Global Outlook 2026 - The shifting world order

Global View: Will next year be merely transitional, or transformative?

We weren’t joking when we said 2025 would be the Year of the Tariff. But 2025 was about much more than just tariffs. Three major shifts began in earnest, and 2026 could be the year that these accelerate. First, the AI boom is beginning to have transformative effects – not so much in the real economy just yet, but certainly in financial markets. While leading institutions have warned of the risk of a potential market bust, we economists will be watching closely for further signs of AI’s impact on labour markets and on productivity. The second major shift we observe is China’s coming-of-age on the global geo-economic stage. While our China economist had rightly flagged that Beijing was much better prepared for Trade War 2.0 than it was in Trump’s first presidency, the degree to which it successfully wielded the leverage of rare earths – and later Nexperia chips – still felt like a milestone in China’s global ascent. The third big shift, and perhaps the corollary of China’s ascent, is the decline in the US’s relative power and influence. While still a clear leader in the AI race, the erosion of US institutions threatens its leadership of the global financial system. For now, the US dollar and Treasuries remain the only game in town as globally safe liquid assets, given the lack of genuine alternatives. With Europe more awake to the challenges of an inward-looking US, but also the opportunities in a world where dollar dominance is no longer a given, next year could – in an admittedly optimistic scenario – be the year that Europe starts to seize the moment.

Against the backdrop of these transformative shifts, our view on the economy next year is a relatively benign one. The US continues to coast along, though solid overall growth figures will mask an increasing divergence between those benefitting from the AI boom, and a Main Street that is feeling the inflationary pinch from tariffs and the immigration crackdown. Europe will still see a slow-burning drag from US tariffs next year, but this will increasingly be offset by stronger domestic demand – helped by the ECB’s rate cuts – as well as Germany’s fiscal bazooka. China is expected to take modest further rebalancing steps, though nothing that is likely to be game-changing as a major global demand impulse. Finally, politics offers as much promise as it does peril. Populism will face key tests with the US midterms in November, and Hungary’s general election (crucial for EU policymaking) in April, while France faces an ongoing risk of a snap election, should the government fail to pass a 2026 Budget (the presidential election will happen in April 2027).

We hope you find reading this year’s Global Outlook as interesting as we did writing it – please do share your feedback. Lastly, wherever developments take us in 2026, we wish our readers a restful holiday period, and a happy new year!

Before we dive into our special topics for this year’s Global Outlook, below is a brief summary of our updated base case for growth, inflation and interest rates in the three key regions (more detail in regional Outlooks; forecasts at the back).

Growth: US solid but vulnerable; eurozone on the up; China struggling with imbalances

Broadly, we see the US continuing to grow at a solid, trend-like pace over the next two years, but as described in our US Outlook, this benign appearance masks considerable variation and vulnerability below the surface. In the eurozone, growth is expected to pick up from the subdued, well below trend pace in recent quarters. Much of the recent weakness has been driven by the unwind of US export frontloading after tariffs were implemented in April, with the domestic economy holding up rather better. Looking ahead, we expect domestic demand to strengthen further, even as the tariff drag persists, although the impact of rising defence spending will be blunted by Europe’s high dependence on imports for defence equipment. All told, growth is expected to gradually accelerate towards trend in the course of 2026, and occasionally grow at an above trend pace later in our horizon. This hinges crucially on how quickly and effectively Germany will deploy its new fiscal firepower. Big picture, the US and eurozone are expected to see a convergence in growth trends over the next year, driven largely by the eurozone’s expected pickup. Within the eurozone itself, we also expect a growth convergence, as Germany finally starts to see some meaningful growth again, while the outperformance of southern Europe is expected to see some headwinds from the end of the EU’s Recovery & Resilience Facility.

Looking beyond advanced economies, we expect China to stay laser focused on being the factory of the world (especially in cleantech). Still, while China has also shown itself to be an equal to the US in global geo-economic clout, recent communications from the CCP suggest there is also some recognition of a need to address imbalances and to raise domestic demand. This could go some way to reducing trade tensions, also with Europe. However, we do not expect any fundamental changes to China’s supply-driven growth model, nor therefore anything dramatic on the demand stimulus front.

Could the imminent US Supreme Court ruling on the legality of Trump’s tariffs be a game changer for the outlook? We think not, as even if the Court rules the tariffs to be illegal, the President has a number of alternatives, especially given support from the Republican majorities in Congress. Still, the Supreme Court’s scepticism revives some of the trade policy uncertainty we thought had withered. With that said, the recent US-China trade truce already by itself offers some relief in the form of lower fentanyl-related tariffs, and in general, we see the downside risks to growth from escalating tariffs to be much lower than before the summer. Still, as discussed in our special Box on page 11, a bigger risk to trade and growth arguably now comes from the increased use of chokepoints rather than tariffs.

