The truth about immigration and wages. By Matthew Yglesias

The truth about immigration and wages. By Matthew Yglesias

A pernicious distraction from the need for fiscal & monetary policy to generate full employment

Matthew Yglesias

Slow Boring.com

January 4, 2021.

Happy New Year!

Two issues that I am passionate about are the merits of immigration to building American prosperity and national power, and the importance of full employment to building worker power and driving economic progress. Oren Cass, a former Mitt Romney aid who is perhaps the leading light of ongoing efforts to construct some kind of actual coherent policy framework around the vague impulses of Donald Trump, says that these two goals are in massive tension and you can’t have tight labor markets and loose borders at the same time.

When I saw his headline my first instinct was to reach for the truckload of academic research disputing this and showing that immigration has either zero or benign impacts on the wages of native workers.

If you read the article, though, you’ll see that Cass is aware this literature exists and just doesn’t care:

The empirical case for the “nothing to see here” approach to immigration economics is paper thin; one can typically guess where the hyperlinks will point before moving the cursor to them. Academics debate endlessly whether a sudden and massive influx of Cuban refugees in Miami in 1980 did or did not depress local wages, as if this deeply unnatural “natural experiment” answers for all times and places the question of how a national economy’s labor market might be affected by the rate at which a given class of workers flows into it.

Instead, he prefers to cite three journalistic accounts:

A Washington Post article arguing that Mexican-born workers returning to Mexico from the United States would likely lower wages in the Mexican automobile industry.

A Wall Street Journal article detailing that in Arizona “wages on farms and in construction jumped 10 percent to 15 percent in just four years” after a crackdown on undocumented workers.

An Economist article which observed that “average wages in occupations that rely heavily on immigrant labor rose especially fast during the Trump years.”

Preferring journalistic arguments to academic ones is a bit odd, and if you’re curious about the relevant academic literature, Noah Smith has produced a wonderful compendium of studies, the vast majority of which are not focused on the Mariel Boatlift (the “deeply unnatural ‘natural experiment’” that Cass mentions). I’d also say that some of these stories don’t seem to be specified correctly. In most cases, an undocumented worker is going to earn less than one with papers, so if you remove the undocumented workers from the labor force, you mechanically push up the average wages without anyone actually benefitting — it’s a pure compositional effect.

But setting that aside, everyone can be right here. The idea that immigration restrictions can increase relative wages in immigrant-heavy occupations is not the same as saying that they can increase overall wages and incomes — relative wages will go down elsewhere as the economy adjusts — and understanding the difference is critical.

Smith says he knows he’s not going to convince anyone about this because people who oppose immigration are going to oppose it anyway. I am not so pessimistic. I don’t think I will convince immigration skeptics to love immigrants, but I really do want to convince them to abandon this particular reason, because I think it’s important to understand that if you want to advance the specific goal of tight labor markets, then you need to look to macroeconomic policy.

If you want to reduce immigration for some other reason, then that’s fine. But don’t expect it to fix the national labor market.

The Trump wage boom is easy to explain

Both Cass and The Economist want to get rhetorical plausibility for their argument by noting that we elected an anti-immigration president and then wage growth went up, so wouldn’t it be weird to deny the existence of a relationship?

That seems to me to posit a mystery where none exists.

Back in 2014, 2015, and 2016, my view was that there was plenty of slack left in the American labor market and it was a mistake to be talking about tighter fiscal and monetary policy. Donald Trump got elected, and then starting right away he implemented the following policy framework:

Taxes went down thanks to TCJA.

He boosted military spending, and in order to get that done in Congress, he also agreed to Democratic calls for higher domestic spending.

Spending on Social Security and Medicare kept rising due to population aging.

He spent a lot of time leaning on the Fed for low interest rates.

There are only three things that could possibly have resulted from that combination of policies.

I was right all along and more stimulus would lead to higher incomes.

The people who thought we were at full employment in 2016 were right and more stimulus would lead to inflation.

Trump’s policies made the economy more efficient in a way that allowed for non-inflationary growth.

Two is simply refuted by the facts. Three seems absurd to me, but that hypothesis is contrary to the immigration answer. The obviously correct answer is one: wage growth accelerated under Trump because he took office with the labor market still depressed, and he delivered massive and welcome fiscal stimulus while minimizing the monetary policy offset.

