Breaking the Laboratories of Democracy

How Immigration Breaks One of Federalism's Greatest Virtues

The phrase "Laboratories of Democracy" refers to how individual states can serve as testing grounds for policies and social experiments, which can then be imitated by other states or taken to the national level. But how can we tell which experiments work? There are a plethora of statistics we could use—life expectancy, HDI, and GDP per capita are some of the more common ones—but how to weigh each of these is not obvious; you can never be sure you’re not missing something important. Furthermore, these things are very heavily influenced by demographics¹ and geography, which are not, in the short term, the result of governance. The acid test is which way people move. Moving shows a revealed preference for the destination over the source from the people who know the circumstances of their own lives best: themselves². The Berlin Wall was far more damning than East Germany’s lackluster GDP per capita³.

Americans Move To Red States

In practice, interstate migration is overwhelmingly from blue states to red states. When making decisions about their own life, as opposed to for the country as a whole, Americans prefer the results of Republican policies. Two ways to see this:

State population changes attributable to interstate migration, as a percentage of existing population, 2020. Per-capita migration statistics are useful because things like jobs and housing scale with the number of people. North Dakota struggles because of the end of the fracking boom there. Source.

Absolute population movement between states, 2022. Absolute numbers are useful because bigger states with more complex economies are less likely to be affected by short-term trends in some particular area. Big losers include the three big blue states, California, Illinois, and New York, as well as Minnesota and infamously corrupt and poorly-governed Louisiana. Big winner is Florida. Source.

To put longer-term numbers on it, between 2010 and 2023, states that Trump won in both 2020 and 2016 gained 5.65 million people through internal migration (39 people per 1000 population in 2023). States that Trump won in 2016 and lost in 2020 gained 410,000 people through domestic migration (8.6/1000 population in 2023). States that Trump lost in both 2016 and 2020 lost 6.06 million people (42.5/1000 population) to red and swing states⁴.

And this isn’t just low-earners forced out of blue states by high housing prices. The difference in incomes between internal migrants to and from a state generally tracks overall population flows—more desirable states attract more productive people.

Difference in earnings between those leaving and those entering the state. This map is very similar to the previous one, but a few things stand out. Minnesota and Illinois are particularly poor performers, while Florida and the “white redoubt” Mountain West are the big winners. Texas, while still a winner, is much less impressive, and Georgia swaps from winner to loser, possibly because Atlanta has become something of a black Mecca. Source.

Mechanisms

So Americans prefer the results of Republican state governance for themselves. But why?

Housing

Housing is the median American household’s single largest expense—33% of total monthly expenditures. But this cost is not evenly distributed around the country. It is particularly expensive in Hawaii, California, and the northeast corridor, with Florida significantly cheaper but more expensive than the rest of the country. The interior of the US, including Texas, is much cheaper. This is the most important cause of emigration from California.

Housing costs are the dominant reason cited for leaving California. Source.

That being said, it would be a mistake to reason from housing prices directly. Housing prices are set by both supply and demand. People moving en masse from red states to blue states will raise prices in the former and lower them in the latter.  

Home affordability, taking into account taxes and insurance, by US county Compare this map with the domestic migration maps above. Note that the biggest winners of domestic migration, Florida and the Mountain West, are reasonably expensive, which makes sense, because demand is high, while huge emigration source Illinois is cheap. Source.

But for a given level of demand, housing prices are lower in red states⁵ because supply is much higher, as a result of fewer regulatory veto points. To put numbers on it: between 2009⁶ and 2023, states Trump won in both 2016 and 2020 constructed about 50% more housing per capita than those he did not⁷. This pro-growth housing environment is the single biggest reason for red-to-blue migration.

Source of data, chart is my own. YIMBYs take note. This is both good governance and good politics for Republicans; homeowners are more likely to vote for them, and low housing prices helps turn renters into owners.

As a gut check, we can see that there is indeed a strong positive relationship between cumulative housing built between 2009 and 2023 and 2022 migration ratios. This should not be taken as evidence of causality; building more means cheaper housing which makes it more attractive to move, but builders will also choose to build more in places experiencing greater in-migration. Sources for data [1][2], chart is my own.

