One lead story in the Wall Street Journal today, How Immigration Remade the U.S. Labor Force, tries to get its arms around the immigrant impact on the US economy using data. On the one hand, the Journal gets points for trying to mine a wide range of sources, such as immigrant court records, a KFF Health News/Los Angeles Times survey to CBO data to Census Bureau surveys.
On the other (and this admittedly reflects the propensities of these sources, to go into drunk under the streetlight mode with the information they gathered) is that these compilations are inherently questionable given how large the population of illegal immigrants is relative to the total, and how most of them understandably are trying to stay under the radar. Even with less than ideal information, the story early on flags why there is a controversy: the inflows have risen considerably under Biden:
Since the start of 2021, net immigration to the U.S. has totaled roughly 9.3 million people, according to estimates by the Congressional Budget Office.
That’s more than three times the net number of people that entered the country over the previous four years
This chart gives a sense of the nature of the problem:
To clarify the categories, from the interactive chart:
.
Lawful permanent residents generally refer to green-card holders. Net inflows of this group have been relatively steady from year to year.
The nonimmigrant group includes temporary workers, student visitors and officials of foreign governments. Negative net immigration in a group means it got smaller from the year before, the result of things like people leaving the U.S. and changes in status.
The biggest net increase has been in the other foreign nationals category, which includes people who entered the U.S. undetected, as well as those who were paroled into the country and are awaiting proceedings in immigration court.
Before we go any further, yours truly is very leery of the pretend precision in this article. Conservatives are not crazy when they contend that the number of undocumented immigrants could well be higher than the estimates. I regard it as a chart crime to include such squishy figures from that taken from populations where the officials have much more solid counts. There’s no way (and no attempt) in any of the many charts and tables to indicate that the illegal immigrant data has vastly higher error bands than that from the other groups.
For instance, in the early 2000s, a unit of American Express asked me to identify long-term trends that they should keep in mind or use in marketing strategies. Obviously there are tons of firms and mavens who pretend to do that for a living. They were asking me to dig through a huge raft of this crystal-balling and pick out forecasts that seemed sound.
One of three I identified was demographic change. This study came right after the 2000 Census results were released. Demographers were shocked at the magnitude of the population increase. They had expected the US to follow other advanced economies who were having declining birth rates and show only a modest increase. The much-greater-than-anticipated rise was due primarily to the heretofore unrecognized magnitude of illegal immigration, and secondarily not allowing enough for the increase in the Hispanic population, which had higher birth rates.
The fact that so many experts had been blindsided by this change told me how bad the information capture had been, and how intrinsically difficult it would be to do much better. Even though there are no doubt concerted efforts to come up with better estimates, they are always going to be very squishy.
Mind you, many of the findings seem intuitively correct. However, when I was at McKinsey, once in great while, the studies would attempt to give theorizing more solidity by representing it as a graphic. I would beef that the charts should be labeled, “McKinsey intuition”.
Again, these tables have some underpinning but should have stronger caveats. So I would treat them as probably directionally correct but would be leery of quoting figures. The magnitude of the undocumented migrant segment is so large and so difficult to dimension with great confidence that it makes it hard to generalize to the entire immigrant population. That problem is worse if the critics are correct and the illegal immigrant numbers are much larger than shown in the estimates here.
Some examples below, based on Census Bureau monthly household surveys. The Journal acknowledges the probable incompleteness:
The number of post-2020 immigrants who participate in the monthly Census survey is small and demographers believe unauthorized immigrants are less likely to respond when the government calls to ask questions.
But looking at the people who do respond to the monthly Census allows some inferences about their characteristics.
Small samples also reduce reliability. Nevertheless:
As the article explains:
Recent migrants are younger and more likely to be of working age than U.S.-born Americans. Of foreigners who arrived since 2020, 78% are between the ages of 16 and 64, compared with 60% of those born in the U.S., according to the monthly census data.
That helps explain why they are also more likely to be in the labor force. Of recent immigrants age 16 or older, 68%—the participation rate—are either working or looking for a job, compared with 62% for U.S.-born Americans. In raw numbers, that likely amounts to more than five million people, equal to roughly 3% of the labor force.
