Who Settled Dodge County?

In 1850, if you walked the section roads of Dodge County and asked people where they were born, four in ten would have told you New York or New England. Thirty years later, in the towns on the east side of the county, eight or nine in ten answered Germany, or had a father who did. This map shows how that happened, town by town, census by census.

Pick a group and a year on each side to compare. The two maps move together. Click any town for its full breakdown.

Explore the full map

The full map lays any census theme over the 1840 Map of Wiskonsin Territory, the original land survey plats and the 1873 and 1890 plat atlases. Use Fade maps to see the old map through the colors, or Compare to swipe between any two layers. For the best view, especially on a phone, open the map full screen. The map key and the credits and sources are in the panels below the map; click either one to open it.

Open the map full screen ↗

A Yankee frontier

The first settlers came west along the Erie Canal and the Great Lakes. In 1850, 29 percent of Dodge County residents had been born in New York and another 12 percent in New England. They set the pattern: the township names, the mills, the courthouse and the first churches. Burnett, Beaver Dam and Oak Grove were close to 60 percent Yankee-born. Only the eastern edge was different. In Theresa and Lebanon, Yankees were barely one in eight. Germans were already there.

The German east

The tide turned fast. By 1870, German-born residents outnumbered Yankee-born ones about two to one across the county, and the census now recorded where in Germany they came from. The answer was overwhelmingly Prussia. In Theresa, 43 percent of all residents had been born in Prussia; in Lebanon, 39 percent.

The 1880 census asked one more question: where was your father born? Counting the immigrants together with their Wisconsin-born children shows how complete the change was. Theresa was 94 percent German. Herman, 91. Lebanon, 85. Lomira, 82. Hubbard, home to Horicon, 74. Meanwhile, the west side held on to its Yankee roots: Fox Lake, Chester and Trenton stayed under a quarter German. Drop the German stock map next to the 1850 Yankee map and the dividing line runs right down the middle of the county.

Pockets of their own

Smaller groups show up just as clearly once you know where to look. In 1850, a third of the residents of Calamus had been born in Wales, and 14 percent of Elba’s. Ashippun had a Norwegian settlement large enough that Norwegian-born residents still outnumbered German-born ones in 1880. The Irish were a fifth of Emmet in 1850 and the largest immigrant group in Elba as late as 1880.

Farm towns shrink, factory towns grow

Between 1890 and 1910 every rural town in Dodge County lost people. Theresa and Rubicon each lost about a quarter of their residents. The cities went the other way. Mayville nearly doubled, from 1,165 to 2,282, in the years of the Northwestern Iron Company. Beaver Dam grew by 60 percent, Juneau by 43 and Horicon by 39. The children of the settlers were leaving the farm for the factory. Turn on Population change 1890–1910 in the full map to see it.

How I built this map

This map started with a simple question: who actually settled the land around Dodge County? The county histories give you the first settlers, the mill builders and the church founders. I wanted everybody: every household, in every town, census after census.

• The census books. Every ten years a census taker walked every road in Dodge County and wrote down every person in every household, with their age and where they were born. Those handwritten pages have since been transcribed into a database of the complete census returns, kept by IPUMS at the University of Minnesota. I requested every Dodge County record for 1850, 1860, 1870, 1880 and 1900: a little over 200,000 lines.
• Finding the town. Here was the catch. The database knows the county, but not reliably the town. For that I turned to the Census Place Project, a team of economists who matched each census entry to the place where it was written down. Their files cover the whole country, close to three gigabytes for 1850 through 1880 alone. I wrote a small program that ran inside QGIS on my own computer and pulled out only the Dodge County people, matching them one by one by record number. All but about 850 of the 1850 records found their town.
• Crunching the numbers. Then came the counting. For every town and every census: how many people were born in New York, in New England, in Prussia, Bavaria or Mecklenburg, in Ireland, Wales, Norway, and two dozen other places. The 1880 census also asked where each person’s father was born, which let me count the immigrants’ Wisconsin-born children along with them, what demographers call “stock.” By the end, every town carried more than 80 numbers.
• Where the census fell short. Some years would not cooperate. The 1860 census recorded post offices, places like Danville, Richwood and Alderley, instead of towns, so I left it off the town map rather than guess. The old census books also lumped cities in with their towns, which is why Horicon is counted with Hubbard and Mayville with Williamstown. Fox Lake and Westford were not separate towns yet in 1850. And the 1890 census, which would have bridged the gap, was lost after a fire in the Commerce Department building in Washington in 1921.
• The printed reports. For the years after 1880 I went to the Census Bureau’s own printed volumes. The 1910 report lists the population of every town, city and village in Dodge County for 1890, 1900 and 1910. The scanned pages are more than a century old and the computer could not read the type reliably, so the table was read line by line, and then every town was added up for each year to make sure the totals matched the county figures printed at the top. They match to the person: 44,984 in 1890, 46,631 in 1900 and 47,436 in 1910.
• Building the map. The town outlines come from the Census Bureau’s town, city and village boundaries, merged where needed to match the old census towns. I built and checked every layer in QGIS, the mapping program I use for all my historical maps, styled each theme, and then built the web map with Leaflet so it loads quickly and works on a phone. The historic maps underneath are the same ones behind my Dodge County history map: the original land survey plats, the 1873 and 1890 plat atlases I tiled for the web, and the 1840 territory map and 1878 county map georeferenced by myself through Allmaps.

Privacy. The census database comes with rules, and one of them is that individual records are not republished. Everything you see here is a total for a whole town. If you are looking for your own ancestors, the original census pages are on FamilySearch and Ancestry.

A word about accuracy. Census takers misspelled places, guessed at ages and sometimes wrote “Germany” where a family would have said Pomerania. Treat these percentages as a strong picture of each town, not gospel. If your family’s story says otherwise, or you have a church record, a letter or a plat book that adds to it, I’d like to hear from you.

Sources: IPUMS USA full-count census data, University of Minnesota (usa.ipums.org); Enrico Berkes, Ezra Karger and Peter Nencka, The Census Place Project (openICPSR 179401); U.S. Census, 1910, Vol. III, Table 1; U.S. Census Bureau TIGER/Line boundaries. Full credits on the map sources page.