Publication

The U.S. economy in 2026: What to watch for

Key takeaways

The U.S. economy has shown remarkable resilience in the face of policy uncertainty and AI’s potential disruption.
Interest rate decisions, the national debt, tariffs, worries about a stock market bubble and the low-hire, low-fire labor market are among the big economic issues that will take center stage in 2026.
Affordability will continue to be a top concern for consumers leading up to the November midterm elections.

It’s a strange time for the U.S. economy. Last year, overall economic growth came in at a solid pace, fueled by consumer spending, rising real wages and a buoyant stock market. The underlying environment, however, was fraught with uncertainty, characterized by a new and sweeping tariff regime, a deteriorating budget trajectory, consumer anxiety around cost-of-living, and concerns about an artificial intelligence bubble.

Looking ahead to 2026, we anticipate that overall growth will continue apace, even as the job market has clearly slowed. We expect this year to bring increased focus on the Federal Reserve’s interest rates decisions, the weakening job market and AI’s impact on it, valuations of AI-related firms, affordability challenges (such as health care and electricity prices), and the country’s limited fiscal space.

In this policy brief, we dive into each of these issues, examining how they may impact the broader economy in the year ahead.

The Federal Reserve’s stagflation challenge

When it comes to the Fed’s forthcoming interest rate decisions, we expect this year to be a continuation of the unusual tensions of late 2025. The Fed has a dual mandate to pursue stable prices and maximum employment. In normal times, these two objectives are roughly correlated. An “overheated” economy typically presents strong labor demand and upward inflationary pressures, prompting the Federal Open Market Committee (FOMC) to raise interest rates and cool the economy. Vice versa in a slack economic environment.

But if the labor market continues to weaken and inflation remains above the Fed’s 2 percent target as President Trump’s tariffs trickle down to consumers, the Fed’s job in the coming year could become more complicated. The big concern is stagflation, a rare condition where inflation and unemployment both run high. Once it starts, stagflation can be hard to reverse. That’s because aggressive moves in response to spiking inflation can drive up unemployment and stifle economic growth, while lowering rates to boost economic growth risks driving up prices.

The stagflation risk is real. Towards the end of last year, the weakening job market said “cut,” while the tariff-induced price pressures said “hold.” In both speeches and votes on monetary policy, differences within the FOMC were on full display (three voting members dissented in mid-December, the most since September 2019). Most members clearly weighted the risks to the labor market more heavily than those of inflation, including Fed Chair Jerome Powell, though he did so while chanting the mantra that “there is no risk-free path for policy.” [1]

To be clear, in our view, recent divisions are understandable given the balance of risks and do not signal any underlying problems with the committee. The inherent ambiguity in the monetary-policy path out of stagflation was amplified by the disruption to the dataflow from the October government shutdown. We will not speculate on when and how much the Fed will cut rates next year, though market expectations are for two 25-basis-point cuts. We do expect that in the second half of the year, the data will provide more clarity as to which side of the stagflation dilemma, and therefore, which side of the Fed’s dual mandate, requires more attention.

Another key 2026 Fed development comes in May of this year, when Powell's term expires. Trump has aggressively attacked Powell and the independence of the Fed, stating unequivocally that his nominee will need to enact his agenda of sharply lowering interest rates. It is important to emphasize two factors that could influence these outcomes. First, even if the new Fed chair does the president's bidding, he or she will be but one of 12 voting members. Second, though Powell's term as chair is up in May, his 14-year term as a Fed governor runs until 2028 so it’s possible he could choose to remain. While very few former chairs have availed themselves of that option, Powell has made it clear that he views the Fed's political independence as paramount to the effectiveness of the institution, and in our view, recent events raise the odds that he'll stay on the board.

Trump’s tariff choice: Stay the course or retreat

One of the most consequential developments of 2025 was Trump’s sweeping new tariff regime. Acting through executive authority — much of which is currently under review by the U.S. Supreme Court — the president increased the effective tariff rate implied from customs duties from 2.1 percent to an estimated 11.7 percent as of January 2026.

Tariffs are taxes on imports and are formally paid by importing firms, but their economic incidence — who ultimately bears the cost — is more complex and can be shared across exporters, wholesalers, retailers and consumers. The best evidence to date suggests that pass-through to consumers now exceeds 50 percent.[2] While this pass-through has been slower and less complete than the near-100 percent pass-through observed under the tariffs enacted during the first Trump administration, it still represents a meaningful burden on households. Consistent with these estimates, Goldman Sachs projects that the current tariff regime will raise inflation by 1 percent between the second half of 2025 and the first half of 2026 relative to its counterfactual path.

