Blog Entry

By Cooper Howes, Marc Dordal I Carreras, Olivier Coibion and Yuriy Gorodnichenko

When the Federal Reserve changes interest rates, the decision is usually announced as if it were the product of a unified voice. The public sees a single policy rate, a carefully worded statement, and perhaps a press conference that generally emphasizes consensus. Yet behind this appearance of unity lies a much messier reality. Monetary policy in the United States is made by committee, and that committee is composed of individuals who often disagree—sometimes sharply—about what should be done and why.

This essay explains how those disagreements arise, how they are resolved, and why leadership matters so much in the process, based on a new working paper that undertakes a large-scale data construction effort (Howes et al. 2025): reading every FOMC meeting transcript over several decades and systematically quantifying what policymakers said, what policies they preferred, and how they justified those preferences. Crucially, these discussions took place in an era when participants spoke candidly behind closed doors, with no expectation that their words would ever become public. It therefore offers a window into the hidden mechanics of monetary policymaking that remain highly relevant today, as disagreements among policymakers have become increasingly visible and the Federal Reserve approaches a major leadership transition.

A committee, not a machine

The FOMC is responsible for setting U.S. monetary policy, primarily by deciding the level of short-term interest rates. Its members include the Chair of the Federal Reserve, other members of the Board of Governors in Washington, and presidents of regional Federal Reserve Banks. All of them receive the same briefing materials before each meeting, including detailed forecasts of inflation, growth, and labor market conditions prepared by the Federal Reserve staff.

One might expect that, armed with the same data and the same mandate to promote stable prices and maximum employment, policymakers would usually reach similar conclusions. What the new evidence shows, however, is that even under these shared conditions, disagreement is widespread, persistent, and quantifiable. Even when economic forecasts are broadly shared, policymakers frequently enter meetings with very different views about the appropriate policy response. Some favor tightening policy, others prefer easing, and many argue for holding steady—often for reasons that go well beyond simple differences in the economic outlook. In many meetings, we see policymakers propose two, three and sometimes more different policy decisions. 

What makes this especially striking is that these disagreements are often invisible from the outside. Formal dissents—votes against the committee’s final decision—are rare. Yet the absence of dissent should not be mistaken for the absence of disagreement.

Figure 1: FOMC members systematically disagree about policy but rarely dissent.

Notes: This plots several measures of participant disagreement. The solid black line reports the total number of dissents per meeting. The dotted blue line plots the number of distinct Bluebook policy options (between A/B/C) preferred. The dashed red line plots the number of distinct target FFR preferences. All series are smoothed using a 13-meeting centered moving average. Shaded areas show NBER recessions.

Why policymakers disagree

If disagreements are not mainly driven by different forecasts, what explains them? The newly constructed data allow this question to be answered directly rather than inferred indirectly from votes or public statements. The historical record points to a central factor: differences in beliefs about how the economy works.

In particular, policymakers have long disagreed about the tradeoff between inflation and real economic activity. Some believe that changes in interest rates mainly affect inflation, with relatively modest effects on output and employment. Others believe the opposite: that monetary policy has powerful effects on real activity but only limited or delayed effects on inflation. These beliefs matter enormously. A policymaker who thinks higher interest rates will quickly and reliably reduce inflation may be eager to tighten policy at the first sign of price pressures. Another who believes the same action would mostly slow growth and raise unemployment may be far more cautious.

These differing views show up clearly in how policymakers justify their preferred actions. By coding and quantifying thousands of such justifications across meetings and individuals, the working paper shows that disagreement originates primarily in differences in perceived economic tradeoffs, not in differences in stated objectives or access to information. Those who emphasize inflationary risks tend to argue for tighter policy. Those who focus on employment, growth, or slack in the economy tend to favor easier policy. Importantly, these are not just rhetorical differences. They reflect genuinely different mental models of the economy—different assumptions about how strongly policy transmits to prices versus output.

Figure 2: FOMC members tend to think monetary policy affects inflation more than output when the economy is “hot” and the opposite when there is economic slack.

Notes: The black line (left axis) shows the average perceived policy tradeoff, which represents participant beliefs about whether the incidence of monetary policy will operate through prices or real activity. This variable takes on values between 1 and 5, with higher values corresponding to larger effects on prices. We take averages across all participants who express them in each meeting and then smoothed using a 13-meeting centered moving average. The blue line (right axis) shows the inverted unemployment rate. Shaded areas show NBER recessions.

Such differences persist over time and across individuals. They are not easily reduced to simple labels like “hawk” or “dove.” Once policymakers’ underlying beliefs about economic tradeoffs are taken into account, traditional distinctions between hawkish and dovish behavior become much less sharp. What looks like a difference in tolerance for inflation often turns out to be a difference in beliefs about how costly it would be, in terms of jobs and output, to bring inflation down.

Disagreement without dissent

Given how pervasive these disagreements are, a natural question arises: why does the FOMC so often present a united front?

The answer lies in the internal dynamics of the committee. While all members participate in discussions, not all voices carry equal weight when it comes time to decide. The Chair of the Federal Reserve plays a uniquely powerful role. Historically, the final policy decision has tracked the Chair’s preferred course of action very closely—much more closely than it tracks the preferences of any other individual member. The new evidence allows this influence to be quantified: the Chair’s preferred policy change passes through almost one-for-one into the final decision, while the preferences of other members translate only weakly into outcomes.

This dominance does not mean that debate is meaningless. Discussions shape how options are framed, how risks are assessed, and how policy is communicated. But when disagreements remain unresolved, the Chair’s view typically prevails.

Why, then, do other members so rarely dissent publicly? One reason is that dissent appears to be costly. Using the newly assembled data, the working paper shows that members who dissent in one meeting subsequently see a measurable decline in how much their preferred policies influence future committee decisions. Their views are less likely to be reflected in future decisions, even when controlling for the substance of their arguments. In effect, dissent can reduce a policymaker’s ability to shape outcomes going forward.

This implicit penalty helps explain why dissent is uncommon even when disagreements are large. It also helps explain why dissent rates vary across leadership regimes. Some Chairs have been more tolerant of open disagreement than others, while some have enforced consensus more tightly.

Why leadership transitions matter

These dynamics have especially important implications for periods of leadership change. As Chair Powell’s term approaches its end, disagreements among FOMC members have become more public than at almost any point in recent decades, with policymakers openly expressing divergent views about inflation risks, labor market slack, and the appropriate path of interest rates. When a Chair’s term is nearing its end, or when it is widely understood that a new Chair will soon take office, the mechanisms that normally suppress dissent may weaken. Members may feel less constrained, disagreements may surface more openly, and the committee may appear more divided.

Leadership transitions also matter because Chairs are not interchangeable. A change at the top does not simply reshuffle personalities; it can alter how disagreements are handled, how much dissent is tolerated, and whose views ultimately shape policy. Each brings their own beliefs about the economy, their own approach to managing disagreement, and their own willingness to tolerate dissent. A new Chair who holds different views about inflation–output tradeoffs, or who places different weight on consensus versus open debate, can meaningfully change how policy decisions are made—even if the formal mandate and institutional structure remain the same.

This helps explain why periods of heightened uncertainty about future leadership are often accompanied by more visible disagreement within the committee. The issue is not just who will set policy tomorrow, but how disagreements will be managed along the way.

Lessons for today

Although the evidence discussed here comes from historical records, the lessons are not confined to the past. Disagreement within monetary policy committees is not a flaw; it is a natural consequence of uncertainty about the economy and the limits of economic knowledge. What matters is how those disagreements are processed and resolved.

