Introduction

Medicaid, Title XIX of the Social Security Act, is a joint federal-state program that finances health care to the poor.^[1] When it was first signed into law, Medicaid eligibility was limited to low-income children, pregnant women, parents of dependent children, the elderly, and people with disabilities. In the sixty years since the program was enacted, however, it has strayed from its mission of providing healthcare for the most vulnerable and has become a steppingstone toward universal government-run health insurance.

This explainer will outline how Medicaid functions, the program’s costs, its influence on healthcare in the United States, and how the proposed policy changes in 2025 could reshape the program.

How Does Medicaid Work?

Medicaid is divided into two groups: traditional Medicaid and the Medicaid Expansion group. Before discussing the differences between the two, it’s important to understand that there are strings attached. For a state to participate in Medicaid (either traditional or expansion), the federal government requires that state to provide Medicaid coverage for certain eligibility groups, including^[2]:

• Certain low-income families, including parents, that meet the financial requirements of the former Aid to Families with Dependent Children (AFDC) cash assistance program;
• Pregnant women with annual income at or below 133% of the Federal Poverty Level (FPL);
• Children with family income at or below 133% of FP;
• Aged, blind, or disabled individuals who receive cash assistance under the Supplemental Security Income (SSI) program;
• Children receiving foster care, adoption assistance, or kinship guardianship assistance under the Social Security Act (SSA) Title IV–E;
• Certain former foster care youth;
• Individuals eligible for the Qualified Medicare Beneficiary program; and
• Certain groups of legal permanent resident immigrants.

Federal law provides two primary benefit packages for state Medicaid programs: traditional benefits and alternative benefit plans (ABPs). These benefit categories (taken from the Congressional Research Service) are recreated in Table 1. States also have some flexibility through Medicaid program waivers, which allow them to be exempt from certain federal requirements. These include research and demonstration projects (Section 1115), managed care/freedom of choice programs (Section 1915(b)), and home and community-based services (Section 1915(c)). To receive a waiver, a state must meet federal financing requirements such as budget neutrality, cost-effectiveness, or cost-neutrality.^[3]

It is also important to note that Medicaid spending is often lumped in with the Children’s Health Insurance Program (CHIP) and similar federal subsidies created under the Patient Protection and Affordable Care Act (Affordable Care Act or ACA). The CHIP program provides health coverage to eligible children in families with incomes above the Medicaid threshold, either through Medicaid or separate state programs. The federal subsidies created under the ACA include premium tax credits (which subsidize the cost of an insurance premium) and cost-sharing reductions (reducing out-of-pocket costs such as deductibles, copays, and coinsurance) for those who purchase health insurance through a government-created healthcare marketplace.

Traditional Medicaid

Traditional Medicaid covers both primary and acute care as well as long-term services and supports (such as care for disabled adults and individuals with chronic illnesses). Eligibility is limited to low-income children, pregnant women, parents of dependent children, the elderly, and people with disabilities. In this program, states are guaranteed federal matching dollars without a cap for qualified services, based on a formula that matches at least 50 percent of state spending. The portion of the federal government’s share of most Medicaid expenditures is known as the Federal Medical Assistance Percentage (FMAP). This matching rate increases as state per-capita income decreases.

Under traditional Medicaid, states define the specific features of each covered benefit within four broad federal guidelines:

• Each service must be sufficient in amount, duration, and scope to reasonably achieve its purpose. States may place appropriate limits on a service based on such criteria as medical necessity.
• Within a state, services available to the various population groups must be equal in amount, duration, and scope (the comparability rule).
• With certain exceptions, the amount, duration, and scope of benefits must be the same statewide (the statewideness rule).
• With certain exceptions, enrollees must have freedom of choice among health care providers.^[4]

Looking ahead to FY 2026 (October 1, 2025 – September 30, 2026), the federal matching rates for state funds are expected to range from 50 percent (the mandatory minimum matching rate) to nearly 77 percent.^[5] Figure 1 shows the federal Medicaid FMAP matching rate for each state.

Figure 1: Federal FMAP Percentages, FY 2026

Sources: KFF estimates of increased FY 2026 FMAPs based on Federal Register, November 29, 2024 (Vol 89, No. 230), pp 94742-94746.

Note: Estimates are rounded to the nearest whole number.

The Medicaid Expansion Group

Under the Affordable Care Act (ACA), states had the option to expand Medicaid to non-elderly adults with income up to 133 percent of the Federal Poverty Level.  When states were initially allowed to expand Medicaid starting January 1, 2014, the federal government promised to cover 100 percent of Medicaid expansion costs to encourage states to participate. With this promise of a “free lunch,” many states rushed to expand Medicaid, sharply increasing enrollment. By 2020, however, the federal match rate for the expansion program was reduced to 90 percent. As a result, states had to increase their own Medicaid spending, on average, $26.7 billion from 2017 to 2022 from their own sources.

As of 2025, all but 10 states have expanded Medicaid.^[6] Those states are shown in Figure 2.

Figure 2: States that Have Not Expanded Medicaid as of 2025

Sources: KFF tracking and analysis of state actions related to adoption of the ACA Medicaid expansion and Searing, Adam. “Federal Funding Cuts to Medicaid May Trigger Automatic Loss of Health Coverage for Millions of Residents of Certain States.” Say Ahhh! Georgetown Center for Children and Families, November 27, 2024

How Much Does Medicaid Cost? Who Pays?

Given that Medicaid is a joint federal and state program, it is important to examine the costs of Medicaid at the federal and state levels. At the federal level, Medicaid, the Children’s Health Insurance Program (CHIP), and other healthcare marketplace subsidies enacted by the ACA cost $759 billion in FY 2024. Put another way, for every dollar the federal government spent, eleven cents of that dollar went to Medicaid, CHIP, and the ACA subsidies.^[7]

At the state level, Medicaid accounts for about 30 percent of total state spending (capital inclusive) and is the single largest expenditure in all state budgets. For every dollar the average state spends, thirty cents go to Medicaid—only ten cents come from state revenue while the remaining 20 cents come from federal transfers.^[8]

Although Medicaid was designed to be a “joint” funding program, state policymakers have found ways to get the federal government to cover the lion’s share of Medicaid spending. This reflects the incentives elected officials face: using accounting gimmicks to offer more generous Medicaid spending while passing the cost to federal taxpayers can help them win reelection.

This problem was exacerbated by Medicaid expansion under the ACA. Figure 3 (recreated from the CRS report) shows the breakdown of federal and state Medicaid spending. The percentages atop each column indicate the federal share of total Medicaid spending.

Figure 3: Federal and State Shares of Medicaid Spending

Sources: Congressional Research Service “R43357: Medicaid: An Overview,” Figure 6: Federal and State Actual Medicaid expenditures CMS, Form CMS-64 Data as reported by states to the Medicaid Budget and Expenditure System, as of May 29, 2024, at https://www.medicaid.gov/medicaid/financial-management/state-expenditure-reporting-for-medicaid-chip/expenditure-reports-mbescbes. CPI-U inflation data collected from US Bureau of Labor Statistics

Notes: CMS, Form CMS-64 Data as reported by states to the Medicaid Budget and Expenditure System, as of May 29, 2024, at https://www.medicaid.gov/medicaid/financial-management/state-expenditure-reporting-for-medicaid-chip/expenditure-reports-mbescbes.

