June 2023

This month, we feature a conversation with Leo Feler and Mine Senses on immigration and the provision of public goods. Mine Senses is an Associate Professor of International Economics at the School of Advanced International Studies (SAIS) at Johns Hopkins University. Her current research focuses on the consequences of globalization on the labor market, with emphasis on quantifying the heterogeneous costs and benefits experienced by different segments of the population.

Leo Feler: Welcome to this edition of the Forecast Direct Podcast. I'm Leo Feler. I'm a Senior Economist with the UCLA Anderson Forecast, and my guest today is Professor Mine Senses. Mine is a professor at the Johns Hopkins School of Advanced International Studies where she teaches on trade and labor market issues. I've been very fortunate to have Mine as a co-author during my academic history. Mine, thanks so much for joining us and welcome to the podcast.

Mine, you have thought through some difficult issues for your paper on immigration and the provision of public goods in the US. In a few words, can you tell us about your paper?

Mine Senses: Thank you for having me. In this paper we're looking at the arrival of immigrants and how it impacts local public finances and the provision of local public goods in the US. The idea of the paper is actually very simple: if immigrants that arrive are different than incumbents in a locality in terms of their skills, their arrival is going to result in changes in average per capita income in that locality. That could be due to just compositional changes, or it could be due to changes in wages and housing prices in the locality. Changes in average incomes are going to have an effect on the local tax base through sales taxes and property taxes. The direction and the magnitude of the change is going to depend on the type of immigrants that come in -- whether high-skilled or low-skilled immigrants. It's also going to depend on the response of the government. The local government can respond by adjusting taxes, adjusting benefits, and choosing the types of services they want to fund. Federal and state governments can also provide transfers to fully or partially offset the impacts on local governments. What we find is that the effects of immigration on local public finances and local public services are on average very small – but decomposing the average, the effects tend to be very uneven across localities in the US.

Leo Feler: Why are effects different and uneven across localities in the US? You mentioned in your paper that you find different effects depending on whether a locality receives low-skilled immigrants or high-skilled immigrants. Why should the effects differ, and can you provide a sense of the magnitude of the differences?

Mine Senses: Let's take high-skilled immigrants as an example. When high-skilled immigrants come into a locality, they tend to increase the average income in that locality. That's going to increase the local tax base. That's because of compositional differences, but it could also be the case that, for example, these high-skilled immigrants have an effect on the housing market, so they demand higher-priced, single-family homes in the suburbs and high performing school districts, and that increases the average price of houses, which increases the property tax base in the locality.

The effect could be the opposite for low-skilled immigrants. They tend to have lower than average incomes, so when they come, they decrease the average income of the locality, which reduces the local tax base on a per capita basis. It could also be the case that low-skilled immigrants affect wages for certain services – so, for example, if immigrants come in and they work as teachers, nurses, bus drivers, and custodians in local schools, that would decrease the price of providing those types of services.

Also, services provided at the local level differ a lot in terms of how much they rely on federal and state funding, so some services might be much more sensitive to local shocks than other services. The state and federal government might provide a buffer. This is all on the supply side.

On the demand side, you can imagine a situation where low and high-skilled immigrants differ a lot in terms of the types of services they demand. If they're poorer on average, for example, the immigrants that come in might demand more public housing. Or they might have more kids, which might increase demand for public education.

Finally, the response of the incumbent population might be quite different depending on whether incoming immigrants are low-skilled or high-skilled. There are many reasons why you might expect the effect to be very different at the local level depending on whether incoming immigrants are low or high-skilled, and US localities differ a lot in terms of the types of immigrants they attract.

Leo Feler: Can you give us a sense of the magnitudes? If you’re a locality that receives low-skilled immigrants, what does this do to your property tax base, your revenues, and your expenditures? If you receive high-skilled immigrants, what does that do to your revenues and expenditures?

And just to confirm, we’re talking here about per capita revenues and expenditures. When we think about per capita revenues and expenditures, that helps us understand how the average might get affected: you're getting more people coming into your locality, and you're dividing the tax base now among more people, and it depends on whether you have high-wage or low-wage individuals contributing to the tax base – that will either pull the average up or down.

Mine Senses: Exactly. Let me give you this number: between 1990 and 2010, the share of low-skilled immigrants has increased on average by about three percentage points. This results in a relative decline in the per capita revenues a locality generates of about 8.4 percent, and it results in a relative decline in per capita expenditures of 5.6 percent. The opposite is true for high-skilled immigrants. The increase in high-skilled immigrants was about 1.8 percentage points. That corresponds to about a 6 percent increase in both locally generated per capita revenues and per capita expenditures. These numbers look big in terms of percentages, but they’re actually very small in terms of dollar values, about $119 dollars annually per person. And on average, for the nation as a whole, these two offset each other so the overall effect of immigrants during this time is about a 0.5 percent decline in general expenditures.

