Geographic Trends Archives - Economic Innovation Group

The plight of college-educated men shows where the job market is going

Stephen Leverton — Sat, 20 Sep 2025 13:27:56 +0000

The post appeared first on Economic Innovation Group.

America’s Heartland Is Coming Back. Can the Recovery Last?

Stephen Leverton — Fri, 11 Apr 2025 13:20:54 +0000

The post appeared first on Economic Innovation Group.

The big government spending Maga voters cannot live without

Stephen Leverton — Thu, 27 Mar 2025 20:52:37 +0000

The post appeared first on Economic Innovation Group.

Raising the SALT cap will not help low-income, rural, or pro-Trump places

Stephen Leverton — Tue, 18 Feb 2025 15:38:33 +0000

��Բ�� on Agglomerations, the Substack newsletter from the Economic Innovation Group.

By Connor O’Brien

The fight over the future of the State and Local Tax deduction (SALT) represents an early test. Will the expansion of President Trump’s coalition of working-class voters, whose support was so critical to his victory in last November’s election, actually translate into a more populist tax agenda in 2025?

The 2017 Tax Cuts and Jobs Act (TCJA) dramatically limited the reach of the SALT deduction, prohibiting each household from deducting more than $10,000 of state and local tax payments from their federally taxable income. Like the rest of TCJA’s changes to the individual tax code, this $10,000 cap on the SALT deduction is set to expire at the end of 2025.

Congress is therefore now considering its options, which include:

Repeal the cap altogether
Raise the cap higher than the current $10,000
Leave the cap in place at $10,000
Eliminate the SALT deduction itself (effectively a $0 cap)

Tax policy experts have already concluded that any reform to expand the SALT deduction will be both expensive and highly regressive, offering little or no benefit to most working or middle-class households. The Tax Policy Center, for instance, that repealing the SALT deduction cap would cost $1.2 trillion over a decade, 93 percent of which would go to households in the top income quintile. Even striking a compromise to raise the cap to $20,000 would benefit of middle-class households.

But SALT reform will also benefit certain kinds of places more than others, in particular high-income, urban, and Democratic parts of the country. Raising the cap, while coming at a steep cost to taxpayers, will also do little to benefit distressed, rural, or pro-Trump communities.

The SALT deduction does virtually nothing to benefit low-income places.

Today, even the capped SALT deduction largely benefits well-off places, particularly those concentrated in high-tax states. Using IRS statistics from individual federal tax returns in 2021, I find that the benefits of the SALT deduction largely flow to more prosperous communities.

Counties in the bottom five median household income deciles, weighted by population take home a quarter of the benefits from SALT, while more prosperous counties combine for three-quarters. (Please see the end of this post for the rationale behind population-weighting and other methodological points.)

The SALT deduction is also tilted against counties that voted for the President in November. Bucketing counties into deciles according to President Trump’s vote share, I find that the bulk of the SALT deduction’s benefits are received by households in counties where President Trump performed worst in November’s election.

The deduction does very little to benefit rural communities. Five percent of state and local taxes deducted on federal returns come from taxpayers in rural counties, which is less than half the share of the national population that resides in rural counties (13 percent).

Raising the SALT deduction cap would be even more tilted against low-income, pro-Trump, and rural counties.

Proposals to raise or eliminate the cap on the SALT deduction would tilt its benefits even further away from the Republican coalition and distressed places.

If the cap on the SALT deduction were eliminated, rather than kept at its current level of $10,000, 87 percent of the benefits would go to taxpayers in counties won by Kamala Harris last November. Virtually none would go to the places that formed the backbone of President Trump’s winning coalition.

Eliminating the SALT deduction’s cap would furthermore do nothing to cut taxes for those living in the lowest-income parts of the country. In fact, it would tilt the deduction even more towards prosperous communities. Less than one-tenth of the expanded deduction would benefit communities in the bottom half of the income distribution.

Finally, eliminating the SALT cap would yield meager benefits to rural communities. An outright majority of the benefits would go to large urban counties. Rural counties, meanwhile, would see a mere one percent of the benefits from eliminating the SALT deduction’s cap, despite accounting for 13 percent of the country’s population.