In the end, federal taxpayers are footing the bill for Medicaid. However, as the national debt continues to strain the federal budget and crowd out other priorities, policymakers in DC are desperate to cut costs. One likely area is federal Medicaid spending. If the federal government were to change the matching rates of either traditional Medicaid or Medicaid expansion, state spending on Medicaid would rapidly increase and crowd out other spending. In more fiscally distressed states, this could spur a fiscal crisis.

How Does Medicaid Impact Healthcare?

The size of Medicaid means that it shapes almost every corner of the American healthcare system, from hospital and acute care to long-term care to medical research. The program covers one in five Americans and finances 19 percent of all health spending in the United States. Here are some of the results of that influence.^[9]

Increasing Coverage with Little to Show for Health Access or Outcomes

Medicaid increases healthcare coverage. Thanks to the Medicaid Expansion under the ACA and more generous federal matching programs created during the COVID-19 era and through the Biden administration’s stimulus packages, enrollment in Medicaid dramatically increased and the percentage of uninsured Americans decreased, reaching an all-time low in 2022.^[10]

Additionally, while use of healthcare services increased, other negative outcomes emerged that decreased access to care, especially for those in traditional Medicaid. Cannon (2022a) notes that the Medicaid Expansion under the ACA creates an incentive for state policymakers to prioritize Medicaid expansion group recipients over traditional Medicaid recipients.^[11] Blase and Gonshorowski (2025) confirmed these findings, noting that Medicaid expansion decreased access to care, crowded out private options, and shifted funds away from the poorest Medicaid recipients.^[12]

In a review of the literature, Sigaud (2025) also finds depressing results^[13] States that expanded Medicaid saw longer wait times and reduced access to care for traditional Medicaid enrollees. Additionally, he notes that symptoms of depression increased among near-elderly adults on Medicaid before and after expansion, especially among rural residents with extremely limited access to mental health providers. He also notes slower ambulance response times and greater delays in the emergency room.

Cementing the Relationship Between Employment and Healthcare

Medicaid expansion under the Affordable Care Act further entrenched employer-sponsored insurance (ESI) as the backbone of American healthcare. The ACA kept the ESI tax structure in place, essentially creating what Cannon (2022b) calls “an implicit penalty on workers who do not (a) surrender control of a sizable portion of their earnings to an employer; (b) enroll in a health plan that their employers choose, control, and revoke upon separation; and (c) pay the balance of the premium directly.”^[14]

In an ideal world, Americans would not need to leave their jobs to change healthcare provider networks. Unfortunately, if Americans want a different health insurance package, they must “fire” their employer, pay a large tax penalty for choosing an employer-sponsored plan, or be stuck with an inferior, public option.

Increasing the Cost of Healthcare

Medicaid costs for healthcare are much greater than the costs of healthcare in the private sector. In my AIER paper “The Work vs Welfare Tradeoff Revisited,” I found that Medicaid paid more per full-year equivalent enrollee than the average annual single premium for an employer-sponsored plan in 43 states.^[15] Despite the higher payments, health outcomes for Medicaid recipients are not better than those of Americans with private insurance.

The reason why Medicaid is so costly comes from the incentives created under the joint federal-state funding relationship, as discussed in the previous section. Cannon (2022a) elaborates, “Spending $1 on police buys $1 of police protection. Spending $1 on Medicaid, however, buys $2 to $10 of medical or long-term care. Medicaid rewards states for spending the marginal dollar on medical and long-term care even when spending it on police, education, or transportation would provide greater benefit.”^[16]State officials have an incentive to maximize Medicaid while cutting basic public services. The open-ended federal matching system allows states to maximize federal matching dollars (especially for expansion populations) through gimmicks such as provider tax loopholes.^[17]As spending on the expansion population increases, traditional Medicaid enrollees are pushed aside, leading to less access to care and worsening health outcomes.

The Government Accountability Office (GAO) regularly lists Medicaid (and its relative Medicare) among the “High-Risk” list for improper payments. The GAO notes that Medicaid program integrity must be strengthened through both legislation and “coordinated effort across multiple entities.”^[18] Additionally, America is one of the most charitable nations in the world. In closing, Mueller opines,

In other words, Medicaid is rife with waste, fraud, and abuse, and fixing it is no small task.

Increased Regulatory Complexity

Medicaid also has a significant impact on the nature and shape of healthcare regulations. Federal rules dictating how states shape their Medicaid policies discourage innovation, research, and flexibility because state policymakers want to maximize those federal matching dollars. Furthermore, states will shape their own healthcare regulations to ensure compliance with federal Medicaid guidelines and maximize federal Medicaid funding. This results in states limiting access to new therapies to control costs.

What Do the 2025 Policy Changes Mean for Medicaid?

In 2025, two major policy changes have impacted Medicaid: proposed changes under the “One Big Beautiful Bill” (H.R. 1) and a Centers for Medicare and Medicaid Services proposed rule to close a provider tax loophole. These changes have the potential to provide immediate fixes to Medicaid, but much deeper reforms are needed.

The largest change comes from the legislative and CMS rule changes toward Medicaid provider taxes. The changes in H.R. 1 phase the Medicaid provider tax rate from 6 percent to 3.5 percent and freeze any new provider taxes created^[19] It would also mandate waiver resubmissions and suspend existing approvals in noncompliant states. These reforms would ensure Medicaid financing aligns with federal intent, helps reduce wasteful spending, and prevents states from misusing federal Medicaid funds for other general fund programs.^[20] It would also mandate waiver resubmissions and suspend existing approvals in noncompliant states. These reforms would ensure Medicaid financing aligns with federal intent, helps reduce wasteful spending, and prevents states from misusing federal Medicaid funds for other general fund programs.

Additionally, H.R. 1 also strengthens work requirements and eligibility checks, ensuring that verification standards are improved and states are allowed to remove ineligible enrollees from Medicaid.

These reforms, unfortunately, only scratch the surface. Deeper changes to Medicaid (as well as healthcare broadly) are needed. One such change is offered by economist David Rose. Rose writes,

“To put it simply, eliminate Obamacare, Medicare, and Medicaid and replace them with a national healthcare voucher system. This transformative change for American healthcare could be limited to the level paid for with a national sales tax, and our unfunded liability problems would simply disappear. While, for practical reasons, this would likely have to start at the national level, the goal could be to then spin it off to the states.”^[21]

There is no shortage of ideas available for healthcare reform. The problem lies in changing the incentives that millions in the healthcare sector face (both in government and the private sector) that keep them maintaining the status quo.

Conclusion

Medicaid was designed to provide a safety net for the most vulnerable Americans. After sixty years, trillions spent, and millions of Americans enrolled, the program has little to show for it. It has strayed from its mission of helping the poor because policymakers prioritize maximizing federal matching rates. Medicaid spends more yet fails to provide better health care access or health outcomes, increases costs, and discourages choice and innovation in healthcare.

The United States—the wealthiest nation in history—and its people deserve health care that delivers access, valuable health outcomes, affordability, and choice. Market-driven solutions can provide such a system.

---

Footnotes

---

[1] Social Security Administration. Medicaid. In Annual Statistical Supplement to the Social Security Bulletin, 2015. https://www.ssa.gov/policy/docs/statcomps/supplement/2015/medicaid.html.