Three lessons stand out. First, differences in beliefs—not just differences in data—are central to understanding monetary policy debates. Second, formal unity can mask substantial underlying disagreement. Third, leadership plays a decisive role, not only in setting policy, but in shaping how much disagreement is expressed and how costly it is to voice it.

As monetary policy confronts new challenges, from persistent inflation risks to shifting labor market dynamics, these institutional features will remain as important as any economic model. Understanding them helps make sense of why policy decisions sometimes seem contentious, why consensus can fray, and why the choice of leadership matters so much.

References:

Howes, Cooper, Marc Dordal I Carreras, Olivier Coibion and Yuriy Gorodnichenko, 2025. “How Monetary Policy Is Made: Lessons from Historical FOMC Discussions,” Manuscript.

By Christoph E. Boehm

What are the effects of the recent increase in Chinese import tariffs on exports from Texas? How do these tariffs impact workers in cities like Austin, Dallas, or Houston? And is there hope that the goods once destined for Chinese markets can instead be sold to other foreign trading partners with whom Texan firms already maintain established relationships? 

These are the kinds of questions that, until recently, were challenging to answer with the U.S. trade data available. The problem lies in the fact that data on U.S. exports and imports is collected at the firm level. Because most exporting firms are large and often operate plants across multiple regions of the country, it is difficult to determine where the production of exported goods takes place and where the workers involved are employed. This lack of geographic detail leaves policymakers in the dark about where the benefits of trade accrue and where the risks associated with fluctuations in foreign demand are concentrated. 

Recent work by Boehm, Flaaen, Pandalai-Nayar, and Schlupp aims to address this shortcoming. The researchers assembled a new dataset that makes it possible to accurately answer questions like those raised above. The idea behind this data resource—available with approved proposals through Census Research Data Centres—is to combine existing firm-level trade data with additional information on where firms operate plants, which products those plants produce, as well as related details, thereby allowing trade transactions to be linked to individual plants within each firm where the product being exported was likely produced. The result is a new dataset that reports trade transactions—including export values, product codes, destination countries, quantity measures, and more—at the plant level for the first time. 

This dataset, called the PLEXT data (for plant-level export transactions), also enables measurement of exports at various geographic levels, since plants have address information. For example, one can aggregate exports from all plants located in a given ZIP code, county, commuting zone, or state. As a result, the dataset can be used to answer questions of the sort raised above, for instance, how workers in Austin, Dallas, or Houston were affected by tariff changes. 

The authors also examine the properties of the new data. A striking finding they document is that exports in the United States are extremely concentrated geographically (see Figure 1, Figure 5 in the paper). In Texas, for example, more than 96 percent of exports in 2021 came from just 10 percent of counties. This pattern contrasts with the distribution of manufacturing sales and employment, which are also geographically concentrated but to a lesser degree. In Texas, the top 10 percent of counties account for roughly 85 percent of manufacturing sales and about 82 percent of manufacturing employment. 

Figure 1: Variation in concentration of employment, sales and exports of establishments for selected states 

This high degree of concentration implies that certain areas of the United States benefit enormously from access to foreign markets. For example, exports per worker in the combined metropolitan areas of El Paso, TX, and Las Cruces, NM, exceeded 35,000 dollars to Mexico alone—an indication of the substantial gains these regions derive from access to the Mexican market. While geographic proximity helps explain this particular case, exports per worker to other countries are also significant in various parts of the U.S. The combined metro areas of Beaumont–Port Arthur, TX, and Lake Charles, LA, for instance, exported more than 14,000 dollars per worker to China. These regions therefore benefit greatly from access to global markets, but the flip side is that they are also highly exposed to increases in foreign tariffs (see Figure 2, which is Figure 3 in the paper). 

Figure 2: Exposure of select Metropolitan Statistical Areas to China, Mexico and Canada through exports 

To illustrate one concrete application of the PLEXT data, the authors analyze the connection between local area exporting and employment during a specific episode of large swings in foreign demand and trade flows.  During the period of the Great Trade Collapse, which took place between 2007 and 2009, U.S. imports and exports fell by roughly 20 percent—a decline far larger than that of overall economic activity. The authors show that U.S. counties more exposed to foreign demand shocks experienced larger reductions in employment, payroll, and wages. This result quantifies how the positive local benefits of access to foreign markets are coupled with vulnerability to swings in foreign demand. The authors also demonstrate that common methods of attempting to infer local exports without detailed data such as the PLEXT data—for instance, by using product-level export data combined with local industry employment shares—cannot accurately reproduce directly measured local-area exports. 

The availability of the PLEXT data opens new avenues for analyzing trade patterns and their role in the U.S. economy. The researchers expect the dataset to be especially useful for research questions that require export data at the local level. How much do individual cities or regions benefit from access to foreign markets? How vulnerable are they to policies such as foreign tariffs? Are local export-promotion programs effective, and if so, how do they work? In addition, the PLEXT data substantially improves the accuracy of plant-level analyses. Researchers can now more precisely examine how many plants export (Answer: More than previously thought, see Boehm, Flaaen, and Pandalai-Nayar, 2023) and how their exporting relationships with foreign partner countries evolve over time. The authors look forward to future research that uses this dataset to answer these and many other important questions. 

References: 

Christoph Boehm, Aaron Flaaen, Nitya Pandalai-Nayar, and Jan Schlupp, “The Local-Area Incidence of Exporting,” NBER Working Paper 34508 (2025) 

Boehm, Christoph, Aaron Flaaen, and Nitya Pandalai-Nayar. 2023. “New Measurement of Export Participation in US Manufacturing.” AEA Papers and Proceedings 113: 93–98 

By Jaedo Choi and Yongseok Shin

“In this game, one American company gets to win. They don’t care if all their US competitors lose. It’s actually better for them. But on the other side, all the Chinese companies win. They all get to step up and create a massive market where none previously existed.” McGee (2025, p.288)

China’s Quid Pro Quo policy, which has mandated foreign multinationals to form joint ventures as a way to promote technology transfers into China, has long been at the center of the US-China economic rivalry. Because joint venture ownership is split between local Chinese firms and foreign multinationals, Chinese partners engage in management and decision-making, which can facilitate technology transfers more effectively than other forms of foreign direct investment (FDI). Although China has not explicitly mandated joint ventures since joining the WTO in 2001, multinationals continued to face implicit pressure. As some executives put it, “voluntary is the new mandatory” (McGee, 2025). Amid rising economic rivalry, the US has imposed restrictions on outward FDI into China in high-tech industries. The CHIPS and Science Act of 2022 is one example.

However, US firms voluntarily entered into joint ventures to access China’s large market and low wages, despite the risks of technology leakage. Is there still an economic justification for restricting joint ventures? If so, did US firms indeed transfer too much technology to China through joint ventures?

In our new research (Choi et al., 2025), we answer these questions. The key issue is that US multinationals do not care about (or internalize) the negative impacts of their JV decisions on other US firms. From the perspective of individual firms, establishing joint ventures is profitable, as they gain access to the Chinese market and lower wages. However, by transferring technology through joint ventures, they make their Chinese partners more productive, and through diffusion, indirectly benefit other Chinese firms. Over time, global competition intensifies, hurting other US firms. Therefore, there could be over-investment in joint ventures relative to the US social optimum. 

