By Just before the election, the Manhattan Institute (MI) released a report titled “Lifetime Fiscal Impact of Immigrants.” MI claims that the recent surge in immigration under President Biden will cost the federal government $1.15 trillion over the lifetimes of these new immigrants and that “mass deportations would significantly reduce the national debt.” Expect this analysis to be cited by the incoming Trump administration.
Today, the Cato Institute has published my response as a working paper, detailing serious errors that, when corrected, reverse MI’s central conclusion. In fact, after corrections, MI’s model predicts enormous fiscal benefits from the recent influx of illegal immigrants: a deficit reduction of $4.9 trillion over the immigrants’ lives. This is consistent with the recent findings of the Congressional Budget Office (CBO).
Here are the nine major issues my paper identifies:
MI assumes that immigrants cause large, immediate increases in military spending;
It excludes significant tax revenues from corporations employing the immigrants;
It inaccurately attributes the costs of the child tax credit solely to parents;
It assumes that low-skilled immigrants are just as unlikely to leave the US before retirement as high-skilled immigrants;
It doesn’t account for how immigrants reduce interest payments on the debt;
It assumes new illegal immigrants are as uneducated as those from a decade ago;
It assumes recent illegal immigrants are as old as other immigrants;
It assumes illegal immigrants are just as likely to use entitlements as other groups; and
It fails to account for interest costs on deportation spending.
Even if you correct only the last item on this list, MI’s estimated deportation costs jump to $1.6 trillion, which exceeds the supposed benefits of mass deportation. The report’s author helpfully shared the model so I could reevaluate its results precisely.
#1 Military spending: MI states that military spending will increase because more people will mean more military personnel, but military personnel has actually halved since the 1950s. Moreover, defense spending has decreased both per capita and per US-born person since the 1950s. MI’s assumption about defense spending is also bizarre when applied to deportation because it implies that Congress will cut defense spending by $40 billion if the recent immigrants are deported—which is unrealistic, to put it mildly.
#2 Tax revenues: MI excludes about $771 billion in tax revenues from its analysis, primarily from corporate income taxes. It is well-established that wage earners create a proportional increase in corporate income (which is why they are hired), yet MI simply leaves out all this revenue. MI claims that it cannot determine the exact percentage of corporate income taxes that should be credited to employers versus workers. The Tax Foundation assesses that most should be credited to workers. Instead of making its own estimate, MI takes the extreme position that all new immigrants generate no additional corporate tax revenue, which is not true.
#3 Emigration: MI assumes that immigrants without a high school degree are as likely to stay in the United States permanently as those with a college degree. This assumption is wrong and critically skews the analysis because most costs accrue during retirement. My suggested adjustment is really an understatement for illegal entrants because many illegal entrants are deported, leading to even higher emigration rates.
#4 Interest costs: To calculate the cost of interest, MI projects future interest costs from new debt and distributes these costs among the fiscally negative people. Two problems here: 1) some new debt is from public goods (like military spending), which immigrants don’t cause; and 2) immigrants who are fiscally positive throughout their working lives accrue a fiscal surplus that reduces the debt in their retirement. Handling this issue correctly flips some immigrants from a fiscal net negative to a net positive.
#5 Tax credits: MI allocates 100 percent of the cost of the child tax credit and the earned income tax credit to the tax filer. This is legally and factually inaccurate and inconsistent with MI’s stated methodology, which apportions all other government benefits for children to the child and distributes household benefits to the whole household. MI could have expanded to calculate the costs and benefits of immigrants’ descendants. But it is biased to include some costs and no tax revenues, especially when the 2nd generation workers are the highest-earning Americans. Finally, MI’s choice here implies that deporting the immigrant tax filer head will zero out these payments when, in reality, it may just shift them to a different tax filer because the US citizen child or spouse cannot be deported and is entitled to these benefits.
Here are how these adjustments affect the results for young, low-skilled immigrants. MI’s results aren’t plausible if any of these changes are made.
These first five issues are all things that affect immigrants in general. But then MI takes the results of its profile of the average immigrant—legal and illegal—and tries to apply them to the recent illegal entrants to the United States.
#6 Immigrant education: The report cites an estimate of the educational attainment of illegal immigrants from 2015, even though we now know that recent illegal immigrants are much more highly educated than the cohort a decade ago. Since education is an important predictor of income, this decision decisively reduces tax revenues from new immigrants.
#7 Immigrant ages: MI inaccurately estimates the age profile of recent illegal immigrants by failing to use the publicly available data that come directly from the Border Patrol. It overstates the number of retirees crossing the border, inflating the share of border-crossing retirees (65+) by a factor of 13. Retirees are much more costly than young immigrants.
#8 Immigrant welfare use: MI assumes that recent illegal entrants will be just as likely as any other immigrants with their demographics to use welfare and entitlement programs like Social Security and Medicare—even though they are legally prohibited from accessing those benefits. Of course, some will receive asylum, but it’s a small minority. This assumption is perhaps the least understandable. MI admits that this inaccurate assumption biases the result, but still uses the $1.1 trillion cost in its headline results.
#9 Deportation interest costs: MI accounts for the interest costs on debt incurred by fiscally negative immigrants over 100 years. But when it is estimating how much cheaper it would supposedly be to deport them, it doesn’t account for the interest costs on that spending, which implicitly assumes a “deportation tax” is used to pay for roughly $500 billion in spending (imagine that!). Once interest costs are accounted for, deportation costs $1.5 trillion, which is more costly than even MI’s fatally flawed $1.1 trillion estimated cost from the immigration surge.
Altogether, MI’s revised analysis shows that mass deportation would cost the US government about $6.4 trillion—$1.5 trillion from the deportations themselves and $4.9 trillion from the lost tax revenue. My paper provides more detail and explanation behind these numbers, and people with an interest in this subject should read the whole report. However, unlike CBO’s results, MI’s results (even after my corrections) do not account for the indirect economic growth effects of immigration, so they should be seen as the minimum possible benefit.
If the purpose of the paper was really to produce the most economically efficient immigration policy, MI should suggest letting immigrants come or stay without benefits. Indeed, restricting welfare benefits makes many low-skilled immigrants more fiscally positive than high-skilled immigrants with benefits. But MI never suggests this policy, focusing more on deportations and exclusions. That doesn’t make sense. Immigrants would happily trade benefits for the right to stay here without fear of deportation, and Americans would benefit from their presence. It is a clear win-win.
MI did take the time to produce an interesting model, but because it incorporates many inaccurate assumptions, its results are flawed. MI’s model—after corrections—reveals significant fiscal upside to immigration—legal and illegal, low-skilled and high-skilled. It should accept these corrections, implementing a more accurate fiscal effects model in the future.
