Interstate Migration Trends in the United States, 2018–2023: Where Are Americans Moving, and Why?

According to Internal Revenue Service (IRS) data, there were 33.9 million interstate moves in the United States for tax years 2018–23.¹ In 2023 alone, 6.7 million Americans packed up and moved across state lines, taking their income, their tax dollars, and their economic gravity with them. Where did all this movement actually take place, and which states gained or lost the most residents?

This policy brief examines the large differences in domestic migration patterns across states between 2018 and 2023 and evaluates a range of possible factors that play a role in those patterns, including tax burden, housing supply, broad economic freedom, labor market freedom, partisan preferences, cost of living, government size, population density, climate, and housing prices. Multiple years of data provide a more reliable picture of migration trends than single-year observations do, while a five-year lookback includes pre-pandemic, pandemic-era, and post-pandemic migration patterns.

Using state-level data, the analysis compares each factor individually before examining how the variables perform when considered together. The results suggest that over the five-year period, interstate migration in the United States has not been driven primarily by cost of living or climate, as commonly assumed. Instead, the data point to a different conclusion: Americans are systematically moving to states with lower tax burdens and more flexible housing supply, suggesting that tax and housing policy together play a central role in shaping where people choose to live. States seeking to sustain population growth need to maintain competitive tax structures and allow housing supply to expand in response to demand.

Domestic Migration Patterns Across States

The migration patterns observed during the five-year period between 2018 and 2023 raise important questions about why some states consistently gained residents while others lost them.² The map in figure 1 shows net migration of individuals between states. On net, Florida gained the most movers (+911,422), while California lost the most (−1,263,365).

Higher-population states would naturally be expected to lose or gain the most individuals. To account for this, a better way to measure the net movement of people between states is to measure the movement of people as a share of the state population. Figure 2 shows the net migration rate—this is the number of individuals moving per 1,000 residents in the state they are moving to or from.

In this case, the states with the highest positive migration rates are Idaho (+63.1), South Carolina (+53.6), Delaware (+42.5), Montana (+42.4), and Florida (+39.7). Meanwhile, the states with the highest negative migration rates are New York (−52.9), Alaska (−37.3), Illinois (−34.0), California (−32.2), and Hawaii (−30.2). Table 1 ranks the 50 states from highest positive migration rates to highest negative migration rates.

Taken together, these patterns suggest that interstate migration is not occurring randomly or evenly across the country. Some states are consistently attracting new residents, while others are steadily losing population. The next section examines which state characteristics are most closely associated with these migration flows.

Comparing the Correlates of Migration

The standard story is that people leave expensive places such as California or New York and move to cheaper ones in the South and Mountain West. Another explanation is that people are chasing sunshine, too (hence big net gains in the Sunbelt region).

To test this theory, I used state-level net interstate migration data covering all 50 states from 2018 to 2023 and examined a range of possible factors including tax burden, housing supply, broad economic freedom, labor market freedom, partisan preferences, cost of living, government size, population density, climate, and housing prices.³ Examining each individually and one at a time produces a clear ranking of the factors most strongly associated with interstate migration. These regressions should be interpreted as descriptive rather than causal. Many of these factors are closely related and tend to be jointly determined with migration. The goal is to identify broad patterns rather than isolate precise causal effects.

One important caveat is that the migration data span a period that includes the COVID-19 pandemic and the associated surge in remote work and geographic mobility. These developments likely amplified interstate migration flows. The current analysis, however, compares differences across states rather than changes within states over time. To the extent that the pandemic shock affected states unevenly, those effects are likely captured by the same structural factors examined in this analysis.

The results of the regressions provide a clear ranking of which state characteristics are most strongly associated with interstate migration. The following scatterplots illustrate the relationship between the empirical variables used to measure those characteristics and the net in-migration rate across all 50 states.

Tax burden

The taxation component of the Fraser Institute’s Economic Freedom of North America index is strongly associated with migration (see figure 3).⁴ This measure captures the overall tax burden by incorporating top marginal tax rates, tax revenues as a percentage of income, as well as property tax and sales taxes as a percentage of state income. Every one-point increase in the tax burden score indicates a lower tax burden and is associated with an 11.5-point increase in the net in-migration rate (95 CI: 6 to 17).⁵

Housing supply

The Building Permits Survey of the US Census Bureau tracks the number of building permits issued per 1,000 state residents, a measure that has the strongest positive association with interstate migration of all the variables examined (see figure 4).⁶ In this context, higher permit rates reflect housing markets that can expand more easily in response to rising demand.⁷ Every additional permit issued per 1,000 residents is associated with an 8.5-point increase in the net in-migration rate (95 CI: 6 to 11).

Using an alternative measure of housing supply—the percentage growth in housing units between 2020 and 2024—yields similar results, although the relationship is somewhat weaker. The weaker relationship is consistent with the fact that housing stock growth partly reflects migration itself, whereas housing permits more directly measure supply conditions.

