Unauthorized and low skilled immigrants are attracted to America’s labor markets, not the size of welfare benefits.  From 2003 through 2012, many unauthorized immigrants were attracted to work in the housing market.  Housing starts demanded a large number of workers fill those jobs.  As many as 27 percent of them were unauthorized immigrants in some states.  Additionally, jobs that indirectly supported the construction of new houses also attracted many lower skilled immigrant workers.

Apprehensions of illegal crossers on the Southwest border (SWB) is a good indication of the size of the unauthorized immigrant flow into the United States.  The chart below shows apprehensions on the SWB and housing starts in each quarter:

Fewer housing starts create fewer construction jobs that attract fewer crossings and, therefore, fewer SWB apprehensions.  The correlation holds before and after the mid-2006 housing collapse. 

What about welfare? 

Here is a chart of the national real average TANF benefit level per family of three from 2003 to 2011 (2012 data is unavailable) and SWB apprehensions:

Prior to mid-2006, TANF benefit levels fell while unauthorized immigration rose.  During the housing construction boom, unauthorized immigrants were attracted by jobs and not declining TANF benefits.  After mid-2006, when housing starts began falling dramatically, real TANF benefit levels and unauthorized immigration both fell at the same time.  If unauthorized immigration was primarily incentivized by the real value of welfare benefits, it would have fallen continuously since 2003.   

The above chart does not capture the full size of welfare benefits or how rapidly other welfare programs increased beginning in 2008.  As economist Casey Mulligan explained in his book The Redistribution Recession, unemployment insurance, food stamps (SNAP), and Medicaid benefits increased in value and duration beginning in mid-2008.  Including those would skew welfare benefits upward in 2008 and beyond, but unauthorized immigration inflows still fell during that time.

In conclusion, housing starts incentivize unauthorized immigration while TANF does not. 

Earlier this month Steven Camarota of the Center for Immigration Studies (CIS) critiqued our Cato Working Paper exploring immigrant welfare use. Our report found that low-income non-citizen immigrants use public benefits like Medicaid, foodstamps, and cash aid less than their poor native-born citizen counterparts.  Furthermore, when immigrants do receive these benefits the amount is less than the amount comparative natives receive. 

Dr. Camarota argues that immigrant-headed households (a once useful term that is now ambiguous) use more benefits than native-headed households, but does not dispute that, when immigrants receive public assistance benefits, the average amount received tends to be less.  All U.S. citizens have access to Medicaid, SNAP, and other benefits.  In contrast, recent legal immigrants who are not yet citizens are often ineligible from the major benefit programs and undocumented immigrants are never eligible. 

The first issue – which is rather wonky – is how to measure immigrant welfare use.  Our approach is to count the benefits used by immigrants individually while Camarota’s approach is to include everyone in a so-called immigrant-headed household regardless of citizenship status – especially U.S.-born children and spouses. 

Our approach of counting immigrant welfare use individually is used by the conservative state of Texas to measure immigrant use of government education and other benefits.  The Texas Comptroller’s Office did not include the children of immigrants who were American citizens when calculating the cost to public services in Texas because, “the inclusion of these children dramatically increased the costs.”  The Texas report continued by stating:

“The Comptroller has chosen not to estimate these costs or revenues [of U.S.-born children] due to uncertainties concerning the estimated population and the question of whether to include the costs and revenues associated only with the first generation or so include subsequent generations, all of which could be seen as costs (emphasis added).”

In other words, counting the cost of the children of immigrants who are born citizens is a bad approach.  If we were to follow Camarota’s methodology, why not count the welfare costs of the great-grandchildren of immigrants who use welfare or public schools today?  Our study, on the other hand, measures the welfare cost of non-naturalized immigrants and, where possible, naturalized Americans.

We focus on comparing poor immigrants to poor natives because it produces an apples-to-apples comparison.  In that way we exclude non-welfare using immigrants like Sergey Brin, the co-founder of Google, and other wealthy Americans.  For instance, the annual cost of Medicaid amongst 100 non-citizen poor adult immigrants was 42 percent below the cost for 100 native born poor adults.  For non-citizen children compared to citizen children, immigrants cost 66 percent less.    

Immigrants’ use of public benefits is important and worth exploring from every possible angle as the immigration reform debate develops.  Individually, poor immigrants are less likely to receive welfare benefits and when they do the value is less than for poor natives. 

This post was co-written by Leighton Ku, PhD, MPH, Professor of Health Policy and Director of the Center for Health Policy Research at George Washington University, Brian Bruen, MS, Lead Research Scientist and Lecturer in the Department of Health Policy at George Washington University, and Alex Nowrasteh of the Cato Institute.  Both Professor Ku and Mr. Bruen have conducted research and policy analyses about the use of medical and other public benefits for needy Americans, including immigrants, for many years. Research articles by Ku and/or Bruen have been published in Health Affairs, New England Journal of Medicine, and American Journal of Public Health.