That’s really my main interest in this. If future Republican presidents want to copy Trump’s immigration policy for whatever reason, then it is crucially important that conservative policy hands not delude them into believing that doing so will reproduce Trump’s macroeconomic results. If you want to replicate Trump’s macroeconomic results, you need to replicate his macroeconomic policies — stimulus, stimulus, stimulus.

So what about Cass’s examples?

Relative vs absolute wages

Christian Dustmann has a good, fairly technical paper explaining that the main reason empirical studies of immigration seem to reach different results is that they are actually measuring different things.

One of my favorite recent studies came when Giovanni Peri, Derek Rury, and Justin Wiltshire looked at a huge influx of Puerto Ricans to the Orlando area after Hurricane Maria. Puerto Ricans are not immigrants per se but the economic impacts are similar. They found that the overall impact on the wages of low-skill people in the Orlando area was positive, but the increased demand raised wages in the retail and hospitality sectors by more than the increased supply of labor lowered wages in the construction sector.

Note that if you sent a reporter to construction sites to ask what’s happening with migration and wages, you would have gotten Cass’s story — a large share of the migrants went to work in construction and it depressed wages.

So if our policy goal was to maximize the wages of Orlando-area construction workers we definitely should not have let the Puerto Ricans come. But if we wanted to maximize the overall prosperity of Orlando, the new arrivals were fine. It’s a relative shift between sectors.

Now shifting wages between sectors is a thing that one might want to do. Cass flags a story indicating that Arizona agricultural wages went up when they cracked down on unauthorized workers. The United States also has a large legal guest worker program for the agricultural sector. If we want to raise the relative wages of fruit and vegetable pickers, we could absolutely try to end that program and invest in enforcement to ensure that the guest workers aren’t replaced by unauthorized workers. But what would happen if we did that? Some combination of the following:

Michael Clemens found that when we tried this in the 1960s, we got less produce variety and lower produce quality as farmers shifted to crops that can be harvested by machines.

We could import more food from abroad, worsening America’s terms of trade and slightly depressing living standards.

We could have inflation, worsening America’s real wages and slightly depressing living standards.

The Fed could offset the inflationary implications with slightly tighter monetary policy, slightly slacker labor markets elsewhere in the economy — again, slightly depressing living standards.

The exact results are a little bit hard to predict because there are so many possible margins of adjustment. But none of them lead to an outcome where the United States employs a large number of people as well-paid fruit and vegetable pickers. If we really want that outcome, however, we could do the following:

Ban the use of immigrant labor in fruit and vegetable picking.

Place high tariffs on important fruit and vegetables to prevent the substitution of foreign produce.

Place regulatory constraints on mechanization.

Add in subsidies so that producers can afford to keep prices low despite high costs.

At that point, the question becomes: what’s the point? Why are we deliberately making it expensive to harvest produce, denying people cheaper foreign-grown produce, and then subsidizing consumption because we don’t want everyone to go hungry? Wouldn’t it be better to let immigrants take jobs they deem desirable, let people buy foreign fruit if they want, and invest our money in a universal welfare state to take care of people?

The goal for native-born workers should be to help everyone who wants one get a well-paying job in a high-productivity sector, not to create a crazy mishmash of subsidies and regulations with highly uncertain secondary impacts.

Labor market complementarities

I think it’s pretty uncontroversial to say that if you kicked out all the immigrant nannies and childcare workers, the average wages for nannies and childcare workers would rise.

If you’re in love with anecdotes about sector-specific wage shifts and wholly neglect the interest of immigrants themselves, that can sound good. But of course one person’s wage is another person’s price. And is there really anyone in America who thinks that child care is too affordable right now?

Patricia Cortés and José Tessada find that when more immigrant labor is available, child care is cheaper and college-educated women are able to do more market work.

Delia Furtado confirms the finding about market work but also finds that women have more children in the presence of more affordable child care.

Peri, Agnese Romiti, and Maria Christina Rossi have a parallel finding about immigrants relieving women of some of the burden of eldercare.

The point in all these cases is the same: cutting off the flow of immigrant workers to a given sector can raise pay in that sector, but that ends up raising costs in ways that ripple throughout the economy.

In general, the economic case against immigration involves forgetting that immigrants buy things that native-born people make while also forgetting that native-born people buy things that immigrants make. Preventing these transactions is a lose-lose proposition for both immigrants and natives.