Leo Feler: It's interesting to know that the effects of immigration on local revenues and expenditures even themselves out.

I want to take a moment to ask how you define low and high-skilled immigrants. I'll give you the example from my family, which immigrated to the US from Brazil. My mother was a doctor in Brazil and my dad was a lawyer. But when they came to the US, their degrees weren’t valid. They went from being high skilled in Brazil to being low skilled in the US. I feel like we see this a lot, for example taxi and Uber drivers in the US who were engineers and scientists in their home countries. How do we think about the time it takes for someone to go from being a low-skilled recent immigrant to eventually becoming a high-skilled worker in the US?

Mine Senses: We're bound in this case by data. The way we classify immigrants is we define an immigrant as high-skilled if they have at least a college degree and low-skilled if they have less than a college education. So in your example, your parents would be classified as high-skilled although they would not be doing occupations where those skills are used, but this is actually one advantage of our study as opposed to a structural model, for example, because we base our inference on data. So the fact that your parents would be working in occupations that do not fit their degrees, hence presumably earning lower wages than they otherwise would and hence would be contributing less to the tax base than if they were working as a doctor and lawyer, that would be reflected in our analysis. If some of the immigrant skills are not utilized immediately, we would actually see that as they're earning lower wages and contributing less to the tax base than they otherwise would.

Leo Feler: Now let's think about the notion of time. In addition to looking at first generation immigrants, you also look at the second generation. There's work by Leah Boustan and Ran Abramitzky, a book called, “Streets of Gold” that we featured on Forecast Direct. They see that second-generation immigrants tend to do pretty well, much better than their parents and much better than the native population. So the average second-generation immigrant tends to be doing better than the average American. Can you talk about this notion that there are some immediate effects on the local economy of immigrants coming in, but there are also longer run effects, which perhaps are different?

Mine Senses: I think this is very important. The main results I talked about all focus on the first generation. But we also look at whether your parents are foreign-born, and we say you're a second-generation immigrant if your parents, or at least one of your parents, is foreign-born. Then we look at the effects of second-generation immigrants over 10-year periods between 1990 and 2010. It is very important to look at the second generation of immigrants because you don't want to have a short-sighted view about immigration policy.

Here again we have some data constraints that we discuss in the paper. But what we find, after we’re able to identify who might be second-generation immigrants, is very consistent with what you suggested and with earlier work. We find that while second generation immigrants have a positive and significant impact on per capita revenues, their effect is insignificant and small in magnitude on per capita expenditures. So in a way, they're sort of compensating for any negative effects of the first-generation immigrants on per capita expenditures.

Leo Feler: I love that. That's fascinating. It suggests that oftentimes immigration is an investment, and when you have low-skilled or high-skilled immigrants coming into the US, we as a country have this fantastic education system, we have entrepreneurialism, we have well-functioning credit markets, we have a well-functioning legal system, and that means we give opportunities to immigrants that might not exist in their countries of origin that then allow them, or their children, to do well here in the US. It just takes some time. It might take a generation for that to happen. What it seems like you're showing here is that there's maybe an immediate effect -- that we as a society invest in lower-income, lower-skilled immigrants that are coming in, perhaps we take a little bit of a financial hit to do so, but pretty soon it pays off because the future generations tend to do pretty well.

Last question here to wrap things up. I think there’s a risk that your paper can be misinterpreted. What interpretation do you want readers and listeners to take away from your paper?

Mine Senses: It's very important to frame our findings correctly. There are many well-documented benefits of immigration. But we also know there has been increasing public apprehension against immigration and that was very much evident in President Trump's election campaign and in the Brexit vote and the recent votes in Finland and Italy. When you actually ask voters why they don't want immigration, they talk about things like how they worry about wage and unemployment effects on the native population, and they worry about the welfare state and the quantity and quality of the public goods. They talk about overcrowding of schools and higher taxes. For the labor market effects, there is a lot of work in the literature and we have a good understanding. But for the effects on the welfare state and local public goods, the literature isn’t as developed and we don’t have as good an understanding as a profession. As economists, it’s our job to take these concerns seriously and to quantify or see if there's anything there, to see how big these effects are and what are some proposed policies that might address some of these concerns. Some of this nuance gets lost in the public discourse. It's important to figure out whether the effects on the welfare state differ depending on whether immigrants are low-skilled versus high-skilled. Or over what time frame do effects differ? Or are there differences between the first and second generations? All of this varies a lot, as we discovered. I think this is super important because given the pushback against immigration, I worry that if we don't take some of these voter concerns seriously, a more drastic and harmful policy like building walls and banning all immigration is going to happen. So I think it's our job to take these concerns seriously and understand and address them to the extent we can so that we can reap the benefits from immigration, which are very well documented.