Taxes

Housing prices are an obvious mechanism of action for emigration from California and New York, but not Illinois or Minnesota, which have much cheaper housing. However, both of these states have high taxes, and when Illinoisans are asked, taxes are the top reason given for emigration. And it’s not just them.

There’s a significant negative correlation between top tax rates and net migration across states. Source.

According to the Tax Foundation:

A strong positive relationship exists between states with well-structured⁸ tax codes and those that experience net inbound migration. Among the 25 best-ranking states on the 2022 State Business Tax Climate Index, which had a snapshot date of July 1, 2021, 17 states experienced net inbound migration between 2021 and 2022. Meanwhile, among the 25 worst-ranking states on the Index (and the District of Columbia), 18 experienced a net loss of taxpayers to interstate migration (Source).

Regulations

It’s not just housing that’s easier to build in red states. Almost everything is, as a consequence of a more lenient regulatory and permitting environment.

For instance, it’s easier to build renewable energy in Texas than in California. In general, it is easier and faster to get interconnection permits in red states. Source.

Unlike housing and taxes, where we have direct survey data on reasons for moving, it’s hard to definitively prove that the broader regulatory environment matters for interstate migration. But jobs and general economic prospects do, and both are partly downstream of the local legal environments. The harder and more expensive it is to invest in factories, power plants, houses, or roads, the harder it is to do anything.

The Importance of Legacy Control

All of these mechanisms operate on the scale of decades. It takes time to build up housing stock or power generation or factories, or for the incentives of different tax regimes to have broad economic effects. As such, recent past control matters as much or more than present control. Much of the Sunbelt is now trending purple and Virginia is now a lean-blue state, but historically these were solidly red for decades, while the reverse is true for the Rust Belt. 

The Common Theme: Markets Work

The Republican Party is the more ideologically pro-market, pro-private enterprise party of the two and has been since at least the New Deal. That manifests itself in fewer regulatory veto points, especially environmental ones, for constructing both housing and everything else, and in lower taxes. Other issues (such as immigration) or social issues (such as abortion) are much less important for domestic migration.

Partisan Implications

So Americans prefer the results of Republican state governance for themselves. But does this help the Republican party? On a federal level, yes. Longitudinal studies show that people shift their views slightly towards those of their neighbors⁹, and so migration can be expected to (slightly) shift the politics of the movers red. Furthermore, state-level politics are “sticky.” Incumbent parties institute election procedures that help themselves and tend to have strong local turnout machines, meaning that the additional electoral college votes and representatives from larger populations are more likely to benefit the party controlling the receiving area than a naive analysis of voter composition would imply. Consequently, interstate migration can be expected to move the country as a whole towards the Republican party. 

But this relies on migrants being politically representative of the country; since they’re disproportionately from blue states, a common concern among conservative red state residents is that blue state migrants fleeing the consequences of the policies they support will transform their states into copies of the ones they left. It’s easy to imagine a domestic version of failure migration turning the whole country blue¹⁰, but fortunately, the reverse seems to hold in practice. If anything, interstate movers tend to be slightly more Republican than their destinations. For instance, Texans from other states were 1% more likely to have voted Trump in 2020 than native Texans (immigrants were 9% more supportive of Biden), with Californian transplants (the most numerous interstate migrants) being 5% more likely to back Trump than native-born Texans. Domestic migration is pushing back the day that immigration turns Texas purple. And while I haven’t found similar polling data for Florida, that the state is reddening amidst massive internal migration and international immigration (immigrants generally backing the Democrats) suggests the same is happening there.

Due to the nature of the Senate and Electoral College, the movement of Republicans from sapphire blue California, Illinois, and New York to purple-ish Florida and Texas helps the Republican Party. This provides a theoretical mechanism (besides persuasion) by which the laboratories of democracy can actually spread their successful results¹¹. But there’s a catch.

Immigration: Breaking the Laboratories of Democracy

Since domestic migration flows shift the federal balance of power in the United States towards the Republican Party and the two parties are roughly evenly matched¹², we would expect the country to shift rightwards over time¹³, at least until the Democratic Party adopted more pro-market positions. But this isn’t happening. Instead, the traditionally-Republican Sunbelt, the main beneficiary of all of this internal migration, is turning purple. Since this isn’t because of blue state transplants, what’s causing it? In a word: immigration. Proximity to Mexico¹⁴ and the same economic incentives driving domestic migration have attracted huge numbers of immigrants, who disproportionately support the Democrats.

