“Men in Motion: Some Data and Speculations about Urban Population Mobility in Nineteenth-Century America”
Stephan Thernstrom and Peter R. Knights, Journal of Interdisciplinary History 1:1 (Autumn, 1970)

Stephan Thernstrom (UCLA, later Harvard) and Peter Knights (University of Illinois, later York University) began the article by agreeing with Joseph Kennedy, the Superintendent of the
1852 Census, that “the roving tendency of our people” is given too little attention by historians. Rural mobility, they said, had been addressed by Malin in 1935, Curti in 1959, and Coleman in 1962. But the point they made about urban population change in this article may apply equally to the countryside. Recorded “net population changes from census to census,” they said, “though often dramatic, pale into insignificance by comparison with the actual gross volume of in and out movement.” (my emphasis) They elaborated that “Even in the most stable small or medium size community which has yet been examined approximately half of the population was transient within a relatively brief span of years.” And in direct contrast with the frontier thesis, “important as the westward movement was in our national history, a far larger migratory stream moved eastward and city-ward.”

To illustrate their point, the authors examined Boston documents which claimed “the proportion of the city’s 1890 residents who had moved into Boston in the preceding decade [when the city’s population rose from 363,000 to 448,000] was...fully one third.”  In fact, they said, because people were also constantly leaving the city throughout the decade, “Nearly 800,000 people moved into Boston between 1880 and 1890 to produce the net migration increase of 65,179.” The turnover of the Boston population they documented means that just about 700,000 people left the city in ten years. These people all went somewhere. “The turnover of the Boston population in the decade, the total of families moving either in or out, was 296,778, a stunning 4.09 times the total number of house-holds in the city in 1880.” This strikes me in a similar way as the often-repeated claim that the
human body replaces ALL or our cells in 7-10 years. If that's the case, in what ways can we consider ourselves to be the “same person”? And to the extent a city’s identity is based on its citizens, how do we understand Boston?

And it turns out the 1880s were not unique in this regard. Between 1830 and 1890, when Boston’s population increased from 61,000 to 448,000, Thernstrom and Knights reported “the number of migrants entering Boston...was an amazing 3,325,000, eight and a half times the net population increase.” Again, that means nearly three million people
left Boston and went someplace else. Where did they go, and when they got there, did they stop moving about? There’s apparently no reason to suppose they did.

The authors discovered these changes by looking beyond the decennial census data to information collected for and reported in city directories published annually. “Returning to the same dwelling after the passage of only 365 days, the city directory canvasser had less than a fifty-fifty chance of finding its former inhabitants living there,” they said. Of course the rich, who owned businesses and real estate, were much more persistent than the poor. Thernstrom and Knights said, “The wealthy, with over $10,000 in property, were from four to six times as likely to remain in Boston as to move away.” They even speculated that transience might be higher than they could measure, because many poor workers may not have stayed long enough to be counted even once. They observed that “substantial numbers of them must have been legally disenfranchised by the electoral laws…which required a year's residence in the state and six months in the city before voting.” This fact is not reflected in any political histories of the era I’m familiar with.

So the political consequence of short tenancy was being frozen out of participating. Interestingly, there were “no consistent ethnic differences in persistence once occupational level was held constant.” Suggesting the most salient factor in persistence and political participation was class and not ethnicity? This may have led, the authors suggested, to a widespread feeling of alienation from the political process and a corresponding inability to organize effective dissident organizations. It may also have contributed to the growth of regional voluntary organizations (and even the Knights of Labor) that could offer people some continuity in spite of their movements. Bruce Laurie mentioned Thernstrom several times in
Artisans to Workers (which is how I discovered the article), but the extreme mobility of poor people and unskilled workers doesn’t significantly impact his story of the skilled tradesmen unionized by the American Federation of Labor. It might help explain the “ruralization” of the Knights of Labor, though. 

This article was an early demonstration of “New Social History” techniques that focused on demographic data rather than sources previously preferred by historians. “Quantitative studies of grassroots social phenomena,” the authors said, “can help us to grasp basic social processes which left little imprint on conventional historical sources-newspapers, diaries, legislative debates, etc.-thus freeing us, to some extent, from our customary dependence upon materials reflecting the beliefs of the articulate classes who controlled the media of the era.” Thernstrom and Knights' discovery of a high-mobility “floating proletariat” challenged Robert Wiebe’s image of “a nation of loosely connected islands,” (described in
The Search for Order) because these uncounted people would have been moving constantly between these islands. Or perhaps even between the urban islands and the rural sea. Taking ideas, attitudes, and class consciousness with them as they travelled from place to place. This would have had huge implications for popular culture, not to mention populism.