VoxEU @ CEPR

@voxeu.org

6304 Followers 1314 Following

VoxEU – CEPR’s policy portal - promotes "research-based policy analysis and commentary by leading economists". VoxEU columns cover all fields of economics broadly defined and are widely read.

Europe is converging on a definition of economic security as reduced dependence – de-risking, reshoring, buying European. This column argues that the definition is mistaken. Three decades of evidence on financial and production networks shows that resilience is a property of a node’s position in the network – the depth, diversity, and quality of its connections – not of the number of connections it has severed for self-sufficiency. Applying this test to critical input supply chains such as energy and defence – with Turkey, the EU’s largest customs union partner, as the defining case – the author argues that Europe passes the initiative’s ‘Singapore test’ only if it treats integration depth, rather than self-sufficiency, as the union’s security variable.
Can reducing trade make conflict more likely? In the latest #VoxTalks, recorded at the PSE-CEPR Policy Forum, Isabelle Méjean discusses the "Safety Paradox", economic coercion, and why trade can act as insurance. 🎧 Listen now: https://ow.ly/wI5I50ZrqSK #EconSky
European banking remains fragmented, leaving billions of euros inefficiently allocated. Reforms to establish a European deposit insurance system have been proposed to allow cross-border banking groups to operate more freely. This column argues that a common deposit insurance system fundamentally changes the link between risk-sharing and risk-taking. In addition to greater integration, it can encourage more risk-taking, with unclear effects on overall welfare. Therefore, greater freedom to move capital and liquidity across borders should be accompanied by sufficiently strict supervision to address the risk-taking incentives created by such reforms.
Figure: Panel A shows that integrated training programmes combining work-based learning with classroom-based instruction are more consistently associated with positive employment outcomes than stand-alone on-the-job training or internships. Panel B shows that programmes leading to recognised qualifications are, on average, more effective than those leading to certificates that are not quality-assured. As digital, environmental, and demographic transformations reshape labour markets, skills influence who benefits from these changes and who falls behind. This column presents the findings of an International Labour Organization report that shows that socio-emotional and cognitive skills are in high demand and that employers seek rounded skills profiles. However, most countries still lack ambitious lifelong learning systems. The inclusiveness, quality, and adaptability of such systems will influence how economies and societies navigate structural transformation.
As the postwar international order fragments and universal agreement becomes increasingly elusive, policymakers face a pressing question: what form of cooperation can replace the postwar hegemon-led, rules-based system? This column examines two complementary answers. Mark Carney argues that coalitions of middle powers can preserve stability, resilience, and strategic autonomy in an era of great-power rivalry. The author of this column proposes a broader architecture of pivotal power coalitions organised around human flourishing, policy alignment, and polycentric governance. Their dialogue illuminates one of the defining policy challenges of our time: how to combine geopolitical realism with effective collective action. Together, these approaches offer a practical roadmap for building a more resilient, cooperative, and problem-solving world order.
Figure shows GDP per capita vs. proportion of labour force with occupational licences across nations Occupational licensing has attracted growing attention in advanced economies, yet little has been known about its prevalence outside the US. This column presents new evidence from 44 countries using nationally representative surveys. It shows that licensing is widespread around the world and, surprisingly, is often more common in developing economies than in richer nations. Higher occupational licensing is associated with lower per capita incomes and higher informal employment rates. These patterns suggest occupational licensing deserves greater attention as part of the broader debate over labour market regulation, informality, and economic development.
Figure shows how the convenience yield on public dollar safe assets has fallen. Measured relative to currency-hedged euro funding, the premium on Treasuries and agency debt has compressed from its 2022 peak towards zero. The dollar's role as the world's reserve currency has been blamed for an overvalued exchange rate that hollows out US manufacturing. This column quantifies what would happen if foreign investors stopped treating US safe assets as international liquidity. In a calibrated two-country model, the exchange-rate consequences are present but modest – a permanent real depreciation of roughly 9% – while the larger effects fall on US interest rates, which rise by around 90 basis points, and on national wealth, where the lost seigniorage is worth close to one year of GDP. The recent decline in convenience yields and foreign ownership of US public debt suggests the transition may already be under way.
The map shows, for each oil-bearing grid cell in the sample, the predicted decline in conflict probability had oil prices remained at their December 1998 level instead of rising to their July 2008 peak. Cells are grouped at cut-offs of 36% and 45%, the 25th and 75th percentiles of positive predicted effects: 938 cells show predicted declines of up to 36%, 1,874 cells between 36% and 45%, and 932 cells above 45%. Oil price shocks do not simply make conflict more likely everywhere; they change where violence happens. Using monthly georeferenced data for 131 low- and middle-income countries from 1989 to 2021, this column shows that higher oil prices pull organised conflict towards petroleum fields, while protests and attacks rise around refineries and pipelines. Aggregate conflict statistics can therefore look calm while the local map of violence is being redrawn.
