Timothy T. Taylor Home Page
Journal of Economic Perspectives
Articles and Writing
Economics Textbook
Classroom Teaching
The Teaching Company
High School Pedagogy

Articles and Writing

October 24, 1993
"Tally How"
San Jose Mercury News
By Timothy Taylor
<< Back to 1993 menu

"THE LIKELY maximum" number of illegal immigrants in the United States is 3.4 million, according to the General Accounting Office. But how can it know? How can anyone know?

Precise counts are impossible. But by comparing the results derived from a variety of different statistical methods, one can state an answer with reasonable confidence. The GAO report, called "Illegal Aliens: Despite Data Limitations, Current Methods Provide Better Population Estimates," describes five different methods for estimating the number of illegal immigrants. The Residual Method. Using data from the Census Bureau and the Current Population Survey, calculate the total number of people who admit to being "foreign-born." Then, using immigration statistics, calculate how many legal residents are foreign-born. The gap between the two is one estimate of the number of illegal immigrants.

The Death Registration Method. Using data on deaths in the United States, separate out those who are foreign-born. Again, compare the number with what would be expected, based on the available immigration data, and use the difference to estimate the overall population. A similar method is to analyze data on birth rates to find the number of women of child-bearing age illegally in the United States.

The Sex Ratio Method. This method has focused on Mexican data, since Mexico is believed to be the source of half to two-thirds of illegal immigration. Examine the historical Mexican data on number of births, and ratio of males to females. Based on that data, the current Mexican census data show fewer Mexican men in the 15-39 age bracket than would have been expected. If one assumes that some proportion of the missing men ended up in the United States, and makes some assumptions about the characteristics of illegal immigrants from Mexico -- for example, one might assume that they are 65 percent male and 60 percent in the 15-39 age bracket -- one can use the number of missing men to project the overall inflow of illegal immigrants from Mexico.

Non-Immigrant Overstay Method. The U.S. Immigration and Naturalization Service records arrivals and departures for all those tourists, business travelers, students and others who arrive with temporary visas. In 1988, for example, there were 14.3 million arrivals. By year's end, historical data would have predicted 13.7 million departures, but only 12.3 million departures recorded. Some of the problem here is that departure records aren't kept accurately; however, some people with temporary visas are also becoming illegal immigrants.

Repeated Trials Method. Collect data from the U.S. Border Patrol data on the number of people who were caught more than once trying to enter the country, as identified by personal recognition of experienced agents. Assume that anyone who tries repeatedly will get through eventually, so the number of those caught more than once gives a rough estimate of the flow of illegal immigration into the United States.

Obviously, each of these methods has its warts and flaws, discussed in some detail by the GAO. But taken as a whole, it's the best the government statisticians can achieve.

<< Back to 1993 menu