Was I wrong about low-skilled immigration?

A response to critics.

Written by Noah Carl.

My recent article ‘Is "immigration" good for the economy?’ generated a fair amount of critical commentary. This is good – we like debate here at Aporia. In the present article, I’d like to address some of that critical commentary.

A number of my detractors, who presumably didn’t bother to read the article, assumed I must have argued that immigration is bad for the economy. This is not what I argued at all. Rather, I said that the effect varies from “substantially positive” to “substantially negative” (rather than always being “substantially positive”, as immigration proponents would have us believe) and hence that it’s meaningless to say immigration per se is good for the economy.

My more conscientious detractors – the ones who bothered to read the article – took issue with a claim I did make:

Immigrants aren’t a magical category of human that, simply in virtue of moving to another country, are destined to make that country richer. They may be more productive than the people who are already there, or they may be less productive. If they’re less productive, they’ll be bad for the economy.

The Cato Institute’s Alex Nowrasteh said this was “very confused”. And Matt Darling of the Niskanen Center called it “perhaps, the worst economic reasoning” he had ever seen (in a tweet that got nearly 5,000 likes).¹

The first thing to say is that I probably shouldn’t have used the word “productive” since that term has a technical meaning in economics, which is not exactly what I had in mind. A better way of phrasing it would have been, “If they’re less skilled, they’ll be bad for the economy.” Or simply, “If they’re less intelligent, they’ll be bad for the economy”.

Another thing to say is that I was referring to the long run impact of less skilled immigrants who permanently settle in another country. I had assumed this was obvious from the context.

Would Darling and Nowrasteh continue to dispute my claim even with these clarifications? I can’t be entirely sure, but from subsequent discussion on the site formerly known as Twitter it seems they would.

For example, when I clarified to Darling that I was referring to the long-run effect of a permanent change in the share of the population selected from the left-hand side of the skill distribution, he accused me of “repeating the bad economic reasoning”. (On the other hand, he later conceded, “It's certainly possible that your conclusion is true”.)

As for Nowrasteh, when I asked him whether he believes that low-skilled immigration into Europe from Muslim countries has been or will be good for the economy, he replied, “Yes”.

I want to reiterate that Darling, Nowrasteh and I all agree that high-skilled immigration is good for the economy. Where we disagree, it seems, is that they believe low-skilled immigration is also good for the economy, whereas I believe it’s bad for the economy (in the long run, when it involves permanent settlement etc.)

Okay, so why do I think I’m right?

The first reason is that Darling and Nowrasteh’s position seems to imply there should be large economies of scale. Saying that both high-skilled and low-skilled immigration are good for the economy is equivalent to saying that additional people are good for the economy. Do we see that countries with larger populations are richer? Not at all – which was one of the points I made in my original article. The raw correlation between GDP per capita and population is basically zero (slightly negative, in fact).

Now, you can always claim that there is a positive effect of population on GDP per capita but it’s masked by other factors. To which I would respond that if there is, it’s clearly very small. Total population size is simply not an important long-run determinant of living standards. As a matter of fact, most of the richest countries are small (the Nordics, Switzerland, Singapore, Hong Kong etc.) And with the exception of the US, none of the largest countries is particularly rich.

Which raises the question: if additional people are neither good nor bad for the economy overall, how could additional low-skilled people possibly be good? Anthropologist Greg Cochran made this point in a characteristically funny blog post from 2017, which I will quote in its entirety:

Imagine a country with an average IQ of 100, some average amount of education (with some distribution), some average amount of capital per head (with some distribution of ownership of capital). Now add immigrants – 10% of the population – that are the same in every way. Same average IQ, same distribution of IQ, same average amount of capital and same distribution. They speak the same language. They have similar political traditions. In other words, it is as if the US had just peacefully annexed an imaginary country that’s a lot like Canada.

Would the original inhabitants gain economically from this merger? Strikes me that this could only happen from economies of scale – since nothing has changed other than a 10% increase in overall size. There might be some diseconomies of scale as well. I wouldn’t expect a big payoff. Except for Nawapa, of course.

Contrast this with a situation in which the extra 10% is fairly different – lower average IQ, much less education on average, don’t speak English. They don’t bring along a lot of capital. They have and bring along their native political traditions, like everyone, but theirs stink. I can easily see how those immigrants might have improved their economic lot but it’s kind of hard to see how bringing in people with low human capital benefits the original citizens more than bringing in people with considerably higher human capital. Yet it must, because adding more of the same clearly has a small effect, while adding in lower-skilled must have a big positive effect. Practically all the economists say so.

In case it wasn’t obvious, Cochran is being sarcastic at the end.

One possible reply would be that immigrants usually arrive when they’re of working age and after they’ve been educated at public expense, so the effect of an additional immigrant isn’t identical to the effect of an additional native-born person.