Finally, while we do not take a view on equity markets, we note the increasing worries expressed by policymakers and business leaders of a potential big market correction following the boom in AI-related stocks this year. Our base case does not assume any major correction, but this could pose a downside risk to growth – especially in the US – if it leads to a broader tightening of financial conditions.

Inflation: US still stubbornly high; eurozone broadly at target; China still exporting deflation

The main worry on the inflation front remains the US, where continued upward pressure is likely to come from tariffs, the immigration crackdown and – in our view – monetary policy that will end up being too loose for such a supply-constrained economy. A helpful offset is the unfolding massive supply glut of oil, which is also expected to push eurozone inflation below the 2% target over the coming year. The eurozone’s inflation undershoot could however prove fleeting, as later in the horizon we expect German fiscal spending to start providing some modest inflationary impulse. Globally, price growth should be kept in check by China’s continued manufacturing largesse, not withstanding the attempts by Beijing to tackle ‘excessive’ competition, or involution as it has come to be known. The European Commission’s dumping monitor suggests some downward pressure on prices from China’s oversupply, though in the aggregate we do not expect this to be enough to move the needle on eurozone inflation.

Central banks: Fed-ECB rates to converge as Fed prioritises the labour market over inflation

The Fed faces arguably the biggest test of all the major central banks next year. As if setting interest rates were not hard enough based on economic data (a scarce commodity in the US these days), the Fed must also contend with the most unprecedented attacks on its independence in decades. As our first topic will elaborate on, our base case sees the Fed broadly maintaining its independence; that is, we do not expect the Committee to be replaced wholesale with Trump appointees. But Trump will still wield considerable influence. Alongside his existing appointees, who have generally lined up in support of Trump’s calls to lower rates, the President will also choose a successor to Fed Chair Powell. The Fed Chair is but one Committee vote out of 12, but still, one that has an outsized influence. While the labour market in the US has indeed weakened, this has in our view been largely driven by the supply side; i.e. it is not something that monetary policy can do a great deal about. At the same time, inflation is still well above target. Even so, we expect labour markets to dominate the Fed’s thinking, and the more dovish tilt of the Fed in 2026-7 leads us to expect a string of rate cuts over the next year, taking the fed funds rate down to the lower end of neutral estimates (3% in the upper bound) by September.

The ECB, by contrast, is in a much more comfortable position, and is expected to keep policy on hold over 2026-7 amid broadly balanced growth and inflation dynamics. In the near term, the risk is still tilted somewhat to another cut if anything, given the expected inflation undershoot, and the upside risks to the euro exchange rate. However, as we move into 2027, that balance is likely to tilt more in favour of hikes. Where things could get trickier for the ECB is if France’s fiscal problems escalate to something more crisis-like. We do not foresee this in our base case, but a major bond market rout has the potential to test the ECB’s intervention resolve at some point.

Special topics

Over the coming pages, we draw on expertise across our macro research team (and beyond) to delve into some of the key themes that we expect to shape the outlook over the next year.

1. Could attacks on Fed independence lead to a US overheating? What could this mean for Europe?

Rogier Quaedvlieg, Bill Diviney and Nick Kounis

Our base case sees the Fed maintaining its independence, but inappropriate rate cuts could lead to a new boom and bust cycle in the US. This could also lead to tighter financial conditions in Europe – and maybe even renewed ECB rate cuts.

The Fed has eased policy rates by 50bp this year. The October FOMC press conference made it clear that easing in the current environment is becoming increasingly contentious. We have previously argued that easing monetary policy is not the solution to the difficulties the US economy is currently facing. Nevertheless, the Trump administration continues to pressure the Fed. What began as blatant rhetorical demands to lower rates and attempts to dismiss Governors, has evolved into a much more subtle effort to redefine the Fed’s role and mandate. We review these new forms of attack on the Fed’s independence and consider the potential consequences for the US economy.

An incomplete list of Fed attacks

Broadly, attacks on the Fed’s independence fall into four categories:

1. Direct political pressure on policy: From his first week in office, Trump renewed his claim that the Fed “keeps America from growing faster than it should.” Public calls for rapid rate cuts have been frequent and explicit. Trump has openly advocated for rates near 1% and has made personal attacks on the Fed Chair, Jerome “Too Late” Powell.

