Wall-o-nomics: Calculating the cost of refugees to the U.S. economy

The proposed Wall/fence that dominates the news right now is at least partly about economics. People who make it across the border are then entitled to make an asylum claim and live in the U.S. for years of administrative processing and, if successful, live in the U.S. forever. They can collect welfare while doing this. Their children and grandchildren born on the U.S. side of the border can collect welfare as well. Advocates for an open border (“A wall, in my view, is an immorality.” — Nancy Pelosi) say that taxes paid by migrants exceed the welfare cost. Let’s look at this…

A reader of an earlier post cited “The Economic and Social Outcomes of Refugees in the United States: Evidence from the ACS” (NBER) as evidence that we are running a profit on our refugee industry:

“By the time refugees who entered the U.S. as adults have been here for 20 years, they will have paid, on average, $21,000 more in taxes to all levels of government than they received in benefits over that time span, according to a working paper released Monday by the National Bureau of Economic Research that examined the economic and social outcomes of refugees in the U.S.”

So if we believe the expert economists at NBER and don’t dig into the paper, we make a profit of $1,050 per year on every refugee who arrives as an adult. The Federal deficit of $779 billion for FY2018, therefore, could be wiped out if we simply admitted 741,904,761 adult refugees, e.g., by asking most adults in India or China to move here and spin an abuse yarn.

When we dig a little deeper, though, it seems that the economists have had their thumbs on the scales (or somewhere else?). An “adult” is defined by the researchers as 18-45, but refugees are admitted without any age limit. A disabled 70-year-old has the same right to asylum as an able-bodied 22-year-old.

Suppose that all refugees were actually aged 18-45 and the economists had gotten the rest of the analysis correct. Would refugees yield a net profit? They’re paying more than they’re taking so they’re not “takers,” right? The Abstract reveals one question to explore: “After 6 years in the country, these refugees work at higher rates than natives but they never attain the earning levels of U.S.-born respondents.” Even for this cherry-picked age subset, the idea is that we’ll become richer overall by having lower earnings on a per-capita basis (and of course anyone in the U.S. income inequality industry will have an uglier statistic to wave around). Can that work?

The paper looks at six government welfare programs: “There are six social insurance programs that account for the majority of government payments to U.S. citizens: welfare, Supplemental Security Income (SSI), Social Security, food stamps, Medicare, and Medicaid.”

What if we subtract the cost of these programs from the total cost of running local, state, and federal government and then see if $1,050 per year per person will pay for the remainder? If we budget the above subset of welfare at $2 trillion per year and subtract from about $7.5 trillion per year in total spending (source) we get $5.5 trillion. Divide by a population of 328.4 million (popclock) and we find that it takes $16,748 per person to fund our government minus these headline welfare programs.

Each refugee paid a net $1,050 per year and consumed an additional roughly $15,700 in government services (roads, schools, libraries, police and fire protection, etc.). Over a 20-year period, then, the refugee took approximately $314,000 from other taxpayers.

Did the economists even begin to do a full cost accounting, though?

As of 2012, there were 79 Federal means-tested welfare programs (Heritage). The NBER looked at only 6.

Since the refugees never get to the median U.S. income, the typical refugee never gets above the 400 percent of poverty disqualification threshold for Obamacare health insurance subsidies. The typical refugee would also be eligible for public housing, a program that can be worth $60,000+ per year per family in the NBER’s home town of Cambridge, Massachusetts and yet they didn’t think it was worth including.

If we assume $5,000 per year in health insurance subsidies and $20,000 per year in public housing subsidies per refugee, over a 20-year period the best-case refugee now costs $814,000. That best-case refugee showed up during his or her core working years.

Once across the border, of course, the best-case refugee or asylum seeker should live for longer than 20 years. Would it be fair to round up the total cost to $2 million? Therefore if $5 billion is spend on an immoral wall/moral fence it has to stop 2,500 migrants in order to pay for itself in pure economic terms?

[Of course, there is more to life than money. We might have other reasons for wanting an open border, e.g., superior morality, loneliness if the U.S. population remains stuck near 330 million, etc.]

Readers: Did I miss anything? Or is $814,000 over 20 years a reasonable estimate? Also, how can people imagine that someone who pays a net $1,050 per year in tax is going to be of any real help in keeping the U.S. government going?

11 thoughts on “Wall-o-nomics: Calculating the cost of refugees to the U.S. economy

  1. This is a very strange accounting. First, under your assumptions I’m not sure how someone 45 or under uses Social Security ($1.05 trillion, $3,000 per person off). From what I can see that alone, with Medicare and Medicaid is well over your $2 trillion. Second, the authors clearly state that they exclude the cost of education, as none of the refugees will take advantage of it. Their children clearly will, and should be given their own accounting. The paper clearly says that children of refugees achieve parity with domestic citizens in education and taxes, but not sure you have ever cared about the future.