Figure shows treated counties (red) and the group of control counties (blue) in China. There is growing evidence that place-based industrial policies lead to positive economic effects. This column analyses the “Million-Rouble Plants” programme of the 1950s when the Soviet Union helped China build state-of-the-art industrial facilities. By tracking the locations that benefitted, it finds positive economic spillovers in the short and medium run. Although the Million-Rouble Plants themselves continued to thrive, the performance of other local establishments eventually declined, as increasing specialisation weakened knowledge spillovers and constrained local entrepreneurship. Policymakers may achieve more durable gains by fostering diversified local production structures that encourage knowledge spillovers, labour market resilience, and adaptation to technological and market shocks.
It's another #VoxTalks episode! This week, Julia Cagé chats with Tim Phillips about a nationwide experiment exploring whether free access to quality #news and #media education can change teenagers' news consumption. 🎧 Listen: https://ow.ly/C9XN50ZoUUn 🔷 Recorded at the 2025 CEPR Paris Symposium #EconSky
Figure: summarises the evolution of banking regulation over the past century. The left-hand panel shows structural regulation. The middle panel reports prudential regulation. The right-hand panel combines both dimensions into an aggregate index. It illustrates the classic regulatory cycle throughout much of the 20th century: regulation increased sharply after the Great Depression, gradually weakened during the era of financial liberalisation beginning in the 1970s, and then stabilised after 2008, without anything resembling the earlier regulatory backlash. The regulatory cycles view, according to which periods of deregulation encourage greater risk-taking and credit booms, which culminate in financial crises and are followed by renewed regulation, rests largely on case studies of individual countries or particular crises. This column uses a new dataset covering 14 countries over the past century to show that the pattern observed after the Great Depression has been much weaker since the Global Financial Crisis. The authors argue that the reason is not simply a change in political priorities or regulatory philosophy. Increasing financial integration has made effective banking regulation at the national level progressively more difficult, shifting regulatory authority from national governments towards international coordination.
Globalisation has changed economic research profoundly, making it more comparable, more competitive, and thoroughly quantitative. This column argues that AI may unsettle this new order as thoroughly as globalisation unsettled the old one and discusses the technical and geopolitical mechanisms at work. One prediction can be made with high confidence: someone who wakes up thirty years from now is likely to recognise economics as little as someone who fell asleep thirty years ago and just woke up today.
Figure: The unadjusted and adjusted gender pay gap in the 27 EU member states. UGPG: Unadjusted gender pay gap (difference in average gross hourly wages between men and women, expressed as % of average men’s gross hourly wages). AGPG: Adjusted gender pay gap (difference in average gross hourly wages between men and women performing similar work, as % of average male average gross hourly wages). Equal pay for equal work has been a core EU principle since 1957, yet women across Europe continue to earn less than men, even in very similar jobs. As EU countries begin implementing the new Pay Transparency Directive, greater wage transparency is expected to narrow these gaps. This column explores the potential effects on the 27 EU countries. Equal pay could substantially narrow gender disparities, but the gains are greater where unequal pay among comparable workers is a key driver of the gender pay gap.
Are Net Zero Commitments #Greenwash? Simon Dietz discusses research on nearly 2,000 firms. Net zero commitments aren't linked to immediate emissions cuts, but are associated with an early shift in strategic climate management, often before they're announced. 🎧 https://ow.ly/pzx650ZobTO #EconSky #VoxTalks
Figure: Contributions to growth in US output per hour. Each bar is labelled by the calendar year(s) whose Q4-over-Q4 growth it contains — e.g., 2005–19 averages quarterly (log) growth from the beginning of 2005 to the end of 2019. 2025–26* is the average annual growth rate over the five quarters ending 2026Q1 (the latest data). Dashed line: average labour-productivity growth, 2023–26. Capital deepening is the contribution of capital relative to quality-adjusted hours; total factor productivity is measured as a residual.
Figure shows weighted-average yield spreads over the ten-year sovereign benchmark for Credit Suisse's AT1, bail-in, and senior unsecured bonds in the months before the merger, announced on 19 March 2023. The bail-in and senior series are restricted to bonds with at most four years of remaining maturity to keep duration broadly comparable. Trading resumed on 20 March 2023. When Credit Suisse failed in March 2023, Swiss authorities, against market expectations, set aside the resolution plan and arranged a UBS takeover backed by public guarantees, sparing the bail-in bondholders who were next in line to absorb losses. This column traces how that choice reshaped investor beliefs about bail-in across Europe. Spreads on bail-in debt narrowed, the funding penalty on weaker banks shrank, and yields moved less on earnings news. Each points the same way: investors now see bail-in as less likely, a sign that its credibility has declined.