2. Influence on policy input: The administration is increasingly challenging the inputs of monetary analysis. For instance, it argued that the Bureau of Labor Statistics’ (BLS) inflation series “overstate shelter costs” and “understate real wage growth,” with a similar critique of labour market reports. The dismissal of the BLS chair was widely interpreted as an attempt to align the data with the administration’s narrative. Our view remains that in the BLS’ data will stay reliable. Still, a new commissioner could, over time, influence models and weights to an extent that some biases may creep in.

3. Changing personnel: We’ve written before about the attempted firing of Fed Governor Lisa Cook on the grounds of alleged mortgage fraud. The intent was clear: secure a Fed board majority sympathetic to the Trump administration. Such a majority could significantly influence the composition of the FOMC in February 2026, when all twelve regional Federal Reserve Bank presidents are due for reappointment. This could result in the replacement of several members and a more dovish FOMC, shifting policy towards more aggressive easing. Lower courts have ruled that Cook may remain until the Supreme Court decides. The SC will handle the case in January, before the February threshold, but earlier rulings give us hope that the court will ultimately rule in her favour.

4. Reframing the mandate: A subtler, and arguably more concerning, challenge is the debate over the Fed’s core responsibilities and mandate. For example, Treasury Secretary Bessent’s campaign against “mission drift” seeks to reframe quantitative easing (QE) and climate-related analysis as outside the Fed’s “core” remit. In an article for International Economy, Bessent argued against the Fed’s “Gain of Function.” By linking technical tools to distributive and moral outcomes, Bessent advocates for a smaller, simpler Fed. While ostensibly in the public interest, this looks convenient in the context of Trump’s general undermining of Fed independence. Another example is the renewed emphasis on the 1977 Federal Reserve Act’s reference to “moderate long-term interest rates” as a third mandate. This unconventional interpretation is used to justify policies such as capping long-term yields or resisting balance-sheet reduction, ostensibly to protect the Treasury’s borrowing capacity. In effect, this subordinates inflation control to debt sustainability, aka fiscal dominance.

If the mandate’s meaning is flexible, independence becomes conditional: the Fed is independent only within the administration’s evolving interpretation of its role. So while we do not expect an explicit loss of independence and an aggressive easing path fully disconnected from the business cycle, we are keeping an eye on more subtle developments that alter the bounds within it operates.

Aggressive easing will overheat economy

What if the Fed complied with Trump’s demands and eased policy aggressively? For simplicity, assume the Fed cuts rates 300bp by June, taking the upper bound of the fed funds rate to 1.00%, and that financial markets fully price in this move. This would be an aggressive easing cycle, comparable to – or even bigger than – that seen during the global financial crisis.

In an unconstrained economy, model-based estimates suggest 300bp in cuts would boost growth by c1.5pp and raise inflation by c1.2pp. This GDP boost would typically result in a 1pp fall in the unemployment rate. However, these results do not apply to the current economy, where supply is constrained in several ways (see also the US Outlook). The economy is not in recession, the labour market cannot accommodate a full percentage point drop in unemployment, and long-term rates would likely rise sharply due to inflation expectations. Cutting rates by 300bp under current conditions would therefore push the US economy against multiple constraints, leading to overheating. Our modelling indicates GDP would see only a modest boost of 0.4pp, and unemployment would barely fall. Instead, wage pressures and inflation would intensify. Core PCE inflation could rise by 2pp if inflation expectations remain anchored, but by much more, and for longer, if they become unanchored. In such a scenario, the appropriate policy response would be to hike rates aggressively. The timing of all of this would be uncertain, but at best, this would create far more volatility in both inflation and employment, and at worst a recession.

What would this mean for the ECB, Europe, and beyond?

Aggressive Fed easing under political pressure would likely tighten financial conditions in the eurozone rather than loosen them. Long-term US yields could rise as inflation expectations climb, spilling over into higher European bond yields and raising borrowing costs for households and firms. At the same time, the euro would likely strengthen – perhaps significantly – reducing imported inflation and hurting export competitiveness. ECB staff simulations suggest that a 10% euro appreciation could knock around 0.2pp from GDP growth while pushing inflation well below target. In contrast to the undershoot we expect for next year (see eurozone), which is largely driven by oil prices, a much stronger euro would also dampen core inflationary pressure, and is therefore more likely to prompt a policy response from the Governing Council. ECB models suggest up to 100bp of rate cuts could be necessary to correct the inflation undershoot, implying the deposit rate falls to 1%.

