New Estimates of Unauthorized Immigrants in the U.S.

by Sue Weishar, Ph.D.

As the movement for comprehensive immigration gains momentum, it is important for advocates to have the most accurate data possible on the number of unauthorized immigrants in their respective states and the U.S as a whole. A new report published in the International Migration Review, by Robert Warren and John Robert Warren, provides such information.

In the table below are the report’s estimates of the number of unauthorized immigrants in the five Gulf South States and the U.S. as of January 1, 2010. Note how the International Migration Review (IMR) report estimates, in the second column, differ from estimates from a widely used resource for immigration population estimates, the Pew Hispanic Center, in the third column. Except for Florida, the IMR Gulf South state estimates are lower than the Pew Hispanic Center estimates. The IMR estimate of the total U.S. population of undocumented immigrants, however, is higher by more than a half a million than what the Pew Hispanic Center reported.

*From Unauthorized Immigration Population: National and State Trends, 2010 by Jeffery S. Passel and D’Vera Cohn, Pew Hispanic Center, February 1, 2011, pg. 23. Ranges provided by PHC are not listed in above chart. For the 16 largest states, including Florida and Texas, estimates are based on augmented March 2010 Community Population Survey (CPS). For the remaining 34 states, because of small sample size, estimates are averages of augmented March Supplements to CPS for 2009 and 2010. 

U.S. Census surveys ask respondents if they were born in the U.S. or not, but do not ask questions about respondents’ immigration status. Therefore researchers commonly use a “residual method” to determine estimates of the unauthorized immigrant population. This requires subtracting known data on the number of authorized immigrants (obtained from federal administrative data) from Census data on the number of foreign born.  The authors of the IMR report provide a detailed account of why they believe the data and residual methods they use produce more accurate estimates with smaller sampling errors than those used by the Pew Hispanic Center and the Department of Homeland Security’s Office of Immigration Statistics (OIS).[1]  The IMR report authors note that because their methods modestly underestimate the number of authorized immigrants[2] their estimates likely overstate the number of unauthorized immigrants.

A unique aspect of the IMR report is that the authors provide a state by state analysis for each year from 1990 to 2010 of unauthorized inflows (through illegal entry and visa overstays) as well as outflows of unauthorized immigrants (through emigration, removal, graduation to legal status, and death).

Some of the report’s key findings about the U.S. undocumented population include the following:

• Significant numbers of unauthorized immigrants leave the unauthorized population each year. From 1990 to 2009, there was an inflow of an estimated 15.7 unauthorized immigrants (through illegal entry or overstay) to the U.S., but in the same time period 7.5 million left the unauthorized population (by emigration, removal, death, or adjustment of status), resulting in net gain of 8.2 million.

• Since 2007, the U.S. has lost more unauthorized immigrants than it has gained. Although the unauthorized population in the U.S. has essentially reached zero growth, the size of the inflow of unauthorized immigrants is still relatively large—close to 400,000 in 2009.

• More than half of all unauthorized immigrants live in California, Texas, Florida, and New York.

• The number of immigrants entering the U.S. undocumented population grew steadily in the 1990’s. However, from 2000 to 2009, the number of persons who entered the undocumented population each year declined by 72%, from about 1.4 million a year in 2000 to 400,000 a year in 2009.

• During the depths of the recession in 2008 and 2009 the drop in the unauthorized population was NOT due to a substantial increase in the rate of emigration, but rather a substantial decrease in arrivals.

Here are important findings from the report regarding the Gulf South states:

• The state that experienced the fastest growth in undocumented immigrants over the past two decades was Alabama. Whereas the total unauthorized population in the U.S. tripled between 1990 and 2010, Alabama’s unauthorized population in 2010 was 19 times greater than it was in 1990, when only 5,000 unauthorized immigrants lived in the state. In 2010 Florida’s undocumented population was over four times greater than in 1990, Louisiana’s was 10 times greater, Mississippi’s was nine times greater, and Texas’ was about four times greater.

• For every state in the country, except Mississippi, the size of the inflow of unauthorized immigrants in 2009 was smaller than the inflow of unauthorized immigrants in 2000.[3]  Between 2000 and 2009 every state in the country saw increases in outflows. By 2009 29 states experienced net losses in their unauthorized populations. Nevertheless, the only Gulf South State that saw a decline in its unauthorized population in 2009 was Texas.

The IMR report sheds important light on the interaction of inflow and outflow in determining the size of the undocumented population. Seeing data on this play out on a state by state basis is also helpful. With so much talk of “border security” in the growing debate on comprehensive immigration reform, however, I would have liked to have seen the authors break down numbers on the “entered population” into categories estimating the numbers of immigrants who crossed without authorization and the number who fell out of legal status after entering legally by overstaying their visas. According to a 2006 report by the Pew Hispanic Center, almost half (45%) of the total unauthorized immigrant population entered the U.S. legally. Increasing border security will do nothing to address this reality-- and may delay or scuttle real reform. The authors indicate that visa overstays are a component of their estimates on unauthorized “inflows”, so it appears they have such data.

A discussion on Katrina’s impact on the flow of unauthorized immigrants to the Gulf South region also would have been helpful. There is general agreement that the Gulf South could not have recovered from Hurricane Katrina to the extent we have without the influx of immigrant workers, many of whom were undocumented. Estimates of the number of undocumented immigrants to the New Orleans area after Katrina varied widely, from 60,000 to as much as 150,000. The IMR report shows the inflow of unauthorized immigrants to Louisiana increasing by only 2,091 persons between 2004 and 2005. (Hurricane Katrina struck August 29, 2005). The increase in annual inflow between 2006 and 2007 is only 862. Could the U.S. Census Bureau not reach the workers or were local estimates wildly off their mark? Any clarification from demographers would help.

[1] Warren and Warren argue that four conceptual and technical differences make their estimates more useful than Pew and OIS estimates of unauthorized immigrants. Briefly:

a) Pew and OIS do not routinely disaggregate annual net change in authorized populations into inflows and outflows, which makes it difficult to interpret net change.

b) Pew estimates are based on data from the U.S. Census’ Current Population Survey (CPS), not the American Community Survey used by the authors and OIS. The authors note that the relatively small same size of the CPS yields large ranges of sampling errors in the Pew estimates. For example, although Pew reports a decline of almost 1 million unauthorized immigrants from 12.1 million in 2007 to 11.1 million in 2009, because the sampling error was so large, Pew concluded that the decline was not statistically significantly different than zero.

c) Neither the Pew or OIS estimates are disaggregated fully and reliably by state.  OIS only provides estimates of unauthorized immigrants in ten states with the largest such populations. For states with small undocumented populations, Pew has averaged estimates across states and/or employed regression analysis, with the results yielding wide margins of error.

d) Whereas Pew and OIS only go back to the year 2000, Warren and Warren produced annual estimates back to 1990.

[2] This is because there are no accurate estimates for certain groups of aliens in legal status-- including asylees, parolees, and persons with Temporary Protected Status who have been given work authorization.

[3] Estimated inflow to Mississippi increased from 1,418 in 2000 to 2,325 in 2009.