Europe pays for innovative medicines through 27 uncoordinated national negotiations, generating access delays, cross-country pricing externalities, and a weak position vis-à-vis both pharmaceutical firms and US trade and most-favoured nation pressure. This column argues that Europe should procure innovative on-patent medicines through a single European negotiation that fixes one wholesale price for the entire Single Market, while reintroducing equity across member states through differentiated national contributions and side transfers. Decoupling what the innovator receives from what each country pays would end the pathologies of external reference pricing and parallel trade, restore Europe's bargaining power, and guarantee simultaneous access for all European patients.
The EU's fiscal rules in force since 2024 are often criticised for their complexity. The official implementation requires thousands of lines of code and a sophisticated debt sustainability analysis. This column argues that the central result can be approximated by a remarkably simple formula: the required primary balance equals the largest debt-stabilising primary balance in a ten-year period plus roughly 1% of the debt ratio. While this simplification cannot replace the full calculations, it helps demystify the debt rules and provides a shortcut for assessing the impact of policies on debt sustainability.
Figure shows real GDP in Italy, Spain, and Greece compared to the control group. By 2025, real GDP was 10.8% above its 2019 level in Greece, 10.2% higher in Spain, and 6.4% higher in Italy, compared with an increase of only 5.4% in the control group The EU Recovery and Resilience Facility was launched to support post-Covid recovery while accelerating structural transformation through investment and reform. This column provides a preliminary descriptive assessment of the macroeconomic impact of the facility in Italy, Greece, and Spain, three countries with large allocations. Using simple counterfactual comparisons, it presents evidence consistent with a positive impact on real GDP, and an even stronger impact on employment and investment in the three countries. The evidence also points to an expected positive impact on potential output in the years ahead.
The figure shows each country’s fair share of the remaining carbon budget under the allocation principle of equal cumulative emissions per person. Fair shares are computed as the population-weighted claim on the global remaining carbon budget adjusted for each country’s historical carbon debt accumulated over 1870-2022. Historical emissions are discounted at an annual rate of 2%. Positive values indicate remaining emission space, while negative values indicate that a country’s historical emissions already exceed its equal per-capita allocation. Values are expressed in GtCO2. The idea of equal cumulative emissions per person is simple: every individual should have an equal claim on the atmosphere regardless of where they live. Using historical population and emissions data from 1870 onward, this column examines how the world’s remaining carbon budget could be allocated fairly across countries. European states have already consumed more than their share, whereas many developing countries retain positive claims. Allocating the remaining budget fairly can be compatible with development goals if developing countries can access cleaner technologies. Europe can start by redesigning the Carbon Border Adjustment Mechanism to embed explicit fairness principles.
Figure: Summary statistics for cyclically adjusted and smoothed current account changes, by episode type. The sample covers 51 countries over 1980–2024. Current account changes are cumulative in percentage points of GDP; values are medians across episodes. Global imbalances are back in the policy debate. This column examines 70 current account adjustment episodes across 51 economies from 1980 to 2024 to discover how large imbalances have unwound in the past. Three lessons stand out: adjustments are fairly common but slow; deficit and surplus narrowing work through different channels; and larger imbalances – especially deficits – are significantly more likely to correct.
The macroeconomic theory that is taught in graduate schools, and used in the conduct of monetary and fiscal policy around the world today, emerged from a revolution in economics that was led by Ned Phelps. This column discusses Phelps’s theoretical contributions and how they created a micro foundation for macroeconomics.
The figure presents treatment effects on posterior beliefs about domestic and foreign AI investment and on the AI investment rate, comparing control and treatment groups. Confidence intervals are at the 95% level. The figure is based on the Q4–2025 wave of the survey. Europe has devoted substantial resources to fostering innovation and AI diffusion, through both centralised EU initiatives and national programmes. This column combines an experiment on 3,316 euro area firms with data from a large firm survey to demonstrate that that the decision to adopt AI is primarily driven by competition pressures rather than informational spillovers. If firms systematically underestimate their competitors’ AI investment – especially those abroad – informational gaps could be driving underinvestment. Reducing invisible barriers to knowledge sharing across regions may be just as vital as financial support for closing gaps in AI investment and innovation across Europe.
Figure shows the effect of a one-degree origin-temperature anomaly on bilateral emigration, cumulative over twelve months, for all developing-country origins and for origins split at the median of long-run (1990–2018) baseline temperature. Climate-driven displacement is widely expected to push millions across international borders. Drawing on monthly bilateral flows for 127 developing-country origins, this column finds the reverse: a temperature shock reduces emigration rather than raising it, because climate stress cuts origin incomes and poor households cannot finance a costly international move. On this evidence, the ‘climate refugee’ is the exception, and hardship more often traps poor populations in place than sets them moving.