Two important caveats

One thing to note here is that while the case that immigration imposes economic costs on natives is weak, the case that immigration imposes economic costs on other immigrants is strong. Curtailing immigration raises wages in immigrant-heavy sectors, which is bad for (most) natives because natives (mostly) do not work in immigrant-heavy sectors — it simply raises costs for natives.

But immigrants do work in immigrant-heavy sectors so you can benefit them by curtailing flows.

I think this is relevant to refugee policy in particular. More so than other immigrant groups, refugees tend to struggle economically in the United States since they are essentially selected for being hard-luck cases rather than for being go-getters who decide to travel vast distances in search of opportunity. Under the circumstances, I do think it makes sense to worry that having too many refugees — especially refugees from the same country — all living in the same place is going to create a problem for the refugees themselves. You want there to be enough people from Country X living in City Y to have a meaningful community, but it’s not in the interest of the Xs themselves for that community to grow without limit.

The other is that the situation which developed in the late-1990s and early-aughts in which you had a huge number of unauthorized workers permanently relocate to the United States did have a number of genuinely undesirable economic effects. Labor law is supposed to be largely self-enforcing; the idea is that workers themselves have a strong incentive to report violations of minimum wage or overtime rules. But if you have a bunch of workers who lack legal rights, that’s a problem.

To the extent that by “loose borders,” people mean a literal lack of border controls, I agree that loose borders undermine worker power. It is good that the Obama and Trump administrations managed to cut down on this considerably.

But a really good way to further reduce the number of illicit workers in the United States is to ensure that there are pathways for foreign-born people to come work here legally. That includes legal status and a path to citizenship for the long-settled undocumented population.

Whatever we do about immigrants, we need full employment

I don’t think that this will change anyone’s mind about immigration.

What’s more, since there is a segment of the electorate that has an objection to immigration that one could kindly characterize as aesthetic (or one could find a less kind word), there will always be politicians who want to cater to those voters and policy advisors who give them policy reasons to support the conclusion they favor.

All that I really ask is you pick a different reason.

Worry about the fiscal impact on states and cities. Worry about the long-term impact on crime. Worry about something vague like “social cohesion” or “trust.”

Just don’t convince yourself that immigration restrictions are an effective means of creating tight labor markets. It’s not true empirically, and it doesn’t make sense theoretically. Fiscal and monetary policy decide whether we have a robust drive for full employment. Trying to use immigration policy to trick central bankers into allowing a tighter labor market than they’d naturally be inclined to favor is ridiculous. You need to put central bankers in place who embrace tight labor markets. And you need to support them with fiscal policy.

The Trump administration’s greatest policy triumph is that they more or less accomplished this by the winter of 2019-2020, and we need to move aggressively in the spring and summer of 2021 to get that back. People are going to keep disagreeing about immigration policy for both good and bad reasons. But whatever you make of it, don’t delude yourself into thinking that it’s an alternative to good macroeconomic policy.

93

200

← Previous

Write a comment…

JW in GB10 hr ago

Pro-immigration (and pro-trade) lefties are overly fond of using counterfactuals to dismiss real economic pain for some lower-income people. Yes, if we had the Danish welfare state and rational macroeconomic policy aimed at full employment, much of the economic downside of immigration (or trade) for some low-wage native workers could be reduced. But we don't have those things, and we won't anytime soon. Fifty years ago, unionized white dudes slaughtered cows for a living in my Midwestern town for middle-class wages. Now the (largely) non-unionized workforces at those slaughterhouses make poverty wages and are largely Hispanic. If my dad were one of those dudes, I might have non-racist reasons to be upset with the change in the local political economy, although this change might also have made me more racist and more receptive to racist politicians. The argument that beef is now marginally cheaper probably wouldn't matter a lot to me, even if the economy is better off as a whole. Concentrated and visible pain and diffuse and invisible gain make for tough politics.

 5Reply

8 replies by Matthew Yglesias and others

Josh H8 hr ago

This is a good piece, but I think it leaves out one important piece: the medium (say, 2-3 generations down the road) and longer run. Focusing on the short-term effects of what happens when Puerto Rican’s come to Orlando is somewhat myopic imo. But... politically it’s not! The reality is after 50+ years or so these people and their children’s children assimilate into different types of careers and lifestyles. That’s relevant though a harder sell now.

 1Reply

1 reply by Matthew Yglesias

198 more comments…

© 2021 Matthew Yglesias. See privacy, terms and information collection notice

Publish on Substack