[2] Congressional Research Service. Medicaid: An Overview. R43357. Washington, DC: Library of Congress, 2023. https://www.congress.gov/crs-product/R43357.

[3] Ibid.

[4] Ibid.

[5] KFF. “Federal Matching Rate and Multiplier.” KFF State Health Facts. Accessed July 9, 2025. https://www.kff.org/medicaid/state-indicator/federal-matching-rate-and-multiplier.

[6] KFF. “Status of State Medicaid Expansion Decisions.” KFF. Accessed July 9, 2025. https://www.kff.org/status-of-state-medicaid-expansion-decisions.

[7]

[8]

[9] Office of the Assistant Secretary for Planning and Evaluation. The Benefits of Expanding Medicaid Eligibility to Low-Income Adults: Evidence from State Expansions. U.S. Department of Health and Human Services, March 28, 2022. https://aspe.hhs.gov/reports/benefits-expanding-medicaid-eligibility.

[10] Office of the Assistant Secretary for Planning and Evaluation. 2022 Uninsurance Rate at an All-Time Low: New Estimates Highlight the Role of the ACA and Medicaid Expansion. U.S. Department of Health and Human Services, September 2022. https://aspe.hhs.gov/reports/2022-uninsurance-at-all-time-low.

[11] Cannon, Michael F. Cato Institute. “Medicaid and the Children’s Health Insurance Program.” In Cato Handbook for Policymakers, 9th ed., 2022. https://www.cato.org/cato-handbook-policymakers/cato-handbook-policymakers-9th-edition-2022/medicaid-childrens-health-insurance-program#perverse-incentives.

[12] Blase, Brian and Gonshorowski, Drew. “Resisting the Wave of Medicaid Expansion: Why Florida Is Right.” Paragon Institute. May 1, 2024. https://paragoninstitute.org/medicaid/resisting-the-wave-of-medicaid-expansion-why-florida-is-right.

[13] Sigaud, Liam. “Losing Focus: How the ACA’s Medicaid Expansion Left Traditional Enrollees Behind.” Paragon Prognosis, February 10, 2025. https://paragoninstitute.org/paragon-prognosis/losing-focus-how-the-acas-medicaid-expansion-left-traditional-enrollees-behind/#:~:text=A%202021%20analysis%20in%20Health,adverse%20outcomes%2C%20including%20higher%20mortality.e.

[14] Cannon, Michael F. Cato Institute. “The Tax Treatment of Health Care.” In Cato Handbook for Policymakers, 9th ed., 2022. https://www.cato.org/cato-handbook-policymakers/cato-handbook-policymakers-9th-edition-2022/tax-treatment-health-care#the-tax-exclusion-for-employer-sponsored-health-insurance.

[15] Savidge, Thomas. “The Work vs. Welfare Tradeoff Revisited.” American Institute for Economic Research, June 17, 2022. https://aier.org/article/the-work-vs-welfare-tradeoff-revisited/#medicaid.

[16] Cannon (2022a). supra note 11.

[17] Blase, Brian. Medicaid Provider Taxes: A Gimmick that Exposes the Flaws in Medicaid’s Financing. Arlington, VA: Mercatus Center at George Mason University, June 20, 2023. https://www.mercatus.org/research/research-papers/medicaid-provider-taxes-gimmick-exposes-flaws-medicaids-financing.

[18] U.S. Government Accountability Office. Medicaid Financing: Actions Needed to Ensure Provider Taxes Do Not Undermine Federal Oversight. GAO-25-107743, May 2025. https://www.gao.gov/products/gao-25-107743.

[19] U.S. Congress. H.R. 1: “One Big Beautiful Reconciliation Act of 2025,” 119th Cong., 1st sess., § 71115, “Provider Taxes” (2025). https://www.congress.gov/bill/119th-congress/house-bill/1/text

[20] Centers for Medicare & Medicaid Services. Preserving Medicaid Funding for Vulnerable Populations by Closing Health Care-Related Tax Loophole: Proposed Rule. Fact Sheet. Washington, DC: U.S. Department of Health and Human Services, May 2, 2024. https://www.cms.gov/newsroom/fact-sheets/preserving-medicaid-funding-vulnerable-populations-closing-health-care-related-tax-loophole-proposed#_ftn2.

[21] Rose, David C. “Want to Fix Medicaid? Look to Milton Friedman.” The Daily Economy, June 6, 2025. https://thedailyeconomy.org/article/want-to-fix-medicaid-look-to-milton-friedman.

This paper investigates the effect of affordable housing obligations in New Jersey on cost of living, cost of housing, and actual housing production. New Jersey’s Mount Laurel court cases established a doctrine whereby municipalities must permit their “fair share” of affordably priced housing. At various points since then, the legislature has interpreted this doctrine with specific, quantitative targets for municipalities to reach, or it has declined to do so, leaving enforcement up to the courts, which in turn backed away for several years from enforcing affordable housing targets. These policy changes provide an opportunity to investigate whether affordable housing targets are having their intended effect. Based on synthetic control analysis, the results show no effect of affordable housing mandates on housing production and minimal to no effect of affordable housing mandates on housing costs.

1. Introduction

The Mount Laurel court cases in New Jersey established that “a developing municipality may not, by a system of land use regulation, make it physically and economically impossible to provide low and moderate income housing.”^[1]

While the initial ruling merely held that municipal land-use regulations could not exclude everything other than large-lot single-family houses, the legislature has interpreted the Mount Laurel doctrine to say that municipalities have an affirmative obligation to construct deed-restricted affordable housing. Until 1985, the legislature did not act. Municipalities had widely flouted or challenged the 1975 New Jersey Supreme Court decision, leading to a “mass of protracted litigation.”^[2] In 1983, the Mount Laurel II decision affirmed a “builder’s remedy” against municipalities that failed to meet judicially determined affordable housing targets, thereby incentivizing the legislature finally to act.^[3]

The result was the Fair Housing Act of 1985, which established an independent body tasked with calculating each municipality’s “fair share” allotment of new affordable housing units, which would afford them safe harbor from builder’s remedy lawsuits. That body, the Council on Affordable Housing (COAH), issued its first round of fair share calculations in 1987, and they remained in force until 1993. Round II of COAH’s fair-share obligations were in force from 1993 to 1999, and after 1999 COAH went effectively defunct after its Round III calculations were invalidated by the courts. It was finally abolished in 2024 in favor of a new method of calculating fair-share obligations. Since 2007, then, when COAH’s Round III rules were invalidated, New Jersey municipalities have had to revert to the former method of judicial certification of compliance with the Mount Laurel doctrine.

How have the Mount Laurel doctrine and its varied interpretations by the legislature affected housing production and costs in New Jersey? To answer this question, we first need to understand the political economy of affordable-housing targets and inclusionary zoning programs.

2. The Effects of Affordable-Housing Mandates

In order to comply with COAH rules, New Jersey municipalities adopted inclusionary zoning. Inclusionary zoning comes in two forms: mandatory and voluntary. Mandatory inclusionary zoning requires that any new development of a specified size or kind set aside a percentage of units that must be rented or sold at an affordable price to low- or moderate-income consumers. Deed restrictions and income verification enforce the affordability requirement. Voluntary inclusionary zoning offers a developer a regulatory benefit, such as an increase in allowed density, provided a certain percentage of units are rented or sold at below-market rates in the same manner.