Figure shows the composition of international reserves. Panel (a) reports the composition of allocated official foreign exchange reserves. Panel (b) reports the market-valued gold share of total official reserves. Data on the Chinese yuan’s share are available from 2016 onwards. Central banks are rethinking reserve portfolios as geopolitical fragmentation raises concerns about the accessibility of foreign-held assets, while dollar-funding stress keeps the US dollar at the centre of global financial stability. Using high-frequency US monetary policy surprises and minute-level exchange rates for 18 economies, this column shows that dollar and gold reserves are associated with smaller currency depreciations, especially where official dollar-liquidity backstops are absent and dollar liabilities are large. The findings highlight the importance of maintaining liquidity in the currencies where stress is likely, alongside sound domestic funding structures, credible policy frameworks, and prudent limits on unhedged foreign currency debt.
FTTH exposure reduces the probability that a child is developing within normal ranges – as opposed to being flagged for monitoring or at risk of developmental delays – by 21 to 33 percentage points for communication and socio-emotional skills. A dose-response analysis shown in the figure confirms that the negative effects grow monotonically with the intensity of exposure, providing further confidence in a causal interpretation. The rapid expansion of high-speed internet has intensified concerns about how digital technologies affect early childhood. Using the staggered rollout of fibre optic to the home in Uruguay as a natural experiment, this column finds that greater lifetime exposure to high-speed internet reduces children’s scores in communication, problem solving, personal-social and socio-emotional skills. The effects are partly explained by indirect channels, through the displacement of valuable activities for child development and alterations in caregivers’ behaviour.
The figure displays difference-in-differences estimates contrasting individuals with higher and lower EIPC receipt. Panel (a) presents effects on annual pension income, confirming post-reform differences in pension income across groups, while Panel (b) shows the effect on unconditional labour earnings past retirement. The red dotted line marks the timing of the reform; capped vertical lines indicate 95% confidence intervals. Can governments raise pensions for low-income retirees without discouraging them from working? Drawing on Germany's Earned Income Pension Credit, this column finds that higher pension income reduces post-retirement labour supply, but the effect is modest and concentrated among retirees still close to the participation margin. Two-thirds of the response runs through whether retirees work at all, including the often-overlooked decision to return to work after an earlier exit.
The figure plots compound abnormal returns in the days following an award announcement; bars show 95% confidence intervals. Observations are positions (firms x individual) at different dates. Pre-award connection status is defined as attendance at the same grande école, in the same or an adjacent cohort, as a politician involved in the awarding process. State honours are designed as symbols of merit and public recognition. But they may also convey information with economic value. This column explores the market value of the Légion d’honneur. When corporate directors of French listed firms receive the honour, their companies’ share prices rise. This market reaction is concentrated among directors without a publicly observable educational link to members of the incumbent government before the award. The pattern suggests that state distinctions can provide investors with new information about a director’s potential access to policymakers.
Figure shows the EU share of total disbursements by sector, 2021-2026 (%). On the demand side, spending has been concentrated in sectors with strong short-term multipliers. In 2024, about half of RRF disbursements went to the construction sector. The EU’s Recovery and Resilience Facility has become a central test case for performance-based fiscal policy. Its experience offers lessons for the design of the next Multiannual Financial Framework. This column reviews evidence on the facility’s economic impact, implementation, and use of milestones and targets. It argues that the facility has supported demand, helped advance reforms, and strengthened EU fiscal integration, but that complex plans and administrative bottlenecks have slowed disbursement. Future EU budget instruments should link reforms and investment while keeping plans focused, flexible, and aligned with implementation capacity.
Figure summarises the results. Each coefficient is best interpreted as an elasticity. For example, a coefficient of 2.3 on the overcrowding rate implies that a country where housing costs as a share of disposable income are 1% above the EU average tends to have an overcrowding rate about 2.3% above the EU average, all else equal. Housing affordability has moved to the forefront of Europe’s policy debate. While the economic literature has mostly focused on the drivers of housing unaffordability, this column examines its consequences in the EU. The authors find that increases in housing cost burdens have large adverse impacts on housing adequacy, with noticeable impacts on poverty and health. The effects on fertility and labour force participation are smaller.
Figure shows household responses to negative deposit rates. Exposure to negative deposit rates caused a strong decrease in deposits (Panel A): the treatment group exposed in 2019 differentially reduced their total deposit balances in the banking system by DKK 90,000 during 2019–2020 relative to a baseline of DKK 1.1 million. Individuals in the treatment group differentially increased their stock market holdings (Panel B) and contributions to illiquid pension accounts during 2019–2020 (Panel C). The treatment group differentially increased annual consumption by around DKK 40,000 during 2019–2020 (Panel D). How households behave in an environment with deposit rates below zero is important for monetary transmission, but the literature is scant. This column studies household behaviour in the context of a major episode of negative interest rates in Denmark. This authors find that the negative deposit rates led to a strong decrease in deposits. Households appear to react more sharply when interest rates below zero are cut compared to when rates are positive. As depositors seem to consume more rather than hold more cash, policy rate cuts in negative territory may have particularly strong consumption effects.