Broad economic freedom

The Fraser Institute’s Economic Freedom of North America index measures broad economic freedom using three subindices: tax, government spending, and labor market freedom (see figure 5).⁸ State in-migration rates are positively correlated with this broader measure of economic freedom. Every one-point increase in economic freedom is associated with an 11-point increase in the net in-migration rate (95 CI: 5 to 17).

Labor market freedom

Labor market freedom is measured using a subindex of the Fraser Institute’s Economic Freedom of North America index (see figure 6).⁹ This measure includes government employment share, union density, and minimum wage income as a percentage of per capita income. States with higher labor market freedom scores, indicating more flexible labor markets with fewer labor market regulations and restrictions, tend to experience more inward migration. Every one-point increase in the labor market freedom score is associated with an increase of 9 points in the net in-migration rate (95 CI: 4 to 14).

Partisan preferences

Using 2024 presidential election vote margins by state as a measure of partisan preferences, the analysis finds a positive association between votes for the Republican candidate (Trump) and interstate migration patterns (figure 7).¹⁰ The results are significant but small. A 10-percentage-point swing toward Republicans is associated with a positive change in the net in-migration of about 4 points (95 CI: 1 to 8).

Cost of living

Cost of living is measured using the State Annual Regional Price Parities (SARPP) index of the Bureau of Economic Analysis (BEA), which compares cost of living across states relative to the national average.¹¹ States with higher costs of living tend to experience lower net in-migration rates (figure 8). For every percentage point by which a state’s cost of living exceeds the national average, the net in-migration rate declines by roughly 1.1 points (95 CI: −2.1 to 0.2).

Government size

The government size component of the Economic Freedom of North America index measures government spending, transfers, and subsidies as a share of state income.¹² A higher score means a state has a smaller government in size and scope. This “size of government” measure is positively associated with net in-migration (figure 9). Every one-point increase in the government size score is associated with an increase in the net in-migration rate of 4.6 points (95 CI: 0.3 to 9).

Population density

Measuring the association between population density and net in-migration using census data reveals a weak and statistically insignificant relationship (figure 10).¹³ While the estimated relationship suggests that people tend to move to less densely populated states, confidence is low, and the possibility of no relationship cannot be ruled out. Because the estimated effect is both weak and imprecise, confidence interval estimates are omitted here.

Climate

While a weak positive correlation between good climate and the migration rate exists, it is not statistically significant (figure 11). The climate comfort index is designed to reward temperatures close to a comfortable benchmark (around the mid-to-high 60s in degrees Fahrenheit) and abundant sunshine, while penalizing excessive rainfall and snowfall. Scores are then normalized onto a 0–10 scale for ease of interpretation. Climate is not a good predictor of state migration patterns. Because the estimated relationship is weak and statistically insignificant, confidence interval estimates are not reported here.

Housing prices

Similarly, average housing prices by state, using 2023 Zillow data, do not show a statistically significant association with interstate migration (p = 0.35) (figure 12).¹⁴ In this case too, because the estimated relationship is weak and statistically insignificant, confidence interval estimates are not reported.

Table 2 summarizes the strength and statistical significance of the bivariate relationships between each variable and the net migration rate.

While these bivariate relationships provide a useful starting point, states differ across many dimensions simultaneously. States with different tax regimes, for example, also tend to differ in housing supply, labor market policy, and cost of living. To determine which factors remain important when considered together, the next section examines multivariate models that account for these overlapping relationships.

Which Factors Matter Most?

Each of the regressions in table 2 examines one variable in isolation, but states are bundles of characteristics. Low-tax states, for example, also tend to have more flexible housing supply, different labor market policies, lower costs of living, and predictable political alignments. This overlap of characteristics makes it difficult to know whether migration is responding to any one factor alone or to several overlapping characteristics at once. As a result, a simple correlation between, say, migration and politics or cost of living may capture several overlapping influences. Is it politics? Or taxes? Or housing supply? Or all of them together?

To disentangle these overlapping relationships, all variables need to be examined simultaneously. Rather than looking at each variable in isolation, the analysis estimates a set of multivariate regressions, holding constant a common set of factors while varying the policy variable of interest. This allows the comparison of states across several characteristics at the same time to isolate which relationships remain strongest after accounting for other differences.

Using the same state-level data on net interstate migration from 2018 to 2023, four models were estimated, each including the same controls but substituting a different policy variable. The models estimate how strongly migration is associated with each policy variable after accounting for housing supply, climate, population density, cost of living, and partisan preferences. Each model takes the following general form:

migration = housing permit rates + climate comfort index + policy variable + population density + cost of living + partisan preferences + error term (variables affecting migration not included in the model)

where the policy variable is one of the following:

• broad economic freedom (composite)

• tax burden

• labor market freedom

• government size.