There are two counters here. The first is that this will only make a difference if the immigrants arrive relatively early in their working lives. If they arrive in the middle of their working lives – after their period of youth dependency but before their period of old-age dependency – it won’t make much difference. The second is that the small fiscal benefit conferred when immigrants arrive early in their working lives isn’t going to outweigh the costs associated with a permanent leftward shift in the skill distribution.

Another reason I think I’m right about low-skilled immigration is that Darling and Nowrasteh’s arguments aren’t very convincing. Both pointed out that the net fiscal effect of low-skilled immigration is only part of the overall economic effect – which is entirely true. What’s implausible is that the other economic effects are large enough to outweigh the negative fiscal effect. And some of those other effects aren’t even positive.

The non-fiscal effects of low-skilled immigration in the economy are largely distributional. By reducing the relative scarcity of low-skilled labour, it makes those who sell such labour (i.e., low-skilled workers) slightly worse off, while making those who buy such labour (i.e., employers and managers) slightly better off. Many economists insist the latter effect outweighs the former, in which case low-skilled immigration does increase the size of the pie, all else being equal. But this doesn’t seem like a particularly strong point.

Again, if we look around the world, do we see that countries with a higher ratio of low-skilled to high-skilled workers are richer? No, we generally see the opposite. So either those economists are wrong, or the pie-increasing effect that comes from reducing the relative scarcity of low-skilled labour is dwarfed by other pie-decreasing effects of low-skilled immigration.²

Some economists claim that low-skilled immigrants are complements not substitutes for low-skilled natives, so their entry into the workforce doesn’t hurt low-skilled natives and might actually help them. I can see how this might be true in very specific circumstances. Suppose there’s a flood: factories can’t operate because the drains are blocked and there aren’t any drain-unblockers in the country. If we import drain-unblockers from abroad, we can get the factories going again, allowing native factory workers to earn a living.

But I can’t see how low-skilled immigrants could possibly be complements for low-skilled natives in the long run. Sure, when they arrive they might possess skills that happen to complement those of native low-skilled workers. But we’re talking about people whose skills, almost by definition, aren’t very difficult to acquire.

Once the drains are unblocked, the drain-unblockers will have to do something else and at that point they’ll be competing with low-skilled natives. Instead of importing drain-unblockers from abroad, low-skilled natives could be taught how to unblock drains. Indeed, if there were no drain-unblockers in the country, the wages for drain-unblockers would rise, incentivising people to enter that profession.

As an example of how “low productivity workers” are good for the economy, Darling wrote that they “free up time away so that "high productivity workers" do not have to spend time on "low productivity" tasks”. Which, of course, illustrates the principle of comparative advantage. Even if Jeff Bezos is better at running Amazon and cleaning his house than anyone else, since he has a comparative advantage in running Amazon, it makes sense for him to spend time doing that, while hiring someone else to clean his house.

Now, specialisation and trade certainly is one of the ways we get richer. But adding low-skilled immigrants to the economy doesn’t necessarily boost specialisation and trade. It’s not like Jeff Bezos has only two options: bring in a low-skilled immigrant or do the cleaning himself. He can hire a low-skilled native. And if he’s unable to find one willing to work at the prevailing wage, he can offer a higher wage.

I would add that Darling’s argument seems vulnerable to a reductio ad absurdum. If low-skilled immigrants are good for the economy because they free up time for high-skilled workers, would the same be true of even lower-skilled immigrants (who could free up time for low-skilled workers) and so on until we are admitting the very lowest-skilled immigrants? Should we admit immigrants with absolutely no skill so they can free up time for workers with the bare minimum of skill?

What’s more, specialisation and trade isn’t the only way we get richer. Another way is technological innovation. And if low-skilled labour is very cheap, there’s less incentive for labour-saving innovations. If Jeff Bezos can hire a low-skilled immigrant to do his cleaning at a very low wage, there’s less incentive for someone to invent, say, the robot vacuum cleaner.

The next effect of low-skilled immigration that has economic consequences is crime. In Denmark, age-and-sex-adjusted violent crime rates for Muslim immigrant groups vary from 2 times the native rate to >6 times the native rate. (The rates for Western immigrant groups are generally below the native rate.) Remember that Nowrasteh believes low-skilled Muslim immigration into Europe has been good for the economy.

Some economic consequences of crime, such as the cost of incarceration, are taken into account when estimating the net fiscal effect of immigration. So these do not count among the non-fiscal effects we’ve been discussing. However, other economic consequence of crime are not. For example, natives may move away from areas where it would be economically efficient for them to live; they may spend extra money on home security etc.

Yet another effect of low-skilled immigration that has economic consequences is institutional change. Here, even Bryan Caplan (the open borders advocate) admits there might be a problem. “I’ve worked with the data,” he writes, “and I’m honestly troubled by low-skilled foreigners’ views on issues like free speech”.