The Fair Housing Act of New Jersey requires municipalities to develop at minimum a voluntary inclusionary zoning program, but mandatory programs are ubiquitous throughout the state, according to a database maintained at inclusionaryhousing.org.

Inclusionary zoning only functions if the market-rate units in an inclusionary development are expensive. The developer has to make a large profit on those units to offset the losses on the units that are required to be sold or rented at below-market rates. If housing in general becomes affordable, then, inclusionary zoning does not work.

Inclusionary zoning has other curious effects. As a form of price control, it creates shortages and rationing. More people want the below-market units than are available, so they have to be allocated by lottery. Moreover, inclusionary zoning, at least in the mandatory form, reduces the supply of housing, because developers know they will make less profit from building than they would if there were no inclusionary requirements.

Empirical research largely confirms these theoretical expectations. Research in the Baltimore-Washington area finds that mandatory inclusionary zoning increases the cost of market-rate housing, though it might not affect new housing supply (Hamilton, 2021). An earlier study of California found that inclusionary zoning policies decreased the size and increased the price of single-family houses (Bento, et al., 2009). A study of San Francisco and Boston found that mandatory inclusionary zoning increased prices and reduced production in Boston while increasing prices during periods of rising prices and reducing prices during periods of falling prices in San Francisco, and not affecting production there (Schuetz, et al., 2011). Yet another study of California found strong adverse effects on production and prices (Means & Stringham, 2012). There’s only one contrarian study on the issue, finding that reduction in inclusionary zoning requirements in parts of California did not affect house prices (Hollingshead, 2015). It could be that these programs create hysteresis in housing markets once they’re implemented.

While these studies focus on mandatory inclusionary zoning, there are reasons to think that even voluntary inclusionary zoning, which is much more widespread, could have adverse consequences for housing production, particularly in a state like New Jersey or Massachusetts with a “builder’s remedy” available when municipalities fail to meet affordable housing targets.

Suppose that municipal leadership generally opposes new multifamily development. If a developer proposes a new market-rate multifamily development, it generally does not directly help the municipality meet its affordable housing target, because the units are usually not priced affordably enough. Moreover, since housing markets are supra-municipal, corresponding largely to commuting areas, municipal leaders may well realize that building market-rate housing has only weak effects on the overall affordability of housing within municipal boundaries, contributing rather to the affordability of housing throughout the local labor market. Realizing that they must meet their affordable housing targets, but disfavoring multifamily development generally, municipal leadership will have a clear incentive to deny planning permission for market-rate multifamily developments.

Anecdotally, the market-rate nature of a development (i.e., the lack of deed-restricted affordable units) frequently comes up as a justification for planning denials. State law doesn’t expressly protect municipal decisions to deny planning approval to market-rate multifamily projects, but municipalities generally enjoy broad discretion to deny approval for larger multifamily projects, through either site plan review, variance, or conditional use permit processes. When a lack of deed-restricted affordable units is combined with other features that may justify a denial, such as traffic impacts, a municipal land-use board may feel themselves on firmer ground to issue a denial than otherwise.

Anecdotally, the market-rate nature of a development (i.e., the lack of deed-restricted affordable units) frequently comes up as a justification for planning denials. State law doesn’t expressly protect municipal decisions to deny planning approval to market-rate multifamily projects, but municipalities generally enjoy broad discretion to deny approval for larger multifamily projects, through either site plan review, variance, or conditional use permit processes. When a lack of deed-restricted affordable units is combined with other features that may justify a denial, such as traffic impacts, a municipal land-use board may feel themselves on firmer ground to issue a denial than otherwise.

In this way, a state-level requirement that municipalities meet affordable-housing production targets will tend to reduce overall housing production and thereby increase overall housing costs.

Have affordable-housing production targets and inclusionary zoning had these effects in New Jersey? We need a credible empirical test comparing New Jersey to similar states.

3. Empirical Analysis

The empirical strategy here is to compare the rounds of binding affordable-housing obligations in New Jersey to the same periods in other states and to periods in New Jersey when there were not binding affordable-housing obligations. Of particular interest are two outcomes: building permits per capita (the best measure of housing production) and two measures of cost of living in general and cost of for-sale houses in particular. There is only one annual measure of state-level cost of living that goes back before 2008 (Berry, et al., 2000). We do not have annual data on rental costs, but the FHA produces an all-transactions house price index based on resales and appraisals. Their quarterly data are averaged by year to create an annual index.

The outcome variables are as follows: 1) total housing units permitted per capita, from the U.S. Census Bureau, 2) housing units in five-or-more-unit buildings per capita (these housing units are most likely to be affordably priced), 3) annual change in state cost of living, and 4) annual change in the state’s all-transactions house price index. The annual state-level building permits data go back to 1980 and up to 2023. The cost of living data go back to 1960 and up to 2007. House price data go from 1975 to 2024.

Figure 1 shows how total building permits per capita have evolved in New Jersey and two neighboring states since 1980. The vertical lines set off the period of 1985 to 1999, when Mount Laurel obligations were most seriously enforced, and the period 1999 to 2007, when Mount Laurel obligations were barely enforced at all, because the courts were waiting on a legislative solution that never came.

Figure 1: Total Building Permits per Capita in New Jersey and Neighboring States

We see here that before 1985, New Jersey was permitting a lot more units per capita than Pennsylvania and New York were. Shortly thereafter, however, New Jersey permitting plunged, and it remained more or less at Pennsylvania’s level until about 2012, when the state regained a small but steady advantage over its neighbors. These changes in permitting do not line up well with the enforcement of affordable housing obligations.

Figure 2 shows how building permits for units in buildings with five or more units evolved over time in the same group of three states.

Figure 2: Five-Plus Building Permits per Capita in New Jersey and Neighboring States

The results here are similar to those in Figure 1, except that Pennsylvania and New York switch places. Pennsylvania produces more housing than New York, but New York produces more multifamily housing than Pennsylvania, which is not surprising since New York is more urbanized. After 1990, New Jersey’s permitting of larger multifamily buildings closely tracks New York’s, but New Jersey did have a slightly higher rate between 2016 and 2022.

So far, the raw data do not show a strong effect of affordable housing obligations on housing production in New Jersey. New Jersey is the most densely populated state in the country, so it makes sense that New Jersey would produce a larger share of multifamily housing than a state like Pennsylvania. But only in the last few years has New Jersey produced more overall housing than Pennsylvania. There is no evidence that the 1985 to 1999 period specifically was more productive of housing in New Jersey than other periods, or that the 2000 to 2007 period was particularly unproductive, except perhaps relative to the post-2012 period.

Of course, building permits are a measure of quantity supplied, but economics teaches us that both supply and demand jointly determine an equilibrium. We cannot reason from quantity changes alone that New Jersey’s affordable housing obligations did not work, because perhaps New Jersey had abnormally low housing demand, which caused builders to want to build less regardless of regulations. If affordable housing obligations corresponded to periods of slow growth in the cost of living in New Jersey, we could infer that these obligations boosted housing supply after all.

Figure 3 shows the evolution of state cost of living from 1960 to 2007 for New Jersey and its larger neighbors.