Table 3 summarizes the results of the four multivariate specifications. Model 1 examines the composite broad economic freedom index and shows that states with higher levels of economic freedom experience significantly more in-migration. That is not especially surprising, but it does not tell us which component of “freedom” is doing the work. Model 2 produces the strongest overall results. The tax burden variable is large, estimated with a relatively high degree of statistical confidence, and highly significant. It is also slightly stronger than the broad economic freedom index. Labor market freedom, shown in Model 3, has a positive effect, but it is weaker, less stable, and marginally significant at conventional levels. Government size, examined in Model 4, also remains positively associated with migration, although the effect is smaller than that of tax burden.

Across all four models, the dominant factor is not “cost of living” in the way people usually think about it. It is not housing prices. It is not even climate. The strongest and most consistent relationships instead point toward policy, and specifically toward taxes and housing supply. The next section explores the mechanisms behind these relationships in greater detail.

The role of taxes and housing supply

Across all specifications, tax burden is the most powerful and robust predictor of migration. States with lower taxes attract more people. That result holds even when controlling for housing permit issuance rates, climate, population density, and cost of living.

The mechanisms behind this relationship are relatively straightforward. Tax policy directly affects after-tax income. For mobile households, particularly higher earners and business owners, even modest differences in state tax burdens can translate into meaningful differences in take-home pay over time. This creates a persistent incentive to relocate toward lower-tax states, especially when those differences compound year after year. States with larger government sectors tend to experience more out-migration than states with leaner government sectors. The effect of government size, however, is smaller and less robust than that of tax burden. The broader “economic freedom” effect turns out to be largely a tax effect. Once the components are separated, labor market freedom and government size also play a role, albeit a smaller one.

But tax burden is not the whole story. The variable of housing permit rates, which affect housing supply, shows up consistently as well. In the strongest model, permits are not just positive, but statistically significant. Housing supply determines whether a state can accommodate new residents without sharply increasing costs. In states where construction is constrained by zoning rules, permitting delays, or regulatory barriers, increases in demand tend to show up as higher prices rather than more housing. This supply constraint limits in-migration and can push existing residents out. By contrast, states with more flexible housing supply can absorb population inflows more easily, keeping costs lower and enabling continued growth.

Importantly, housing supply is not independent of the tax burden variable. States with more favorable tax and regulatory environments are also more likely to permit new construction, suggesting these factors reinforce each other rather than operate in isolation. This combination both attracts new residents and sustains those inflows over time.

Cost of living and migration

The relationship among housing supply, tax policy, and migration also helps explain why commonly cited factors such as cost of living or house prices lose significance in the multivariate analysis. High prices are often the result of underlying policy constraints, particularly limited housing supply and higher tax burdens, rather than independent drivers of migration decisions. Cost of living, measured using the BEA’s regional price parity index, is negatively associated with migration, as expected, but the relationship is not statistically significant. The same is true of population density. Even partisan preferences, proxied by 2024 vote margins, have no independent explanatory power.

Housing prices are excluded from the multivariate specifications due to their strong overlap with the broader cost-of-living measure and their lack of statistical significance in the bivariate analysis. High-cost states tend to be places with restrictive housing supply, higher taxes, and tighter regulatory environments. Those underlying factors both raise prices and push people out. Once those underlying factors are taken into account, prices lose much of their explanatory power. In other words, it is tempting to say people move to cheaper states. But “cheap” is an outcome, not a cause.

The migration patterns from 2018 to 2023 are not primarily about sunshine or sticker prices. They reflect something deeper: how states structure their economies. Lower-tax states with more flexible housing supply are gaining population, while higher-tax states with constrained supply are losing it. Everything else—climate, politics, even cost of living—is secondary.

The standard narrative focuses on affordability and climate. But the data point elsewhere: toward tax policy and housing supply. That story is less simple, but it is more accurate and ultimately more useful for understanding where Americans are going and why.

Policy Implications

The results suggest several implications for state policymakers.

First, tax policy plays a central role in shaping migration patterns. States with higher tax burdens risk losing residents, particularly higher-income and more mobile households, to lower-tax jurisdictions. Efforts to raise revenue through higher marginal tax rates may therefore face important tradeoffs, including the gradual erosion of the tax base over time.

Second, housing supply constraints significantly limit a state’s ability to attract and retain residents. Restrictive zoning laws, lengthy permitting processes, and other regulatory barriers not only raise housing costs, but also reduce in-migration by preventing supply from responding to demand. States that want to accommodate population growth must ensure that housing markets can expand accordingly.

More broadly, the findings suggest that attracting and retaining residents depends less on geographic advantages than on policy choices. States that prioritize competitive tax structures and allow housing supply to respond to demand are more likely to experience sustained population growth.

About the Author

Jack Salmon is a Gibbs Scholar and research fellow at the Mercatus Center at George Mason University, where he focuses on economic and fiscal policy, with an emphasis on federal budgets, taxation, economic growth, and institutional analysis. His research and commentary have been featured in a variety of outlets, including The Hill, Business Insider, RealClearPolicy, National Review, the American Institute for Economic Research, and Reason Magazine. Salmon has provided expert analysis on fiscal and economic issues in various policy forums, including testimony before Congress on the risks of debt accumulation, deficit spending, and inflation.