Globally, the stakes are even higher. The Fed anchors the world’s reserve currency and the most liquid safe asset: US Treasuries. A loss of confidence in the Fed’s independence would undermine these pillars, triggering market volatility, weaker investment, and ultimately tighter financial conditions worldwide. In a severe scenario, erosion of trust in US assets could gum up the global financial system, echoing the instability of the 1970s. In short, an “over-easing” Fed risks not only US overheating but also a destabilised global monetary order. In our Special on US institutions, we also discuss the potential upside for Europe in this scenario.

----------------------------

Box: The Fed Five: The shortlist to succeed Powell

As Chair Powell’s term nears its end, Treasury Secretary Scott Bessent has named five candidates to succeed him. Two are current board members: Christopher Waller, a long-serving Fed official known for his loyalty and early support for rate cuts and for criticizing the Fed’s “mission drift,” especially on climate policy; and Michelle Bowman, the vice president overseeing banking supervision, who is working to ease bank regulations in line with the administration’s goals. Both have advocated for policy easing this year but are considered unlikely to be chosen.

The three other candidates are: Kevin Hassett, Chair of the National Economic Council and a key figure in the first Trump administration (and currently seen as the frontrunner); Kevin Warsh, a former Fed Governor now at Stanford’s Hoover Institution; and Rick Rieder, Chief Investment Officer at Blackrock. Notably, these candidates have publicly discussed not only the need for lower rates but also significant changes to the Fed’s governance, such as reducing its size and scope, limiting public communications by central bankers, holding fewer meetings, and scaling back quantitative easing, echoing Bessent’s editorial stance. Frontrunner Kevin Hassett has been relatively quiet on the institutional makeup of the Fed, but is widely viewed as loyal to President Trump.

While the Fed Chair wields significant influence, their power is not absolute. As shown by Stephen Miran’s experience, a single vote cannot override the technocratic FOMC members without strong economic justification. The risk of a policy shift increases if Lisa Cook and the regional Fed presidents are replaced by members more aligned with the administration. However, deeper institutional changes would require reinterpretation or amendment of the Federal Reserve Act, typically involving Congress.

Traditionally, the outgoing Chair steps down from the Board, but Powell’s term as Governor runs until 2028, even if his Chair role ends in May 2026. If Powell decides to step down, his replacement will tilt the Board of Governors balance to a majority sympathetic to the Trump administration, but crucially past the February reappointment window. If Powell decides to stay on the board, he might serve as somewhat of a Shadow Chair, and potentially quite an influential one.

----------------------------

2. Will global trade’s remarkable resilience in the ‘Year of the Tariff’ continue in 2026?

Arjen van Dijkhuizen

Annual global trade growth is expected to slow materially in 2026, following a remarkable acceleration so far this year.

In our 2025 Global Outlook, The Year of the Tariff, we analysed the potential impact of the anticipated rise in US import tariffs on the global economy. Our projections for the tariff increase, as set out in our 2025 Outlook, have so far proved quite accurate on balance, but the tariff impact on the global economy has generally been less severe than was widely anticipated earlier. Let us take global trade as an example. The two leading institutions publishing global trade forecasts, IMF and WTO, both slashed their 2025 global trade forecasts in April 2025 following the announcement of global high ‘reciprocal’ tariffs on ‘US Liberation Day’, but raised them again in subsequent revision rounds (see chart).

Why is that? Several specific factors (beside shifts in macro policies) contributed to the resilience of global trade this year:

1. Trade frontloading by US importers and foreign exporters in the run-up to anticipated higher US tariffs. This put the base for global trade higher in early 2025. Imports into tariff ‘epicentre’ the US surged in Q1, driving a spike of three full points in CPB’s world trade volume index in March (see chart). This index fell back somewhat in April-June as the effects of trade frontloading unwound, but nowhere near as much as we had expected, and imports even began to rise again in July. The global CPB index shows an average of 4.9% y/y growth in world (goods) trade volumes in January-August 2025 (2024: 2.2%). Due to these shifts, the forecasts for 2025 have been revised up, but the forecasts for 2026 have been generally revised down, meaning that the tariff impact is more spread out than previously expected (albeit also generally lower on aggregate).

2. Restraint with retaliation and trade deals leading to lower tariffs. Most countries(except China) have exercised restraint in terms of retaliation, while trade deals with the US resulting in lower tariffs have also helped to mitigate the ultimate impact. According to Bloomberg data, the US effective tariff rate has fallen from ±27.5% mid-April to ±16% currently, although staying much higher compared to the pre-Trump 2.0 episode.

