A common complaint amongst genealogists is ancestors who are “missing” in a census or other document in which they “should” appear. And it doesn’t really help that America is a really, really big place – theses ancestors, theoretically, could be anywhere.
An article by Michael P. Conzen, a geography professor, provide clues in his article “Local Migration Systems in Nineteenth-Century Iowa.” In this piece Conzen explores the migration fields in Iowa using the 1895 Iowa State census. A migration field, is simply, the area from which a destination draws its migrants.
What Conzen found in Iowa is that many people moved along the Rivers and later west along the railroad lines. In fact the connection between westward movement and the railroad was so strong, he basically said it wasn’t worth talking about. He also found that the larger the city, the larger the migration field.
On the ground, these means smaller cities in Eastern Iowa, like Keokuk sent plenty of migrants westward and only pulled in-migrants from a few surrounding counties. The two largest cities Sioux City and Dubuque pulled people from all over the state. In Ohio, where I do much research, Wellsville, on the Ohio River, would only pull from a small surrounding area. On the other hand, Cincinnati and later Cleveland, would draw people from a larger area. Similar patterns are seen in Great Britain.
How might Conzen’s study help you find your missing ancestors? For starters if you're trying to figure out where someone was before look eastwards. If you are looking for where they went next, always look west. Yes, Americans did move move in both directions, but the general trend was an east-west migration.
Follow the major transportation networks: rivers, turnpikes, canals and eventually, the railroads. I suppose your ancestors could have gone orienteering with a map and a compass to read their destination, but it’s not very likely.
If your ancestor is located in a small town in one census but not in the previous one, check all the surrounding counties in an expanding radius. If they are in a large town, follow the same procedure, but with bigger radiating circles. Basically, it is more likely that your ancestor would have moved, for example, from Brush Creek Township in Jefferson County, Ohio to either sizable towns of Wellsville or Steubenville than to have moved from Cleveland to Brush Creek Township.
Conzen, Michael P. “Local Migration Systems in Nineteenth-Century Iowa,” Geographical Review, vol. 64, no. 3, (Jul 1974), pp. 339-361. Stable URL from JSTOR is here, which provides access to the first page unless you have subscription to JSTOR. You may be able to access JSTOR through your local library.