From Orrenius et al. (2025), GDP growth is down 0.81% in 2026, 0.49% in 2027.
These estimates based on a VAR(2) using annual data, net unauthorized immigration as a percent of the previous-year U.S. population, annual growth rate of nonfarm payroll employment, the PCE annual inflation rate, and yearly GDP growth.
Source: Orrenius et al. (2025).
PCE inflation is 15 bps and 6 bps higher in 2026 and 2027 under baseline.
It’s interesting that the attributed positive effect on GDP from illegal immigration coincided with an economic rebound of pent-up domestic demand from the COVID shutdowns. I wonder how that was separated in the analysis. The old saying that correlation is not necessarily causation survives for a reason. Other voices say differently with regard to the net effect of illegal immigration.
https://budget.house.gov/imo/media/doc/the_cost_of_illegal_immigration_to_taxpayers.pdf
I accept that there may be different viewpoints and data used when attempting to estimate either the positive or negative impact of illegal immigration. Golf courses and slaughter houses love compliant, low-paid workers. Poor Americans might not like the competition for subsidized, low-income housing.
https://www.axios.com/2023/09/23/housing-crisis-migrant-immigrants-homeless
Bruce Hall It’s interesting that the attributed positive effect on GDP from illegal immigration coincided with an economic rebound of pent-up domestic demand from the COVID shutdowns
The authors used data going back to the 1950s, so the relationship is not just COVID rebound effects.
The old saying that correlation is not necessarily causation survives for a reason.
The authors used a structural VAR, so your comment really doesn’t apply.
I accept that there may be different viewpoints
The House budget stuff is garbage economics. Virtually every intellectually respectable study finds that immigration increases GDP. FYI, the non-partisan CBO found that immigrants increase GDP. The House budget “analysis” is just intended to provide people who don’t have a deep understanding of economics a kind of fig leaf to hide behind…as though there’s some kind of equivalence between MAGA economics and actual economics.
What you said.
Brucie has Dunning-Krugered himself again.
Bruce Hall Your link on the cost of illegal immigration is shot through with errors. This Bozo has it completely wrong regarding the net cost of immigrants. According to CBO https://www.cbo.gov/publication/60569#_idTextAnchor019
In its analysis of the effects of the surge, CBO considered the effects on the budget stemming from taxes paid and benefits received by the surge population and their U.S.-born children as well as the budgetary effects of the changes in the economy caused by the surge.
The largest effects on the federal budget stem from individual income and payroll taxes paid by immigrants. In CBO’s baseline projections, the people in the surge population pay $788 billion in taxes over the 2024–2034 period. The participation of those immigrants and their children in federal programs adds $177 billion to mandatory spending over the period.
As to the overall effect on GDP:
Some of the projected budgetary effects of the immigration surge stem from broader changes in the economy that the surge is expected to bring about. In CBO’s projections, the surge boosts total nominal gross domestic product (GDP) by $1.3 trillion (or 3.2 percent) in 2034 and by $8.9 trillion over the 2024–2034 period. The surge increases the total amount of wages paid each year by a percentage that grows steadily over that period and reaches about 3 percent in 2034. Those additional wages are a major contributor to the boost in revenues because they are subject to both payroll and income taxes. In addition, two main factors resulting from the surge—faster growth of the labor force and greater demand for residential investment—boost the rate of return on capital and put upward pressure on interest rates. The increases in interest rates are a major contributor to the boost in federal spending.
And keep in mind, these numbers only refer to those immigrants who entered the country during the “surge.” Immigrants who have lived here for many years add even more to GDP.
You really need to find some better (i.e., less clearly racist) sources. And your source isn’t even an economist, so he really should shut up about macroeconomics.
U.S. economic and weather data is going to be much more difficult to find or nonexistent under the Trump admin. One example – the Trump admin stopped collecting the economic impact from flooding – some estimates from recent Texas flooding is that it caused $15-22 billion in damages. https://www.bloomberg.com/news/articles/2025-07-10/trump-s-cuts-to-federal-government-are-making-data-disappear
Hmmm.
If I take the Orrenius data (tab Data 4), and I add up the annual net unauthorized immigration, we come out with just under 60 million in total through 2019. No estimate, even the most virulent ones, comes anywhere close to this. Through 2019, PEW Research, generally considered the gold standard of unauthorized population estimates, puts the total around 10-11 million, with really high right wing estimates around 22 million.