Figure 3: Cost of Living in New Jersey and Neighboring States

This chart suggests as well that affordable housing obligations did not work. Cost of living in New Jersey grew more rapidly than in neighboring states between 1980 and 1990 and since then has remained consistently higher. This evidence suggests that demand for housing in New Jersey was growing rapidly in the 1980s, and more housing units would have been supplied to meet market conditions even in the absence of affordable housing obligations. It’s even possible that affordable housing obligations suppressed supply through the regulatory incentives established by inclusionary zoning.

Finally, Figure 4 plots the annual change in the house price index for all three states. It is important to note here that the house price index is specific to each state, so levels are not comparable across states, only changes are. Every state’s house price index is set to 100 in the first quarter of 1980.

Figure 4: Changes in House Prices in New Jersey and Neighboring States

New Jersey house prices mostly change in lockstep with New York’s. But between 2015 and 2020 there is a brief period when New Jersey’s house prices rose less rapidly than New York’s. We probably shouldn’t overinterpret six years of data in this time series, but in concert with the evidence from Figure 2 it might suggest that the period of court supervision of Mount Laurel obligations was more productive for housing than either the period of legislative supervision or the period when these obligations were mostly not enforced.

Still, these charts are not conclusive. We can investigate the effects of New Jersey’s affordable housing obligations using synthetic control analysis (Abadie, et al., 2015). Synthetic control creates a weighted average of similar units to the treated unit, then compares the actual results in the treated unit to the counterfactual results represented by the weighted average of similar units. The treatment effect of the intervention equals actual New Jersey’s value minus synthetic New Jersey’s value on an outcome.

I used the immediate lag of the dependent variable, population, personal income, land area, and number of local governments per square mile from Ruger and Sorens (2023) to create a “synthetic” New Jersey from other states. In the change in cost of living and change in house prices analyses, I also used the one-year lag of the level of each index to capture any scale effects occurring in long time series of price indices. This synthetic New Jersey is as similar as possible to the real New Jersey, allowing us to investigate the counterfactual building permits, cost of living, and housing costs that would have occurred in New Jersey in the absence of the Mount Laurel affordable housing obligations.

I look at two treatment periods, the first when the legislature made Mount Laurel obligations effective from 1985 to 1999, and the second when neither the legislature nor the courts were enforcing Mount Laurel obligations (2000 to 2007). Separate synthetic control analyses are run for each period.

Outcome:	Total units	Total units	5-Unit permits	5-Unit permits	Δ Cost of living	Δ Cost of living	Δ House price index	Δ House price index
Treatment period:	1985-1999	2000-2007	1985-1999	2000-2007	1985-1999	2000-2007	1985-1999	2000-2007
Massachusetts	0.66	0.58	0.82	0.22	0.6	0.29	0.44	0.2
Pennsylvania	0.24				0.17	0.17	0.22	0.52
Connecticut		0.13		0.46	0.14	0.38
New York	0.02	0.23	0.17	0.28	0.09	0.16		0.01
Florida	0.09							0.27
Nevada		0.07		0.04
Arizona			0.02
Hawaii							0.25
California							0.1

Table 1: Construction of Synthetic New Jersey in Each Analysis

Because I’m running eight different synthetic control analyses (two treatment periods combined with four outcomes), the precise content of synthetic New Jersey varies from analysis to analysis. Table 1 shows how synthetic New Jersey is constructed in each analysis; the numbers represent the weights on each state.

Unsurprisingly, neighboring states New York and Pennsylvania contribute a lot to many of these analyses. Massachusetts is the only state that contributes to synthetic New Jersey in every analysis. The biggest surprise is to see Hawaii enter as a significant contributor to synthetic New Jersey in the 1985–1999 house price index analysis. Otherwise, unusual states make up only a very small proportion of synthetic New Jersey in each instance.

Predictor balance was largely good, with treated and synthetic New Jersey matching closely on population, personal income, and effective competing jurisdictions per square mile. However, treated New Jersey land area was generally about a half to a quarter of synthetic New Jersey’s. New Jersey is abnormally small for its population and personal income (it is the most densely populated state), and it’s hard for the algorithm to replicate that pattern with other states.

Figure 5 shows the results of the synthetic control analyses of total building permits per capita for the two treatment periods.

Figure 6: Synthetic Control Analysis of New Jersey Five-Unit Building Permits

The results are basically the same. The only statistically significant results based on standardized p-values are in 1986 and 2000, but they go in the “wrong” direction, suggesting that New Jersey had abnormally few five-unit permits in 1986 and abnormally many five-unit permits in 2000.

Figure 7 moves on to general inflation. Here, we would expect legislative enforcement of Mount Laurel to reduce state inflation rates and no enforcement of Mount Laurel to increase them, if the system actually promoted housing supply.

Figure 7: Synthetic Control Analysis of New Jersey Inflation

Here, we are able to model New Jersey’s pretreatment inflation extremely closely and accurately. Once again, we find essentially no effect of the Mount Laurel regime. None of the treatment effects are close to statistical significance, except in 2002, where the result suggests that the lack of Mount Laurel enforcement may have added a tenth of a percentage point to the state’s inflation rate. But for most years, the data suggest that Mount Laurel has made no difference to general inflation in New Jersey.

Next, Figure 8 looks at the results for change in the all-transactions house price index.

Figure 8: Synthetic Control Analysis of Change in New Jersey House Prices

The results here are all over the map, suggesting that Mount Laurel enforcement raised house prices in 1986 and 1987 and cut them in 1989, 1990, and 1991. The average treatment effect over the first period is -1.2, corresponding to about a 0.5 percent change in house prices over the whole 1985–1999 era. After enforcement was removed, house prices rose, but only in two years was that increase statistically significant: 2002 and 2006. The average treatment effect over this period is 11.7, corresponding to about a 2.7 percent change in house prices during that era. There is inconsistent evidence that Mount Laurel enforcement kept house prices down slightly.

When we put all this evidence together, it suggests that Mount Laurel enforcement did not increase housing supply. It may have cut housing prices slightly without increasing permitting, suggesting a reduction in demand for housing, but those changes were too slight to make any impact on broader state-level inflation.

4. Discussion

The results from the foregoing analyses suggest that the Mount Laurel doctrine and multiple rounds of affordable housing obligations dating back at least 39 years have done nothing, or next to nothing, to make housing more abundant or affordable in New Jersey. This is a disappointing result, perhaps, but it’s hard to see it as a surprising one, since New Jersey remains a costly state for housing and relatively slow-growing compared to Sunbelt states that make it easy to build.

Some of the evidence suggests that New Jersey built a lot of housing during 2015 to 2020, and that this period also corresponded to a moderation in house price increases. This was a period of judicial enforcement of Mount Laurel obligations without any legislative framework. At least, the results still support pessimism about the legislature’s ability to come up with criteria for Mount Laurel compliance that foster growth in housing supply.

These results also shouldn’t be surprising since the bulk of the scholarship on the question finds that inclusionary zoning policies tend to make housing less affordable, less abundant, or both. The Fair Housing Act’s mandating of voluntary inclusionary zoning and encouragement of mandatory inclusionary zoning has, at minimum, counteracted the intent of the law to make housing more available to families of modest income.