It’s not just free speech. Low-skilled Muslims in Europe vote overwhelmingly for left-wing parties. At the last presidential election, 69% of French Muslims backed the explicitly socialist party of Jean-Luc Mélenchon. Would it be good for France’s economy if the country moved in a more socialist direction?

Of course, the net effect of low-skilled immigration isn’t necessarily an increase in support for the left because natives may simultaneously increase their support for right-wing parties. But in the case of Europe, are natives increasing their support for the sort of right-wing parties that Nowrasteh believes would be good for the economy? Not really. They’re voting for “far right” anti-immigration parties, some of which have fairly left-wing economic agendas. And in the US, they’re voting for Donald Trump.

Economist Garett Jones has penned an excellent book, The Culture Transplant: How Migrants Make the Economies They Move to a Lot Like the Ones They Left. He presents evidence that immigrants tend to bring their culture with them, and that they do not rapidly assimilate to the host culture. In some cases, like European colonisation of the Americas, they massively transform the places they settle. Jones’s findings raise the question of how making France more like North Africa will be good for the French economy.

To summarise: some non-fiscal effects of low-skilled immigration may be positive and others may be negative, and it’s far from clear that the former outweigh the latter. In fact, they almost certainly don’t. The negative impact of low-skilled immigration on institutions is likely enough, by itself, to outweigh any positive effects. And before Caplan suggests that we deny immigrants the vote and “make their non-voting status hereditary”, that is never going to happen.³

Even if you ignore the negative impact on institutions and insist that the non-fiscal effects are positive, those non-fiscal effects will not compensate for the negative fiscal effect – which can be absolutely massive. A 2017 analysis found that Somalis who’d been in Denmark for up to 25 years had an employment rate of just 40%, meaning the other 60% were either unemployed or economically inactive. This helps to explain why in 2018, they made an average fiscal contribution of –130,000 krone. They’re also convicted of violent crimes at 6x the rate of native Danes.

What’s more, even if low-skilled immigrants gradually converge to the level of low-skilled natives (which they don’t, in the case of Muslims in Europe) that level is still going to be negative. After all, low-skilled natives use more in government services than they pay in taxes. Which isn’t surprising: the whole point of the welfare state is to transfer money from people with high incomes to people with low incomes. So adding low-skilled people to the population means increasing the share of people who’re likely to be net recipients of public money.

In our discussion, Nowrasteh pointed out that the net fiscal effect of low-skilled immigration when using static accounting may be different from the net fiscal effect when using dynamic accounting – which again is entirely true. For example, capital owners may pay more taxes than they otherwise would have thanks to the extra low-skilled labour. However, the difference between static and dynamic accounting won’t drastically change the net fiscal effect of, say, low-skilled Muslim immigration into Europe. Here’s what Jan van de Beek had to say:

Your point may be valid (in the case of the Netherlands) for those low skilled who are actually working, like most labour migrants from Central and Eastern Europe. In those cases, there are positive economic effects beyond the fiscal effects, which may compensate (partly) the fiscal effects. But also negative economic effects: cheap import labour has slowed innovation in agriculture for example. And population growth has negative externalities in a densely populated country like the Netherlands. Absolutely not sure whether the net sum of all those effects will be positive.

Your argument is absolutely not true for other low skilled groups, like asylum seekers, former guest workers, and family migrants from Africa and the Middle East, who score extremely bad on all indicators of integration. Net fiscal contribution over the life course is the ultimate integration indicator, correlating strongly with all other indicators of integration. There is no way that for those groups any economic effect could offset the huge fiscal cost of such bad integration of first, second, and even third generation.

All else being equal, low-skilled immigration increases total population, while shifting the skill distribution to the left.⁴ In the long run, a larger population is neither good nor bad for the economy, but a left-shifted skill distribution is bad. Therefore, low-skilled immigration is bad for the economy. The extent to which it is bad varies with factors such as the exact skill level of the immigrants, as well as their cultural distance from the host population.

Noah Carl is an Editor at Aporia Magazine.

---

Support Aporia with a $6 monthly subscription and follow us on Twitter.

1. The quoted text does not actually contain any reasoning; just a conditional proposition. I did not explain why less productive immigrants will be bad for the economy; I merely said they will be. The statement, “if the dog is wet, then he will be happy” doesn’t contain any reasoning. But the statement, “if the dog is wet, then he will be happy because he will have been swimming” does. In particular “he will have been swimming” is a reason why the dog will be happy.

2. It also seems to imply that high-skilled emigration is good for the economy since that too reduces the relative scarcity of low-skilled labour.

3. And what if they marry a native: do their children get to vote or not?

4. In other words, it shifts the IQ distribution to the left.