3. Trade rerouting/diversification: China in particular has offset falling direct exports to the US by rerouting trade (exporting to US through ASEAN) and diversifying to other destinations, including Europe. Rerouting can actually inflate trade volumes, as reported export flows may be double-counted when trade becomes ‘triangular’. Europe, by contrast, has so far been less successful in offsetting the sharp decline in exports to the US through diversification (see left-hand chart below).

4. AI boom: The WTO found that stronger demand for AI-related goods was a key driver of global trade in 1H-25.

5. Limited pass-through in the US: The pass-through of higher tariffs into US consumer prices appears to be taking longer than earlier anticipated, which has softened the drag from tariffs on US consumption and imports so far.

In our base case, we still expect tariffs to have a dampening effect on global trade growth as we move into 2026. The global manufacturing PMI’s export sub-index fell back deeper into contraction territory in October. While US tariff policy uncertainty has clearly declined since April, it remains higher than in the pre-Trump 2.0 period (and some legal uncertainties also remain). The recent flare-up of tensions between China, the US and other countries – including the Netherlands – around chokepoints (see box) is a reminder of how fragile the current equilibrium is. That said, the recent US-China deal prevents a harmful re-escalation for now, and gives both countries time to continue decoupling in sectors deemed strategically important. Nevertheless, trade spats between China and the US/West are likely to flare up from time to time, particularly around chokepoints. Other factors likely to dampen global trade growth in 2026 include the expected fading of trade rerouting (as US-China trade stabilises and tariff differences between China and ASEAN narrow), and the delayed pass-through of tariffs into US consumer prices, which will weigh on US consumption and imports going forward (see US chapter). On the other hand, policy support in the key economies, and higher defence imports from Europe, will likely cushion global growth and trade.

All in all, we expect an end to some of the special factors that boosted global goods trade this year to result in slower growth next year. Taking into account base effects from 2025, we expect annual growth in global goods trade to slow materially in 2026, in line with the recent IMF and WTO forecasts.

---------------------------

Box: Chokepoint tensions between US and China put on hold, but chokepoint related risks remain material

The escalation of US-China trade/tech tensions in the course of this year initially was centred around new US import tariffs, but following the truce agreed in Geneva mid-May this focus shifted more and more to ‘chokepoints’ (see here). Both countries are trying to increase leverage over each other regarding products where one country has a dominant position in production/exports and the other has a critical dependence. These chokepoints in bilateral trade are concentrated in advanced semiconductors (US versus China) and rare earth metals and related products (China versus US). Over the past few years, the US has spun a broad web of export restrictions on semiconductors, which are used to contain China’s access to the most advanced chips and related machines. This escalated in late September, after the US Department of Commerce decided to expand the entity list used for these export controls with foreign subsidiaries that are at least 50% owned by companies already on that list. This led to a massive expansion of the number of companies that are affected by this policy, with the number of Chinese firms rising from around 1,500 to around 20,000. This ruling also affected the Dutch semiconductor producer Nexperia, due to its Chinese owner.

China responded in October with a considerable tightening of its export controls for rare earths and related products, presenting a host of new measures on top of the controls that were already rolled out in April (see here). New rare earth elements were added to the list, raising the total number of restricted elements to 12 (out of 17), and control measures for foreign producers and foreign users of semiconductors were also sharply tightened. Beijing also threatened to block exports of rare earths for military purposes. These measures were primarily aimed at mirroring the US export control regime for semiconductors. In this way, Beijing created proportional leverage in future trade negotiations. In addition, it gave Beijing potential leverage over countries that are willing to team up with the US against China (this partly explains why the controls were not only for US-bound exports).

China’s matching strategy proved effective. A key outcome of the Xi-Trump meeting end-October was the postponement of the expansion of the US entity list by one year in exchange for the postponement of China’s new export regime for rare earths (and other concessions, see China chapter). This has settled one of the most thorny issues in US-China relations for now, and gives both countries more time to work on decoupling from each other in these areas. It looks as if China is more ahead in that respect. China has already been working for years to reduce its dependence on semiconductor imports from the US. The US also stepped up efforts to diversify its supply chains of rare earths and related products. Still, China remains the dominant global player, with a 60% share in mining, a 80-90% share in processing/refining, and almost half of total global reserves.

The truce also reduces the risk of massive disturbances in global industrial supply chains for now, also for European producers. However, chokepoint-related risks are still material, as is for instance shown by the Nexperia case. Expanding policies to reduce risks in this respect has also gained greater priority for the EU, which aims to step up defence spending (see Eurozone chapter). Brussels is putting more and more effort into risk-mitigating its rare earths supply strategy, in line with the European Critical Raw Materials Act (see here).