New Jersey is currently undergoing a fourth round of affordable housing obligations, and the process has generated much controversy. The law, Act 2 of 2024, requires the Department of Community Affairs to develop municipality-specific affordable housing obligations under a detailed, precise formula. Those obligations go into force this year.

Parts of the law’s formula are worth questioning. First of all, “Qualified Urban Aid” municipalities are exempted entirely from the law. This might make sense if all of these municipalities already offered abundant affordable housing, but that is not necessarily the case. Municipalities can qualify for this list if they have a high proportion of substandard and deficient housing or simply a high population density.

Second, Prospective Need – the new affordable housing obligation – is calculated on the assumption that 40% of the housing demand in every region and in every municipality will come from low and moderate income (LMI) households. This ignores the likelihood that LMI households prefer to live in some places rather than others (for instance, places with access to public transit or walkable to employment).

Communities end up with a higher quota if they have had more rapid commercial valuation growth over the previous decade. This provision encourages communities to squelch commercial development. Moreover, commercial valuation growth does not necessarily imply employment growth. Smart-growth principles suggest using employment growth instead to determine housing need.

The formula also includes an income capacity factor, which punishes communities not just for being wealthy, but also for having a small population, because it averages a purely income-based measure with a household-weighted income-based measure. To avoid penalizing communities for having a small number of households, the measure should simply be the household-weighted measure.

The law does take into account land capacity of the municipality, which is supposed to reduce new housing obligation for already built-out municipalities. However, the calculation depends in part on Geographic Information Systems (GIS) land-cover and tax map data that are often not sufficiently granular, accurate, and up-to-date for the purpose.

Finally, the “allocation factors” are simply averaged to produce the fair share calculation, with no justification. For example, the small community of Monmouth Beach has complained about its positive Prospective Need number, given that it has zero developable land.

Given these problems, it is no wonder that the renewed affordable housing obligations are facing strong resistance around the state. Sometimes state governments do need to put guardrails on the municipal zoning power, but in general, mandates on the number of units to produce do not work well, because they encourage wasteful litigation, and municipalities are often able to avoid them through sub rosa methods like delaying permits, requiring extra studies, and zoning land for development that is not actually developable.

5. Conclusion and Recommendations

Is the Mount Laurel doctrine working for New Jersey? Not in its legislative interpretations to date. The legislative framework based in 1985’s Fair Housing Act has demonstrably failed to solve the affordable housing problem in New Jersey. In fact, there’s no evidence it has increased permitting at all, and precious little that it has brought down the cost of housing in the state.

Three recommendations for reform follow.

1. As mentioned, Act 2 exempts urban municipalities from affordable housing mandates. This exemption makes little sense if one’s goal is to provide more affordable housing where there is demand for it. Act 2 could therefore be amended to apply to all municipalities in the state, and the quantitative housing mandates recalculated. Some lawmakers may question this approach on the grounds that higher affordable housing mandates for suburban and rural municipalities will be more likely to reduce residential segregation. But it is a generally valid maxim that when one aims at two targets at once, one risks hitting neither. A better way to deal with segregation would be to address its negative consequences directly.
   
2. A modest reform would shift affordable housing targets from municipal to regional bodies, either the existing county planning boards or new regional authorities created by legislation. These bodies could review development applications specifically for affordable housing developments, in consultation with the municipalities involved. A regional target system would be more flexible than the existing municipal target system, and could allow market demand to have a greater say in where these projects occur.
   
3. New Jersey lawmakers should consider repealing Act 2 entirely and discuss ways to leverage private property rights and the free market to grow housing supply, such as by conferring definite development rights for certain types of projects based on site and infrastructure conditions.
   
4. The state can also speed up permitting with shot-clocks, third-party permitting, and broadened exemptions from special permit requirements. The Mercatus Center publishes an annual report on the state legislative “toolbox” for housing, a useful source of ideas on this front (Furth, et al., 2024).

Instead of detailed affordable housing mandates that have spawned an entire litigation industry, the state should move toward regulatory reforms aimed at general housing abundance. The state could still financially reward towns that actually permit a lot of new building, but those rewards should come after the permits have been issued, not after mere rezonings. This approach would build a more collaborative relationship between the state and municipalities and also make it easier for builders to know what they can build and where, without having to go through years of legal rigmarole. These reforms would make housing more abundant where it’s needed, bringing down the cost and doubling down on New Jersey’s existing economic strengths, like a skilled, urbanized workforce and strong industrial and port infrastructure.

Acknowledgments

The author thanks Audrey Lane for editorial support and guidance and anonymous referees for helpful feedback. All remaining errors are the author’s responsibility.

References

Abadie, A., Diamond, A. & Hainmueller, J., 2015. Comparative Politics and the Synthetic Control Method. American Journal of Political Science, 59(2), pp. 495-510.

Aten, B. H., 2017. Regional Price Parities and Real Regional Income for the United States. Social Indicators Research, 131(1), pp. 123-143.

Bento, A., Lowe, S., Knaap, G.-J. & Chakraborty, A., 2009. Housing Market Effects of Inclusionary Zoning. Cityscape, 11(2), pp. 7-26.

Berry, W. D., Fording, R. C. & Hanson, R. L., 2000. An Annual Cost of Living Index for the American States, 1960-1995. Journal of Politics, 62(2), pp. 550-567.

Furth, S., Hamilton, E. & Gardner, C., 2024. Housing Reform in the States: A Menu of Options for 2025, Arlington, Va.: Mercatus Center at George Mason University.

Hamilton, E., 2021. Inclusionary Zoning and Housing Market Outcomes. Cityscape, 23(1), pp. 161-194.

Hollingshead, A., 2015. When and How Should Cities Implement Inclusionary Housing Policies?, Berkeley, Calif.: Cornerstone Partnership.

Means, T. & Stringham, E. P., 2012. Unintended or Intended Consequences? The Effect of Below-Market Housing Mandates on Housing Markets in California. Journal of Public Finance and Public Choice, 30(1-3), pp. 39-64.

Ruger, W. & Sorens, J., 2023. Freedom in the 50 States: An Index of Personal and Economic Freedom. 7th ed. Washington, D.C.: Cato Institute.

Schuetz, J., Meltzer, R. & Been, V., 2011. Silver Bullet or Trojan Horse? The Effects of Inclusionary Zoning on Local Housing Markets in the United States. Urban Studies, 48(2), pp. 297-329.

---

End Notes

[1] Southern Burlington County N.A.A.C.P. v. Township of Mount Laurel, 67 N.J. 151 (1975).

[2] Monaghan, Justin M., and William Penkethman Jr. “The Fair Housing Act: Meeting the Mount Laurel Obligation with a Statewide Plan.” Seton Hall Legis. J. 9 (1985): 585–619, 587.

[3] Southern Burlington County NAACP v. Mount Laurel Township, 92 N.J.

158, 456 A.2d 390 (1983).

As New York City Mayor-elect Zohran Mamdani prepares to take office, tax-happy progressive groups are eager to let you know that the idea that rich people move because of taxes is all a big myth. There are no consequences to raising taxes on rich people, they argue, because rich people will be rich no matter what. 

It’s a pretty picture, and a convenient one for those who have never met anything economically productive that they didn’t want to tax. The only problem is that the data proves it just isn’t true.

