A Response to Keith Hennessey on Immigration

COMMENTARY Immigration

A Response to Keith Hennessey on Immigration

May 23, 2013 15 min read
Derrick Morgan

Executive Vice President

Derrick Morgan is the Executive Vice President of The Heritage Foundation.

Recently, Keith Hennessey, a former economic advisor to President George W. Bush, pointed out what he thinks are eight problems with the Heritage cost estimate of unlawful immigration and amnesty. During the course of his post he notes that “making 8-9 million people here illegally into U.S. citizens would increase future deficits” and “[t]here is a significant fiscal effect from making…them legal taxpayers and eventually beneficiaries eligible for the full panoply of government subsidies.”Those comments are refreshing and affirm that Heritage’s work reaches a valid conclusion: amnesty will cost taxpayers. While this seems commonsensical, some have implied or even posited that economic growth associated with amnesty would somehow make the huge fiscal costs vanish. To his credit, Hennessey does not do that.

While acknowledging amnesty will be costly, Hennessey takes issue with the $6.3 trillion tabulation. We welcome input from him and others. Heritage is the only one so far to try to quantify the lifetime fiscal costs in this debate. Heritage put our entire methodology as clearly as possible in the nearly 100-page study in part for the purpose of inviting comment and, as is our custom, making improvements.

Below, I address each of Hennessey’s critiques and note that several of the points he makes are in harmony with our own views.

Baseline Critique

First, Hennessey says that the $6.3 trillion number is too high because it includes costs that unlawful immigrants now impose. The title of the paper is “The Fiscal Cost of Unlawful Immigrants and Amnesty to the U.S. Taxpayer” (emphasis added). The purpose of the paper, in part, is to add up all the benefits and services received by unlawful immigrants and subtract taxes paid by them to come up with their net fiscal cost. Rector, following the methodology used by the National Academy of Sciences, includes costs of unlawful immigrants on “congestible” goods such as firefighters, emergency rooms, and schools.

The number Hennessey wants to know is the difference between the status quo (allowing unlawful immigrants to stay) and the costs we would have under amnesty, not the costs of unlawful immigrants versus having no households headed by those unlawfully present. Rector clearly provides this figure in the report. Having estimated the status quo costs at around $1 trillion, Rector writes, “The net increased fiscal costs generated by amnesty [alone] would be around $5.3 trillion ($6.3 trillion minus $1 trillion).”

Both of these numbers are useful to know. The $6.3 trillion figure combines the costs of status quo and amnesty. Amnesty alone costs $5.3 trillion.

Household Data

Second, Hennessey objects that the figure does not equal the costs of illegal immigrants because it includes citizen-children in unlawful immigrant households. Heritage, like others who have studied the issue, uses the most recent household data from the U.S. Census and includes costs of children in those households in its calculations. Heritage uses this same methodology (counting the costs of children toward their parents’ fiscal balance) to measure the impact of government benefits and taxes on economic equality in the whole U.S. population.

To do otherwise when looking at the U.S. population as a whole would effectively ignore that government policy is highly redistributive; the elderly and disabled might be net fiscal beneficiaries, but almost everyone else would be seen to pull their own weight (except, of course, children, who would individually be huge net liabilities). This does not represent actual conditions in the U.S.

Hennessey asserts the study is misleading because it employs an unusual analytic approach. Heritage is actually following what others who have examined the fiscal impact of immigrants have done by counting citizen-children. Prior to the Heritage study there have been three major studies on the fiscal cost of immigration or low-skill immigration at the national level and all three, including perhaps the most important of these, a report by the National Academy of Sciences, included the native-born children in cost estimates.

Steve Camarota analyzed the fiscal cost of less-educated immigrants to the federal government for Center for Immigration Studies, and included the costs of children.

Even Julian Simon, who was for years the leading immigration authority at the libertarian Cato Institute, used this approach in analyzing the fiscal consequences of immigration. Simon asserted that the fiscal impact of immigration must include both the immigrant and the family “he brings or acquires.” After all, the children are here only because their parents have managed to enter or remain in the country unlawfully.

Hennessey would prefer to exclude the cost of the children of immigrants from fiscal analysis because, in his view, most of the costs associated with the U.S.-born children of unlawful immigrants are “sunk costs.” Hennessey implies that since there is no way to avoid these costs, there is no point in analyzing them. The Heritage analysis estimates that the net fiscal cost of unlawful immigrants under the status quo (allowing unlawful presence) will be around $1 trillion; all the “sunk costs” concerning immigrants’ children are contained in that sum. The analysis estimates that amnesty alone will add $5.3 trillion in additional fiscal costs.