Y-axis is share of Republicans. There is a strong correlation between immigrant status and support for the Democratic Party. Source.

In numbers: states that Trump won in both 2016 and 2020 received 4.09 million immigrants between 2010 and 2023. (28.2/1000 people). In that same time period, states that Trump won in 2016 and lost in 2020 received 1.11 million immigrants (23.5/1000 people), and states that Trump lost in both 2016 and 2020 received 5.19 million immigrants (36.4/1000 people)¹⁵.

In the absence of immigration, massive flows from blue to red states would weaken the Democratic Party federally. Immigration breaks this link. The populations, and thus electoral power, of blue states are propped up by millions of immigrants. New York and California don’t have to be better than Florida and Texas to retain national influence; they just need to outdo Mexico and India, a much easier task. And the economically-booming recipients of domestic migration are slowly turned purple by an influx of foreigners. The laboratories break; successful experiments do not spread. Rather than the pro-market, pro-development politics that Americans choose for their own lives winning, we get an inexorable shift towards California¹⁶.

The net effect of immigration is to make the United States much less white, and therefore more anti-capitalist. Source.

Conclusions

One of American federalism’s biggest strengths is the ability to locally experiment and spread the best results. In recent years, internal migration flows show that, when it comes to their own lives, Americans view the red state model of low taxes and light regulations as superior to the alternative. In the absence of immigration, this would strengthen the Republican Party federally and force the Democratic Party to adopt pro-market policies to stay competitive. But immigration breaks this link by propping up blue state population and power and turning booming red states purple. We should end this distortion, and return to a model where good state governance wins.

1. A common Twitter interaction is someone posting a map showing that the American South has very high levels of X social problem, followed by someone else pointing out that this is because the American South has a lot of blacks.

2. As it so happens, international migration flows track GDP per capita + legal barriers to migration almost perfectly. For all of its well-known problems, GDP per capita really is a great measure of a country’s quality of life. But this doesn’t apply to internal migration.

3. Which, on paper, grew faster than that of West Germany for the entire existence of the DDR, though GDP statistics in a command economy and Communist statistics in general should be taken with a grain of salt.

4. Source is author’s own calculation with census data from here and here.

5. While housing regulation is often done at the local level, the state level is still the correct unit of analysis, because local governments are not sovereign and can have their powers stripped by their states at any time.

6. 2009 was not chosen arbitrarily; the US housing market was very different pre-Great Financial Crisis.

7. North Carolina is particularly impressive, having permitted about as much housing in 2023 as all of California. Strongly encourage people to check out this data explorer.

8. In this context, “well-structured” means low and minimally distortionary.

9. This is a much smaller driver of partisan segregation than selection, but it is real, and US elections are close enough for it to matter.

10. Or, rather, shifting both parties left to remain competitive, in the same way that the Republican party has moved left on entitlements to survive in the face of an increasingly nonwhite electorate.

11. And incidentally provides an argument for the electoral college. It’s a good thing that good state governance leading to higher state populations translates to more political influence, even without the movers necessarily changing any of their own positions.

12. Electorally. The faction aligned with the Democratic Party dominates most non-electoral sources of power, like the civil service, though there are intriguing signs that tech, which has massive economic and narrative control power, may be shifting red.

13. Specifically, through the mechanism of destination red states getting larger and therefore more influential on the national stage while remaining red.

14. And in Florida’s near future, Cuba and Haiti. ~10% of Cuba’s population emigrated in 2022-23, mostly to the US, and about 6% of Haiti has immigrated to the US since 2020, very disproportionately to Florida. These people can’t vote yet, but it’s only a matter of time (though how Cuban migration will shake out politically is not clear. Historically, Cubans have been slightly Republican, but recent waves are more Democratic, as is the second generation).

15. Source is author’s own calculation with census data from here and here. Caveat that census data should be taken with a grain of salt with regards to illegal immigrants.

16. Immigration also weakens the link between good local governance and federal power in another way: it lowers internal migration as a whole, and is in fact entirely responsible for the decline in US internal migration since the 1970s. Given the importance of internal migration to national unity, personal opportunity, and allowing better governance to spread (the focus of this article), this is bad.