While narrowly targeted tariffs can be a useful tool to push back on unfair trading practices, sweeping tariffs do more harm than good. In addition to raising consumer prices, they raise production costs for domestic producers. Since roughly half of our imports are inputs into domestic production, they also undermine the administration’s goal of reversing the decline in manufacturing employment, which continued last year, with the sector dropping 68,000 jobs.

Despite denying any negative impacts, the administration may soon be offered an off-ramp from its tariff regime. The Supreme Court is expected to rule shortly on whether the administration had the legal authority to impose the majority of its tariffs, and the consensus is that the court will rule against the administration (betting markets currently put those odds at around 75 percent). Given the tariffs’ contribution to business uncertainty and higher costs at a time when Americans are concerned about affordability, the administration could use a negative SCOTUS decision as cover for a wholesale tariff rollback.

However, we suspect the administration will not take this path. There have been multiple junctures where the administration could have reversed course on tariffs. While it retreated from the most extreme “Liberation Day” tariffs and rolled back duties on salient foodstuffs like coffee and bananas following the party’s poor performance in the November 2025 elections, those moves appear calibrated. With reports that the administration is preparing backup options, we do not expect an about-face on tariff policy in 2026.

Moreover, as 2026 begins, the administration continues to use tariffs to gain leverage in international disputes, most recently through threats of a new 10 percent tariff on several European countries in connection with negotiations over Greenland. While these specific tariffs may not ultimately be imposed, the broader pattern suggests that tariff threats will remain a recurring feature of Trump's policy agenda this year.

AI’s workforce impacts: A mixed bag

In remarks last year, AI executives built up 2025 as an inflection point, with OpenAI CEO Sam Altman predicting AI agents would “join the workforce” and materially change the output of companies,[3] and Anthropic CEO Dario Amodei forecasting that AI would be able to match the capabilities of a PhD student or an early career professional within the year.[4]

Looking back, these predictions were directionally right: Firms did begin to deploy AI agents and notable advancements in AI models were achieved. However, 2025 also delivered a reality check. Agents can make costly errors, requiring careful risk management.[5] Many generative AI pilots remained experimental, with only a small share moving to enterprise deployment.[6] And the pace of business AI adoption, which accelerated throughout 2024, stagnated.[7]

Figure 1: AI usage by firm size 2024-2025. 4-week rolling average

Source: U.S. Census Bureau, Business Trends and Outlook Survey. Share of firms responding "yes" to the question "In the last two weeks, did this business use artificial intelligence (AI) in producing goods or services?"

Last year also saw the emergence of new evidence on AI’s labor market impacts. Taken together, this research finds little indication that AI has impacted aggregate U.S. labor market conditions so far.[8] Although unemployment has increased, it has risen most among workers in occupations with the least AI exposure, suggesting that other factors are at play. That said, small pockets of disruption from AI may also exist, including among young workers in AI-exposed occupations, such as customer service and computer programming.[9]

The limited impact of AI on the labor market to date should not be surprising. Historically, firms take many years to integrate new technologies into their existing business operations. For example, in 1900, 5 percent of installed mechanical power was provided by industrial electric motors. It took 30 years to reach 80 percent adoption. Considering this timeline, we should temper expectations regarding how much we will learn about AI’s full labor market impacts in 2026. Still, given substantial investments in AI technology, we anticipate that the topic will remain of central interest this year.

Labor market: ‘Meaningful’ downside risks

Last year was a tough one for job seekers. Job openings fell, hiring was sluggish and employment growth slowed to a crawl. Indeed, Fed Chair Jerome Powell stated recently that he believes payroll employment growth has been overstated and that revised data will show the U.S. has been losing jobs since April. The slowdown in job growth is due in part to a sharp decline in immigration, but that was not the only factor. Unemployment rose from 4.1 percent to 4.4 percent in 2025, indicating labor demand weakened by more than labor supply.

At the same time, for those with a job, the labor market was fairly strong. Wage growth, although softening, still outpaced inflation by around 1 percent, and layoffs remained historically low. Taken together, the labor market has settled into a low-hire, low-fire equilibrium.

Looking ahead to 2026, most forecasters expect modest job growth and a stable unemployment rate at around its current level. Some forecasters have the labor market picking up in the latter half of the year as stimulus from tax cuts and easing monetary policy takes effect.

