Behind the Pew Study on Illegal Immigration, Part II: How Many Illegals?

The recently released Pew report on the decline of the illegal population in the United States has garnered a lot of attention, though a great deal of it for the wrong reasons. Obama administration officials are already touting their policies to account for the decline. A DHS spokesman went even further, saying that "the downward trend in border crossings has continued since Obama took office in January 2009, citing Homeland Security statistics that show decreases in illegal-immigrant apprehensions during the past two years" (emphasis mine).

Of course "decreases in illegal immigrant apprehensions" can stem from several causes, among them fewer attempts by illegal immigrants to cross the border or more success in doing so. The use of the metric "decreases in apprehensions" also assumes that the rates of apprehension bear some meaningful relationship to the number of attempted illegal border crossings, and that requires that we know what that figure is. But we don't. Welcome to the murky world of illegal immigration statistics and claims.

At the heart of the Pew report about the decline in the number of illegal immigrants is the actual number of illegal immigrants. While several news reports and administration officials have accepted Pew's figures as real and accurate, the truth is that they are estimates, as the Pew report makes very clear (pp. 22-28).

What, exactly, does it mean to have an "estimate" of illegal immigration? How are those numbers derived? And how much confidence can we have in them?

The purpose of this entry is not to denigrate Pew's estimates. Indeed, in the next installment of this series of entries I will, as the lawyers say, "stipulate" their accuracy in the service of inquiring more deeply in that what the report actual says that goes beyond the headlines that have been the primary focus of public attention.

Nor is my intention here to provide a technical analysis of the statistical assumptions and properties of the models that are used to make these estimates. That is better left to those with the training and inclination to do so.

My point here is to simply lay out, in understandable language and logic, the basis of the estimates, what they do and do not include so that readers will have a better grasp of the meaning of the numbers that capture the immigration headlines.

Let us begin with Pew's figures. The Pew report clearly calls attention to the fact that this report is based on "new estimates." It then states unequivocally that, "this sharp decline has contributed to an overall reduction of 8% in the number of unauthorized immigrants currently living in the U.S. – to 11.1 million in March 2009 from a peak of 12 million in March 2007" (p. i).

But one page later the report says, "… there may have been a decline in the unauthorized population between 2008 (11.6 million) and 2009 (11.1) million, but this finding is not conclusive because of the margin of error in these estimates" (p. ii, my emphasis).

One page before the 11.1 million illegal immigrants was presented as established and legitimate fact, but on the very next page the difference between the 2008 and 2009 figures are called into question, leaving one to wonder which of the two years are suspect, and whether if one of those years is and the other is not, what is the difference?

There is no easy way to discern the answer, as will become clear in my next entry.