Core Inflation in Advances Economies

 The Decomposition Methodology (DLM), or Dynamic Linear Model, is the core analytical tool utilized in the sources to explore differences in the dynamics of core inflation between Europe and North America. This method is a parsimonious multivariate Bayesian time series filter used to decompose the level of core inflation in major advanced economies into three distinct categories: regional, global, and country-specific components.

Structure and Specification of the DLM

The DLM is tailored to compare inflation dynamics across Europe and North America. It is applied to monthly core inflation rates for 12 advanced economies, with data spanning back to the 1970s.

The decomposition is formally specified such that the core inflation series ($Y_{i,t}$) for a given country ($i$) at time ($t$) is the sum of these components: a global factor, a regional factor, and a country-specific component.

1. Global and Regional Components: The model assumes that the global and regional level factors ($\mu^{Global}_t$, $\mu^{Regional}_t$) follow random walks with time-varying growth rates ($\beta^{Global}_t$, $\beta^{Regional}_t$). These growth rates, in turn, are modeled as first-order auto-regressive (AR(1)) processes.
   • The regional factors specifically divide the sample into a European component (shared by nine European countries) and a North America component (shared by the U.S. and Canada). Japan is included but only decomposes into global and country-specific components, as it does not share a regional component with the others in the sample.
2. Country-Specific Component: The persistent country-specific factors follow a moving average (MA) representation. These components generally contribute less than the regional or global components and capture higher-frequency noise, reflecting differences in core inflation levels even among countries within the same region (e.g., the U.K. versus other European countries).

Estimation and Analytical Approach

The model's deep parameters (variances and moving average parameters) are estimated via Maximum Likelihood Estimation (MLE). Certain parameters are calibrated to ensure the global component maintains a smooth trend.

The DLM leverages a Bayesian approach and state-space form, which offers several methodological advantages:

• Filter and Smoother: The process involves filtering the rich monthly time series through a Bayesian state-space filter to estimate the posterior means and variances of the time-varying components. The Kalman Filter is used for recursive forward estimation.
• Retrospective Analysis: Since the analysis is focused on explaining past movements in inflation, the researchers utilize the forward-filtered-backward-smoothed estimates of the state variables (Kalman smoother). These estimates incorporate information from the full sample (past and future observations) and are thus smoother and more precise.
• Interpretability and Alternatives: The DLM's parsimonious structure facilitates easily interpretable global, regional, and country-specific components. The DLM was chosen over the alternative dynamic factor model, and the authors found that their baseline DLM incorporating regional factors was favored by the data.

Role in Core Inflation Dynamics

By using the DLM to isolate these components, the sources draw conclusions regarding the drivers of inflation in advanced economies, particularly concerning the post-COVID surge:

• Regional Factors: The analysis identified a prominent role for regional factors in shaping inflation dynamics. Historically, the North American and European regional components have at times diverged.
• Drivers: The decomposition allows the authors to associate different economic drivers with different components. For instance, in the post-pandemic surge, the global component was largely associated with global supply frictions and past energy shocks. Conversely, the North American regional component was largely associated with labor market tightness in the region and lagged changes in households’ savings behavior.
• Policy Implications: Understanding these distinct components has implications for monetary policy, especially when global and regional components move in opposite directions, potentially masking underlying trends in overall core inflation. The Bayesian framework also allows for computing predictive distributions, which, given the strong persistence in the components, could facilitate future out-of-sample forecasting.

The DLM functions like a forensic tool, meticulously separating a country's observed core inflation (the overall signal) into its constituent origins—whether global economic forces, specific regional characteristics, or idiosyncratic domestic noise—allowing researchers to identify and study the unique drivers behind each layer of inflation dynamics.

The Inflation Component Analysis described in the sources is the use of a Dynamic Linear Model (DLM), a parsimonious multivariate Bayesian time series filter, to systematically decompose the level of core inflation in major advanced economies. This analysis aims to explore the differences in core inflation dynamics, particularly between Europe and North America.

This decomposition analysis separates the observed 12-month core inflation rates for 12 advanced economies into three distinct categories: regional, global, and country-specific components.