Figure shows perceived inflation and CPI inflation. Perceived inflation closely tracks CPI inflation over the long run, but it exhibits substantially greater short-run volatility. The divergence becomes particularly pronounced during periods of economic stress, including the Global Crisis 2008-09, the COVID-19 recession, and the inflation surge of 2021–22. Households often report inflation perceptions and expectations that differ markedly from official measures. This column proposes a new measure of perceived inflation, which uses disaggregated inflation data with indicators for household attention to capture the prices that people actually notice. The difference between perceived and official inflation varies over time and becomes particularly pronounced during periods of economic stress. Heightened attention to inflation is persistent and can remain elevated even after official inflation measures improve. As central banks recognise the importance of household beliefs for macroeconomic outcomes, it becomes important to also understand the gap between measured and perceived inflation.
Map shows defence production plants in TL3 regions in Europe. Defence policy is usually discussed as a national security issue, but its economic effects are often highly local. This column presents new evidence from 1,011 regions in 15 European countries suggesting that places hosting defence production and supply chain activity have a larger economic footprint than comparable places without such activity: they enjoy higher income levels, and the premium is especially large in more innovative regions. As Europe raises defence spending, the geography of where that money ultimately lands may matter almost as much as the total amount.
The figure plot the cohort decomposition of male labour force participation and lifetime labour market experiences. Labour force participation among prime-aged American men has fallen for over half a century, and standard explanations leave much of the decline unaccounted for. This column combines panel and survey data to show that the labour markets these men experienced while growing up shaped their willingness to work decades later. These effects survive childhood moves across states and likely operate through men’s beliefs about whether work will pay off. If temporary downturns in labour demand can persistently influence labour supply, anchoring expectations about employment, not just inflation, may have a place in macroeconomic policy.
In the latest #VoxTalks Economics Episode, recorded on 19 June 2026 at the PSE-CEPR Policy Forum in Paris, Gita Gopinath (Harvard University) and Philip Lane (European Central Bank) joined Tim Phillips to discuss the third wave of Global Imbalances. Twice before, global imbalances ended in major disruption: first with the Plaza Accord, then with the global financial crisis. So what makes this time different? 🎧 Listen to the full conversation now: https://cepr.org/multimedia/addressing-global-imbalances #CentralBanking #Finance #International #Regulation #Economics
The use of sanctions has grown steadily over the past seven decades, with well-documented effects on bilateral trade. This column analyses the effects of sanctions on cross-border mergers and acquisitions from sanctioning to target countries over the period 2006-2023. It finds that sanctions reduce the number of deals by around 50%, with stronger effects as time passes. The impact is stronger in the mining, manufacturing, and energy sector than in the services sector. The results reinforce the broad-based potential effects of sanctions on private sector economic activity.
The figure shows the youth surge in emerging markets and developing economies. The left panel shows the number of young working-age people in EMDEs by decade, the right panel shows youth cohorts by EMDE region. EAP = East Asia and Pacific; ECA = Europe and Central Asia; EMDEs = emerging market and developing economies; LAC = Latin America and the Caribbean; MNA = Middle East, North Africa, Afghanistan and Pakistan; SAR = South Asia; SSA = Sub-Saharan Africa. Bars show the number of young working-age people (aged 15–24), living in EMDEs (A) or EMDE regions (B) in each year. Between 2025 and 2035, the largest-ever cohort of young people in emerging market and developing economies will reach working age. Slowing growth, overlapping crises, tight fiscal conditions, and weak job-creation capacity in many countries make creating enough good jobs for this generation one of the defining development challenges of the next decade. This column argues that, with determined policy action, this demographic wave can become a powerful engine of growth, poverty reduction, and development progress.
The figure plots the New York Fed Global Supply Chain Pressure Index and 12-month US core CPI inflation. Supply chain disruptions have become a recurring feature of recent economic dynamics and an increasing concern in inflation policy discussions. This column argues that inflationary effects of these disruptions operate through more than higher import and transportation costs: when supply-chain bottlenecks make products harder to replenish, firms’ pricing power increases, and these price increases are further amplified by strategic responses to competitors’ disruptions. Using new data that link consumer prices to US maritime import shipments by consumer goods producers, the authors show that these competitive pricing spillovers substantially amplified the price effects of supply chain disruptions in the aftermath of the pandemic in the US.
Figure shows the historical demographic divergence across Japanese municipalities between 1980 and 2010. Older municipalities in 1980 lost population and aged further, while younger municipalities tended to grow and remain younger. Policy debates about ageing and shrinking populations often treat the problem as a national one. This column presents evidence from Japan showing that it is also profoundly spatial: ageing and depopulation hit rural regions first. Outmigration from shrinking regions today reduces births tomorrow, which makes the origin region older still. However, policies that reduce spatial inequality also reduce aggregate efficiency, presenting a trade-off that leaves policymakers with the challenge of choosing among different goals.