---------------------------

3. From Steam to Silicon: How AI is transforming the economy

Rogier Quaedvlieg and Philip Bokeloh

AI’s impact on the economy is more comparable to the industrial revolution than recent tech innovations.

The pace of development in artificial intelligence (AI) is astonishing. In just a few years, AI companions have become commonplace in our lives and workplaces. Analyses, which used to take hours, now take mere seconds. The generation of images, and videos, for which one would have to hire professionals, can now be done by anyone with a computer. This technology has already impacted the way we work, and early signs of its effects on the broader economy are emerging. As highlighted in our US Outlook, investments by, and valuations of, major companies are driving much of US economic growth. Here, we document further evidence of AI’s influence on the US economy and speculate on its longer-term trajectory. A separate box considers the implications of the global AI race for Europe.

The impact of technological change on the economy

Technological innovation is typically expected to boost productivity and therefore destroy jobs. However, as the chart on the left below shows, the introduction of recent innovations such as computers and the internet did not lead to significant job losses. After computers, productivity grew slightly above trend, as did employment. The internet era saw productivity rise faster, but not at the expense of employment growth. While headline figures are not a substitute for rigorous analysis, academic studies generally reach similar conclusions: productivity gains are cumulative but modest year-on-year, while adoption is gradual enough for workers to transition and for new sectors to emerge. Indeed, MIT economist Autor (2015) finds that automation often complements human labour, creating jobs even within industries where technology enhances productivity.

Since ChatGPT’s launch in November 2022, the US has seen a decline in four-year average productivity but rising employment growth. While some employment measures have stalled this year, much of this is due to supply factors. Meanwhile, as the right-hand chart below shows, adoption of AI is much faster than for other technologies, which could accelerate job displacement and test the economy’s ability to create new complementary roles.

A more detailed assessment of AI’s impact on the labour market

Technological change inevitably affects jobs. Recent innovations have typically increased demand for highly educated workers, while reducing it for the less educated. With AI, the impact could be broader. This technological innovation is capital-biased rather than labour-biased, and could therefore potentially affect all jobs, albeit to varying degrees. AI is highly capital-intensive: S&P estimates the four largest hyperscalers (builders of infrastructure) will spend over $2 trillion in capital expenditure by 2030. This makes AI more akin to the Industrial Revolution than to the computer or internet eras. The capital intensity suggests that gains will accrue mainly to investors, putting pressure on wages and jobs, especially if AI investment crowds out other sectors. On the bright side, the lesson from the industrial revolution is one of long-term prosperity for everyone. Unfortunately, it might take a while for all to reap the benefits, however.

The fact that the impact of AI on labour markets is broader has some silver linings. Previous technological shifts often hit local labour markets hard, wiping out most jobs in specific towns, amplifying local downturns. With AI, job losses are likely to be more dispersed, making them easier to absorb. Recent developments in the US labour market suggest that job losses are indeed not concentrated geographically, but rather demographically. Young graduates are struggling to find work, with highly educated but inexperienced workers most affected. This is not just an AI story. College educated workers have been losing their relative advantage, also before ChatGPT, and the downturn can also partially be attributed to an unwind of the post-pandemic labour hoarding. Still, sectoral data using AI exposure measures (Eloundou et al., 2024) does show slower job growth, or even losses, in sectors most exposed to generative AI. Overall employment is still growing, but this is just the start of adoption. AI developments put jobs in the 90–99% and 50–90% exposure buckets at particularly high risk.

Can we expect a productivity boost?

Will productivity gains appear as quickly as labour market effects? Rapid adoption does not guarantee immediate productivity improvements. Other bottlenecks (e.g. regulatory, energy, or human) may slow progress. For example, medical innovations may be delayed by regulatory approval, and energy or grid constraints could become limiting factors. It can also simply be humans. It could be due to slow adoption, or it could also be that humans are simply holding AI back being the worst performer in the team.

The longer-term trajectory on productivity is highly uncertain. It could be that AI leads to a one off-boost in productivity, or, more likely, lead to a faster pace of innovation, perpetually boosting productivity growth. A big difference with previous technological innovation is the potential of AI to further its own development. In the near term, the scale of investment suggests the innovating firms expect significant productivity gains, and an ability to monetise their AI products. Estimates of GenAI’s economic impact vary widely: we’ve seen estimates ranging from about $72 billion of revenue by 2028, up to a staggering $1.1 trillion by that same time. The higher figure would represent nearly 1% of global GDP. In competitive markets, Nordhaus (2004) suggests value extraction is capped at about 25%, implying GenAI would, in the most optimistic revenue estimates, have to boost global GDP by almost 4% over the next three years. It’ s clear that AI will transform the economy, but it’s far from clear that we will see such a generational leap in productivity within the next three years.

