美联储-美国疫情后的价格灵活性_价格设定模型的证据和启示（英）-2025.3_60页_1mb

Post-Pandemic Price Flexibility in the U.S.: Evidence and Implications for Price Setting Models

Core Content

This paper investigates firm price-setting behavior in the U.S. during and after the Covid-19 pandemic, focusing on the increase in inflation and its implications for price-setting models. Using microdata from the Bureau of Labor Statistics (BLS) Consumer Price Index (CPI) Research Database, the authors analyze how firms adjusted their prices in response to economic shocks and how this behavior compares to theoretical models.

Main Findings

• Frequency of Price Changes:

  • The frequency of price changes increased significantly during the pandemic, especially in 2022, with the median frequency of price changes more than doubling compared to pre-pandemic levels.
  • As inflation declined post-2022, the frequency of price changes also decreased.
  • The frequency of price increases rose sharply, while the frequency of price decreases remained relatively stable.
  • This suggests that firms were more responsive to inflationary pressures through price increases than through decreases.

• Size of Price Changes:

  • The average size of price changes increased during the pandemic, contributing to the overall inflation.
  • The absolute value of price changes (i.e., the magnitude of increases or decreases) remained relatively constant.
  • The dispersion of price changes (i.e., the variability of price changes across firms) did not decrease with inflation, contrary to predictions of state-dependent models.

• Inflation Decomposition:

  • The increase in the intensive margin (size of price changes) accounted for almost all of the inflation increase.
  • The extensive margin (frequency of price changes) played a smaller role in inflation dynamics, even though it increased significantly.

Key Insights from the Analysis

• State-Dependent vs. Time-Dependent Models:

  • State-dependent models (e.g., menu cost models) predict that the frequency of price changes increases with inflation and that the size of price changes remains stable.
  • These models align with the observed patterns of price changes during the pandemic.
  • However, they fail to match the large increase in the frequency of price changes, which was more than 10 percentage points.
  • A re-calibrated menu cost model with smaller menu costs and larger idiosyncratic shocks can match the frequency increase but not the dispersion.

• Implications for the Phillips Curve:

  • The findings suggest that the Phillips Curve may have been steeper during the post-pandemic period, as the re-calibrated model predicts a faster pass-through of shocks to inflation.
  • This supports the idea that non-linear Phillips Curve models, which allow for a steeper slope during high inflation, may be more appropriate for understanding recent inflation dynamics.

• Sectoral Analysis:

  • The paper also examines how price-setting behavior varied across sectors.
  • Some sectors with a high frequency of price changes pre-pandemic continued to show similar behavior post-pandemic.
  • The authors highlight the importance of sectoral dispersion and shocks in explaining inflation.

Methodology

• Data Source:

  • The BLS CPI microdata, which includes over 90,000 products from more than 80 metropolitan areas.
  • This dataset covers approximately 70% of consumer goods and services, offering broader coverage than scanner or web-scraped data.

• Statistical Approach:

  • The authors use a two-step process to calculate statistics: first by ELI (Entry Level Item) category, then by aggregating across categories using weights from the 2018 Consumer Expenditure Survey.
  • They account for data limitations, such as bi-monthly observations and sales, by applying adjustments and excluding temporary sales.

• Counterfactual Inflation Analysis:

  • They construct two counterfactual inflation series: one with fixed frequency and varying size, and one with fixed size and varying frequency.
  • A third counterfactual fixes both frequency and size, while allowing the share of price increases to vary.

Conclusion

• The paper finds that the post-pandemic inflation was largely driven by the intensive margin of price changes, with the frequency of price increases playing a key role.
• Standard menu cost models, when calibrated to pre-pandemic data, struggle to match the magnitude of the frequency increase, suggesting that either the nature of price rigidity changed or that omitted shocks were significant.
• The re-calibrated model implies a steeper Phillips Curve and faster inflation response to shocks, which aligns with some recent theoretical developments.
• The findings contribute to the growing body of research using micro price data to understand inflation dynamics and the role of sticky prices in economic models.

Implications for Future Research

• The paper highlights the need for more flexible models that can account for the observed patterns of price-setting behavior.
• It also encourages further investigation into the role of sectoral dispersion and the potential impact of non-linear Phillips Curve models.
• The results may inform future monetary policy decisions by highlighting the importance of understanding how firms adjust prices in response to inflationary shocks.