The bars show the distribution of publications for graduates of high-tax schools (MIT, Harvard, Stanford, Princeton, and Yale), other elite schools (Chicago, Northwestern, and Berkeley), other US schools, foreign schools, and applicants who did not graduate. The left panel (WDI) shows co-authorship- and journal-influence-weighted “deep impact” publications, grouped into 0, (0, 1], (1, 2], and 2+; the right panel shows top-five journal counts, grouped into 0, 1, 2, and 3+. The sample is limited to 1996–2010 MIT application cohorts. The 2025 US university endowment tax hike, together with other financial pressures, is pushing the wealthiest universities to shrink their doctoral cohorts. This column exploits a novel sample of MIT Economics PhD applicants to shed light on the research value-added of elite doctoral programmes and the possible consequences of programme cutbacks. Holding applicant ability fixed, graduates of high-tax and other elite schools produce 60-75% more impact-weighted publications than comparable applicants. Neither peer effects nor home-journal favouritism explains this advantage. If elite PhD programmes add similar value in other disciplines, the endowment tax may well slow the pace of scientific progress.
Panel A shows the financial indicators typically used to limit firm borrowing (y-axis) and the fraction of firms that face a financing constraint based on these indicators. Panel B plots the distribution of the number of tight financing constraints across firms. A longstanding puzzle in monetary economics is that policy tightening has stronger effects than easing. This column shows that the multiplicity of financing constraints notably dampens the transmission of expansionary monetary policy to firm borrowing and investment, while amplifying the transmission of policy tightening. This asymmetry arises because, when policy tightens, the most responsive constraint binds; whereas, when policy eases, the least responsive constraint limits the expansion of borrowing. A standard New Keynesian framework incorporating this mechanism shows that the decline in aggregate investment following contractionary monetary shocks is twice as large as the increase following equally sized expansionary shocks.
Figure shows observed behaviour versus LLM-recommended behaviour over the life cycle. Red: survey respondents’ reported behaviour; green: one-shot LLM advice at current circumstances; blue: full life cycle simulation following LLM advice each year. The rise of generative AI has raised hopes that high-quality, personalised financial advice might finally become cheap and universally accessible. This column develops a method to quantify the impact of following AI financial advice over a lifetime. The authors find that on the decisions that matter most for long-run wealth, such as investing in diversified equity or building a savings buffer, today’s leading models give advice broadly consistent with what economists would prescribe. But what people get from AI depends as much on the model's ability as on how they write their prompts, with the advice varying by gender, financial literacy, and prior AI experience.
The figure shows the different exchange rate mentions in the ECB’s monetary policy statements over time, alongside the EUR/USD exchange. ‘Technical assumptions’: Exchange rate is mentioned as a technical assumption in the staff projections. ‘Monitoring’: Statements regarding the need to monitor the exchange rate due to its impact on the economy and inflation. ‘Economic effect of the exchange rate’: Statements regarding the exchange rate’s impact on the economy and inflation. ‘ECB concerned about volatility’ and ‘Unwelcomed move of the exchange rate’: statements expressing the ECB’s concerns about exchange rate volatility or direction/level of the exchange rate. Evidence is mixed on whether central banks can influence exchange rates simply by talking about them. This column explores this question by analysing more than two decades of ECB communications. The authors find that exchange rate communication has only limited effects on the euro exchange rate, and any effects that do emerge tend to disappear quickly. This stands in contrast with monetary policy communication, which is a major driver of exchange rate movements. Financial markets seem to pay much more attention to what central banks do or say about monetary policy than to what they say about exchange rates.
Figure shows the evolution of the combined interest rate risk exposure of bonds and swaps alongside the ECB deposit facility rate. As the tightening cycle begins in mid-2022, the net exposure of the banking system becomes progressively less negative. This occurs despite the continued accumulation of bond holdings, implying that banks actively increase derivative-based protection as interest rate risk materialises. This is consistent with a dynamic view of risk management. Valuation losses can build quickly in response to higher interest rates, as demonstrated by the collapse of Silicon Valley Bank. This column shows that Italian banks do hedge against interest rate risk and, crucially, increase hedging when rates rise. This suggests banks actively adjust hedging intensity through derivatives, consistent with a state-dependent model of risk management. Institutions with lower capital ratios, a larger share of wholesale funding, and a smaller deposit base hedge more aggressively. These findings highlight the importance of considering derivative exposure in stress-testing frameworks, as derivatives are a key channel through which banks absorb shocks.
This figure displays the daily Economic Policy Uncertainty Index in the week before and after August 1. The solid black line is the raw daily index and the dashed lines on either side of August 1 represent the average EPU index in the week before and week after August 1 respectively. Trust in official economic statistics has become an increasingly salient policy issue, including in the US where the Commissioner of the Bureau of Labor Statistics was dismissed in August 2025 amid allegations that agency data had been manipulated. This column shows that the events produced a sharp increase in economic policy uncertainty, with existing estimates linking uncertainty to macroeconomic outcomes implying that the resulting loss of confidence may have reduced US GDP by roughly $20 billion. While some economic activity postponed during periods of uncertainty may eventually occur later, there are reasons to believe that at least part of the effect could be lasting.
