If immigrants don’t quickly catch up in earnings to the US born—neither today nor in the past—then where did the myth of rapid immigrant mobility come from? One possibility is that boosterish stories like Rocco’s tale of quickly rising in America have been accepted uncritically, especially in how Americans view the past.
As time goes by, family stories also soften into myth, with the immigrant generation remembered as penniless—arriving with “a goat and one Bitcoin,” as the comedian Desus put it. With distorted memories suggesting a lowly start, any later success seems all the more impressive.
Once we got access to all the records—and, crucially, once we were able to follow individual immigrants over time—we found that the rags-to-riches story is nothing more than a data illusion.
Another possibility is that immigrant success was a genuine hallmark of an earlier era, before the Age of Mass Migration, and that this national myth stuck. Scholars have found that the last time that rapid immigrant mobility was a feature of the American economy was before the Civil War. Immigrants who arrived in the 1840s and 1850s, soon after the invention of the steamship, actually did move up the occupational ladder faster than the US born.
But there’s a third possibility that we want to suggest here: the nostalgic view of immigrant success a century ago persisted for so long because we simply didn’t have the data necessary to check the facts. What’s more, the data that was available to researchers led scholars to draw the wrong conclusions from what they saw in the numbers. Once we got access to all the records—and, crucially, once we were able to follow individual immigrants over time—we found that the rags-to-riches story is nothing more than a data illusion.
To begin to understand this illusion, you need to remember that the most accurate way to measure upward mobility requires following a person over time—from childhood to adulthood or from the beginning of his career to its end. And that requires being able to link data across censuses.
But previous generations of researchers did not have the linked data because of census privacy rules. (Recall that the Census Bureau releases identifying information such as first and last name only after seventy-two years.) Researchers were also unable to digitize what amounted to literally millions of census manuscripts.
Without access to linked data across censuses, scholars tried to learn as much as possible from the data that they did have. The most famous study of this type concluded that immigrants could rise quickly from rags to riches by comparing immigrants in the 1970 census who had arrived in the United States in different years.