Working Paper: NBER ID: w9436
Authors: Kenn Ariga; Kenji Matsui
Abstract: In this paper, we investigate several key problems in Japanese economic statistics. We use CPI mismeasurements and biases as an example to explore the roots of the problems and also to offer guidelines for improvements. We emphasize 3 major shortcomings shared by many official statistics in Japan: (1) long delays in adjustments, (2) lack of proper coordination, and 3) insufficient information disclosure. In the analysis of CPI bias, we limit our focus to potential biases due to aggregation, survey methodology and sample selection procedures. We estimate that, in recent years, the commodity CPI inflation rate is biased upward by at least 0.5% per year, even if we assume away the potential bias associated with the quality adjustment, delay in incorporating changes in consumption basket, and other important unresolved problems.
Keywords: No keywords provided
JEL Codes: F4; F1
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| CPI mismeasurement (E31) | inflation rate accuracy (E31) |
| biases in aggregation, survey methodology, and sample selection (C83) | CPI mismeasurement (E31) |
| CPI mismeasurement (E31) | upward bias in inflation rate (E31) |