--------------------------

Box: Has Europe already lost the AI race?

We have already discussed the impact of AI on the US economy. But the rest of the world is not sitting still. China, too, is advancing, though the scope of its investment in the AI race is harder to discern in macroeconomic figures. Indirectly, there are indications that AI is leaving its mark on the economy. For instance, it seems likely that the rise of energy-intensive AI is driving up energy consumption. In China, the process is more actively state-led compared to in the US. For instance, the government is promoting “AI+” policies that embed AI into industries and education, there is a national AI Fund, and a strong government push with industrial policy to advance China’s chip-making capabilities. A big win this year was DeepSeek, which offered a low-cost and open-source alternative to US competitors.

Meanwhile, progress in Europe is substantially slower. In specific areas, there are stand-out leading companies, such as ASML, but in terms of scale and impact, European companies cannot compete with their American or Chinese counterparts. Part of the explanation lies in the less-developed capital market, which makes it more challenging to secure funding for risky investments. Another contributing factor is the stringent regulatory environment in Europe. Consider, for example, the Digital Services Act, the Digital Markets Act, the Data Act, and the AI Act, along with additional regulations from individual member states. The protection of privacy and copyrights complicates the training of AI models using large datasets, causing delays in their development.

The fact that Europe is not a frontrunner in the AI race does not however mean that it won’t be able to reap the rewards. Europe possesses two critical conditions for the application of AI: an educated population and an extensive knowledge infrastructure. Moreover, there is considerable potential economic gain from the application of AI. A key finding of the Draghi report is that the potential of the common market is underutilised due to trade barriers arising from regulatory differences and language barriers. Even if the EU may struggle to further reduce these barriers, AI may very well lower their negative impact.

--------------------------

4. Which sovereign debtor should we worry more about: France, or the UK?

Bill Diviney and Sonia Renoult

France is a much bigger worry than the UK in the near-term.

France and the UK’s fiscal woes have been a regular theme in the financial markets news cycle in recent years. If the various headlines – and bouts of bond market stress – are taken at face value, both are in a dire fiscal situation, perhaps even a doom loop of low growth and unsustainable debt. Or are they?

Actually, France is in a much worse shape – and not simply because of its merry-go-round governments over the past year (four and counting, lasting 3 months on average). France has struggled much more than the UK to get a grip on its deficit, while debt dynamics are made worse by its very low inflation rate (c1%) and interest rates that are clearly too high for France’s economic circumstances. Given the political paralysis that has made relatively modest reforms next to impossible – such as raising the pension age to international standards – we have little confidence that France will be able to implement the necessary fiscal adjustment to return debt to a sustainable trajectory in the near future. Elections, whether of the snap variety or as scheduled in 2027, bring as much risk as they do promise: if the far right or the far left come to power, France may end up with a Liz Truss-like government that cares little for bond markets, unless a crisis erupts. This could well end up bringing things to a head. A more benign scenario would see the return of a centrist majority that seeks to more aggressively rein in France’s high deficits. The middle ground would be a continuation of what we see now: a slow-burning muddling through, with France’s debt ratio on a gently rising trajectory, but not alarming enough that it causes a crisis.

Depending on your perspective, France’s membership of the eurozone is both a blessing and a curse. Blessing, because it affords France financial market stability, but a curse because it insulates France’s politics (and its public) from the pressure that is arguably necessary to make hard choices around government spending. Still, although France is not eligible for the ECB’s Transmission Protection Instrument, investors expect – rightly or wrongly – that the ECB probably would step in should a genuine crisis erupt in French bond markets. It is this implicit backstop that is likely to keep French bond markets relatively calm, albeit with a continued elevated spread over German bonds over the coming years. An exception to this is in the scenario where a far left or far right win a majority in the Assembly, which may well test the bond market’s – and the ECB’s – resolve to its limits.

For the UK, the risk is bigger in 2029

As we will argue in the upcoming Part II of our Ugly Contest series, the UK has been broadly sticking to its self-imposed fiscal rules, and for this reason we see the risk of a bond market crisis over the coming years to be relatively low. Part of the reason for this is that the UK already went through a mini crisis two years ago, following the infamous mini budget of the ill-fated Liz Truss government. The trauma from that episode, when bond yields and – importantly for voters – mortgage rates surged, caused lasting damage to the centre-right Conservative Party’s reputation for sound economic management. But the ghost of Liz Truss is also what hangs over Chancellor Rachel Reeves whenever she makes fiscal pronouncements. Even the smallest of market ructions, driven by speculation that the government might water down its commitment to the fiscal rules, have to led immediate responses from the government reaffirming said commitment. While the UK government has – like France – struggled to rein in spending (for instance, it failed to pass reforms to disability benefits earlier this year), it also has a lot more space on the revenue side to bring the deficit back down. Given this, we expect the UK to comfortably meet market expectations in announcing new revenue-raising measures at the Autumn Statement budget announcement on 26 November, and in our base case, we expect debt to broadly stabilise as a share of GDP, if not decline slightly, over the coming years.