The article also shows how different data sources, using different sampling methods, come to different conclusions:
The 12 largest source countries for newcomers assigned immigration-court hearings since late 2020 are in Latin America or the Caribbean, the TRAC data show, led by Venezuela at 14%, Mexico at 13% and Honduras at 8.5%.
Monthly census data paint a slightly different picture, suggesting that Mexico is the most-common country of origin, followed by Venezuela and India.
Admittedly, the comments on this story present sample bias of their own. But it was instructive to see some arguing against illegal immigration in order to support US wages and thus provide for stronger social programs:
Claude King
Misleading article again from WSJ, illegal immigration hurts the low income groups and degrades the quality of workers available.
Legal immigration is what we need, along with an energy policy that supports on-shoring production.
Our government has favored the Globalists profits over what’s good for the country.
Hiding behind cheaper prices and low quality instead good jobs and reduction of entitlements which are used to buy votes.
Some objected to not including other information sources (note I have not verified that the claim below is accurate but it didn’t elicit pushback in a pretty feisty se of comments:
Micky Murray
The article is incorrect because they list total immigration in 2024 at less than 4 million. The correct YTD number is over 8 million according to the department of homeland security. and the real number is even higher when you include the known and unknown gotaways. So it is easily over 10:million.
The article depicted the immigrants on average having higher levels that natives of those lacking a high school degree, but also doing a smidge better on the proportion with a college eduction. A comment on skill levels:
JoNg TiffMia
I’ve hired many of the illegal immigrants that have moved into my area. Many are hard working but unskilled to a level that is unbelievable. They have no idea how to paint a house or dig a hole, much less a trade that requires some level of skill and training. If there are millions of new workers with limited English and very low skill, who is going to pay for the training and upskilling of these people? On top of this, the undocumented nature of these folks means they have endless problems navigating life in the USA. One example, a Honduran worker I hired needed some medicine for his diabetes. He thought it was a prescription that he couldn’t get in the US so he had his daughter ship it from Honduras, only to have UPS hold the package because they couldn’t verify who the recipient was. He couldn’t navigate the UPS customs process so I helped him do it, only to find out that they were paying hundreds of dollars to ship an herbal creme that was available over the counter on Amazon for $25 and could be shipped to me by tomorrow. Just that small experience made me realize having a life as an undocumented worker is completely untenable to have millions of people under this status, permanently. Therefore, I conclude that Americans will be forced to realize this is reality at some point and we will have to accept that ALL of these people will eventually become documented Americans at some point. It’ll just take until the next Dem president, house and senate. That’s the reality.
I have to differ with the claim that illegal immigrants will be regularized at some point. I had a friend back in the day who had gotten partway through the legal immigration process and then stalled out due when his US wife stopped cooperating. Obama had developed an amnesty program, and he joked that he would be the only person to qualify, since it required paying all taxes due. He’d gotten as far as getting a SSN and so was compliant. I suspect the tax demand will remain a sticking point.
And a reader pushed back, objecting to likely labor exploitation:
Marc Jones
Take a look at Europe. It is tearing itself apart as it tries to do what you wish.
Look in the mirror. Is this about you being magnanimous or just being greedy? Those illegals you hire, do you pay them enough such that they can live without housing and food subsidies? Do you have them all enrolled in a healthcare program, so they can pay for what is necessary when they and their families are sick or injured. Are you making donations to your local school district so the extra educational services their children require are not dropped on the backs of taxpaying citizens?
I have been listening to this rap my whole life. It always sounds great. But when I turn it over and look closely, the picture changes.
These comments do not quite amount to a sea change, but they do seem to represent a much lower level of defense of migration as a presumed good thing than I recall seeing in the Journal in the past. The comments included observations that were shades of Brexit, on how high and low wage immigrants both were contributing to high housing prices. So if nothing else, the level of cheap sloganeering to at least somewhat reasoned arguments was an improvement from what I’ve often seen at the Journal before. One can take that as progress of sorts.
A huge problem with assessing immigration is the lag time in getting firm statistics (usually census data), and the extreme difficulty in determining if foreign residents in a country are permanent or semi-permanent, or just part of a floating population of people going where the jobs are at any one time or another (especially in Europe, where moving within the EU is quite easy). These people are often ‘semi-legal’ – for example, using student visas to work. Add this to the huge reluctance of governments to admit openly to either a huge immigration (or emigration) issue.