Three Empirical Findings: Joint Ventures Appear to Generate Direct and Indirect Benefits for Chinese firms, but Negative Consequences for US Firms

We combine multiple micro datasets containing firm-level balance sheet and patent information for both US and Chinese firms. Our analysis provides three key findings:

1. Direct effects for Chinese partner firms: Chinese firms that formed joint ventures with foreign multinationals experienced faster growth and their patenting activities became more similar to their foreign multinational partners, suggestive of direct technology diffusion.
2. Indirect effects for other Chinese firms: In industries with more FDI into China, even Chinese firms that were not directly involved in joint ventures grew faster and technologically more advanced.
3. Negative US outcomes: In these industries with more FDI, US firms experienced more negative outcomes in terms of size and innovation.

Higher US welfare and more US innovation when restricting joint ventures in 1999

Using a quantitative two-country open economy growth model with endogenous joint venture and innovation decisions, we evaluate the welfare effects of restricting joint ventures starting in 1999. The key trade-off of US firms’ joint venture decisions is gaining access to larger market and low wages on the one hand, and the risk of technology leakage on the other. This leakage makes Chinese firms, both joint venture partners and other firms, more productive, which we refer to as the leakage effect. The model is calibrated to our micro-level empirical facts 1, 2, and 3 above.

Figure 1: Average Productivity Ratio between the US and China. Baseline vs. Restricting Joint Venture in 1999

We find that shutting down joint ventures slows China’s catch-up with US (Figure 1), which in turn improves US welfare by 1.2%. Figure 2 illustrates consumption path when joint ventures are shut down, compared to the baseline with joint ventures. In the short run, consumption declines because US firms lose profits due to reduced access to Chinese market and cheap labor. Over time, however, consumption begins to rise as the long-run benefits of limiting technology leakage outweigh the short-run costs.

Figure 2: Relative Consumption over Time. Baseline vs. Restricting Joint Venture in 1999

One of the long-run drivers of higher consumption is an increase in US innovation. When joint ventures are restricted, US innovation rises, as shown in Figure 3, which reports the difference in innovation rates between the counterfactual and the baseline. Joint ventures have two opposing effects on US innovation. On the one hand, a higher risk of technology leakage to China reduces the payoff from successful innovation; this discourages innovation. On the other hand, the option to form joint ventures can stimulate innovation, because of larger market size and an increase in the bargaining fees US firms receive from their Chinese partners. Our model shows that, in the long run, the negative leakage effect dominates, so restricting joint ventures ultimately boosts US innovation.

This innovation result is consistent with the observation made by Andy Grove, a former CEO of Intel: “Our pursuit of our individual businesses … often involves transferring manufacturing and a great deal of engineering out of the country … We don’t just lose jobs—we lose our hold on new technologies and ultimately damage our capacity to innovate” (Bloomberg, 2010).

Shutting down all joint ventures is not an optimal policy. Targeting restrictions only to industries with large technology gaps between the US and China result in greater US welfare gains, because these industries are most vulnerable to the leakage effect. However, this state-dependent policy reduces US innovation by reducing firms’ incentives to maintain large technology leads over Chinese firms.  

Figure 3: Differences in US Innovation Rates over Time. Baseline vs. Restricting Joint Venture in 1999

Coordinating joint ventures improves US welfare

From the US perspective, coordinating JVs can improve US welfare, while preserving gains from joint ventures. We consider an alternative scenario in which US firms are required to compensate other domestic firms’ profit losses caused by their joint venture activities. In this scenario, fewer joint ventures are established, and shutting down joint ventures reduces US welfare, because the technology leakage effect is internalized.

Timing matters: US welfare decreases when restricting joint ventures in 2025

However, shutting down joint ventures does not always improve US welfare. We consider shutting down joint ventures in 2025, instead of 1999, in light of more recent policy debates. In contrast to the 1999 case, restricting joint ventures in 2025 reduces US welfare by 0.7%, rather than raising it. By 2025, the US-China technology gap became much smaller, so technology diffusion and the resulting leakage effects on other domestic firms has become weaker. Although the restriction still slightly widens the US-China productivity gap, the loss of forgone JV profits and market access outweigh the modest gains from reduced technology diffusion.

Policy implications

As technology rivalry between the US and China deepens, our findings have important implications.

1. There is potential over-investment in joint ventures because multinationals do not take other firms’ profit losses into account.
2. Technology leakage from joint ventures may undermine US innovation.
3. Shutting down all joint ventures may not be optimal; restricting them only in industries with large technology gaps between the US and China leads to greater US welfare improvements.
4. Shutting down joint ventures does not always improve US welfare; the outcome depends on the technology gap between the US and China.

This article is based on the paper, “The Dynamics of Technology Transfer: Multinational Investment in China and Rising Global Competition, EMPCT Working Paper 2025-07

Note: The views expressed herein are our own and do not represent the views of the Federal Reserve Bank of St. Louis, the IMF, its Executive Board, or its management. They also do not necessarily represent the views of the University of Texas at Austin.

Reference
Choi, Jaedo, George Cui, Younghun Shim, and Yongseok Shin. 2025. “The Dynamics of Technology Transfer: Multinational Investment in China and Rising Global Competition.” EMPCT Working Paper 2025-07

Grove, Andy. 2010. “How America Can Create Jobs.” Bloomberg, July 1. https://www.bloomberg.com/news/articles/2010-07-01/andy-grove-how-america-can-create-jobs

McGee, Patrick. 2025. Apple in China: The Capture of the World’s Greatest Company.

By Javier Cravino, Andrei A. Levchenko, Francesc Ortega and Nitya Pandalai-Nayar

The U.S. is in the midst of an unprecedented wave of deportations. Between January and July 2025, nearly 150,000 individuals were deported, and the provision of a total of $170 billion in new federal funding for immigration enforcement suggests these efforts will continue. This raises several key economic questions and concerns: How will this policy reshape the economy? Will consumer prices of everyday items such as food soar? Will American workers benefit?

In new research, we quantify the economic consequences of large-scale deportations using fresh data on workers’ legal status combined with a rich, multi-region, multi-sector quantitative model of the U.S. economy with heterogeneous workers.

Where unauthorized workers are concentrated

We use a novel algorithm to impute the legal status of foreign-born workers in the American Community Survey (ACS). Importantly, our approach accounts for migrant arrivals in 2024, and distinguishes between workers with temporary work authorizations (TPS, DACA, parole, asylum seekers), and workers fully unauthorized to live and work in the US. Our analysis focuses on this latter group only, though our model can be used to study other policies affecting the temporary work authorization group as well. Unauthorized immigrants make up about 3% of the U.S. workforce, but their presence is far from uniform:

• By state: California, Texas, Nevada, New Jersey, Washington and Florida each have around 5–6% of their workforces unauthorized. In contrast, states like Montana and West Virginia have negligible shares.
• By industry: Farming stands out, with more than one-third of workers unauthorized. Construction, Food & Drink Services, and Forestry & Fishing also rely heavily on this labor.
• By occupation: Farmers, construction workers, and personal service providers (such as cleaning and yard work) have some of the highest shares of unauthorized workers.

Figure 1: (adapted from Figures 1,2 and 3 of our paper) Shares of unauthorized workers across states, industries, and occupations.

States

Industries

Occupations

What happens when deportations rise?

Using a stylized version of our model, we show analytically that native workers’ (US born and naturalized citizens’) average real wages fall in all regions in response to deportation shocks. Analytically, the fall is in proportion to the initial presence of unauthorized workers in the geographic location. However, despite this overall decline, native wages in some immigrant-intensive occupations may increase if native and foreign-born workers are sufficiently substitutable. Finally, the removal of unauthorized workers raises the relative prices of immigrant-intensive sectors, in proportion to how much labor these sectors use in producing output, and to the share of unauthorized workers in these sectors’ wage bills.

