Working Papers
2025
June 6, 2025
From Text to Quantified Insights: A Large-Scale LLM Analysis of Central Bank Communication
Description: This paper introduces a classification framework to analyze central bank communications across four dimensions: topic, communication stance, sentiment, and audience. Using a fine-tuned large language model trained on central bank documents, we classify individual sentences to transform policy language into systematic and quantifiable metrics on how central banks convey information to diverse stakeholders. Applied to a multilingual dataset of 74,882 documents from 169 central banks spanning 1884 to 2025, this study delivers the most comprehensive empirical analysis of central bank communication to date. Monetary policy communication changes significantly with inflation targeting, as backward-looking exchange rate discussions give way to forward-looking statements on inflation, interest rates, and economic conditions. We develop a directional communication index that captures signals about future policy rate changes and unconventional measures, including forward guidance and balance sheet operations. This unified signal helps explain future movements in market rates. While tailoring messages to audiences is often asserted, we offer the first systematic quantification of this practice. Audience-specific risk communication has remained stable for decades, suggesting a structural and deliberate tone. Central banks adopt neutral, fact-based language with financial markets, build confidence with the public, and highlight risks to governments. During crises, however, this pattern shifts remarkably: confidence-building rises in communication to the financial sector and government, while risk signaling increases for other audiences. Forward-looking risk communication also predicts future market volatility, demonstrating that central bank language plays a dual role across monetary and financial stability channels. Together, these findings provide novel evidence that communication is an active policy tool for steering expectations and shaping economic and financial conditions.
May 30, 2025
Cyclical Fiscal Multipliers: Policy Mix and Financial Friction Puzzle
Description: This paper investigates dynamic relationships between U.S. government expenditure multipliers and the economy's cyclical position from 1949 to 2018 using a Time-Varying Parameter Vector Autoregression (TVP-VAR) model. We challenge the existing literature, which predominantly relies on predefined economic regimes and assumes a stable relationship between fiscal multipliers and business cycles. Our findings identify two distinct periods: fiscal multipliers were counter-cyclical from 1949 to the late 1980s, followed by a significant decline in their effectiveness during recessions thereafter. These variations are attributed to the prevailing fiscal-monetary policy mix; with higher fiscal multipliers during earlier recessions resulting from sharp shifts toward a fiscally led policy stance, followed by a decline after the Dot-com recession due to a transition toward a monetary-led policy mix. We find particularly low multipliers during the global financial crisis, which provides new insights into the evolving role of financial frictions in the transmission of fiscal policy.
May 30, 2025
Interest Rate Sensitivity Scenarios to Guide Monetary Policy
Description: This paper examines the challenges of formulating monetary policy in the face of heightened uncertainty. We develop a framework to assess the optimal monetary policy path under uncertainty, focusing on four key dimensions: the expectation formation process, inflation persistence, the measurement of the neutral interest rate, and the slope of the Phillips curve. Our framework provides a flexible tool for policymakers to address uncertainty and enhance decision-making in pursuit of economic stability. This framework is helpful to improve the risk management approach to monetary policy by showing how scenarios can quantify different sources of uncertainty faced by the ECB and give market participants an idea of how the ECB would react if those scenarios materialize.
May 30, 2025
Digitalization: A Catalyst for Intergenerational Occupational Mobility?
Description: This paper analyzes the link between digitalization and intergenerational occupational mobility in Africa. We use a probit model estimated on a large sample of 28 million individuals aged 14 and higher and co-residing with at least one individual from the older generation. We find that digitalization could help boost upward mobility and limit the risks of downward occupational mobility, thereby improving job opportunities. While strong institutions, political and social stability, better access to adequate digital infrastructure, and education are important to increase and accelerate upward mobility, new technologies and digital tools can intensify these positive effects and contribute to creating jobs and enhancing living standards in Africa. Similar results hold for downward mobility.
May 30, 2025
Scenario Synthesis and Macroeconomic Risk
Description: We develop methodology to bridge scenario analysis and risk forecasting, leveraging their respective strengths in policy settings. The methodology, rooted in Bayesian analysis, addresses the fundamental challenge of reconciling judgmental narrative approaches with statistical forecasting. Analysis evaluates explicit measures of concordance of scenarios with a reference forecasting model, delivers predictive synthesis of the scenarios to best match that reference, and addresses scenario set incompleteness. This provides a framework to systematically evaluate and integrate risks from different scenarios, aiding forecasting in policy institutions while supporting clear and rigorous communication of evolving risks. We also discuss broader questions of integrating judgmental information with statistical model-based forecasts in the face of as-yet unmodeled circumstances.