The latest media blitz comes in response to Mamdani’s campaign proposals to raise the income tax rate for top earners in the city from 3.9 percent to 5.9 percent. That’s in addition to statewide rates, which currently run as high as 10.9 percent. That means that, under Mamdani’s proposal, the wealthiest Big Apple residents would face state and local income taxes as high as 16.8 percent, even before federal taxes.

But never fear, say progressive groups such as Patriotic Millionaires — Zohran can tax to his heart’s content without fear of millionaire tax flight. They attempt to fortify their claims with research by the Center for Budget and Policy Priorities and tax-happy academics who make points that are technically true, yet entirely miss the point.

For instance, Patriotic Millionaires cites data showing that the millionaire population in New York grew in the wake of recent tax increases on the wealthy at the state level. But of course it did — the population of millionaires is constantly growing across the country due to economic growth and inflation. The more important thing, as the New York-based Empire Center shows, is that New York’s share of the nationwide millionaire population has dropped precipitously in recent years, from 12.7 percent in 2010 to 8.7 percent in 2022.

Others point to a spike in sales in the New York City luxury real estate market to suggest that “there is no Mamdani effect.” But that actually is an indication of the ongoing exodus, not a rebuttal. The New York City housing market has such a severe shortage of housing that when some wealthy New Yorkers pack up and leave, it’s no surprise that remaining millionaires snap up those luxury properties quickly. It’s no coincidence that inquiries from New Yorkers to the Miami Beach Ritz-Carlton for beachfront penthouses worth $10 million or more nearly tripled in the wake of Mamdani’s election.

Looking at the impact of net migration, the highest-tax states lose big among the wealthy every year. In the most recent IRS data, New York lost the second-most wealthy residents (shocker: California lost the most). On the other hand, Florida gained the most new wealthy residents from other states, followed by Texas.

If pressed further, progressive tax advocates may fall back on another true yet ultimately irrelevant point: that specific tax increases, generally speaking, raise more money than they lose in tax flight. And, indeed, Zohran’s two-percent income tax surcharge would likely leave the city with more revenue in the short term. But the cost comes in the long term, and has been coming for spending-addicted cities and states for some time. 

The National Taxpayers Union Foundation estimates that New York will have $3.8 billion less tax revenue to work with at both the state and local levels in 2025 because of out-migration. New York and New York City are losing that revenue year after year, shrinking the tax base and making future spending binges even harder to finance. 

As the cash cows in the top income brackets leave for greener pastures, there are only two options for politicians who treat the idea of “reining in spending” as an odd foreign custom. One is to increase taxes further on the wealthier New Yorkers who are left, which only exacerbates the problem. The other is to start to shift more and more of that tax burden onto the middle class.  

And guess what? A lot of those wealthy emigrants take their businesses — employers who provide jobs and pay a lot of tax revenue — with them. No state is losing firms to other states faster than New York. 

Even long-time New York City staples are looking elsewhere, as Mamdani’s election has managed to accelerate the already exploding growth of the Dallas counterpart to Wall Street (affectionately known as “Y’all Street”). Big names such as Goldman Sachs and JPMorgan Chase continue to shift more and more of their operations to the Lone Star state, and Texas now boasts more jobs in the financial services sector than New York does.

Progressives should not stick their heads in the sand about the consequences of their policies. Many wealthy New Yorkers will choose to stay after yet another tax hike from Mayor Mamdani, and some of those will stay after the next tax hike as well. But with death by a thousand cuts, it’s the steady bleeding that kills you.

In 1980, at Dartmouth College, psychologists Richard E. Kleck and Angelo G. Strenta set out to study how people perceive subtle social cues. In their own mischievous words, “Individuals were led to believe that they were perceived as physically deviant in the eyes of an interactant.”

Over a series of studies, dozens of volunteers (mostly females) pretended to have a physical ailment or disfigurement — often a facial scar — during an interview. The rub was this: the person pretending to have the disfigurement was the true test subject.

To make volunteers believe they bore a scar, each subject sat down with a professional makeup artist, who carefully applied a facial “injury.” The participants were shown their fake scar in a hand mirror and given a moment to absorb their appearance. But before meeting the stranger, the makeup artist returned under the pretense of “touching up” the scar  —  and quietly removed it. Participants were unaware of the subterfuge and entered their interview convinced their face still bore the disfigurement.

Even though no scar was present, nearly all subjects reported that the strangers they met seemed uneasy by their appearance — avoiding eye contact, speaking awkwardly, or looking at them with pity.

A Distorted ‘Social Reality’

What Kleck and Strenta had uncovered was how easily human expectations about how we’re perceived can color — and distort — our reading of other people’s behavior.

Though the study was groundbreaking, previous studies had treaded similar ground. This included Beatrice A. Wright’s “Physical Disability: A Psychological Approach,” which had found that once a person acquires a physical disability, their perception of social reality becomes filtered through that disability.

While Kleck and Strenta noted that their research differed from Wright’s in key ways, they also saw a clear connection, noting “that persons who are permanently physically deviant make the same kinds of disability-linked attributes to a natural stream of behavior as did the subjects in the present studies.”

The findings of Kleck and Strenta are highly relevant half a century later. In modern America, it’s not uncommon for people to see their group identity as a kind of social handicap. The idea that certain identity groups face discrimination has taken root in the American mind. For example, a May 2025 Pew Research survey asked Americans how much discrimination they think various groups face. Their responses are instructive:

• Illegal Immigrants: 82% say some (57% say “a lot”) 
• Transgender people: 77%
• Muslims: 74%
• Jews: 72%
• Black people: 74%
• Hispanic people: 72%
• Asian people: 66% 
• Legal immigrants: 65% 
• Gay/Lesbian individuals: 70%
• Women: 64%
• Older people: 59%
• Religious people: 57%

Discrimination certainly exists, at both the individual and collective level. No doubt some individuals in all these groups have faced discrimination, just as some individuals have in groups less commonly associated with it, such as rural Americans (41%), young people (40%), white people (38%), and atheists (33%).

Kleck and Strenta’s experiment shows that people often believe they’re being discriminated against because they expect to be, not because they are. Edgier comedies of the 1990s explored this idea.

In one Seinfeld episode, Jerry is on a date and stops to ask a mailman (whose face is obscured) if he knows where a certain Chinese restaurant is. “Excuse me, you must know where the Chinese restaurant is around here,” the comedian says.

Things get awkward, however, when it turns out the mailman is Asian. “Why must I know? Because I’m Chinese?” the mailman says in a heavy accent. “You think I know where all the Chinese restaurants are?”

The scene is comical and a bit absurd, but it reflects a phenomenon Kleck and Strenta observed in their experiments: Once humans begin to feel they are being treated differently based on their physical appearance or inherent attributes, they will believe people are treating them differently even when that is not the case.

“… if we expect others to react negatively to some aspect of our physical appearance,” the authors wrote, “there is probably little those others can do to prevent us from confirming our expectation.”

The Cost of Victimhood Culture

I learned about the Kleck-Strenta as an undergraduate in an introductory psychology class nearly 30 years ago. But I had rarely thought about it since — until Konstantin Kisin, a British-Russian political commentator, brought it up in recent podcast appearances and connected the study to the rise of victimhood culture.