 We then calibrate our full model with standard parameter estimates. We match the shares of native, authorized and unauthorized workers across states, sectors and occupations, as well as national sectoral expenditures and intra-region and international trade flows.  We simulate the removal of 50% of unauthorized workers nationwide – about 1.5% of the workforce. While the shock is aggregate, the impact is heterogeneous across regions, sectors and occupations, as the distribution of unauthorized workers varies at this granular level. The results validate our analytical results, and challenge some conventional expectations:

1. Native workers lose overall, but average real wage declines are modest

We find average real wages for natives fall by 0.3% nationally and fall in every state. The declines are steepest in California, Washington, New Jersey, and Texas (−0.4 to −0.5%). But, as predicted by the theory, not all natives lose: those in immigrant-heavy jobs, like farming, see wage increases of as much as 3–7%.

Figure 2: (adapted from Figure 4 of our paper) Native wage changes by state and occupation.

2. Immigrant workers gain

Authorized immigrants experience a 3% increase in real wages. Unauthorized immigrants who remain in the country see the biggest gains—12% on average— due to their relative scarcity and low occupational substitutability with other categories of workers.

3. Prices rise in certain sectors, but modestly in most

Sectors most exposed to the shock see modest price increases. For instance, farm goods become 1.6% more expensive relative to the overall price index. Forestry & Fishing see similar increases, while most other sectors show little change. Consumer prices overall rise by less than producer prices, as intra and international trade and substitution towards less affected regions cushions the consumer price impact.

Figure 3: (adapted from Figure 6 of our paper) Producer and consumer price changes by sector.

Who bears the burden?

We find the cost-of-living impact is uneven. States with larger shares of unauthorized workers (California, Washington, Texas, New Jersey, Florida) face consumer prices 0.2–0.3% above the national average, while states with minimal exposure (North Dakota, West Virginia) see relative price declines. The gap between the most and least exposed states is around 0.7%.

It is well known that lower-income households have larger expenditure shares on staples like food items, which are sectors seeing relatively larger price increases from the policy. As expected, looking across households, deportations are slightly regressive. Lower-income households—who spend a larger share on food—face somewhat higher price increases than wealthy households. Yet the gap is limited: just 0.02 percentage points between the bottom and top 5% of earners. This reflects both the size of relative consumer price increases by sector: the most affected sectors see modest price increases, and the insufficiently large gap in expenditure shares on these sectors between the top and bottom earning households.

Figure 4: (adapted from Figures 8 and 9 of our paper) Regional and household variation in consumer price changes.

Regional Consumer Price Variation

CPI of top 5% vs bottom 5%

Bottom line

Large-scale deportations redistribute income in ways that are not always intuitive. They hurt the majority of U.S. households due to slightly higher living costs. These costs are borne disproportionately by regions with a higher initial share of immigrant workers. The primary beneficiaries of these policies are concentrated in the few native workers in the most impacted occupations like farming or construction, authorized workers, and the unauthorized immigrant workers remaining in the US.

This article is based on the paper, “The Economic Impact of Mass Deportations”, EMPCT Working Paper 2025-09

Note: The views expressed herein are solely those of the authors and should not be reported as representing the views of the institutions they are associated with. They also do not necessarily represent the views of the University of Texas at Austin.

By Yuriy Gorodnichenko and Olivier Coibion

What do the wars in Ukraine and the Middle East mean for your grocery bill, your savings, and your spending decisions? For most of us, they feel like tragedies happening far away. But new evidence shows they reach into our households in subtle yet powerful ways — by reshaping how people view the economic future.

The Hidden Front

Wars don’t just destroy cities and lives where they are fought. They also ripple outward, creating uncertainty in energy markets, disrupting global trade, and rattling financial markets. That much is obvious.

What’s less obvious is how they change the way ordinary consumers think about the economy — and therefore how they behave. When people fear the future will be worse, they spend less today. And when millions of households do this at once, recessions can follow.

Economists have long suspected this. The Gulf War coincided with the 1990–91 U.S. recession, as consumer confidence fell sharply. After the Brexit referendum, U.K. households pulled back spending. And when Russian tanks rolled into Ukraine in 2022, European confidence indicators plunged. But were those just correlations, or did the wars cause people to spend less?

That’s the question my co-authors and I set out to answer.

Running a Thought Experiment at Scale

Together with colleagues at the European Central Bank, we fielded a large-scale survey of euro-area households. We couldn’t randomly assign people to live in a world with or without war. But we could do the next best thing: randomize how long respondents were told the current wars might last.

Some were told the conflicts would end within months. Others were told they might drag on for years. Then we asked how they thought that scenario would affect prices, growth, stock markets, government debt, their own financial situation, and their spending.

Because the scenarios were randomly assigned, differences in responses capture the causal effect of perceived geopolitical risk.

The results were striking.

A Stagflationary Future

When people imagine a longer war, their expectations shift sharply.

• They expect higher inflation.
• They expect lower growth.
• They expect rising unemployment.
• They expect falling stock prices and weaker trade.
• They expect governments to pile up debt and raise taxes.

In other words: prolonged geopolitical risk looks stagflationary to households.

Figure 1 (adapted from our paper’s Figure 3) shows this clearly: the longer the expected conflict, the darker the collective outlook.

From National Outlook to Personal Wallets

But this isn’t just about “the economy.” People also bring it home. Asked about their own financial situation, households facing the “long war” scenario were much more likely to say they’d be worse off. That pessimism translated into expected cuts in spending — especially on big-ticket items like appliances, cars, or vacations.

Using panel data from the survey, we can actually see the same households spending less when their geopolitical concerns rise. On average, a modest increase in geopolitical worries led to a nearly 1% drop in nondurable spending (food, clothing, etc.) and a 4–5% drop in durable spending (furniture, electronics, cars).

Figure 2 (adapted from our paper’s Figure 4) shows how consumption expectations fall as the expected duration of war rises.

What This Means for the U.S.

Although our data come from Europe, the lessons extend across the Atlantic. American consumers also react strongly to geopolitical shocks. After Iraq invaded Kuwait in 1990, U.S. consumer confidence collapsed, helping tip the economy into recession. More recently, spikes in geopolitical risk indexes have coincided with jumps in gas prices and drops in consumer sentiment in the U.S.

The timing matters. The United States is already in what looks uncomfortably close to a stagflationary environment: growth has slowed, productivity is weak, while inflation remains above target. Wars abroad that reinforce these stagflationary expectations — higher prices, lower growth — can make an already delicate balance much harder to manage.

This is where central bank independence comes in. If households start doubting that the Federal Reserve will be able — or willing — to keep inflation in check when geopolitical shocks hit, expectations can become unanchored. And once inflation expectations drift upward, they feed into wage demands, pricing decisions, and borrowing costs, making stagflation more entrenched.

The Danger of Normalizing Crisis

One lesson from our findings is that uncertainty fatigue can have real costs. If consumers come to see endless war as the “new normal,” their lower spending could become self-fulfilling, dragging down economies already strained by climate change, technological upheaval, and populist politics.

That may help explain why leaders like Christine Lagarde of the ECB warned early on that “a long-lasting war in Ukraine remains a significant risk” for European confidence. She was right — not just politically, but economically. And the same message should resonate in Washington: foreign policy doesn’t just shape geopolitics. It shapes Main Street wallets, and it tests whether central banks can maintain credibility in the face of political pressure.