Even so, meaningful downside risks remain. The recent rise in unemployment, which most forecasts assume will stabilize, may continue. AI, which has had minimal impact on labor demand so far, could begin to weigh on hiring. More subtly, optimism about AI could act as a drag on the labor market if it gives CEOs greater confidence — or cover — to reduce headcount.

Figure 2. Change in employment 2025, by industry

Health care costs back in the spotlight

Health care costs moved to the center of the political debate in the second half of 2025. The issue first surfaced during summer negotiations over the budget bill, when Republicans declined to extend enhanced Affordable Care Act (ACA) exchange subsidies, despite warnings from vulnerable members of their caucus. Concerns about costs returned in the fall, when Democrats made reinstating ACA subsidies their primary demand during a 43-day government shutdown. Although Democrats failed, many observers argued that they benefited politically by elevating health care costs, a top issue on which voters trust Democrats more than Republicans.

The policy consequences are now becoming tangible. As a result of the decrease in subsidies, an estimated 20 million Americans are seeing their insurance premiums roughly double starting this January. Many political observers will remember how, in the 2018 midterms, Democrats capitalized on the backlash to Republicans' failed attempt to repeal the ACA to pick up 40 House seats. With health care costs top of mind, both parties are likely to press competing visions for health care reform. Democrats will likely emphasize restoring ACA subsidies and rolling back Medicaid cuts, while Republicans are expected to tout premium support, expanded Health Savings Accounts, and related proposals that emphasize consumer choice but shift more financial responsibility onto households.

Figure 3. Percent change in gross and net ACA premium payments, 2026

A ‘dangerous’ fiscal path

While tax cuts from the budget bill are expected to support growth in the first half of this year — through refund checks driven by withholding changes — rising deficits and debt pose growing risks for two reasons.

First, in recent years, the historical relationship between economic growth and the federal deficit has weakened. Previously, when the economy reached full capacity, the deficit as a share of gross domestic product (GDP) typically improved. In the last two expansions, however, deficits failed to narrow even as unemployment fell, with relatively high deficit-to-GDP ratios occurring alongside low unemployment.

Figure 4: Federal deficit or surplus as percentage of GDP

Source: Office of Management and Budget. Shaded areas indicate recessions.

Table 1: U.S. fiscal and labor market outlook (2023-2026)

Year	Budget deficit (% of GDP)	Unemployment (%)
2023	-6.2	3.6
2024	-6.3	3.9
2025	-6.0	4.2
2026 (projected)	-5.5	4.5

Notes: Data are reported on for the fiscal-year. For FY2026, the deficit-to-GDP ratio reflects projections from the Congressional Budget Office, and the unemployment rate reflects projections from Goldman Sachs.

Second, as Bernstein et al. wrote in a SIEPR Policy Brief,[10] the U.S. “budget math” has become more dangerous. For many years, even as federal debt increased, interest rates remained below the economy’s growth rate, keeping debt service costs stable. Today, interest rates and growth rates are now much closer. While no one can forecast the path of interest rates, most projections suggest they will remain elevated. If so, debt servicing will become a heavier lift, increasingly crowding out more public spending and private investment.

To be clear, we view the risk of a “Liz Truss”-style episode in the U.S. — where global creditors would abruptly pull back — as very low. But fiscal risk lies on a continuum between a sudden stop and complete disregard of the fiscal trajectory. We are already seeing higher risk and term premia in U.S. Treasury yields, complicating our “budget math” going forward.

The ongoing AI bubble debate

A core question for financial market participants is whether the stock market is experiencing an AI bubble. As two of us (Bernstein and Cummings) have written,[11] there are certainly bubbly features: Valuations of AI-exposed firms have risen sharply even as revenue from AI-specific products and services remains limited. As the figure below shows, the market-cap-weighted index of the "Magnificent Seven” — firms heavily invested in and exposed to AI — has substantially outperformed the rest of the S&P 500 since ChatGPT’s November 2022 release.

Figure 5: S&P 493 vs. Mag 7 since ChatGPT launch
Index (Nov 30, 2022 = 100)

Source: Bloomberg Finance, L.P.
Note: Indices are market-cap weighted. S&P 493 uses all firms in the S&P 500 ex Mag 7 at each given date, which does not always equal 493 firms exactly.

