The Kenan Institute kicks off 2026 with a rundown of five issues that are top of mind for business leaders and policymakers, including our analysis of recent data and discussion of key questions that define the economic year ahead. This annual exercise is never easy, and estimating 2026 is especially tricky. The challenges to forecasting are many and include major uncertainty about labor markets and inflation, the potential effects of artificial intelligence on short- and long-run economic trends, government shutdown-induced data gaps and resulting informational fog, and the continuing transformation of workplace culture.
We start our examination of the broader economy by sticking our proverbial finger in the air and assessing the prevailing winds. The American consumer powers the US economic engine, so taking stock of how the country’s people are faring is essential to gauging the near-term outlook.
How Miserable Are We?
The so-called misery index, which is the sum of the unemployment and inflation rates, was developed in the 1970s when both measures were high and rising. While not representative of individual household challenges, this index provides a simple, equal-weight measure of the most elementary economic problems affecting businesses, workers and consumers. By this standard, the level of economic misery in the US is generally low, in part because unemployment and inflation are in tension with each other: When one is high, the other is usually low. When both are high, a phenomenon known as stagflation, it creates widespread misery, as wage growth falls behind the rise in prices.
Relative to the miserable economic days of the 1970s and early 1980s, the misery index today looks low at 7.1%. Inflation, however, remains stubbornly elevated and unemployment is creeping up. Tariff-induced goods inflation has offset slowing housing inflation. America’s recent experience, when pandemic disruptions spurred nearly double-digit inflation and the loss of purchasing power, has made people (and policymakers) especially sensitive to even modest increases in prices. In the year ahead, will there be additional price shocks that inflame this combustible environment? Will the inflation increases we saw in 2025 prove transitory and begin to roll back in 2026? Or have firms until now shielded consumers from the full cost burdens brought about by new duties and trade restrictions? If the latter proves true, businesses did not merely pass through price hikes last year but acted as a dam. If tariffs persist or escalate in 2026, the floodgates could open.
Meanwhile, weak job creation has pushed the unemployment rate up. Given the hiring challenges businesses have experienced since the COVID-19 pandemic, firms have been hesitant to lay off workers, a phenomenon known as “labor hoarding.” Over the past year, however, hiring has slowed. It is not certain whether this hiring slowdown is due to cyclical or structural reasons (we discuss the potential impact of AI later), yet we know that labor hoarding imposes inefficiencies on firms that can tax their nimbleness and growth potential. If margin pressures increase or revenue growth slows in 2026, labor hoarding could turn into labor shedding, as businesses scramble to streamline operations amid fierce competition, technological shifts and geopolitical uncertainties.
Is AI Exacerbating Uncertainty?
Fueled by the AI boom, investment in information technology hardware and software was a major driver of GDP growth last year. Based on trends for the first three quarters of 2025, IT investment is estimated to account for 0.4 percentage points (or roughly 20%) of the 2.2% GDP growth in 2025. While much discussion and angst have focused on the data centers and power plants being built to support the technology, their contributions to overall economic growth have been much smaller (an estimated total 0.05 percentage point). As the graph shows, IT-energized growth has a precedent in US economic history. During the 1990s internet boom, IT made similar contributions to GDP growth, although the overall growth rate then was higher than today because of both faster growth in both the labor force and productivity.
A key question for today’s boom is what stage are we now in? An early stage of a multiyear AI-supported period of growth? Or the bubble stage, analogous to the 1999-2000 dot-com bubble, with a subsequent slowdown in investment? Are today’s tech companies overvalued?
The mid-to-late-1990s analogy is also pertinent to productivity. Following a 20-year period of weak productivity growth, much like what we have had up to today, productivity growth surged in the mid-1990s, as decades of technology uptake came together to increase efficiency across the economy, from inventory management to airline reservations. Today, there are high hopes for AI technologies to produce a similar surge in productivity. We are, however, in the early stages of its integration, corresponding more to where computers were in the early 1980s than in the late 1990s. The “J-curve” of change suggests we will see productivity declines as employers and employees figure out how to best use AI. In this early period, firms and investors are making large capital investments in technological infrastructure, and it could be many years before these investments pay off with productivity growth.
As with IT investment’s impact on economic growth, parallel questions apply to productivity: Where are we along the J-curve? How many years will it take to broadly realize productivity gains from AI integration? Does AI’s transformational potential mean this technological boom will skip over — or zoom through — initial losses? Or will this J-curve be especially pronounced, with unprecedentedly large upfront investments and commensurate losses to be expected?
Meanwhile, uncertainty abounds about AI’s effects on jobs. The fear that technology will take jobs away is not new. Yet historically, after a new technology is introduced, its uptake results in some job displacement while ultimately creating new sources of demand and additional jobs that complement the innovation. Who would have imagined the job title of “social media influencer” 20 years ago? Some observers think that this time is different for AI and the job market. I am skeptical, however, that this time is indeed different (I have a very high bar for declaring unprecedented events). Are new graduates having a tough time finding jobs because of AI? The evidence to date is inconclusive, but the implications are huge, not only for entrants but also for firms and for the very nature of work.
What Should the Fed Do?