What To Do About It

Policymakers can’t stop wars on their own. But they can mitigate how wars shape expectations. Clear communication about energy security, fiscal discipline, and long-term resilience can help counter the drumbeat of pessimism. So can targeted transfers to households most vulnerable to shocks.

For central banks, the challenge is sharper: to remain committed to price stability even when geopolitical risks push growth lower. That is precisely why central bank independence is so critical. If monetary policy bends too easily to short-term political pressures in the face of stagflationary shocks, the costs of regaining credibility later will be even greater.

Conclusion

Wars abroad may feel far away. But their shadows stretch into our supermarkets, our job prospects, and our wallets. As our research shows, the fear of protracted conflict alone is enough to slow spending and growth.

In a world where geopolitical crises seem permanent, the battle for economic stability won’t just be fought with tanks or tariffs. It will also be fought in central banks, in fiscal policy debates, and in the expectations that guide the everyday decisions of households.

By: Christoph E Boehm and Niklas Kroner

It is well-known that changes in US interest rates can have a major impact on global financial markets, but how does other US economic news affect the rest of the world? In new research, Christoph Boehm and Niklas Kroner find that global stock prices jump immediately, and in a synchronized way, to news releases about changes in the US economy and that US monetary policy can have a stabilizing role following such releases. In contrast, foreign economic news releases have little to no effect on US markets.

In today’s interconnected world, countries’ stock prices often rise and fall in tandem. For instance, the world saw a global decline in stock markets during the COVID-19 pandemic, followed by swift recoveries for most countries. This co-movement has existed for several decades and is a broad phenomenon which largely holds globally. Further, it is not limited to stock prices. Research has documented that countries’ gross capital flows, leverage, and credit all tend to move together. Financial conditions in different countries thus have an important common component: at certain times, they are favorable around the world, while at others, they become globally unfavorable. This co-movement of financial conditions has been termed the global financial cycle.

What is less clear, however, is what drives the global financial cycle. With the central role of the US financial system and the dollar’s status as the dominant global currency, the policy of the US Federal Reserve (which sets US interest rates) is known to affect financial conditions globally. In new research, we show that the US’ special role extends well beyond monetary policy: news about US economic activity itself plays a large and distinct role in driving global financial conditions. Hence, when the US economy does well, global financial conditions also tend to be good.

High-frequency analysis of US economic news

Although commentators often attach plausible narratives to financial and macroeconomic developments, it is typically very difficult to accurately determine the causes and effects in macroeconomics. This is because most financial and macroeconomic outcomes are interdependent. For example, if an economy’s growth is accelerating, then this acceleration is likely to lead to higher stock prices. The higher stock prices, however, will then feed back into overall economic growth, for instance, by affecting consumption and investment decisions. To deal with this “simultaneity problem”, in new research, we studied the effects of macroeconomic data releases in the US on global asset markets. Since data collection efforts for a particular macroeconomic outcome have been completed at the time of the announcement, the published release cannot be impacted by such feedback anymore. This leaves only one possible direction for causality: news releases of macroeconomic data can affect stock prices, and we show that this is indeed the case.

Our research shows that after US macroeconomic news releases, global stock prices respond immediately and in a synchronized way. For instance, higher-than-expected nonfarm payroll employment – one of the major US economic news releases – leads to an increase in stock prices in essentially all countries in our sample (see Figure 1 for a select subset). These effects are also large. Foreign countries’ stock prices respond with magnitudes like that of the US stock market.

Figure 1 – Effects of US news on foreign stock markets, by news type and country

Notes: Bars show the effect of a one standard deviation surprise in the data release. Thin red lines represent 95% confidence intervals. All measures the average effect of 27 countries. Sources: Authors’ calculations using data from Bloomberg and LSEG Workspace.

Why is it that global stock prices respond so much to surprises about US macroeconomic data releases? To answer this question, we studied a variety of additional asset prices and related measures – such as the VIX index, which is often interpreted as a gauge of risk and uncertainty. It turns out that whenever global stock prices rise after US macroeconomic news releases, investors’ perceived risk and uncertainty falls and so do prices of relatively safe assets such as government bonds. This behavior suggests that investors are more confident in holding riskier assets when the US economy is doing well, and that positive news about the US macroeconomy triggers the onset of such “risk-on” periods. The flipside, of course, is that bad news about the US economy can lead to a global stock market panic as was the case during the global financial crises in 2008 and 2009.

Interestingly, US monetary policy, which has often been blamed for fluctuations in global asset markets, appears to have a stabilizing role after US macroeconomic news releases. When bad news about the US economy becomes available, markets expect the Federal Reserve to lower interest rates, which partially offsets the decline in global stock markets and thus stabilizes both asset markets and the economy. In the absence of the Fed’s decisive policy, which combats downturns and prevents economic overheating, the effects of US news on global stock prices would likely be worse.

The central role of the US

One of the most striking findings from our study is that the US is a clear outlier in terms of the effects of its news releases. While US macroeconomic announcements have large effects on foreign stock markets, the reverse is not true. Foreign economic news releases have little to no effects on US markets. This is the case for Japan and Germany, two relatively large economies, and the UK, an important financial center, as well as several other countries. In comparison to the US, the effects of these countries’ news releases are minimal. This difference highlights the special role the US economy plays in the global financial and monetary system.

With the recent turbulence in global tariff policy and the deteriorating US fiscal outlook some have speculated that the transition towards a new center of the global financial system has accelerated. In such a scenario the dollar would ultimately lose its dominant currency status and be replaced by the euro, the renminbi, or by multiple currencies in a multipolar system. Given how central the US economy is to most market participants and that there is no alternative that matches the US in the relevant dimensions, such a transition is still very difficult to imagine. Should it ultimately happen, however, it is unlikely that many countries would emerge from such a transition unscathed: news about the US economy has a strong tendency to move asset prices and financial conditions globally.

This article is based on the paper, ‘The U.S., Economic News, and the Global Financial Cycle’, in The Review of Economic Studies.

Note: The views expressed herein are solely those of the authors and should not be reported as representing the views of the Federal Reserve Board, the principals of the Board of Governors, or the Federal Reserve System. They also do not necessarily represent the views of the University of Texas at Austin.

By: Erwan Gautier, Frédérique Savignac, and Olivier Coibion

As inflation surged across the globe in the aftermath of the COVID-19 pandemic and the war in Ukraine, economists and policymakers worried that rising inflation expectations could become entrenched and drive a wage-price spiral. While much of the attention on expectations focused on households and financial markets, relatively little was known about the expectations of firms—the agents directly responsible for setting prices and wages, due to the absence of data on their beliefs.

Recent work has begun to fill this gap. Baumann et al. (2024), Abberger et al. (2024), and Candia et al. (2023), for example, have examined firms’ inflation expectations in the Euro area and the U.S., revealing both the promise and pitfalls of using survey data to better understand inflation dynamics. In a recent study, we contribute to this growing literature with a uniquely rich dataset: a quarterly survey of French firms conducted by the Banque de France between 2020 and 2024. This period covers the full cycle from low inflation to peak and back to near-target levels, offering rare insight into how inflation expectations form, evolve, and affect decisions. Furthermore, the survey includes information on firms’ beliefs about the evolution of future wages within their firm, thereby providing a unique setting to study the potential passthrough of inflation into wages.

