2011年-BIS国际清算银行_Measuring_disagreement_in_UK_consumer_and_central_bank_inflation_forecasts_49页_716kb

Summary of "Measuring Disagreement in UK Consumer and Central Bank Inflation Forecasts"

Core Content

This paper explores the nature of disagreement in inflation expectations between UK consumers and the Bank of England's Monetary Policy Committee (MPC). It uses a novel method involving adaptive kernel density estimation and bootstrap multimodality tests to analyze the full probability distributions of inflation forecasts, going beyond traditional summary statistics like mean and standard deviation.

Main Views and Key Findings

• Disagreement in Inflation Expectations:

  • UK consumers showed substantial disagreement in their one-year-ahead inflation forecasts.
  • The MPC's forecasts were more concentrated, indicating a narrower range of expectations.
  • There was no convergence in consumer forecasts over the sample period, suggesting that inflation expectations were not well-anchored.

• Methodology:

  • The authors used the GfK NOP consumer survey data from 2001 to 2007.
  • They applied three resampling schemes to convert the histogram data into continuous probability distributions:
    1. Uniform distribution within each bin.
    2. Non-truncated normal distribution centered at each bin's midpoint.
    3. A combination of the first two schemes.
  • They also used an adaptive bootstrap multimodality test to identify multiple modes in the distribution, which are interpreted as evidence of disagreement.

• Consumer Forecast Characteristics:

  • The consumer forecasts had multiple modes, with the main ones around 0% and higher than the central mode.
  • The 90% confidence intervals of the MPC's fan charts covered less than 50% of the probability mass of consumer forecasts, highlighting the disparity in dispersion.
  • The increased probability mass around the higher mode towards the end of the sample period may be linked to rising energy prices.

• Policy Implications:

  • The MPC may have an information and analytical advantage over consumers, which could explain the narrower range of its forecasts.
  • The MPC's ability to anchor inflation expectations is evaluated by comparing its forecast densities with those of consumers.
  • The paper suggests that disagreement measures based on multimodality provide more comprehensive insights than conventional quantile-based measures.

Key Information

• Data Sources:

  • The GfK NOP consumer surveys provide data on inflation expectations, with a focus on the February surveys between 2001 and 2007.
  • Each survey includes around 4,000 respondents, which allows for more accurate estimation of forecast densities.

• Survey Structure:

  • The survey includes 14 questions, but only 9 are asked quarterly.
  • The remaining 5 questions are asked annually in February.
  • Bins are centered at 0.5%, 1.5%, 2.5%, 3.5%, and 4.5%, with ±0.5% bands.
  • The extreme bins are open-ended, covering values below 0% and above 5%.

• MPC Forecasting:

  • The MPC publishes probabilistic inflation forecasts in the form of fan charts, which show different levels of uncertainty.
  • Since 1997, the MPC has used CPI as the inflation target, but the paper focuses on RPIX-based forecasts due to historical data and public recognition.

• Empirical Results:

  • The analysis found that consumer inflation forecasts were more dispersed than MPC forecasts.
  • The presence of multiple modes in consumer forecasts indicates greater disagreement.
  • The MPC's forecasts were generally more accurate in capturing the central tendency of inflation, but overestimated forecast uncertainty.

Conclusion

The paper contributes to the literature on inflation expectations by emphasizing the importance of analyzing the full probability distribution of forecasts rather than relying on summary statistics. It provides evidence that UK consumers have more diverse views on future inflation than the MPC, which may reflect differences in information, analytical capacity, and policy influence. The use of nonparametric density estimation and multimodality tests offers a more robust and detailed measure of disagreement, which can be useful for assessing the effectiveness of monetary policy and the anchoring of inflation expectations.