mi2. The average density for the 388 non-ocean-fronting counties was 5.4 people/mi2.
  • Population density in ocean-fronting counties ranged from less than 1 people/mi2 to slightly more than 118 people/mi2 with seven of the most populated ocean-fronting counties being in Massachusetts and Rhode Island.

So what are we left with? First, the census mapping exercise is not the slam dunk I hoped it could be because the fine-grained data I need does not currently exist in tabular form. (This is one of those few times that I truly hope I’m wrong and a diligent reader will set me straight.) Second, I realized that population density is the tip of the proverbial iceberg so I went back to NHGIS for more census data. Preliminary results: the population characteristics captured in the 1800 census do not reveal any statistically significant differences between the 17 ocean-facing counties and the population as a whole. Dive a bit deeper and there’s some interesting differences, particularly regarding race. But we’ll save that for another post. Finally, I started to (re)learn some valuable tech skills that I’m going to be using much more than I probably anticipate. I look forward to sharing.

[1] https://nhgis.org/documentation/gis-data

Advertisements

2 Comments

Filed under Notes from the Field

2 responses to “Mapping the American Coastal Frontier ca. 1800

  1. that program is really sweet. interesting to see that the coast was equal to the land locked in population; wouldn’t have expected that, especially then! i wish it had epidemiology data. best wishes in your research!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s