    It’s also very strange to consider that the interest payments on debt ($363 billion) is the responsibility of a refugee. Going forward, additions to that can be averaged to a refugee, but servicing existing debt is not somehow caused by a refugee. Additionally, strange accounting to think that we should use an average of national defense (~$1 trillion). I guess you could make a marginal cost argument in trying to keep them out, but letting in more refugees may actually lower the national defense cost. The same with infrastructure spending, I’d imagine the average refugee uses considerably less infrastructure than citizens, and while there may be some marginal projects because of refugees, I’d like to see the data on how much.

    Then you start making asinine assumptions about public housing, etc. and try to extrapolate that to the entire population. Surprisingly, public housing is already included in the government spending numbers you have used.

    Finally, it’s incomprehensible to argue purely from a government spending argument. If you are making that argument, given our massive deficits, then every citizen must be a drain on the country on average. The deficit this year is expected to be ~$985 billion. Meaning that every citizen this year is in the hole for about $3K. Fortunately, people better at this than you recognize that the economy extends outside of the government. Adding people to an economy, if they are productive, adds to the economy. This is the strangest argument by people scared of immigration, that there is some fixed number of jobs that will be stolen. Increasing economic activity, unsurprisingly, increases economic activity. Additional jobs, payroll taxes (which I didn’t see in the paper), businesses and business taxes, and an increase in income with that increased activity.

    You can be against refugees, and you can even think your wall will keep people out (it won’t), but you should be a little less clueless in your analysis.

    Finally, if you are following international law the wall should be entirely irrelevant to our decision on refugees. It may prevent illegal immigration, but that wasn’t what you discussed at all.

    • A young immigrant can’t use Social Security? Why not SSDI? (the paper accounts for SSI separately) And eventually won’t the refugee be collecting conventional Social Security just the same as anyone else who has lived in the U.S. during his or her working years?

      How is the consumption of public housing accounted for in the paper? (I don’t think it is accounted for anywhere, actually, since the modern practice is to force apartment building owners to give up 10 or 12 percent of their units for allocation by a housing ministry. So the cost of public housing ends up being hidden in higher rent for Americans who are irrational enough to work and pay market rent.)

      Children of penniless welfare-dependent migrants ultimately earn the same as the children of native-born? That would be awesome! It means the U.S. is a model of class mobility! How can people simultaneously say that the U.S. system is rigged in favor of the privileged and that young people who grew up without any privilege end up doing just as well?

    • Continuing to look at the claim that the children of refugees will be earning and paying taxes at the same level as children born to native-born Americans…

      Aren’t most of the migrants/refugees/asylum-seekers coming over the southern border Hispanic under the U.S. government’s definition? https://www.brookings.edu/blog/social-mobility-memos/2015/11/10/how-upwardly-mobile-are-hispanic-children-depends-how-you-look-at-it/ says “Hispanic children fare quite badly in the U.S. compared to other Americans … One way to look at how second-generation immigrants are faring is to compare them to other groups in the U.S., including whites. Through this lens a familiar picture can be seen. Hispanic Americans, along with African Americans, have lower median earnings than both whites and second-generation Asian Americans … College completion rates are also lower for 2nd generation Hispanic immigrants compared to their white peers”

      The report from the well-meaning folks at Brookings says “The children of poorly educated adults are much less likely to end up highly educated themselves.”

      So they are saying that low-skill folks who stream over an unfenced border will tend to have children with low education and income levels.

    • Scott: Finally, I don’t think that it makes sense to use the phrase “illegal immigration” as you do at the end of your comment. Anyone who is present on U.S. soil, regardless of how he or she came to be here, is entitled to make an asylum claim. Therefore, the idea of “illegal immigration” no longer makes sense. Every migrant is legal as soon as the words “I am seeking asylum” are uttered.

  2. Whew. Yes, I think you’re missing a lot of things and you’re understating the badness considerably.

    At the very top you note we could pay off the FY2018 Federal deficit if we just brought in 741,904,761 refugees. That’s true if you use the $1,050 figure, but from the paper: “Note that fiscal costs are larger than benefits for the first eight years in the U.S. Starting in year 9, refugees contribute more in taxes than they cost to the government in social insurance costs.” So the first 8 years with our 700+ million new neighbors are going to be a real doozy. Plus, that’s just the Federal budget deficit. What about the State and Local budget deficits?

    Reading the paper, they also completely ignore costs to the criminal justice system, which they estimate at $32,000 per year for incarceration of those who “interact” with it, but then decide that’s unimportant based on the refugee data they use, and out the window it goes. No criminal justice system costs. There are no criminals. Since they ignore those costs, they must also be ignoring the costs borne by their victims. How can there be any crime victims if there are no criminals?

    Then they don’t account for what happens after age 65 when the refugees who are still alive become eligible for Medicare and Social Security. Why? Because: “With the synthetic cohorts, we only observe people in the analysis that are alive at the time of the survey. Without estimates of mortality patterns of refugees, it is difficult to build in expenditures after 65.” So they just don’t do it.

    Oh, and also, from page 27: “One cost that is excluded from this analysis is the cost of public education for refugee children.” Why? “If the investment in the education of U.S.-born children results in an economic benefit to society in this calculation, then we would expect that the investment in the education of refugee children to result in a similar benefit.” So they just state that’s it is true because they expect it to be true, and that’s that.

    Also your $16,748 per person figure isn’t adjusted for the time value of money?