Some random anecdotes:
An in-law who has a small accountancy business in NY told me that every restaurant has two sets of books – an ‘official’ one, and a second one that includes all the illegal staff payments. Needless to say, the differential between the two are pretty significant.
In the 1980’s I was doing some student work for a consultancy in London, which included looking at census returns. I was told that while the census population of Greater London was 8 million the ‘rule of thumb’ in the industry was to add a million. When I later did door to door survey work in some north London boroughs, I found that if anything this was an underestimate.
Back in the early 00’s in Ireland I was looking through some census returns that indicated that there were 12,000 native mandarin speakers and 80,000 native Polish speakers in Dublin. A friend who worked in print advertising (this was shortly before the internet destroyed the industry) laughed at that at showed me circulation data for newspapers used by the industry. At the time there were three mandarin language newspapers in Dublin with a combined readership of over 70,000 and a Polish newspaper with a circulation of 200,000. So either the one of those sets of figures was wrong or there is a hell of a lot of Irish people learning Mandarin and Polish. As a follow up, after the 2008 crash there was a massive outflow of immigrants, either going home or going to wherever there was work (and, for that matter, many Irish emigrating). This was extremely obvious just wandering into any restaurant or cafe where those particular groups socialised. But there was barely a blip in the ‘official’ figures before, during, or 20 years after the crash (more recently there has been a big inward surge). You have to dig into employment or other indirect measures to get a real indication as to migrationflows.
I later did a deep dive into census data for parts of the north inner city I know well. According to the ‘official’ data, the overall population in most wards has remained fairly stable for 50 years. Nobody – absolutely nobody – who has lived in the area believes that. The reality is that most immigrants avoid filling out official forms unless they absolutely have to, even when they are fully legal. Sometimes its suspicion about giving out personal information, sometimes its just down to language/literacy issues. One multi-occupancy building a friend lived in had at least (by her count) 30 people living in 8 flats. According to the census there was 10. And Irish census taking is generally considered very good by international standards. I’ve talked about this several time with some senior civil servants I know who work in education planning. I’ve pointed out how their school projections are completely inadequate if they are based on census figures (and they are), but its like talking to a brick wall. They simply refuse to believe that census figures can be wrong – most likely because to ask the simple question ‘are these figures accurate?’ within government would raise a host of issues that nobody wants to own up to.
Another twist to US Census data is that probably many illegal immigrants are making themselves scarce when the census taker comes a knocking. If they show up on the census data, it could very well happen that that data would be used to locate, arrest and deport them at a later date. Outgoing President Donald Trump had a plan to exclude undocumented immigrants from the census totals used to apportion congressional seats which in itself made the Census a political device rather than a simple headcount so why should those illegals put any trust in taking part?
Your comment as to advertising data brings up one way we could estimate immigrant numbers fairly accurately now: web usage statistics. How many Spanish language, Mandarin language, etc browser-language users this year vs last? Is anyone tracking this?
“Is anyone tracking this?”
I am certain that one or more of the “Three Letter Agencies” is, but they do not tell anyone what they are up to, period. “Need to know.”
For many, just having the data is an end in itself. To them, someone else can figure out what to do with it.
An excellent piece. Kudos for not using the phrase ‘scapegoating immigrants.’
“It may appear to be the interest of the rulers, and the rich of a state, to force population, and thereby lower the price of labour, and consequently the expense of fleets and armies, and the cost of manufactures for foreign sale; but every attempt of the kind should be carefully watched and strenuously resisted by the friends of the poor, particularly when it comes under the deceitful guise of benevolence…”
T.R. Malthus, “An Essay on the Principle of Population”, 1798
It’s important to remember that the human beings who are working illegally in the US, not to mention the many who are working legally while on humanitarian parole, largely are coming from countries in desperate circumstances at least in part because of US policy, Venezuela and the Central American and Caribbean countries being the prime examples.
Another unaired problem with these high illegal immigrant numbers will be the normalization over time of the common Latin American social custom, and I read of examples of similar social customs in Asian populations, of bribery of lower level “public servants.” That is a large population of more vulnerable people who are used to bribing their way through “official” encounters. What this will do to the already shaky ‘integrity’ of local governments is hard to predict, but it will have an effect.
Third World here we come! Wheee!