“If you preach to people constantly that we’re all oppressed, that we’re all being discriminated against, then that primes people to look for that,” says Kisin, “even where it doesn’t exist.”

As a concept, victimhood culture wasn’t articulated until the 1990s. But by 2015, the phenomenon was widely discussed in academic literature and mainstream publications, likely due to the rise of political correctness that preceded it. 

“‘Victimhood culture,’ Arthur Brookes wrote in the New York Times a decade ago, “has now been identified as a widening phenomenon by mainstream sociologists.”

Though relatively new in scholarship, the psychological phenomenon itself has been around for ages. In his masterpiece The Brothers Karamazov, Fyodor Dostoevsky explores the human proclivity to manufacture grievance by taking offense. 

“A man who lies to himself is the first to take offense. It sometimes feels very good to take offense, doesn’t it? And surely he knows that no one has offended him, and that he himself has invented the offense and told lies just for the beauty of it, that he has exaggerated for the sake of effect, that he has picked on a word and made a mountain out of a pea — he knows all that, and still he is the first to take offense, he likes feeling offended, it gives him great pleasure, and thus he reaches the point of real hostility.”

These words are uttered by Father Zosima, a wise monk tasked with settling a dispute among the Karamazov family. Fyodor Karamazov, the lecherous patriarch, practical purrs in agreement.

“Precisely, precisely—it feels good to be offended,” he replies. “You put it so well, I’ve never heard it before.”

Dostoevsky was observing a psychological tendency: our ability to see ourselves as victims. (Anyone who has watched The Sopranos has seen this idea explored in brilliant artistic fashion.) 

Combine that impulse with postmodern philosophies that have turned oppression into a kind of fascination, and you get a potent—and corrosive—worldview.

None of this is to deny that real oppression exists or that people are sometimes treated differently because of their appearance. But the research of Kleck and Strenta suggests that, oftentimes, the perceived discrimination exists only in the minds of those who believe they’ve been wronged.

“Regime uncertainty” should be our bywords for understanding the economy of 2025. Trump’s push for “state capitalism,” ranging from tariffs to taking federal stakes in companies to industrial policy to jawboning companies to fire executives to targeted regulatory carveouts, has created a chaotic, pay-to-play environment in which firms find they can get favorable treatment by contributing to Trump’s political success, but the basic rules of the economic game are unpredictable and open to constant negotiation. That unpredictability has in turn deterred private investment and brought on stagflation.

Economist Robert Higgs developed the concept of regime uncertainty to explain why American recovery from the Great Depression was so slow. Investors feared for the security of their contracts and their private property rights as FDR turned explicitly anti-business during the 1936 presidential campaign. As a result, private investment stagnated and the economy tipped back into recession, prolonging the Great Depression. Investor confidence didn’t return until after the war, when it launched an economic boom.

Regime uncertainty was on my mind when I watched the fateful “Liberation Day” press conference in the Rose Garden on April 2, when President Trump held up the schedule of so-called “reciprocal tariffs” that would apply to imports from other countries. The tariff rates themselves were extremely high, but more than that, they were absurd and irrational. They had no logical basis in law or economics. I immediately moved my retirement savings out of US equities into global equities and bought physical gold with all my family’s free cash. After most of the tariffs were rolled back, I gradually started shifting back to a 60-40 US-global equities mix, which is still overweighted to international stocks compared to most Americans’ portfolios.

I didn’t shift to gold or global stocks because I thought the tariffs themselves would be so destructive, but because I lost all confidence in this administration’s economic policy team. I figured we were in for a wild ride, and unfortunately, we have been.

Regime uncertainty explains not just why global stocks have outperformed US stocks (Figure 1), but why gold has performed so well, despite reasonably moderate inflation (Figure 2). Gold is hedging not just ongoing inflation, but policy uncertainty more broadly.

Figure 1: S&P 500 Index vs. S&P Global Index, Jan. 2025 to Dec. 2025 (source: WSJ)

Figure 2: Gold Spot Price, Jan. 2025 to Dec. 2025 (source: TradingView)

Regime uncertainty explains another curious fact about the current US economy: the return of mild stagflation. With private investment lagging, unemployment has risen along with inflation. Macroeconomists would interpret this combination as an “adverse supply shock,” that is, a loss of productivity growth.

As Figure 3 shows, US job growth has slowed markedly since April. Monthly job growth since then has averaged a measly 35,000 workers, compared to over 100,000 every single month going back to October 2024.

Figure 3: US Nonfarm Jobs, Monthly Change, Jan. 2022 to Nov. 2025

The unemployment rate has also now risen a full percentage point since its 2023 low. Note that there is a gap in the series because of the government shutdown: the October figure is missing. The November figure is 4.6%, which means that the unemployment rate has risen a full half a percentage point just since June.

Figure 4: US Unemployment Rate, Jan. 2022 to Nov. 2025

Figure 5 shows monthly personal consumption expenditures (PCE) inflation rates up through November (excluding October). While this series is highly noisy, an upward trend since March is plainly apparent. Moreover, the disinflationary trend that we saw from the beginning of the series up until March has definitively come to an end. For the months through September, we had six straight months of PCE inflation near or above the Fed’s target. The last time that happened was the six months ending in June 2023.

Figure 5: US PCE Inflation, Monthly, Jan. 2022 to Sept. 2025

So far the data suggest the US economy is slowing, investors are hedging against something, and investors prefer to park their money abroad rather than in the US. But the smoking gun that suggests regime uncertainty is at fault is private investment. Real gross private investment is the series that Higgs himself uses to establish a “capital strike” during the late New Deal period.

What do we see when we look at US real gross domestic private investment in 2025? In Q2 2025, the quarter that includes Liberation Day, we see the largest single-quarter decline in US private investment since Q2 2020, the quarter most affected by the global pandemic (Figure 6). To see another decline as large, we have to go all the way back to 2009, during the throes of the Great Recession. Indeed, to find another quarterly decline as large outside the immediate period during or around an official recession, we have to go all the way back to Q1 1988. The just-released figures for Q3 show another decline in real private investment, of 0.3%.

Figure 6: US Real Gross Domestic Private Investment, Quarterly Growth, Q1 2020 to Q3 2025

All of this has been happening at the same time as an AI-fueled boom in capital investment for electricity generation and data centers has taken hold. Had this technologically driven boom not been ongoing, the effects of regime uncertainty on the US economy presumably would have been much worse.

Indeed, if we look at the components of private investment, “information processing equipment,” “software,” and “research and development” have contributed significantly to growth this year. But their growth has been more than offset in the last two quarters by declines in nonresidential structures and residential construction.

Why has regime uncertainty become such a problem? After all, the Trump Administration has made a verbal commitment to deregulation, and the tax cuts passed in the One Big Beautiful Bill should have incentivized business investment. But all of the deregulation enacted by the Trump Administration has come through executive orders and the rulemaking process, not legislation. A future administration could easily undo it. As a result, businesses lack the confidence that large capital investments will pay off in the long run.

The evidence suggests that to turn the American economy around, the Trump Administration needs to work through Congress to pass statutory deregulation, end its experiments with industrial policy, government ownership, and tariffs, and shift from a “deal-making” posture of transactional politics to a firm, credible commitment to enforcing a level playing field for private business. Without a believable shift in strategy, this administration risks incurring an economic malaise that could last for another three years.