Figure shows that exports and air pollution rose together in China following WTO accession and plots trends over time. During the sample period 2000–2007, exports increased sharply, in lock-step with average and median particulate matter concentrations. This plot strongly suggests that China’s export boom coincided with worsening air quality. As countries consider trade deals to foster development, the question of how trade impacts pollution keeps resurfacing. This column examines how China’s rapid entry into world markets in the early 2000s affected pollution through three channels. The authors find that (1) when firms export more, pollution near their manufacturing plants tends to fall; (2) pollution among local suppliers to exporting firms also tends to fall, suggesting that cleaner practices spread along production networks; and (3) because China relied heavily on coal-fired power plants, export growth increased pollution around those plants and in downwind localities. Policies that promote trade should be paired with cleaner electricity and environmental safeguards to avoid the downsides of the Chinese experience.
Figure reports reasons for not investing among firms that had not invested in AI in 2025. Large firms are those with at least 500 employees. Reasons in decreasing order are: immature, untrained workforce, privacy, uncertain capabilities, not applicable in industry, expensive, bias, no data, unimpressed, and restrictions. Real-time evidence from users can inform the current policy debates around artificial intelligence. This column reports on a survey of executives at typical US firms to reveal that AI adoption is already widespread but shallow. Executives report positive productivity effects and expect larger gains, especially in high-skill services and finance, although measured gains lag perceived productivity improvements. The gains appear to come mainly through innovation- and demand-oriented channels rather than through cost cutting. Capital deepening is low, with most AI adopters renting intangible capital from upstream providers, and near-term employment effects are small on net, with larger reductions concentrated among larger firms and a shift away from routine clerical work towards skilled technical roles.
The large US tariffs announced in April 2025 led other countries to debate the best policy response. This column uses a multi-country model to study the implications of the US withdrawal from free trade for that cooperative equilibrium. It finds that even a complete US exit from global trade cooperation has limited quantitative impact on other countries’ optimal tariffs and on the sustainability of trade cooperation. Trade cooperation becomes harder to sustain when policymaker patience is lower and when punishment horizons are shorter. Still, conditions for continued trade cooperation remain strong.
The figure tracks organisational capital around CEO turnovers. It compares organisational capital before and after a change in leadership while controlling for firm characteristics and aggregate shocks. The way firms coordinate activities, motivate employees, accumulate knowledge, and adapt to changing environments evolves gradually over time, and is shaped by managerial decisions. This column proposes a new way of measuring this organisational capital using employee reviews on Glassdoor. The measure shows that CEO-related variation accounts for roughly one-third of organisational capital and that CEO turnovers are immediately followed by significant declines in organisational capital. The findings reveal how listening to employees’ voices can help firms detect organisational weaknesses before they show up in financial performance.
The figure plots regression coefficients with 68% and 90% confidence intervals for one-month, three-month, six-month, and 12-month rates, in response to over-delivery and under-delivery monetary policy surprises. Financial markets pay close attention to monetary policy, and surprise policy decisions can strongly affect asset prices and the real economy. This column shows that financial markets care not only about the size of monetary policy surprises, but also whether the central bank over-delivers or under-delivers on market expectations. Short-term interest rates react up to ten times more strongly when central banks over-deliver. Moreover, it finds that information effects emerge primarily in over-delivery episodes. In an era in which central banks increasingly rely on communication and expectation management, decisions that exceed expectations may communicate more than the policy move itself.
Effects of global versus regional energy shocks: Impact on EU GDP. Simulations based on a multi-country multi-sector DSGE model with production linkages and trade. The regional energy shock arises from trade frictions on EU imports of oil and gas from the RoW; the global shock from a decline in productivity in the RoW oil and gas sector. The shocks are both scaled to imply a 10 per cent rise in the EU energy price on impact, with a half-life of eight quarters. Not all energy shocks are alike. This column uses a multi-country, multi-sector DSGE model with production networks and trade linkages to compare a Europe-centric regional energy shock — akin to the 2022 Russian gas crisis — with a broader global energy shock like the current Strait of Hormuz disruption. The global shock worsens the EU's terms of trade more severely and triggers exchange rate depreciation rather than the appreciation seen under a regional shock. Crucially, while the regional shock still allows firms and consumers to substitute towards cheaper foreign goods and inputs, the global shock closes off this escape route — amplifying both the GDP contraction and the indirect inflationary effects through global supply chains.