Underreaction, Overreaction, and Re-anchoring

Figure 1 plots the trajectory of French firms’ perceptions and expectations of inflation at different horizons (1 year ahead for the short term, and 3 to 5 years for the long term). As inflation began rising in 2022, firms initially underreacted: their short-term expectations rose more slowly than actual inflation. But this gave way to a period of persistent overshooting—firms expected more inflation than actually materialized, especially during the disinflation phase in 2023–2024.

Figure 1: Inflation expectations and perceptions

Note: Average perceptions and expectations of French firms over time. Source: Banque de France survey as shown in Gautier, Savignac and Coibion (2025)

This pattern—a delayed and amplified response to inflation surprises—is consistent with theories of imperfect information or rational inattention (Angeletos, Huo & Sastry, 2021). But despite this overshooting, inflation expectations gradually re-anchored by the end of 2024. Both short- and long-run expectations converged back to the ECB 2% target, suggesting that fears of a permanent de-anchoring were overstated—at least in France.

Limited Scope for a Wage-Price Spiral

One of the most policy-relevant questions is whether rising inflation expectations feed directly into wage-setting. Our survey provides a direct answer: firms were asked how much they expected to raise wages in the coming year. These wage expectations, while rising in 2022–2023, remained far more subdued than inflation expectations and tracked actual wage growth relatively well.

Figure 2 plots the correlation between firms’ inflation expectations and their own wage expectations. Only short-term expectations (and to a lesser extent, inflation perceptions) are meaningfully correlated with expected wage growth. Long-run inflation expectations have essentially no predictive power—consistent with models where wage contracts are short-lived and expectations beyond the contract horizon are irrelevant (Werning, 2022).

Figure 2: Inflation expectations versus wage expectations

Note: Relationship between inflation expectations (at 1-year horizon (left-panel); 3 to 5 year horizon (right panel) and expected wage growth (1-year ahead). Source: Banque de France survey as shown in Gautier, Savignac and Coibion (2025).

Moreover, we find that the pass-through of inflation expectations into wage expectations—and into actual decisions like prices or employment—was notably weaker during the high-inflation period. Why? A key factor was the emergence of firms expecting what we term an “inflation disaster”—very high inflation rates (above 10%). These firms grew more numerous in 2022–2023, but paradoxically showed less responsiveness in their behavior. Their expectations were detached from actual outcomes and from their own actions.

Among firms with more moderate expectations, pass-through remained relatively stable. This pattern suggests that high inflation doesn’t necessarily amplify expectational effects across the board—if anything, it may weaken them for some firms.

Expectations Matter—But Less Than Feared

Taken together, our results complicate the standard narrative. Yes, inflation expectations rose. Yes, they correlate with wage and price-setting. But their power to drive actual decisions weakened during the inflation surge, especially among firms with the most pessimistic outlooks. This likely dampened the feared feedback loop of inflation expectations fueling further inflation.

The French experience stands in contrast to that of other countries such as the U.S., where expectations appeared to become more sensitive to news (Weber et al., 2025) and where energy price pass-through was more direct. The French government’s efforts to shield households and firms from energy price shocks may have helped limit the unmooring of expectations (Coibion and Gorodnichenko 2025).

Implications for Policymakers

Our findings offer cautious reassurance to central bankers. While anchoring expectations remains crucial, not every inflation surge leads to runaway expectations or self-fulfilling dynamics. Structural features—such as wage bargaining institutions, policy interventions, and even the extent of inattention—can shape how firms respond to shocks.

They also suggest that surveys of firms’ expectations are an essential complement to household or market-based measures. Firms see the world differently. They can be slow to update, occasionally alarmist, but often pragmatic when it comes to their own wage and pricing decisions.

By building out these types of high-frequency, multi-horizon surveys across countries, we can gain a much clearer view of the inflation process—before the next surge surprises us again.

Disclaimer: The views expressed in this Vox column are those of the authors and do not necessarily represent those of the Banque de France, the Eurosystem or any other organization with which they are affiliated.

References

Abberger, K., A.-K. Funk, M. Lamla, S. Lein and S. Siegrist (2024), “How inflation expectations pass through to prices and wages in the Euro Area”, VoxEU.org, 19 February.

Angeletos, G.-M., Z. Huo and K. Sastry (2021), “Imperfect macroeconomic expectations: Evidence and theory”, NBER Macroeconomics Annual 35.

Baumann, U., A. Ferrando, D. Georgarakos, Y. Gorodnichenko and T. Reineilt (2024), “How firms update their inflation expectations: Evidence from the ECB’s new survey”, VoxEU.org, 11 March.

Candia, B., O. Coibion and Y. Gorodnichenko (2023), “Firms’ expectations and the macroeconomy: Lessons from surveys”, VoxEU.org, 16 January.

Coibion, O. and Y. Gorodnichenko (2025), “Inflation, expectations and monetary policy: What have we learned and to what end?”, manuscript.

Dong, D., Z. Liu, P. Wang and M. Wei (2024), “Inflation disagreement weakens the power of monetary policy”, Federal Reserve Bank of San Francisco Working Paper 2024-27.

Gautier, E., F. Savignac and O. Coibion, (2025), “Firms’ Inflation and Wage Expectations during the Surge,” NBER Working Paper 33799.

Weber, M., B. Candia, H. Afrouzi, T. Ropele, R. Lluberas, S. Frache, B. Meyer, S. Kumar, D. Georgarakos, O. Coibion, G. Kenny and J. Ponces (2025), “Tell me something I don’t already know: Learning in low and high inflation settings”, Econometrica 93: 229–264.

Werning, I. (2022), “Expectations and the rate of inflation”, NBER Working Paper No. 30260.

By: Yuriy Gorodnichenko and Olivier Coibion

Economic policymakers spend a lot of time dissecting different measures of inflation expectations in the economy, since these affect prices, wages, consumption and many other outcomes. Over the last year or so, a major gap has appeared between two of the leading surveys of U.S. consumers’ inflation expectations: the New York Federal Reserve’s Survey of Consumption Expectations (SCE) and the Michigan Survey of Consumers (MSC). The latter displays high levels of expected inflation since the inflation surge whereas the former reports low “anchored” levels, as shown in Figure 1. Which is correct? The answer is the Michigan survey.

Figure 1: Surveys of Household 1-Year Ahead Inflation Expectations

How do we know this? We use two strategies. The first is to bring in other surveys of consumer expectations to determine whether it is the MSC or the SCE that is the outlier. The second is to dig into the surveys to understand why they seem to point to different answers. Both approaches lead us to the same conclusion.

Other surveys of inflation expectations of households confirm the Michigan survey. Figure 1 plots results from surveys of Nielsen Homescan panels we run every quarter, covering around 20,000 representative households per quarter (much larger than Michigan or NY Fed). It also plots the Indirect Consumer Inflation Expectations survey from the Cleveland Fed. Both give approximately the same dynamics as Michigan. What is clear from this figure is that it is the reported data from the NY Fed survey that is the outlier, not the Michigan survey.

The NY Fed applies different methodological choices: The NY Fed survey results reported in Figure 1 are different from those of the other surveys in three ways, each of which we think is a problem.

a. Medians vs means: The NY Fed series is for the median instead of the mean: this throws away all the information in the right tail, which is where the action is when it comes to inflation expectations. The mean is more volatile but the resulting volatility is not noise. See Figure 2 for example: most of the high-frequency variation in mean forecasts is associated with gasoline (oil) price variation that consumers observe. We should not be throwing away this information by relying on the median. Since the NY Fed provides individual-level responses prior to the last 12 months, we can easily construct the mean, with the exception of the most recent data.