Methodology

Estimates in this blog are based on IRS SOI tax stats from 2021, the most recent available data. Taxpayers are assigned to counties and eight different income buckets. Within each unique county-income bucket, the IRS provides the total number of tax returns and the aggregate value of state and local taxes reported. The IRS also reports the number of taxpayers taking the SALT deduction and the combined value of such deductions.

The estimates of SALT reform’s benefits to particular places are measured in changes to Adjusted Gross Income (AGI) since the data does not allow us to apply individuals’ specific federal marginal tax rates to such changes. All references to benefits in these estimates are to changes in AGI, not fiscal impact.

To estimate benefits by geography from eliminating the cap, we calculate the average state and local tax burden in each county-income bucket, subtract $10,000 to reflect the current cap, and then multiply the excess state and local tax payments (which would be fully deductible) by the number of taxpayers in each bucket. These estimates are then aggregated according to population weighted deciles of 2024 county-level Trump vote share and 2018-2022 median household income, as well as 91PORN’s county-level geographic typology.

We weigh Trump’s vote share and median household income deciles by population throughout this analysis. This is due to the highly skewed distribution of county population sizes. There are a little more than 3,100 counties in the United States. More than 2,000 are home to 50,0000 people or less. Weighing deciles by population gives us a more representative picture of the SALT deduction’s effects.

Acknowledgments: Big shoutouts to Ernie Tedeschi of the Yale Budget Lab for advice on the methodology for this analysis and Jiaxin He for data help.

The post Raising the SALT cap will not help low-income, rural, or pro-Trump places appeared first on Economic Innovation Group.

Economy in Counties Trump Won Points to Governing Challenges

Reuben Francis — Fri, 03 Jan 2025 21:55:19 +0000

The post appeared first on Economic Innovation Group.

Mark Rozell: Trump’s cost-cutting plans threaten rural Virginia most

Stephen Leverton — Thu, 19 Dec 2024 15:35:23 +0000

The post appeared first on Economic Innovation Group.

It Was a Haven for New York Families. Now They Can’t Afford to Stay.

Stephen Leverton — Thu, 19 Dec 2024 15:18:54 +0000

The post appeared first on Economic Innovation Group.

The economic geography of the 2024 elections

Stephen Leverton — Fri, 15 Nov 2024 16:08:50 +0000

��Բ�� on Agglomerations, the Substack newsletter from the Economic Innovation Group.

By Sarah Eckhardt, Connor O’Brien, and Ben Glasner

The geographic breadth of Donald Trump’s gains in the 2024 presidential elections was remarkable. Trump received a larger share of the vote, relative to his 2020 performance, in nine out of every 10 counties. And the shift in his direction was most common in the large urban counties that typically favor Democrats.^[1]

Which economic and demographic variables were correlated with a rising vote share for Trump in a given county?

We already noted in a that voters were most likely to move towards Trump in the most populous, most expensive counties. We also found, surprisingly, that inflation rates for housing and food were not at all correlated with the change in vote shares by county.

In this post we dig much deeper into the red shift and identify more of the factors associated with Trump’s gains in 2024. Here is what we found:

A leading indicator for how far a county swung towards Trump is the size of its immigrant population share.

The higher a county’s share of immigrants, the more likely it was to shift votes towards Trump. We found that a county’s immigrant population share had one of the strongest correlations with a shift towards Trump of the demographic and economic factors we explored. (See the Appendix below for the full regression analysis.)

Queens County, New York is a good example. Nearly half (47.1 percent) of Queens residents are immigrants — and Trump received 10.4 percent more of the vote in Queens than he did in 2020, one of the biggest shifts among large urban counties.

Notably, this relationship does not extend to changes in a county’s immigrant population. When we looked at swings in the immigrant population share between 2016 and 2022,^[2] we found a tiny negative correlation with voter shifts towards Trump (and thus a corresponding positive correlation with a shift towards Kamala Harris).^[3]

Trump gained the most in the least educated counties.

