Putting the "Cost" of Illegal Immigration in Perspective

By David Karjanen,

Federal legislation to reform immigration policies often focus on the costs associated with undocumented immigration to the United States. These are not new concerns, there are studies regarding the cost of immigrants going back as far as the 1920s.

The debate about immigration costs are particularly heated in southern border states. In California, Arizona, New Mexico and Texas, newspaper editorials and pronouncements from civic leaders and elected officials often use phrases like “being overrun,� “under assault,� and “inundated� as if from a tidal wave of unauthorized immigrants—migrants who are seen as sapping the life from these states. Indeed, this is part of the motivation for the lawsuit against the federal government seeking compensation for health care and other costs. For the past few years I have been looking at the labor market dynamics of immigrants and low-wage workers in California, and quite frankly, the cost/benefit debate fundamentally misses the point. Costs associated with immigration should not be seen as costs, per se, but rather subsidies to employers.

The costs/benefits of immigration have been estimated by nearly 30 different studies over the past 10 years, and the results have largely been mixed—some find a net gain from immigration, others a net loss. A recent and widely cited report from the Center on Immigration Studies (CIS: http://www.cis.org/articles/2004/fiscal.html) finds that households headed by undocumented immigrants generate $26.3 billion in costs for the federal government in 2002 and paid only $16 billion in taxes, creating a net fiscal deficit of almost $10.4 billion, or $2,700 per undocumented household.

When the public hears this number, it is largely seen in black and white terms: immigrants cost, it is unjust for us to be subsidizing them, kick them out. In policy debates, this also leads to a fairly simple set of assumptions: immigration is a net cost, to control costs we must have urgent policy reform. Unfortunately for those who are quick to use these figures as reason to clamp down on immigration, the estimation of costs and benefits is far more complex.

There are many assumptions people have about undocumented immigrants that are simply incorrect—that these are low-skilled workers—for instance. The reality is that there are many undocumented immigrants in the country who are highly skilled—they overstay work, study, or tourist visas. What I am more concerned with here are the assumptions regarding how to fix the problem of immigration costs. The implicit assumption of most cost/benefit studies is that shutting off the flow of lesser-skilled (less than a high school education) labor, and only allowing H-1b visa holders (highly skilled) immigrants to arrive, would fix, or at least move towards fixing the problem. This assumption follows general economic theory, and is held not only be economists, but the general public. The underlying logic runs as follows: shutting off the flow of low-skilled labor would force employers to hire native-born labor. This, in turn, would raise wages and reduce unemployment (as employers would have to pay more to hire fewer workers). This in turn would reduce the public sector costs of undocumented immigrants, because we would have native born workers earning good wages and not relying on public service programs. The problem, however, is that there is not strong enough evidence that removing immigrants raises the wages dramatically or reduces the public sectors program dependence of native-born workers. Could we really imagine retail sales clerks, cashiers, janitors, and landscapers seeing wages increase from $8.00/hr to $12.00-14.00/hr (the rates at which full time workers become ineligible for most public programs)? Even the most extreme labor market models do not suggest that there would be such a dramatic wage increase. The other problem is that most low-wage workers on public assistance in the United States are native born.

What people fail to recognize is that regardless of nativity or citizenship status, low-wage workers simply earn too little to move off of many forms of public assistance. Hypothetically, even if we could expel the estimated 11 million undocumented immigrants in the United States, and then kept any more from arriving, we would still need a growing pool of labor to work in the lowest paid sectors of our economy—retail sales persons, customer service representatives, janitors, waiters/waitresses, food service workers, and so forth. These are just some of the occupations with the greatest job growth projected by the Bureau of Labor Statistics for the next 10-15 years, yet none of them pay adequate wages to move people off of public assistance, even when working full time.

Using 2002 data, I worked on a study that found that low-wage workers—those households earning less the minimum wage are eligible for up to $23,417 per year in public assistance through programs like food stamps, supplemental insurance, and section 8 housing, while paying $3,454 in total state and federal taxes.(http://www.onlinecpi.org/article.php?list=type&type=62 ) This leaves a potential gap of approximately $20,000 per minimum wage household. This is an amount far greater than the estimated $2,700.00 per household that undocumented immigrants are estimated to cost in public assistance. Fortunately, very few workers actually earn the minimum wage in the United States, and those that do are not taking full advantage of the public services available to them. If we move higher up the wage scale, however, the rates of public service utilization and the costs remain very high for low-wage workers.

Using a similar methodology, the Institute on Labor and Employment at UC Berkeley found that in California, two million working families received public assistance in 2002. The price tag for this assistance was $10 billion per year, with most support going to families with full-time workers who earned near the minimum wage. (http://repositories.cdlib.org/ile/scl2004/01) This is a cost of approximately $2000.00 per working family—a significant subsidy on average per household, but far less than the amount that the working poor qualify for. Significantly, the majority of households have a high-school degree or higher and working full time—that is to say, they are not dependent on public assistance because of less than a high school education or failure to work, but rather because of the low wages and benefits available through employment. What this means is that the other side of the coin on the immigration costs/benefits debate is the costs of the native-born working poor. In fact, the costs to subsidize native-born working poor are higher than undocumented immigrants who are working and poor.

As the CIS study notes, on average, the costs that illegal households impose on federal coffers are less than that of other households (the reason they are a net cost is because their tax payments are only one-fourth that of other households). However, is this cost differential because the immigrants are undocumented, or because they are employed in low-wage occupations that do not pay enough to move beyond public assistance? Even CIS acknowledges that if undocumented immigrants were given amnesty and began to pay taxes and use services like households headed by legal immigrants with the same education levels, the estimated annual net fiscal deficit would increase from $2,700 per household to nearly $7,700, for total net cost of $29 billion. What this figure represents, assuming their model is accurate, is the total subsidy to low-wage households would increase because of their greater access to public sector programs. In other words, it is far cheaper to keep workers undocumented! What this study and other fail to recognize is that we already subsidize low-wage households by billions of dollars each year who are citizens—welfare reform has not eliminated the need for health care, shelter, subsidized meals and other vital programs.

So then are undocumented immigrants really such a great burden? Looking at Current Population Survey data for 2003, if we compare immigrant men to native born men, they tend to have higher rates of labor force participation, earn less, and use far less in public services. This represents a boon to employers—a cheaper labor force, one with a lower rate of unemployment than native-born workers, and they use less in public services than native born workers. The total costs for taxpayers—assuming it is roughly $2,700 per household—is really a subsidy to the employer to the extent that these are households which are working, but don’t earn enough to be disqualified for public assistance programs. We could add to this the massive subsidy to the economy that lower labor costs provide in keeping inflation down, and the additional economic activity generated by immigrants.

What the public should be concerned with is not that they are footing the bill for undocumented immigrants, but for low-wage work, that is, for employers. Hiring native-born workers would not fix the problem—the costs would still largely be there because low-wage workers do not earn enough to move beyond the need for public assistance regardless of citizenship. But as rates of employer provided health insurance continue to fall, wages for less than high-school educated workers are stagnant and falling, yet we have massive growth in the low-wage sector of the economy, it is time to take a hard look at the direction of the US economy overall—immigrants have just become a convenient scapegoat, and a smokescreen for the problem—an economy that isn’t working well for the working poor.
__________

David Karjanen is a Professor at the Institute for Global Studies at the University of Minnesota

About this Entry

This page contains a single entry by Dan Ott published on October 16, 2007 4:20 PM.

A Journey Across Our America: This land is your and mine - Part II was the previous entry in this blog.

Legal Rights of Illegal Immigrants is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.