    What I really want to know is: How does anyone get away with publishing a paper like that, with all the assumptions, omissions, “synthetic cohorts”, etc., and then have it used to explain and justify policy decisions?

  3. https://www.nytimes.com/2019/02/16/us/border-migrants-asylum-mexico-aclu.html

    Pushed beyond their limits by prolonged waits in dangerous and squalid conditions in parts of Northern Mexico, thousands of caravan members who had been waiting to seek asylum in the United States appear to have given up, Mexican officials said, dealing President Trump an apparent win after a humbling week for his immigration agenda.

    About 6,000 asylum seekers who had traveled en masse, many of them in defiance of Mr. Trump’s demands that they turn around, arrived in Northern Mexico in late November as part of a caravan that originated in Honduras. Since then, more than 1,000 have accepted an offer to be returned home by the Mexican government, the officials said.

  4. @Scott, @Alex, and anyone who is in support of migrants, let us ignore the $$ numbers for a minute.

    If this NBER finding is so true as such legal and illegal migrant add so much value to the country they are migrating to, then why they are not a positive to the country they are migrating from? Shouldn’t the government of their native country be doing everything in their power to stop them from migrating? And why other countries, such as Canada to name one isn’t welcoming them without any restriction [1] given the finding of NBER? And why every country has a yearly quota on number of migrant it will accept?

    As a side note, it can take up to 20 or more years for someone to migrate legally to the US and that person need to provide an Affidavit of Support [2] in order to get his / her application approved. Who ever signed the Affidavit of Support on behalf of the migrant is on the hook until when the migrant is US Citizen.

    [1] https://www.canada.ca/en/immigration-refugees-citizenship/services/immigrate-canada.html
    [2] https://www.uscis.gov/greencard/affidavit-support

  5. How about just escorting them to Canada? I keep bringing up this idea because Trudeau keep saying he will welcome all people to his country. I say we take him up on his offer. Every time there are a few hundred “refugees” that appear at the border, there are buses that basically drive them to Canada. Let’s provide bathrooms on the bus and some snacks. They get off long enough to stretch their legs in a remote area. Maybe there are more bathrooms and food there. After a quick rest break, they are counted and the same number go back on the bus. Military is there, armed, to remind them that they are not in America but a holding zone. Maybe the UN could oversee it. We deliver them to Canada, and they are Trudeau’s problem. While we have them at a rest stop, maybe we could vet them and offer asylum to regugees who actually have needed skills.

  6. Sadly, all the money we spend on this problem could actually solve it if we actually were smart about it.

    Politicians want to keep trying the same things with the same rules which keep failing and that is the definition of insanity.

    First, I think most refugees are honest people who want to work. I think they just need a “jump start”. Since employers enjoy the benefits of having lower wage workers, responsibility should fall to them to train them and be responsible for them.

    When an immigrant arrives at a border post, we interview them. We find out if they lived city or country. If they lived city, then what job did they do? We match employers with similar language skills and cultures, as much as possible. We basically get them a job.

    In the interview, we explain to them that they get “one chance” in this country. If they protest, commit crimes, drink and disorderly, etc. they are deported. We expect them to be “model citizens”. That includes family members as well. Employer must arrange for housing. We keep families together.

    If teens misbehave, the family is deported. They pay taxes and those taxes pay for strict law enforcement. They wear tracking drives while here. So they cannot disappear.

    There is a carrot for the immigrant and their family. They will get a green card in 5 years if they learn English and obey the laws. None of this 20 years bs. I have an Indian friend who came here in 1998 with his infant daughter, and when I left the company in 2017, he still was working on a Visa. That is ridiculous. He was worried that she was going to have to go back to India with him, with no skills to live there. No Hindi, nothing.

    The focus needs to be on making every immigrant and US citizen between the ages of 18 and 65 a productive working member of society. We need to provide incentives to get the homeless out of cities, and into education centers to teach them to be citizens. Those who refuse should be offered $50,000 cash to renounce their citizenship and provide a list of countries who will accept them and their money, and they can leave.

    We need a border to keep out people who come here to mooch. That, and reducing military spending by about 50% will fix our budget problems. I have no problem supporting “widows and orphans” and retired people who worked most of their lives in the US. I have a huge problem supporting people who think the US “owes them” anything.

    • Many good ideas, and since they are good, have zero percent chance of being enacted.

      You are promoting matching authority and responsibility, and in general proposing a “skin in the game” approach.

      I must point out that the annual budget deficit is ~$1T, whereas the defense budget is $700 billion, so reducing defense spending 50% only solves 1/3 of the deficit problem.

  7. Did I miss anything?

    The main thing that you missed is that there is no plan to build a solid wall. There will still be ports of entry. Asylum seekers will still be able to present themselves to the CBP and request asylum.

    Also, I’ve never heard anyone at all suggest that a wall be build along the borders with Canada. So if a border without a wall is an open border, every American is in favor of open borders. There’s also the issue of maritime borders, the lines on the map that mark the boundary between the part of the ocean is under American jurisdiction and the part that is international waters. Securing that those borders would require many thousands of new coast guard cutters.

Comments are closed.