Having indulged an unhealthy interest in human migration for decades, I’ve been intrigued by how the number of illegal immigrants that journalists cite as living in the US never changes. For years on end, I’ve read that the population of America’s ‘undocumented’ — a euphemism that seems to upbraid the receiving country’s bureaucrats for failing to issue its woefully overlooked residents the proper papers — is 11 million. With the artificial precision that often attends these unverifiable figures, which derive their authority from sheer repetition, some journalists will quote the number as 11.3 million. Oh, there was a brief period about 15 years ago when that standard statistic rose to 13 million, but then it popped right back down to 11 again, without any headlines about a stampede of two million suddenly ploughing over the Mexican border in the southern direction.
But how do our well-informed correspondents know how many immigrants live illegally in the US? That 11.3 million figure originates from the Census Bureau’s annual ‘American Community Survey’. But these are the very residents notoriously averse to responding to census inquiries, regardless of assurances that being counted won’t lead to an unwelcome knock on the door from the cold-hearted immigration heavies at ICE. ‘Irregular’ migrants have a vested interest in remaining invisible. Should you ask a crowd of such people ‘Who’s not supposed to be here?’, funnily enough nobody’s likely to raise a hand.
A trio of academics affiliated with MIT and Yale University set about trying to reproduce that oft-quoted statistic by other means. Their results were published in 2018. Employing a range of operational stats (such as deportations and visa overstays) and other demographic data (such as death rates and immigration rates) from 1990 to 2016, they first fed their computer model the most conservative assumptions possible. As the study’s authors anticipated arriving at a number well below that eternally unchanging 11.3 million, confirmation bias alone would have helped produce a lowball estimate. Instead, even with all the inputs rigged at improbably low levels, the model still generated a population of illegal immigrants in the US of 16.7 million — 50 percent higher. The modelers were astonished. As an on-again, off-again New Yorker, I’m not.