While the tabulation of all costs, including those Hennessey believes are “sunk,” may or may not be of immediate concern to policymakers with respect to granting amnesty to this current population of unlawful immigrants and their children (compared to keeping the status quo), it is still valuable information. Cataloguing the total fiscal costs of unlawful immigration helps us understand the additional taxpayer burdens our citizens and legal residents face because of unlawful immigration.

Granting amnesty to current unlawful immigrants may indeed encourage more unlawful immigration in the future. After all, when Congress passed the first amnesty for unlawful immigrants in 1986, the sponsors of the bill said it was a one-time only event. Should an amnesty pass, many would-be unlawful immigrants might conclude that the United States will engage in serial amnesties in the future and decide to enter or remain in the United States unlawfully. It is therefore important for policymakers to understand all the fiscal costs of unlawful immigration, which include the costs to taxpayers of educating and otherwise providing for citizen-children of unlawful immigrants. (Of note, this report did not try to estimate any future costs from future unlawful immigrants or the vast increase in visas for family members of those receiving amnesty.)

It is also worth noting that amnesty does add extra costs for those children. Unlawful-immigrant parents do not currently receive the earned income tax credit and the additional child tax credit. Rector’s calculations show that after amnesty, the parents will become eligible for both credits at a net cost of $10 billion per year.

Because our methodology looks at households, it does not count the costs of about 20 percent of unlawful immigrants who live in households headed by lawful immigrants or U.S. citizens, which in that way would undercount the cost of all unlawful immigrants. This is only one way in which the $6.3 trillion figure could be too low.

This also raises an important related point. Any study of this kind includes estimations and simplifying factors. In a study of nearly 100 pages, Rector has gone into great detail as to his methodology and listed the most important 10 ways the costs in the study could be higher or lower. Ultimately, the figures Rector calculates must be understood to be estimates that roughly balance the effects of these various assumptions. Even if some would do particular elements of the study differently, there can be no doubt that amnesty is costly to the American taxpayer on the order of trillions of dollars.

What About Citizens and Lawful Immigrants' Deficits?

Third, Hennessey argues that “The Heritage logic would apply equally to legal immigrants and to babies born to low-income U.S. citizens.”

In fact, Hennessey is correct that the paper is an indictment of our bloated welfare and overburdened entitlement programs. Heritage certainly agrees with that assessment, as do, apparently, our colleagues at other think tanks. For example, Chris Edwards at the Cato Institute recently noted, “To me, the study provides a very useful exploration into how massive the American welfare state has become.”

Nicholas Eberstadt at the American Enterprise Institute wrote an entire book examining the redistributive nature of government policy in the U.S. In that work, he noted that there is “very little empirical research [] available on the overall incidence of taxation and transfer recipience in modern America….One admirable exception…is a study by the Heritage Foundation’s Robert Rector and Christine Kim.” Rector uses the same basic methodology in examining the costs of unlawful immigration.

Undoubtedly, U.S.-born citizens with lower education receive on average far more in benefits and services than they pay in taxes. Our government provides high levels of benefits to the least advantaged U.S. citizens and legal immigrants while requiring they pay comparatively little in taxes. Previous research by Rector has demonstrated this. Although there is disagreement about the appropriate magnitude of redistribution – and nobody has done more than the study’s author, Robert Rector, to try to reform our means-tested welfare programs – Americans seem to generally accept the notion that government should support and protect the most vulnerable U.S. citizens and lawful immigrants.

Applying this same system of generous support, however, to unlawful immigrants whose only claim on taxpayer resources is that they have broken U.S. laws by unlawfully entering or staying in the country is another question entirely.

Entitlements Are the Problem

Fourth, Hennessey notes that the real problem is our unbalanced entitlement programs, not unlawful immigrants. “Illegal immigrants are being promised far more in benefits than they will pay in taxes not just because they are on average low income, but because almost everyone is being promised old age benefits that will exceed the taxes they will pay,” he wrote.

Hennessey is right that our unsustainable entitlement programs are a major problem. But how is adding millions more predominantly low-skilled individuals – those who entered or stayed in the United States unlawfully – to an unsustainable system helping at all? The answer is that it does not help; it makes the problem worse. (For the first decade or two it may appear to make our Social Security and Medicare problems better as younger unlawful immigrants pay into the system, but in the long run most will receive significantly more in benefits than they pay in taxes, just as all low-income Americans will.)