Figure 2: Household Inflation Expectations and Gasoline Prices

b. Distribution questions vs point forecasts: Whereas the other surveys use point forecasts, the NY Fed presents responses from a “distribution” question where respondents assign probabilities to a pre-defined set of possible outcomes. This induces framing effects (it’s centered around zero) and introduces a bias as the inflation rate rises (because the bins are fixed). Fortunately, the NY Fed also asks respondents for point forecasts, so this is easy to address.

c. Panel conditioning effects: The NY Fed has respondents participate every month for a year. This is a problem because people learn from participating in the survey: their responses in the wave are usually ~2p.p. higher than in their third wave (Kim and Binder 2023). In other words, after the first wave, they are no longer representative and should not be used. Since the NY Fed provides individual data with information about how many times they have participated (albeit with a one-year delay), we can easily address this issue by focusing on the subset of respondents who are responding for the first time.

If we correct for these three differences using the NY Fed micro-data, we get the results in Figure 3 below, which mirror what we get from the other surveys. In other words, the surveys actually all tell the same message: year-ahead inflation expectations are currently around 5-6%. The same logic applies to LR expectations: you should rely on mean 5-year ahead expectations of the Michigan survey, and these are currently very high and unambiguously unanchored.

Figure 3: Uncorrected and Corrected NY Fed Survey Expectations

Conclusion:

Expectations of households and firms, following the inflation surge, have settled at new higher levels after the surge than they were prior to the surge. This suggests that the inflation surge experience has further unanchored their expectations. Since these feed into consumer spending as well as firms’ pricing, employment and investment decisions, we should expect inflationary pressures to be particularly high, complicating the job of achieving the “last mile” of inflation reduction and responding to new inflationary pressures from tariff policies.   

By: Barthélémy Bonadio, Zhen Huo, Elliot Kang, Andrei A. Levchenko, Nitya Pandalai-Nayar, and Hiroshi Toma

Related Working Paper: Playing with blocs: Quantifying decoupling

As geopolitical tensions rise between major powers, concerns intensify over the economic implications of trade fragmentation. Policymakers worldwide are concerned that the US-China trade war, Brexit, and the breakdown of economic ties between the West and Russia are all signs of a hazardous trend toward “deglobalization.” Yet despite these challenges, global trade remained surprisingly resilient. A recent paper, Bonadio et al. (2024), presents new evidence that during this period, the global trading system reconfigured itself into distinct blocs. The authors find that, contrary to popular belief, this realignment prior to 2024 did not result in significant welfare losses. In contrast, the 2025 US tariff increases and retaliation by trade partners announced as of mid-April, are likely to raise trade barriers both within and across the pre-2024 blocs, potentially reducing welfare.

Figure 1: World Trade / GDP Ratio

The Rise of Trade Blocs

The post-2016 era has been marked by significant policy shifts that potentially threaten international trade, such as the Brexit vote, the US-China trade war, and more recently, Russia’s invasion of Ukraine and subsequent sanctions. These events have sparked concerns about the possible unraveling of globalization.

However, an analysis of global trade data shows that trade relative to economic activity has not decreased since the onset of these shocks (Figure 1). In fact, the ratio of world trade to GDP has partially reversed the downward trend that started following the Global Financial Crisis.

What explains this resilience? As Antràs (2020) and Goldberg and Reed (2023) have documented, we are not experiencing deglobalization but rather fragmentation or decoupling – a reconfiguration of trade links across the globe. As countries involved in trade conflicts disengage from each other, they ramp up trade with other partners. Table 1 confirms this pattern: even as trade between the US and China fell, both economies increased trade with the rest of the world.

Table 1: Change in Trade (2015-2023)

Measuring Bloc Formation and Its Economic Impact

Our research employs a data-driven approach to identify which countries are aligning into which trade blocs. Using bilateral trade data from 2015 to 2023, we identify decoupling patterns and quantify their economic impacts.

We categorize countries into three groups: 1) The US bloc, consisting of countries that experienced reduced trade costs with the US and increased trade costs with China, 2) The China bloc, comprising countries with heightened trade costs with the US and decreased trade costs with China, and 3) Unaligned countries, where trade costs with both the US and China moved in the same direction[PT1] . We identify revealed trade costs from observed trade flows, by running a standard gravity equation. The residual of this equation is then revealed bilateral trade costs, up to the trade elasticity. Our approach, based on observed trade flows, contrasts to alternatives where the bloc structure is imposed on the data using observations like U.N voting patterns.

In our analysis of 187 countries, we find that roughly one-quarter of countries have aligned with the US, another quarter with China, and about half remain unaligned. Russia, Saudi Arabia, Israel, and Hong Kong are clearly aligned with China, while European countries, along with India, Korea, Japan, and Singapore, are part of the US bloc.

Surprising Consequences from Decoupling

Utilizing a quantitative multi-country, multi-sector global production network model, we evaluate the economic effects of observed changes in trade costs. Contrary to the widespread belief that fragmentation reduces welfare, our findings indicate that the median country experienced a marginal increase in real income from trade cost changes between 2015 and 2023, 0.6% respectively.

Thus, the process of decoupling has not yet diminished global GDP . As expected, nonaligned countries see larger income changes (0.8%) compared to those aligned with the US or China, since they do not consistently face increased trade costs with either bloc . However, even countries within the US and China blocs generally experience small gains in real income due to trade cost changes.

This finding stems from the fact that global trade remained quite stable relative to global activity from 2015 to 2023. Lower trade flows between some countries (such as the US and China) were more than compensated by higher trade flows between others. In other words, trade data reveal no significant rise in  trade costs on average, which explains the lack of a negative impact on average real GDP.

Are Countries in the “Right” Blocs?

According to the factual simulation above, most countries are not worse off in the fragmented world, as trade costs decreases within blocs more than offset trade cost increases across blocs. A natural question arises: are countries aligning with blocs based on the benefits arising from international trade? To answer this question, we conduct counterfactual exercises where we shift countries to different blocs and compare their outcomes to the baseline. Our counterfactual analysis examines whether countries would benefit economically from changing bloc alignment. Surprisingly, we find that economic self-interest does not fully account for current bloc formations. The median country in the US bloc would benefit from switching to the China bloc, and vice versa.

This suggests that geopolitical factors, rather than purely economic considerations, are influencing bloc alignment decisions. By investigating the determinants of trade changes, we find that geopolitical considerations indeed increased their role in international trade. After 2015, countries increased trade with their pre-2015 political allies (measured by UN voting patterns) at the expense of non-allies.

Policy Implications

As the global economy continues to fragment, our findings have important implications. Firstly, fragmentation need not necessarily lead to economic losses. In fact, the reconfiguration of trade in the lead up to 2023 benefited marginally most countries by allowing them to redirect trade toward partners with lower costs. Secondly, regional trade integration may offset some negative effects of cross-bloc decoupling, as evidenced by declining trade costs within blocs. Thirdly, when economic and geopolitical incentives diverge, policymakers may face complex trade-offs. Our research shows that many countries are not part of the bloc that would be economically optimal for them.

Conclusion

The reorganization of global trade into blocs results in both winners and losers, but up until 2023, most countries have marginally benefited from the observed changes in trade costs. The increased integration within blocs has compensated for the negative impacts of fragmentation between blocs.