At the same time, some analysts contend that today’s valuations may be justified. For example, Joseph Briggs of Goldman Sachs estimates[12] that generative AI could create $8 trillion of value for U.S. firms through labor productivity gains. If productivity gains of this magnitude are realized, current valuations may prove conservative.

Ultimately, whether current valuations represent a bubble will depend on the scale and speed of AI adoption and on firms’ ability to monetize those AI investments. If 2026 features a notable move towards greater AI adoption and profitability, then current valuations will be perceived as better aligned with fundamentals. For now, however, less favorable outcomes remain possible.

For the real economy, one way the possibility of a bubble matters is through the wealth effects of changing stock prices. As Bernstein and Cummings estimated in the piece referenced above, wealth effects are currently contributing $100 billion (or 0.4 percent) to GDP growth. A market correction driven by AI concerns could reverse this, putting a damper on economic performance this year.

Affordability: On voters’ minds

One of the dominant economic policy issues of 2025 was, and continues to be, affordability. While the term is imprecise, it has come to refer to a set of policies aimed at addressing Americans’ deep dissatisfaction with the cost of living — particularly for housing, health care, child care, utilities and groceries.

Part of the intellectual foundation for this affordability agenda was the release of Abundance, by Ezra Klein and Derek Thompson, which argues that expanding the supply of many essential goods and services has become unnecessarily costly and burdensome. The book highlights what various SIEPR scholars have termed “procedural sludge”[13]: federal and sub-federal rules that constrain supply expansion with limited regulatory justification, such as permitting requirements that function more to block construction than to address genuine problems. A central aim of the affordability agenda is to remove these outdated constraints.

These issues resonated politically in 2025, with a number of candidates campaigning successfully on affordability concerns and pledging to continue pursuing related reforms. The central question now is whether policymakers will be able to enact legislation that meaningfully advances this agenda — and, if so, whether such policies will reduce costs or at least slow the pace of cost growth. If they don’t, expect more political fallout in the November midterm elections.

Electricity prices: A complex problem

Since the pandemic, consumers across much of the U.S. have experienced rapid increases in electricity prices, with most states seeing electricity costs rise well above overall inflation. California, in particular, has seen electricity prices nearly double.

Figure 6: Percent change in real residential electricity prices 2019–2025

**Source:** EIA, BLS and authors' calculations

While energy-hungry AI data centers often draw criticism for rising electricity prices, the underlying causes are interrelated and multifaceted. Analysis suggests that higher wholesale power costs, investment to replace aging grid infrastructure, extreme weather events, state policies such as net-metered solar and renewable energy standards, and rising demand from data centers and electric vehicles have all contributed to higher prices.[14]

In response, policymakers are exploring solutions to ease the burden of higher prices. In New Jersey, Governor Mikie Sherrill has proposed freezing utility costs. Implementing such a policy will be challenging, however, because a large share of households' electricity costs is passed through by the Independent System Operator, which serves multiple states. Other approaches — such as expanding electricity generation and increasing the capacity and efficiency of the existing grid[15] — could help over time, but are unlikely to deliver near-term relief.

Conclusion

The U.S. economy has continued to show remarkable resilience in the face of increased policy uncertainty and the potentially disruptive force of AI. How well consumers, businesses and policymakers continue to navigate this uncertainty will be decisive for the economy’s overall performance. Here, we have highlighted economic and policy issues we think will take center stage in 2026, although few of them are likely to be resolved within the next year.

About the Authors

Jared Bernstein is a Distinguished Policy Fellow at SIEPR. He was Chair of the White Council of Economic Advisers under President Biden and also served as Chief Economist to Vice President Biden in the Obama administration.

Ryan Cummings is Chief of Staff at SIEPR. He is a former Staff Economist in the White House Council of Economic Advisers.

Neale Mahoney is the Trione Director of SIEPR and the TG Wijaya Professor of Economics at Stanford University. He previously served as a Special Policy Advisor in the White House National Economic Council.

Erika McEntarfer is a Research Scholar and a Distinguished Policy Fellow at SIEPR. She served as Commissioner of the Bureau of Labor Statistics from 2024 to 2025 and previously held roles in the U.S. Census Bureau, the U.S. Department of the Treasury and the White House Council of Economic Advisers.

The Stanford Institute for Economic Policy Research (SIEPR) catalyzes and promotes evidence-based knowledge about pressing economic issues, leading to better-informed policy solutions for generations to come. We are a nonpartisan research institute, and SIEPR policy briefs reflect the views and ideas of the authors only.