Heritage is hardly the only policy group to note the huge fiscal imbalance of Social Security and Medicare. Eugene Steuerle at the Urban Institute has published a piece looking at how much an average wage earner pays into and receives in benefits from these two programs. Even an average wage-earning couple who have worked all their lives and paid payroll taxes receive a huge net benefit from these programs, especially Medicare. Anyone with a lower wage who does not work for their entire lifetime in the U.S. gets an even bigger net gain courtesy of the taxpayer.

Hennessey argues that “Heritage’s calculations mistakenly assume Social Security and Medicare benefits will be paid in full to newly legal immigrants [in the future]” even though current law requires benefits to be cut before the amnesty recipients reach retirement. Specifically, Hennessey argues that in 2032, the Social Security trust fund will reach zero balance and Social Security benefits will be immediately cut across the board by 27 percent. Similarly, Hennessey predicts that Medicare benefits will be slashed by 17 percent in 2024. All too true, but what would Hennessey propose that the analysis assume instead?

Heritage is on solid ground in assuming for purposes of this analysis that these cuts will not occur. Look, for example, at the “sustainable growth rate” or “doc fix” that Congress changes every year. It is supposed to reduce reimbursement rates for government health care entitlement programs; instead, each year Congress puts off the cuts. At this stage no elected politician publicly supports massive cuts in Social Security and Medicare benefits such as will necessarily occur under current law. It would be interesting if comprehensive immigration reform proponents were to argue publicly that the long-term costs of amnesty will be reduced because they plan to cut Social Security benefits by 27 percent or more in the future.

Based on the benefits currently received by low-skilled legal immigrants, the Heritage analysis assumes that amnesty recipients will have future Social Security benefits only slightly above the minimum benefit level. As the Social Security trust fund and Health Insurance trust funds incrementally run out of money, and as Medicare more generally places an ever-increasing burden on the rest of the budget, changes will occur: more money could be taken from general revenue, benefits for the more affluent may increasingly be taxed, means-testing of benefits may be increased, and FICA taxes may be increased. The least likely change is that future benefits will be slashed across the board, including for the poorest beneficiaries; amnesty recipients will overwhelmingly fall in that group.

Amnesty increases the number of future low-income beneficiaries in Social Security and Medicare. By increasing the long-term unfunded liability in both programs, amnesty increases the pressure for future tax increases and/or benefit cuts.

What About Mobility?

Fifth, Hennessey says, “as best I can tell, they assume that a poor, low-skilled, poorly educated illegal immigrant will remain poor, low-skill, and poorly-educated, and that he will draw government subsidies his entire life.” At Heritage we have been studying mobility and continue to work on understanding the challenges of low-income, low-education workers trying to climb the economic ladder. We hope to be able to increase the chances for mobility and education for all Americans and those who come here looking to work.

Contrary to Hennessey’s suspicions, the analysis increases unlawful immigrants’ wages. A good way to project what will happen, on average, with low-skilled unlawful immigrants is to look at what happened after 1986. The amnesty enacted in 1986 did lead to an increase in wages among former unlawful immigrants. Based on academic research of the 1986 amnesty, the Heritage analysis assumes that amnesty recipients would receive an immediate boost of five percent in wages and there would be increases in reported incomes and taxes paid by boosting the portion of work they do “on the books.”

Admittedly, a one-time boost does not completely answer Hennessey’s concern. The data on less-educated immigrants indicates that wage increases will, on average, be modest and will mostly occur before age 40 (the average unlawful immigrant is 34). Moreover, as low-skill workers age, their labor-force participation declines and disability and medical costs increase. As a consequence, although wages go up with age (as reflected in the Current Population Survey data used in the report), the fiscal deficit per household still remains high. The bottom line is that even if unlawful immigrants’ wages were considerably higher than we already adjusted them, such households would still have a large and, Rector’s work shows, substantially similar fiscal deficit.

In addition, once unlawful immigrants earn access to the full panoply of welfare, they will face the same dilemma as U.S. workers: earning more means less government benefits. James Pethokoukis of the American Enterprise Institute calculated that a mother of two would be as well off with a job making $29,000 a year as a job making $69,000 a year after you include her government benefits with the lower-paying job.

Unfortunately, low-skill wages have not been increasing much, if at all, in the United States for some time, even before our recent recession, so dramatic increases in the average wages of unlawful immigrants, who on average have a tenth grade education, is unlikely.

Hennessey also notes that there are “people who arrive in the U.S. illegally [who] may initially take jobs well below their skill level because of language barriers that they later overcome. Others will get further education or build skills over time.”