The cyclical and structural questions laid out above complicate the Federal Reserve’s already difficult job. The Fed cut rates three times in 2025, giving more weight to labor market concerns than inflation concerns. Researchers at both the Federal Reserve Board and the New York Fed suggest that the current policy setting is about neutral. This means, in theory, that the Fed is letting the economy coast on its own, neither stepping on the accelerator (by lowering interest rates) nor braking (by raising interest rates). Yet, as the graph below shows, measuring neutral interest rates is an imprecise science. The gap between these widely cited estimates is between one-quarter and one-half of a percentage point, but it has been much wider in the past.
The Fed’s outlook depends on whether the sticky inflation and rising unemployment we have recently experienced is a temporary phenomenon or the result of structural factors. Here are four potential scenarios based on the questions posed above about unemployment, inflation and AI:
- Short-lived inflation: The Fed should focus on employment. This approach reflects the Fed’s current consensus, yet the Federal Open Market Committee voter dissents and significant variation in interest-rate projections indicate that the view is not unanimous.
- Stubborn inflation: Expect some misery as the Fed will need to inflict labor market pain. Whether this means rate hikes or holding rates at neutral would depend on the labor market/inflation nexus. If inflation does not increase much and the labor market remains weak, the Fed may hold interest rates at or near neutral and let the forces of higher unemployment bring down inflation. Alternatively, the risk is that inflation accelerates, which could cause inflation expectations to become unanchored. This scenario occurred in the late 1970s into the early 1980s, when the Fed’s only recourse was to cause a deep recession to bring down inflation in emphatic fashion.
- AI raises productivity: This development would mean a favorable growth and inflation scenario, as workers produce more with less. This outcome, however, implies higher neutral interest rates as the return on capital has risen. Counterintuitively, the Fed might have to raise rates to keep the economy coasting rather than unnecessarily applying the accelerator. This scenario is what happened in the mid-to-late 1990s.
- AI gains result in higher structural unemployment: This case would also lead to counterintuitive policy, as the Fed would need to accommodate higher unemployment.If AI causes mass unemployment, cutting rates would not create jobs but instead raise inflation. This scenario likely would occur in concert with higher productivity, which would mean the Fed should raiserates. In this case, fiscal policy would need to play an active role in retraining workers and providing income support.
What Challenges Do Local Economies Face?
Our American Growth Project aims to measure and understand the varied drivers of economic growth across metro areas in the United States. The official measures of local GDP growth are released with a one-year lag, however, imposing a major challenge that we must work around to build out meaningful economic indicators. Complicating matters, the 2024 data is delayed until February 2026 due to the recent government shutdown. Worse still, government statisticians will no longer produce metro area GDP statistics.
The good news is that we have solutions for these data impediments. Using innovative data aggregation techniques, we have built sophisticated econometric models to estimate GDP growth at the local level. This month, we will release our updated 2024 and 2025 estimates of GDP growth for the 150 largest Extended Metropolitan Areas as well as a forecast for 2026. We anticipate slower growth in all our EMAs, with 18 shrinking.
Local economies experience economic challenges differently than states and nations do and distinctly from one another. These distinctions stem from differences in industry mix, demographics and other locally specific structural factors. For instance, many of the country’s fastest-growing EMAs have large and growing tech sectors. Questions about AI and its cyclical impacts therefore have outsized weight for these microeconomies. Similarly, government funding cuts for science will reverberate through these areas more than in economies that are less reliant on basic research. Historically, government research and development spending has buffered local economies from cyclical forces. To what extent can private sector spending offset government R&D cuts? How will this affect long-term growth prospects?
One structural factor affecting every area is housing. Everyone needs an affordable place to live, and housing affordability is at the forefront for all EMAs, especially as metros compete for talent. While overall inflation estimates include a measure of housing inflation, these metrics are tied to rents rather than home prices. This means that the misery index includes the cost of owning a home but not buying a new home. Meanwhile, home prices for the US as a whole have appreciated roughly 30% more than rental costs in the past decade. Targeting this data gap, we have started a research program focused on housing affordability at the local level. As we build out our housing affordability model this year, look out for our results and discussions on local affordability measures in our 150 EMAs.
How Will Workplace Demands Be Balanced?
The COVID-19 pandemic disrupted workplace culture, indelibly changing the unwritten rules binding workers and firms. The expectations of mutual commitment that once anchored work have frayed. In 2026, through our annual Grand Challenge, we will explore this phenomenon, using three key framing questions to guide a yearlong investigation:
- What is the future of work arrangements, and how can organizations balance flexibility with performance expectations?
- What does leadership look like in this new era of work, and how is the next generation of leaders being developed?
- Is work more transactional these days, and is this good or bad for firms?
The workplace culture of the future will be defined by a host of macro-factors, many of which we discuss as part of our other trends. For example, if labor hoarding becomes labor shedding and AI uptake upends the job market, will workers’ preferences lose out? If so, in the near term, the future of work may resemble the past more than it does the present. In the longer term, demographic and immigration factors will make it harder to find workers, likely forcing employers to be more flexible.
We will explore this topic in depth throughout the 2026 Grand Challenge, convening business leaders and academics to contribute to our investigation. Visit the Kenan Institute website to keep up with Grand Challenge research and events, and how this year’s decisions will determine the future of work.
