• Price Rigidity and Market Structure: Evidence from the Japanese Scanner Data
  Takayasu Matsuoka

  Abstract

  This paper investigates price rigidity arise out of the specific market structures, such as degree of market concentration and pricing decisions of retailers and manufacturers. Using Japanese scanner data that contains transaction prices and sales for more than 1,600 commodity groups from 1988 to 2008, we find statistically significant negative correlation between the degree of market concentration and the frequency of price changes, including both bargain price changes and regular price changes. The results of two-way analysis of variance suggests that the variation of the frequency of price changes depends on the dierences among manufacturers as well as those among retailers.

  Introduction

  The relationship between price rigidity and market structure has been discussed since the American economist, Gardinar C. Means suggested that the downward rigidity of price during the Great Depression had a relationship to industrial concentration in a Senate Document in 1935. The implication of Means' findings is that the prices of less competitive markets tend to be sticky. This is referred to as the "administered prices" hypothesis (Domberger, 1979) and still attracts considerable attention. This is partly because empirical literature in this field found strong heterogeneity in price stickiness across commodity items and is interested in the determinants of item-levels variation in the frequency of price changes.

  2011.04

  PDFファイルはこちら

• Goodness-of-Fit Test for Price Duration Distributions
  Takayasu Matsuoka

  Abstract

  Is the actual price-setting behavior of an individual commodity item consistent with the assumptions of a sticky-price model? Part of the question may formally be addressed by performing a goodnessof-fit test for price duration distributions. For each of the 429 items in the Japanese retail price data for 2000–2005, we fitted the standard parametric models with or without unobserved heterogeneity to the data and tested the goodness of fit. We found that 8.6 percent of the tested items cannot reject the hypothesis that the underlying distribution is exponential, which corresponds to the time-dependent pricing model of Calvo (1983).

  Introduction

  This paper examines the distributional assumption of the duration of price spells. It forms part of an attempt to construct a formal theory dealing with sticky prices, because existing sticky-price models in macroeconomics explicitly formulate the mechanism of a firm’s price change by assuming that the length of price spells follows a certain distribution. One example is the Calvo (1983) model, which assumes that the probability of a firm’s price change is determined exogenously and does not change over time. This assumption implies that price spell durations have an exponential distribution with a constant hazard rate. The other example is the Dotsey, King, and Wolman (1999) model, which assumes a fixed cost of adjusting price. This model predicts a monotonically increasing hazard function when the general level of prices continues upward.

  2009.03

  PDFファイルはこちら

• Unobserved Heterogeneity in Price-Setting Behavior: a Duration Analysis Approach
  Takayasu Matsuoka

  Abstract

  There is strong empirical evidence that the degree of price stickiness differs across commodity items, and that the nonparametric hazard function of price changes is downward-sloping with some spikes. We introduce item-specific heterogeneity into the standard single-sector model of Calvo (1983) and estimate a hazard function of price adjustment, by applying duration analysis. We present the appropriate form of heterogeneity for the data structure, and show that the decreasing (population) hazard function is well described. In the presence of item-specific heterogeneity, the probability that prices remain unchanged is predicted to be higher than in the single-sector model.

  Introduction

  Previous studies (Bils and Klenow, 2004; Dhyne et al., 2005; Saita et al., 2006) have shown that the degree of price stickiness differs across commodity items. The time-dependent pricing model (Calvo, 1983), in which one single parameter represents price stickiness, cannot reproduce the strong empirical evidence in such a way that the nonparametric hazard function of price changes is decreasing (Álvarez, Burriel and Hernando, 2005). Each item has specific factors related to its survival experience. These specific factors, whether observable or not, change the shape of the (individual) hazard function. If the variability in hazard is not fully captured by covariates, it is necessary to model unobserved heterogeneity.

  2007.10

  PDFファイルはこちら

• ｢名目価格の硬直性に関するミクロ計量分析」
  Takayasu Matsuoka

  Abstract

  本研究は日本における小売物価の硬直性を Calvo（1983）で与えられた定義に従って計測し、その価格改定パターンを解明することを目的とする。分析には総務省統計局が作成している小売物価統計調査の価格データ（2000 年 1 月～2005 年 12 月）を用い、消費者物価指数のバスケットを構成する銘柄の価格硬直性を計測した。計測結果から価格変更のハザード率が 21.1％であり、平均して 4 ヶ月間は価格が同じ水準で維持されることが明らかになった。また、本研究では生存時間分析の枠組みを応用し、Calvo 型の価格設定行動の妥当性を検証した。Weibull ハザードモデルによって推定を行い、ハザード率一定の Wald検定を行った結果、Calvo 型価格設定行動では一般物価の価格改定パターンを十分に説明できないことが明らかになった。

  Introduction

  総需要の変化に対して雇用や産出量といった実物変数がどのように反応するかという問題は、古くからマクロ経済学の中心的な争点であった。総需要のショックが実物変数に波及していく経路があるとするならば、その有力な論拠の一つとなるのが名目価格の硬直性である。財・サービスの名目価格が即座に調整されないとき、総需要の変化に対して企業は所与の価格のもとで生産量を変化させる。価格調整に関するコストや制度的要因といった「摩擦」があることにより、総需要ショックが実物変数に影響する経路が生まれるのである。したがって、そうした価格調整にともなう「摩擦」とは何であり、どういった要因にもとづいて価格が硬直的になるのかを実証的に明らかにすることは、短期的なマクロ経済の変動を描写するにあたって、また総需要に影響を与える様々な経済政策を評価するにあたってきわめて重要な意義をもっている。

  2007.10

  PDFファイルはこちら