The Three Components of Core Inflation

The model assumes that a country's core inflation ($Y_{i,t}$) is the sum of these factors, allowing researchers to isolate the influence of different drivers.

1. Global Component ($\mu^{Global}_t$): This factor is common to all countries in the sample. The analysis estimates that the global component explains a sizable fraction of core inflation variation in major advanced economies. The analysis found that the global component rose sharply after the COVID pandemic, remaining at still-elevated levels as of July 2025.

   • Drivers: Amid the post-COVID inflation surge, the global component was largely associated with global supply frictions (like changes in sea freight shipping costs and manufacturing backlogs) and past energy shocks/oil price fluctuations.

2. Regional Components ($\mu^{Regional}_t$): These factors are shared by countries within specific geographic blocs—Europe (nine countries) and North America (U.S. and Canada). The analysis identifies a prominent role for regional factors in shaping inflation dynamics.

   • Dynamics: Historically, the two regional components have often diverged. For example, the European component was higher than the North America component through the 1970s and 1980s. Amid the COVID inflation surge, the regional components diverged again: North America’s component peaked earlier and at a higher level than Europe’s. Furthermore, while the North America component was declining in 2022, the European component continued to rise.
   • Drivers: The North American regional component, specifically, was found to be largely associated with labor market tightness in the region (measured by changes in the job-openings-to-unemployment ratio and average weekly earnings) and lagged changes in households’ savings behavior.

3. Country-Specific Components ($\epsilon_{i,t}$): These components capture higher-frequency noise and idiosyncratic differences in core inflation levels across countries. They generally contribute less than the regional and global components. For instance, the analysis showed that the U.K. had a large positive country-specific component because its core inflation was higher than that of other European countries in the sample, while France had a smaller, negative component. The component for the U.S. rose sharply in 2021, reflecting a sharper core inflation increase relative to Canada.

Significance and Policy Implications

Accounting for regional components is essential for a deeper understanding of core inflation dynamics. Ignoring the regional factor could lead to overstating the importance of global and domestic components.

The results of this analysis highlight that global and regional inflation components have moved in different directions and appear to have different economic drivers in recent years. This differentiation is vital for monetary policy:

• Monetary policy can be particularly challenging when global and regional components are moving in opposite directions. For instance, in the second half of 2022, the leveling off and decline of the U.S. regional component was masked by continued increases in the global component, leaving overall core inflation largely unchanged at elevated levels.
• The strong persistence observed in the levels and growth rates of these estimated components could facilitate out-of-sample forecasting, suggesting regional-level information may be useful for predicting future inflation.

The Inflation Component Analysis, by separating core inflation into these three layers, acts like a multi-lens telescope, allowing researchers to see beyond the national aggregate rate and identify whether inflationary pressure originates from broad global forces, shared regional economic trends, or specific domestic policy and market quirks.

The sources provide a detailed analysis of the economic drivers behind the core inflation surge experienced in advanced economies following the COVID-19 pandemic, specifically by analyzing changes in the estimated global and North American regional components of core inflation. This approach highlights the different origins of inflationary pressure during this period.

The analysis uses reduced-form regressions focusing on the post-pandemic surge and subsequent pullback to identify the main correlates of these estimated components.

1. Drivers of the Global Component

The global component, which is common to all advanced economies in the sample, peaked after the COVID pandemic and remained at elevated levels as of July 2025. Its post-COVID surge was largely explained by factors related to global commerce and energy prices.

The specific drivers associated with the change in the global component of core inflation are:

• Global Supply Frictions: The global component is associated with factors measuring supply chain stress. The post-COVID surge was largely explained by global supply frictions. Specifically, increases in the 12-month change in the global component were positively correlated with changes in sea freight shipping costs and global manufacturing backlogs.
• Past Energy Shocks/Oil Prices: The global component is also associated with past energy shocks. The positive correlation between the global component and the level and lagged changes of Brent crude oil prices is consistent with "second-round effects" of higher energy prices passing through to core consumer prices.
• Global Labor Market Tightness: The component is associated with overall labor market tightness across the advanced economies in the sample. This is indicated by a negative coefficient on the unemployment gap (the 12-month change of the 12-month average unemployment rate), suggesting that a shrinking (tightening) gap is associated with increases in the global component.

