Free Exchange challenges my interpretation of European vs. American population density:

But high American wages also made the country an attractive place for European emigrants, and millions did opt to make the trans-Atlantic journey, swelling the size of American cities, increasing American urban population density, and reducing upward pressure on densities in Europe. In fact, by the beginning of the 20th century, there were not significant differences between the density of urban populations in America and those in Europe. The massive, dense industrial towns of the American Northeast and Midwest didn't look all that different from the massive, dense industrial towns of the Midlands and the Ruhr Valley. In cities on both continents, transportation technologies were similar, and so limitations to the size and scope of cities were also similar.

The important differences result from what came next. America systematically starved and disassembled its public transportation and rail infrastructure and spent rather extravagently on highways, both within and between cities. While Europeans also increased their spending on roadways, they nonetheless maintained strong national commitments to rail and public transit. Americans also chose to leave their petrol taxes at levels far below those in European nations. These differences are key to explaining recent urban experiences. Certainly the differences in urban area and density in cities like Atlanta, Houston, and Chicago relative to Berlin, Madrid, or London are not due, in any direct sense, to patterns in medieval agriculture

I don't think the relevant criteria is the density of cities at the turn of the century; it's the distance between them. Much of Europe looks like the denser bits of the American east coast because those densely populated urban centers were relatively close to each other, limiting the scope of urban expansion. Both are areas that were fully agriculturally exploited pre-industrialization, and have the resulting small distances between major cities. On the Richmond to Boston route, ten hours long you've got seven or eight cities that were major population centers in 1830. And just as in Europe, everywhere you look along the train line, there are people.

Most American cities, on the other hand, had huge, practically unlimited, tracts of farmland to expand into outside of the East Coast, and forthwith did. One reason that the East Coast and San Francisco are the places seeing a return to urban living is that those places are, like Europe, rapidly running out of space to build new suburbs.

This also affects the cost-effectiveness of trains. The way that Americans use land obviously makes trains less cost-effective. But while the favourite culprit of much of the left is federal highway money, this lacks sufficient explanatory vigor to account for the massive differences in density. It seems obvious to me that Americans used land profligately in the 1950's and beyond because they could, and Europeans didn't because they couldn't: because they were too poor to buy cars; because there wasn't that much land to be had; because the Common Agricultural Policy, along with impoverishing third-world farmers and producing mountains of butter and lakes of wine, encouraged the preservation of farmland near cities. This has nothing to do with virtuous development policy.

That matters not because I'm interested in debating which continent has the nicer people, but because it goes directly to the assumption that if we just had more virtuous development policy, we'd get Americans out of their cars. I'd be more than happy to be proven wrong, but both Americans and Europeans seem to really like tooling around in their cars, and living in detached houses with yards, and it seems to take more than the gentle bonds of development policy to restrain them.