A new study provides a reminder of the fluidity of racial and ethnic identities as a number of Americans reported different identities on the 2000 and 2010 Census:

The report showed that 1 in 16 people — or approximately 9.8 million of 162 million — who responded to both the 2000 and 2010 censuses gave different answers when it came to race and ethnicity.If extrapolated across the entire population, that would mean that 8.3 percent of people in the United States would have made a change in their racial or ethnic identity in that decade, according to the paper authored by Sonya Rastogi, Leticia E. Fernandez, James M. Noon and Sharon R. Ennis of the U.S. Census Bureau and Carolyn A. Liebler of the University of Minnesota.

The largest change was from Hispanic (some other race) to Hispanic white, with 2.38 million people making that change on their census forms. But the next greatest change was from Hispanic white to Hispanic (some other race), with 1.2 million people deciding that designation fit them better. Put together, these two changes make up more than a third (37 percent) of the race/ethnicity changes in the report…

The groups most likely to change were people who were children and/or living in the West in 2000. That region also had a higher rate of interracial marriage, and multiple race reporting, the report said. The census defines the West as being Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington and Wyoming.

The most stable groups were single-race, non-Hispanic whites, blacks and Asians, with those who checked those boxes staying with them in both censuses. People were also consistent with their Hispanic/non-Hispanic choices.

Those who make strong predictions about the demography of the United States in the coming decades have to contend with changes like this. It isn’t as easy as suggesting that the proportion of whites will continue to decline. What if more Hispanics see themselves as white? White as a category changed quite a bit in the past to include new immigrant groups and will likely continue to change in the future. All of this introduces uncertainty in thinking about how this could play out with contemporary debates, like with immigration.

It would also be interesting to compare the responses provided to the Census versus an everyday understanding of one’s racial and/or ethnic identity. The Census categories have their own history and may not always match lived realities.