Undoubtedly there will be some who will advance in skills and income more than average. As opportunity conservatives, we should do our best to help everyone to advance economically. It is important to remember that among unlawful immigrants, more than half do not have a high school diploma and another 27 percent have only a high school diploma. The vast majority of this group will work in relatively low-skill, low-income occupations for the rest of their working lives.

As Hennessey points out elsewhere in his critique, the fiscal deficit of unlawful immigrants is not unlike the fiscal deficit of lawful residents and U.S. citizens. On average, people with low education levels – of whatever immigration status, ethnicity, or national origin – will receive significantly more in government services and benefits than they pay in taxes.

It should also be noted that the analysis matches amnesty recipients’ benefits (means-tested welfare, etc.) to those of legal immigrants of the same age and education level.

More Labor From Unlawful Immigrants

Sixth, Hennessey notes that “the fear of discovery and deportation has to constrain the labor supplied by those here illegally.” He uses the example of an engineer working as a cab driver since “green card verification is weaker.” As just pointed out, a majority of unlawful immigrants lack a high school diploma and another 27 percent have only a high school diploma; the vast majority, therefore, are not engineers. Moreover, the critique is at least partially answered by the increase in wages immediately following amnesty and the fact that the study moves most work “on the books,” both of which lead to higher taxes immediately after initial amnesty.

His point about increased labor supply, however, is generally valid, though by definition one can only speculate as to the extent of the effect. We did not include a macroeconomic modeling effect of increased labor supply due to amnesty, but total output and incomes likely would go up a little since an increase in labor necessarily means an increase in GDP. The unlawful immigrants are already participants in the economy, so it would only be a modest increase in hours worked among a few million workers in a $15 trillion economy. The effect would most likely be only a very small increase that would overwhelmingly go to the formerly unlawful immigrants themselves, not to those here lawfully.

The proper economic lens through which to consider immigration and immigration reform is not whether GDP goes up or down – it will always go up with more labor – but does it increase the after-tax income of American citizens and lawful residents?

Summations of Costs

Seventh, Hennessey points out that the numbers are summations of 50 years of costs. We chose this time frame because the point of the paper was to look at the lifetime costs of amnesty for a specific group of people. The bulk of these costs will be outside the 10-year window typically used by the Congressional Budget Office. This is especially so because legalized immigrants under the descriptions of the immigration reform bill in the Senate, for example, would not immediately have access to many means-tested welfare benefits. (In the paper we assume many benefits will be unavailable during the “interim phase” of amnesty that lasts 13 years.) Looking at long time horizons is important when we are examining costs of programs that promise expensive retirement benefits.  The government, for example, looks at 75-year time horizons for these programs.

In some ways, a constant-dollar approach presents a smaller cost than it otherwise could; health and welfare inflation and spending have historically increased faster than overall inflation, but in our study we assume these costs will increase equal with overall inflation.

We chose a legitimate way of adjusting for inflation, by using constant 2010 dollars, believing this is a more understandable way to present costs. Hennessey’s preferred approach of net present value is also a valid way to look at costs and can be helpful to compare future costs to current costs. Looked at another way, net present value is the amount of money we would have to put aside today in order to pay for the costs over the next fifty years or more.

Cost to Government Not Enough to Know

Eighth, Hennessey faults the study for focusing only on the cost to government. Others have raised this same concern. The cost to government (or, more accurately, the cost to the taxpayers) is the point of a fiscal cost analysis; it looks at the fiscal, i.e., tax and spending, consequences of a policy. We have elsewhere pointed out that there could be economic gains from a properly functioning immigration system. Some reforms, like increasing high-skilled immigration, could help the economy and actually lower, on average, the fiscal burden of U.S. citizens and lawful immigrants.

The key question is: if we can get the bulk of economic benefits that those who support comprehensive immigration reform promise without absorbing the costly incremental fiscal impacts of amnesty, why wouldn’t we do that?

Our nearly 100-page study takes an exhaustive and detailed look at our modern redistributive welfare state as it operates in the context of illegal immigration. Proceeding with immigration reform without considering fiscal impacts is unwise.


Heritage and all those who want to know the cost of amnesty have benefited from the questions Keith Hennessey has raised. While not everyone will be completely happy with a complex analysis of 34 categories of taxes (including lottery ticket sales) and 73 expenditure categories, Rector’s work provides critically important information that must be considered by policymakers at this time of extremely high debts, bloated welfare spending, and clearly overburdened entitlement programs. We welcome questions about the report and additional work by others that properly accounts for our modern, massive redistributive state in considering the costs of amnesty.

-Derrick Morgan is the Vice President of Domestic and Economic Policy at The Heritage Foundation.

First appeared in Real Clear Policy.