The time series behaviour around the reform shown in the figure suggests a clear turning point in employment composition following the PWL. Relative to the pre-reform period, the share of informal workers declines steadily by 16.1 percentage points, a drop that mirrors the surge in the share of formal contractors by 16.8 percentage points. By contrast, the share of formal employees hardly changes and nor do relative wages vary. Chile’s 2022 Platform Work Law was the first reform to address concerns about independent contractors of platform delivery and transportation services, a sector in which informality is pervasive across Latin America. This column takes advantage of Chilean survey data on platform work to evaluate the impact of the law. The findings suggest that so far it has failed in regulating minimum pay and working hours and in increasing the share of employees. However, it has led to some substitution of informal self-employed by formal ones and to a moderate rise in social security coverage among these workers.
Figure: China-sensitive goods account for around 40% of the euro area non-energy industrial goods Harmonised Index of Consumer Prices (HICP) basket and have contributed to the moderation of industrial goods inflation since mid-2024. Authors’ calculations based on Eurostat data. ‘China-sensitive’ items are those most responsive to China import unit value shocks. Chinese goods exports have expanded sharply since 2020, with Chinese firms increasingly competing in advanced manufacturing and technology sectors. This column shows that the dominant factor driving this export growth is weak domestic demand, while government subsidies and technological upgrading play a smaller role. For the euro area, lower import unit values from China reduce consumer prices and lower capital accumulation, while stronger Chinese competition in third markets weakens export growth. The long-term challenge is to ensure that the gains from lower import prices are accompanied by policies to support competitiveness and de-risk affected upstream sectors.
Weekly mean calls per hour for the control (‘no office days’) and treatment (‘one office day per month’) groups. Both groups averaged 10.8 calls per hour before the intervention; post-intervention means are 11.5 and 12.4, respectively. Shaded areas are 95% confidence intervals. Remote work has become a permanent feature of labour markets. But the question remains how to balance the benefits of remote work with the workplace interactions that support performance and retention. This column uses a randomised controlled trial in a large customer service multinational in Türkiye to explore the effects of assigning some employees to work together in the office one day per month. Monthly office days raised productivity by 8% without loss of service quality, improved communications, and cut attrition by one-third. The findings suggest that a small amount of coordinated in-person interaction can strengthen both performance and retention in otherwise remote work teams.
Figure shows the evolution of risk for selected central banks. Balance sheet risk was modest and stable prior to 2008, but then rose sharply as central banks deployed asset purchases and other policies that expanded their balance sheets. The fiscal dominance view holds that politically captured central banks are more likely to be pushed into taking risky positions, while independent ones resist. This column presents new estimates of the value at risk in the balance sheets of 18 advanced economies’ central banks over two decades which suggest the opposite: more independent central banks take more financial risk, and they do so most aggressively when fiscal policy tightens. Independence, it seems, frees central banks to deploy their balance sheets in pursuit of their mandates, whatever the cost to profits.
European defence ministries and intelligence services run on infrastructure they do not control and cannot independently audit – a situation highlighted by the US administration’s directive of 12 June 2026 ordering Anthropic to suspend access to its most powerful AI models for all non-US persons. This column argues that it will take ten to fifteen years for Europe to develop AI capability sufficient to negotiate as a peer, and proposes three steps to buy time while Europe closes the gap.
Monthly counts of newly released iOS applications (left) and total usage accumulated by each monthly cohort of new apps in its first three months as proxied by total ratings (right). Shaded region marks the agentic-coding era (February 2025 onward). The paper reports analogous series for Google Play, the Chrome Web Store, and SourceForge; the pattern is cleanest for iOS. Generative AI now writes a substantial share of the world's code, but aggregate software output has changed far less. This column uses data on more than 100,000 GitHub developers, combined with data on their AI tool usage, to trace the productivity effects of three successive generations of AI coding tools. Each generation raises coding activity more than the last, but the gains shrink sharply as work moves from writing code to shipping software. Human bottlenecks in reviewing, integrating, testing, and releasing code, together with scarce consumer attention, help explain why large task-level gains have not yet produced a software output boom.
As women gain political ground around the world, there is hope that the election of more women might also mean cleaner government. But is this the case? This column examines thousands of close mixed-gender elections in Brazil across two decades and finds that electing a woman mayor neither increases nor reduces corruption. Instead, incumbents are consistently more corrupt across various metrics. Women continue to face substantial barriers on the path to office and dismantling these remains a fundamental democratic objective, but on the grounds of equity in representation, not because it will reduce corruption.
@johnvanreenen.bsky.social joins @talknormal.co.uk for the latest #VoxTalks Economics to discuss whether defence spending can be pro-growth. John argues that as spending rises, governments should focus on making it innovation-rich to maximise benefits. 🎧 Listen now: https://ow.ly/8mN550ZeFEb #EconSky
Graph showing trade flows between countries with and without trade agreements, 1827-1939. Authors’ dataset based on the Consolidated Treaty Series (Parry 1969–1980), United Nations Treaty Collection, merged with TRADHIST dataset by CEPII.