2. Drivers of the North American Regional Component

The North American regional component (shared by the U.S. and Canada) showed different dynamics, peaking earlier and at a higher level during the COVID inflation surge compared to the European component. The drivers for the North American component are primarily focused on regional domestic dynamics.

The specific drivers associated with the change in the North America regional component of core inflation are:

• Regional Labor Market Tightness: The North American regional component is largely driven by tightness in U.S. and Canadian labor markets. This is measured by the change in the job-openings-to-unemployment ratio and average weekly earnings within the region.
• Consumer Spending Patterns/Savings Behavior: The surge was also linked to lagged changes in households’ savings behavior. The sources suggest that excess savings accumulated during the pandemic boosted consumption spending. This is correlated with lagged changes in the personal savings rate.

Policy Implications of Divergent Drivers

The component analysis reveals that, in recent years, the global and regional inflation factors have moved in different directions and appear to have different economic drivers.

• For example, in the second half of 2022, the leveling off and decline of the North American regional component was masked by continued increases in the global component, which left overall core inflation largely unchanged at elevated levels.
• This divergence underscores that monetary policy can be particularly challenging when global and regional components move oppositely. The isolation of these drivers helps policymakers understand whether the inflationary pressure they face is due to global supply shocks or regional demand/labor market pressures.

The decomposition process, therefore, acts as a diagnostic tool, confirming that the post-COVID inflation surge was not monolithic but resulted from distinct global supply shocks (driving the global component) and strong regional demand and labor market tightness (driving the North American regional component).

The sources draw several key Policy Implications and Conclusions from the inflation component analysis, emphasizing the importance of isolating global and regional drivers of core inflation for effective monetary policy in advanced economies.

Policy Implications for Central Banks

The primary policy challenge highlighted by the analysis arises from the potential divergence and distinct drivers of the estimated inflation components:

• Challenging Policy Environment: Monetary policy decisions can be particularly challenging when the global and regional components of core inflation are moving in opposite directions. This makes reading the overall core inflation signal difficult.
• Masking Underlying Trends: The sources provide a concrete example from the post-COVID period: in the second half of 2022, the regional component of U.S. core inflation had leveled off and was starting to decline, but this decline was masked by continued increases in the global component. This counteraction left overall U.S. core inflation largely unchanged at elevated levels. This scenario shows how focusing solely on the aggregate core inflation rate could obscure underlying disinflationary or inflationary forces specific to a region or market.
• Understanding Drivers: Extracting a regional component is crucial for understanding the drivers of inflation, which, in turn, has monetary policy implications. The decomposition confirms that the post-COVID surge was driven by distinct factors: global supply frictions and past energy shocks primarily drove the global component, while labor market tightness in the region and consumer spending patterns drove the North American regional component. This knowledge is essential for policymakers aiming to address the correct source of inflation (e.g., supply-side versus demand-side pressures).

Key Conclusions and Future Research Avenues

The research concludes that both regional and global factors play a prominent role in shaping core inflation dynamics in advanced economies. The study documents significant differences in regional inflation experiences between Europe and North America, both historically and amid the COVID inflation surge.

The analysis also points toward future utility for the DLM framework:

• Forecasting Potential: The DLM's Bayesian framework easily allows for the computation of predictive densities for the state variables and observables. Given the strong persistence in both the levels and the growth rates of the estimated global, regional, and country-specific components, this suggests that regional-level information could potentially be useful when forecasting inflation.
• Model Optimization: While the current model is optimized for decomposing and explaining past movements in inflation, further research could explore how to optimize these types of models to produce reliable out-of-sample forecasts.
• Relevance in Fragmentation: The existence of regional inflation components raises the possibility that regional-level information could potentially be useful when forecasting inflation. This depth of understanding could be particularly important if the world grows increasingly fragmented.
• Model Extensibility: Although the focus was on 12 advanced economies with similar long-term inflation experiences, the model could easily be extended to include more countries in future analysis.

In essence, the decomposition methodology provides central banks with a refined lens, moving beyond the simple national inflation rate to see if price pressures originate from shared global challenges (like supply shocks) or distinct regional economic conditions (like labor market tightness), which is critical for formulating timely and appropriate monetary policy responses.