So what explains these discrepancies between religious affiliation inside prisons and outside them? I’ll set out a couple of possible theories.

Maybe it’s income. We know that Muslims and Americans who identify with historically black Protestant churches tend to have lower incomes than the national average. And we know that 44 percent of federal prisoners earned less than $1,500 in the month prior to their arrest (I’m afraid the last time prisoners’ income was surveyed was in 2004, though, and those are 2004 dollars).

Maybe it’s race. We know that even though African-Americans represented only 13 percent of the U.S. population in 2013, they represented 36 percent of the prison population. And we know that Muslims are almost twice as likely to be African-American as the U.S. population as a whole. (I think race is also probably relevant when we’re thinking about the over-representation of Rastafarians and Santerians in prison too.)

Maybe it’s immigration. We know that most American Muslims immigrated to the U.S. sometime after 1992. And we know that 12 percent of the federal prison population is described as “non-citizens” (a term that means they’re foreign-born and, according to Ed Ross, a spokesman at the Bureau of Prisons, entered the country illegally).

Maybe it’s conversion.