Our research suggests that factors beyond international trade are primarily driving these bloc alignments. This indicates that policymakers should aim to balance economic opportunities with geopolitical considerations in an increasingly divided world, strategically navigating geopolitical realities alongside economic prospects. That said, the 2025 trade war is already raising trade costs within the blocs that formed prior to 2023, and therefore might offset some of the marginal welfare gains this far.

References

Aiyar, Shekhar, Andrea Presbitero, and Michele Ruta. 2023. Geoeconomic Fragmentation: The Economic Risks from a Fractured World Economy. CEPR Press.

Antràs, Pol. 2020. “De-globalisation? Global value chains in the post-COVID-19 age.” NBER Working Paper No. 28115, National Bureau of Economic Research.

Bonadio, Barthélémy, Zhen Huo, Elliot Kang, Andrei A. Levchenko, Nitya Pandalai-Nayar, Hiroshi Toma, and Petia Topalova. 2024. “Playing with blocs: Quantifying decoupling.” Working Paper.

Clayton, Christopher, Matteo Maggiori, and Jesse Schreger. 2023. “A framework for geoeconomics.” NBER Working Paper No. 31852, National Bureau of Economic Research.

Goldberg, Pinelopi K and Tristan Reed. 2023. “Is the global economy deglobalizing? If so, why? And what is next?” Brookings Papers on Economic Activity 2023 (1):347–423.

Gopinath, Gita, Pierre-Olivier Gourinchas, Andrea Presbitero, and Petia Topalova. 2024. “Changing Global Linkages: A New Cold War?” IMF Working Paper 2024-076, International Monetary Fund.

---

 

By Rodrigo Adão (Chicago Booth), Martin Beraja (UC Berkeley Haas), and Nitya Pandalai-Nayar (UT Austin)

As artificial intelligence reshapes the workplace, policymakers and business leaders are confronted with a pivotal question: how quickly will workers adapt to AI-driven changes? To understand what shapes these transition speeds, in a recent paper Adão et al. (2024), we examine two major technological revolutions of the past century: the manufacturing innovations of the early 1900s and the emergence of Information & Communications Technology (ICT) from the 1980s onward. We find that the skill requirements of new technologies critically determine how quickly economies adapt. The adjustment was faster for manufacturing because it required skills more similar to existing occupations, whereas the ICT revolution required different skill sets, resulting in a slower transition. While older workers found it challenging to transition into ICT-intensive occupations, younger generations gradually filled these roles.

Comparing Technological Shifts

Historically, innovations such as steam engines, electricity, and computers have led to fundamental shifts in economic activity. Looking to the future, artificial intelligence and robotics promise similar transformative impacts. When innovations favor specific skills, inequality may rise rapidly (Katz and Murphy, 1992). The reorganization of labor markets might take decades as workers shift roles and newer generations acquire necessary skills (Chari and Hopenhayn, 1991). But do economies adjust at different paces depending on technology? The literature mainly explains common patterns across adjustment episodes rather than examining how specific technologies shape them (Helpman, 1998; Herrendorf et al., 2014).

We present new insights into how exposure to technological innovations affected employment and wages across different occupations in the U.S. during two eras: the manufacturing-enhancing technologies of the early 20th century and the ICT innovations later that century. The contrast between these historical episodes is striking. Panel A of Figure 1 shows that the relative employment growth in more exposed occupations was both faster and stronger overall following the manufacturing innovations of the early 1900s compared to the ICT innovations of the late twentieth century. When assembly lines and electrical machinery transformed manufacturing in the early 20th century, workers rapidly shifted into manufacturing-intensive occupations. Panel B indicates a quick surge in demand for ICT-intensive occupations between 1980 and 1990, with a corresponding labor supply response after 2000.

Figure 1

Panels C and D of Figure 1 depict the employment response for older and younger workers, respectively. The ICT expansion post-1980 was almost solely driven by younger workers, as older workers did not shift to ICT roles—the black dots in Panel C show negligible and statistically insignificant changes for seasoned workers. However, the black dots in Panel D illustrate a significant increase in young workers entering ICT roles over forty years. Overall, the data indicate that the labor supply adjustment to ICT innovations was limited and delayed due to the minimal reallocation of older generation.

The Role of Skill Specificity

What explains these contrasting patterns?  Our research shows that a key factor is how different the required skills are between old and new jobs – what we call “skill specificity.” When new technologies demand skills that differ from those used in existing jobs, the transition tends to be slower and more reliant on new generations entering the workforce.

Figure 2

To investigate this, we develop systematic measures of skill similarity between occupations. Figure 2 depicts two histograms, one for each episode, showing the task distance measure across occupations. The task distance distribution for ICT exposure in the latter episode has more mass on higher distance values than the distribution for manufacturing exposure in the earlier episode. For instance, the analytical and programming skills needed in software development had limited overlap with most existing jobs in 1980. This high skill specificity made it difficult for experienced workers to transition into ICT roles.  In contrast, manufacturing jobs in the early 1900s required skills more similar to existing occupations. A craftsman could more readily adapt his manual and technical abilities to operate new industrial machinery. This lower skill specificity enabled the faster reallocation of incumbent workers.

Dynamics of Adjustment

The degree of skill specificity shapes transition dynamics through two key channels: First, when skills are more specific, experienced workers find it harder to switch to new occupations, even when wages are higher. This creates a direct barrier to workforce reallocation. Second, this limited mobility of incumbent workers leads to higher wage premiums in new occupations. These wage increases provide strong incentives for young workers entering the labor force to invest in new skills. This explains why technological transitions with high skill specificity are driven primarily by new generations. We develop a theoretical model that formalizes these mechanisms and demonstrates how they can quantitatively account for the different adjustment patterns observed in the manufacturing and ICT transitions.

Implications for the AI Revolution

Our findings have important implications for how economies might adapt to artificial intelligence. Early evidence suggests that some AI applications, like large language models, may be relatively easy to use with existing skills. As discussed in Acemoglu et al. (2023), AI could augment rather than replace current capabilities, suggesting a potentially faster adaptation than during the ICT revolution. However, more advanced AI applications may require highly specific skills with limited transferability from existing occupations. If this proves true, the transition could be slow, primarily driven by new workers, similar to the ICT experience. This scenario could exacerbate generational inequality and slow down aggregate productivity gains.

Conclusion

The lessons from past technological revolutions highlight that while the nature of jobs might evolve, the fundamental challenge of equipping workers with relevant skills remains constant. As we face potentially one of history’s most significant technological shifts with the rise of AI, a proactive approach will be essential to ensure all segments of the workforce can benefit. Our findings provide valuable guidance in anticipating and addressing these challenges, enabling a smoother transition into an AI-driven future.

References

Acemoglu, D., Autor, D., & Johnson, S. (2023), “How AI can become pro-worker”, VoxEU.org, 4 October.

Adão, R., Beraja, M., & Pandalai-Nayar, N. (2024). Fast and slow technological transitions. Journal of Political Economy Macroeconomics, 2(2), 183-227.

Chari, V. V. and Hopenhayn, H. (1991). Vintage human capital, growth, and the diffusion of new technology. Journal of Political Economy, 99(6):1142–1165.

Helpman, E. (1998). General purpose technologies and economic growth. MIT press.

Herrendorf, B., Rogerson, R., and Valentinyi, Á. (2014). Growth and structural transformation.  Handbook of economic growth, 2, 855-941.

Katz, L. F. and Murphy, K. M. (1992). Changes in relative wages, 1963–1987: supply and demand factors. The Quarterly Journal of Economics, 107(1):35–78.

For more details, please see the associated working paper.