May 29, 2025
Europe’s National-Level Structural Reform Priorities
Description: Europe has a large and persistent per capita income gap with the United States. Deficiencies in total factor productivity, labor utilization, and capital intensity all play a role. While deeper intra-Europe integration is one key element towards closing these gaps, remaining structural domestic policy gaps with respect to most growth-friendly regulatory settings highlight the scope for complementary policy action at the national level. This paper compiles IMF staff’s structural policy priorities for European countries to lift output over the medium term, reviews key implementation challenges, and discusses complementarities with EU-level efforts. While addressing these domestic reform priorities will require overcoming long-standing political economy and—in some cases—technical obstacles to reform, successful implementation could entail sizeable medium-term gains of around 5, 7 and 9 percent for advanced European, Central Eastern and Southeastern European, and Western Balkan economies, respectively.
May 23, 2025
Fiscal Multipliers in Mongolia
Description: Fiscal policy plays an important macroeconomic role in Mongolia. On the one hand, fiscal expansion is seen as a measure to boost aggregate demand. On the other hand, import leakages mitigate the impact of fiscal expansion on growth. Applying a structural vector autoregressive model, this paper finds that Mongolia’s total spending and revenue multipliers are below 1, peaking at 0.3 and -0.1, respectively. The lower than 1 multiplier can be explained by import leakages in Mongolia. Capital spending multiplier peaks at 0.6, exceeds and remains more persistent than the current spending multiplier, suggesting that public investment is more efficient in boosting growth than current spending. Tax revenue and non-tax revenue multipliers peak at -0.1 and -0.2, respectively, and are short-lived. Revenue multipliers are broadly comparable in size, but their assessment is challenging due to lack of sizeable tax policy measures in Mongolia.
May 23, 2025
Labor Market Consequences of Homicides: Evidence from Mexico
Description: This paper explores how fluctuations in crime rates influence labor market outcomes in Mexico. Using detailed survey data and an individual-fixed effect estimation, the analysis reveals distinct gender dynamics in response to rising homicide rates. Men are more likely to exit the labor market due to reduced demand for their labor, while women increasingly join the workforce, mainly in the informal sector, to offset this decline. This outcome is largely driven by the presence of drug trafficking organizations, which primarily employ men in their operations. Escalating violence also increases labor mobility, leading to higher job separations, particularly among women seeking safer employment. Our results highlight that while increasing crime in the form of homicides may not induce large changes in the aggregate level of employment, there is evidence of labor reallocation across and within sectors. This suggests an increase in labor market misallocation.
May 23, 2025
Macro-Financial Policies and Vulnerabilities in IMF-Supported Programs
Description: We construct a unique dataset by collecting macro-financial commitments data using textual analysis of the Memorandum of Economic and Financial Policies (MEFPs), a document outlining, inter-alia, policy commitments by member countries, in the context of an IMF-supported program. We combine this data with information on structural conditionality. Using a staggered difference-in-differences methodology, we show that IMF-supported programs with macro-financial policy commitments are followed by periods of lower non-performing loans and in some cases lower credit-to-GDP ratios, relative to IMF-supported programs without macro-financial commitments, mostly for the post global financial crisis (GFC) period before the COVID-19 pandemic. The NPL-to-loans ratio does not seem to decrease as a result of credit expansion. The results point to stronger and more abrupt declines in credit-to-GDP following ex-post macro-financial policies, those implemented after a crisis occurs (e.g., restructuring), and milder and more gradual declines following ex-ante policies, those implemented before risks materialize (e.g., regulatory requirements). The responses are also larger when countries have positive credit gaps at the start of the program than when credit gaps are negative. These results point to the importance of considering the country’s position in the credit cycle in program design and in addressing vulnerabilities preemptively to reduce the need for abrupt corrections when risks materialize. Finally, macro-financial policies targeting financial inclusion tend to increase credit-to-GDP ratios in low credit-to-GDP program countries.
May 23, 2025
From Banks to Nonbanks: Macroprudential and Monetary Policy Effects on Corporate Lending
Description: The growing role of nonbanks in corporate credit intermediation raises important but underexplored questions about how both monetary policy (MP) and macroprudential policies (MaPP) affect lending and the real economy. Using syndicated loan data, we examine the joint impact of MP and MaPP shocks on credit supply to nonfinancial firms. Our findings show that nonbanks act as shock absorbers, cushioning firms—particularly those with existing nonbank relationships—from policy tightening. We also find that these shocks drive credit away from weaker banks toward nonbanks, raising concerns about credit quality. Finally, we provide evidence that MaPPs on banks can lead them, especially weaker banks, to shift lending to nonbanks and away from nonfinancial corporations. This allows nonbanks to expand their role in corporate credit markets. Overall, our findings highlight that tighter MP and MaPP may unintentionally push credit intermediation into a sector largely outside the regulatory perimeter, posing new financial stability risks.