But the UK faces its own significant risks come the late 2020s, the biggest being the possible election of a radical right-wing government in 2029. Nigel Farage’s Reform party is currently leading in the polls, and due to the UK’s antiquated first-past-the-post electoral system, it could well get a parliamentary majority with just 1/3 of the vote. Reform has threatened to exert the same kind of political pressure on the Bank of England that we are seeing now in the US with the Federal Reserve. With the UK lacking the exorbitant privilege of the dollar, the main risk in the UK would then transfer from bond markets to the currency. With the current moderate Labour government having a sizable majority, this risk remains remote for the time being – and much can happen in the meantime politically. But it will be a risk we will no doubt return to in future.

5. What are the European implications of the Dutch election outcome?

Jan-Paul van de Kerke, Aggie van Huisseling and Anne de Clercq Zubli

While coalition formation is expected to be lengthy and difficult, the incoming Dutch government will likely be more centrist, pursuing a more pro-European course.

“Europe risks stagnation if we fail to deepen integration. The Netherlands helped found the Union, now we should help shape its future,” were the words of election winner and possible next PM Rob Jetten of the liberal progressive D66, exclaimed following the Dutch snap elections held at the end of October. Pro-EU centrists D66 and the Christian Democratic CDA won a large number of seats in parliament, and this signals a more constructive stance towards the EU in the fifth biggest economy in the eurozone. Given the Netherlands' traditional conservative stance on EU integration and treaty reform, its preference for frugal spending, and reputation for punching above its weight in EU policymaking, we examine the implications of the Dutch election outcome for EU policy direction.

Despite the strong showing of the centrist parties, the message to Europe from this election is not entirely clear. Yes, far-right firebrand and Eurosceptic Geert Wilders and his PVV party lost 11 seats, but the (far-)right did not decline in aggregate, it just became more dispersed (FvD and JA21 gained seats). The PVV’s losses align with those of his former coalition partners under the Schoof 1 cabinet; the VVD (centre-right), BBB (right-wing) and NSC (centre-right) all lost seats.

Moreover, political fragmentation in the Netherlands remains high. D66 is the election winner but is the smallest winner in Dutch history, with no party securing more than a fifth of the votes. This signals arduous and probably lengthy coalition negotiations ahead. The resurgence or rather concentration of the two winning centre parties, D66 and CDA – which are almost certainly forming the engine bloc of the next coalition – means the next Dutch government is expected to be more centrist. While this holds true, the exact extent depends on the outcome of coalition talks that have just begun. Currently, two coalition options are being considered. The first is a centrist coalition, where the CDA and D66 engine bloc is complemented by GL-PvdA (centre-left) and the VVD. This option is arithmetically favoured, as it would have a comfortable majority with 86 seats in parliament. The second option is a centre-right coalition, including VVD and JA21 (Eurosceptic, right-wing). However, this coalition falls short of a majority, with only 75 seats, and may require additional parties to complement it.

On the policymaking front, the Dutch elections do send an important message to Europe. The incoming Dutch government is expected to adopt a significantly more constructive approach towards European integration than its predecessor. While D66 has consistently been pro-European, there now appears to be a majority among the six largest parties in favour of advancing integration on key EU policy issues. The table below summarises the manifesto positions of these parties on key EU policy areas. There is broad consensus on enhancing the Single Market and advancing the Capital Markets Union. However, there is less agreement on common financing, such as Eurobonds.

Unlike previous years, however, there is now room for negotiation. D66 and GL-PvdA are supportive of Eurobonds, while CDA has softened its stance and shown openness on this sensitive issue during the election. The specific coalition and ensuing negotiations will ultimately shape the level of constructive engagement and outcomes on joint financing; for instance a centre-right coalition with JA21 may be more outspoken against common financing. Nonetheless, the incoming government is expected to adopt a more constructive position than the previous administration. This could portend a change in the EU policy landscape at a time of negotiations on the multi-year EU budget, the debate on common EU financing for innovation and defence, and decisions on long-term financial support for Ukraine.