Further, I have no clue why Orrenius stops the data at 2019, given that we have monthlies to date, I believe.
I may have to spend more time on the matter. But there is a notably missing graph above, and that’s the unauthorized immigrant share and absolute numbers of the labor force. That should be graph 2, where the GDP growth graph sits now.
Perhaps if you’d read the study:
“We choose 2019 as the end of our estimation sample to avoid including the pandemic years.”
If the authors do math based on public sources, and then you do some math you don’t show us and come up with a number you don’t find credible, perhaps you should check your own math before questioning theirs. For instance, you seem to have assumed that no immigrants have died since 1956? We’d better stick with the authors numbers.
First, the post title should read GDP Impact of 87.5K Removals/Month. The authors make this clear: “This methodology allows us to consider several scenarios for the future evolution of unauthorized immigration, including a mass-deportation scenario in which the number of people removed from the country annually rises to 1 million by year-end 2027.” (87,500 * 12 = 1,000,000) An annual removal of 87.5k people would not be detectable in the macro data.
I think Orrenius stops at 2019 for a different reason. During the Biden administration, migrants crossed the border illegally, but immediately claimed asylum status. Many of these people still have pending cases.
Indeed, the authors lead with this right from the top:
The U.S. experienced an unprecedented immigration surge from 2021 to 2024, driven mainly by an influx of immigrants crossing the Southwest border without visas. Many were released into the country after encountering the Border Patrol.
These immigrants were issued humanitarian parole, allowed to enter the U.S. to seek asylum or given notice to appear in immigration court. Others entered the country illegally without encountering enforcement personnel. The Congressional Budget Office (CBO) projected that 7.3 million “other foreign nationals,” unauthorized immigrants or immigrants with some form of quasi-legal status, but not valid visas, were added to the U.S. population on net during 2021–24.
So we added during 2021-2024 some 7.3 million “other foreign nationals”, quasi-legals, on a 2019 PEW unauthorized population estimate around 11 million. That’s a two-thirds increase! You can’t just leave that out of the statistics.
https://www.pewresearch.org/short-reads/2024/07/22/what-we-know-about-unauthorized-immigrants-living-in-the-us/
Notwithstanding, I agree with the general tone of the conclusions. If we deported all illegals and quasi legals, that would be the equivalent of deporting almost the entire population of New York, that is, all but the population of the fourth most populous US state. Do you think that would show up in the macro data? You bet.
Declining immigration weighs on GDP growth has the feel of the Harvard SPH Maria thing, but I think the overall conclusions in this case are solid, and I do not think they represent the worst case scenario in terms of GDP growth.
I would remind all readers, no doubt including Orrenius et al, that the US has never defeated a single black market of any sort without turning to a legalize-and-tax approach. Therefore, the scenarios painted in Chart 2 are all implausible, except in the case that democracy has failed in the United States. Other than that, I will take the ‘over’ on illegal immigration. History makes clear that, in illegal immigration as with other black markets, market dynamics inevitably dominate laws.
I’m always happy to come to the Dallas Fed, or indeed the Fed in DC, to present more plausible scenarios and make the case for the inescapable necessity of a market-based solution to end illegal immigration. The timing for such a solution is actually pretty favorable.
Steve Kopits: For Pete’s sake, can’t you read the linked-to-article. It says:
Having presented at the Dallas Fed, I would certainly like to watch the audience’s reaction to a presentation you would make there.
First, you didn’t know why the authors stopped at 2019. Then, when I pointed out that the authors had given a clear reason, you essentially called them liars. Sure, Stevie. The problem is with them, not you.
You told Menzie he “should” have used 87.5 thousand per month, when the article gave him a clear reason to use 87.5k per year. You ignored that reason and dug up one of your own. Soooo eager to correct what doesn’t need correcting.
This sloppy reading followed by arrogant posing that you do, so very often, is one reason you have to go around begging to be heard. Real expertise is valued whenever real analysis is wanted. Posers and fakes are valued when preconceptions are all that’s wanted. Here we are, showering in the felon-in-chief’s preconceptions and you still can’t get face time in DC! Something seriously wrong with whatever you’re doing.
The horizontal axis is years or months? Or?
mike: years.