Political polarization across education groups is growing. The correlation between the share of a county’s population without a bachelor’s degree and the vote going to the Republican presidential candidate has been climbing ever since the 2012 election. And the opposite trend applies to the share of a county’s population with a college degree, which is correlated with Democrats gaining votes.

Non-college voters shifted towards Trump across all geographies, but Trump’s gains within this group were biggest in large urban counties.^[4]

The larger a county’s Hispanic share, the more likely Trump was to gain votes in the county.

In both 2020 and 2024, Trump gained votes in counties with large shares of Hispanic residents. This relationship was strongest in large urban counties and in counties on the southern border.

The largest shift in the Trump voting share — of any county across the entire United States — took place in Maverick County, Texas, which swung towards Trump by 14.1 percentage points. Roughly 19 out of every 20 Maverick residents are Hispanic.

This trend is a reversal from the results of the 2016 election, when higher shares of voters in counties with large Hispanic populations turned out for Hillary Clinton in 2016 than they had for Barack Obama in 2012.

Trump gained more in counties that are in economic distress.

Following historical precedent, counties that are relatively worse off in terms of unemployment, poverty, and median household income were more likely to choose the non-incumbent candidate, Trump in this case.^[5]

The figure below splits counties into five quintiles, according to the gains and losses in the Trump vote share. Counties in Quintile 5 had the highest increase in the share of votes going to Trump between 2020 and 2024, and counties in Quintile 1 had the lowest.^[6]

The relationship is again clear — the more a county has been suffering economic distress, the bigger its voting shift to Trump. (We measure indicators of distress using our Distressed Communities Index.) For counties in the highest quintile, the average poverty rate was 5 percentage points higher than for counties in the lowest quintile, prime-age adults were 6 percentage points more likely not to be employed, and a much smaller share of residents lived inside prosperous zip codes.

Place-based federal spending did not help Harris.

Democrats may have hoped that high-profile federal investments in local communities over the past few years would sway voters in their direction. There is no evidence that this spending had an impact on the election results. At the county level, we found no relationship between voting shifts to Trump at the county level and spending per capita on projects from the Bipartisan Infrastructure Law, CHIPS and Science Act, or Inflation Reduction Act.

Counties with a high dependence on transfer income favor Trump.

Places that rely more on government transfers tended to vote for Trump, while less reliant places tended to vote for Harris. (Transfers include programs like Social Security, Medicare, Medicaid, Unemployment Insurance, and the Earned Income Tax Credit.)

Earlier this year we published The Great Transfer-mation, an analysis about the growing reliance of American communities on government transfers. We found that as of 2022, 18 percent of total personal income in the United States came from government transfers.

Trump received 63 percent of the vote among significantly reliant counties, whose residents receive at least a quarter of their personal income from transfers. Harris received 56 percent of the vote among minimally reliant counties, where less than 15 percent of total personal income is from transfers.

—�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ�Ĕ

APPENDIX

To explore the relative importance of economic and demographic factors, we estimated a series of regression models: with demographic characteristics only, economic characteristics only, economic and demographic characteristics combined, and economic and demographic characteristics with state fixed effects. All independent variables are z-scored, and population weights are applied. It is clear that the low-educated population share, the foreign-born population share, and the Hispanic population share, poverty, and unemployment are the most strongly associated with how far a county swung right, though the relative importance of these variables depends on the model specification.

Notes

The data is updated as of Tuesday, November 12, 2024. We exclude counties for which voter turnout was <80 percent that of the 2020 election.
The most recently available data goes through 2022.
This correlation is -0.04.
The correlation coefficient between the share of a county’s population without a high school degree and Trump’s gain in the vote share is In large urban counties this was 0.76 for large urban counties, and 0.39 for rural counties.
Counties that shifted towards Harris (away from Trump) were also included in the rankings. They are part of Quintile 1.

The post The economic geography of the 2024 elections appeared first on Economic Innovation Group.

Manufacturing already has made a comeback

Stephen Leverton — Tue, 12 Nov 2024 15:58:01 +0000

The post appeared first on Economic Innovation Group.

Opinion – The voter shift that should alarm Democrats most

Stephen Leverton — Sun, 10 Nov 2024 15:49:47 +0000

The post appeared first on Economic Innovation Group.