1. Главная
  2. Ресурсы
  3. Библиотечный поиск
Scale economies withregard to priceadjustment costs andthe speed of priceadjustment inAustralianmanufacturing Michael Olive2008 год

Scale economies with regard to price adjustment costs and the speed of price adjustment in Australian manufacturing
статья из журнала

Полный текстСтраница публикацииПубликация в OpenAlex
Аннотация: Abstract This paper aims to examine the impact of firm size, industry concentration and the length of production on industry speed of price adjustment. To motivate the paper, an industry pricing model in error correction form is derived from firm pricing behaviour. As a new development, firms are assumed to have price adjustment costs that are a function of their size. The empirical model is estimated using two‐digit Australian manufacturing industry data for the period 1994:3 to 2006:1. The results suggest that the industry speed of price adjustment is positively related to firm size and negatively related to industry concentration and the production lag. Implied values for industry speeds of price adjustment are generally small when compared to other country industry studies. However, the industry average median lag of 7.1 quarters indicates a slightly faster speed of price adjustment than the estimate for the Australian consumer price index by Dwyer and Leong (Citation2001). Keywords: speed of price adjustmentadjustment costsAustralian manufacturing Acknowledgements The author wishes to thank Harry Bloch, Chris Heaton, Roselyne Joyeux, Ellen Young and two anonymous referees for their helpful comments and also to thank Abdel Joubaili and George Milunovich for their research assistance. Notes 1. It should be noted that the number of price changes per year by a firm is not the same thing as the speed of price adjustment to a target price. However, given changing economic conditions, price changes are a precondition for a positive speed of price adjustment. Indeed, Calvo (Citation1983) models a direct positive link between the number of price changes and the speed of price adjustment, although recent evidence by Bils and Klenow (Citation2004) indicates that Calvo's model does not fully explain US data. 2. Zero fixed costs are assumed for simplicity. This does not affect the analysis. 3. For small changes, Δpit /p it−1 ≈ lnpit − p it−1 and . 4. This method for obtaining an error correction model through aggregation could be contrasted with that outlined by Nickell (Citation1985), where an agent minimizes a loss function with respect to two components of an aggregate variable. 5. When the domestic market four‐firm concentration ratio is correlated with the average size of the four largest industry firms measured by sales for the nine industries in 1998–1999, the correlation ratio is 0.43. This moderate level of correlation indicates that there is enough information for the differing influences of these variables on the industry speed of price adjustment to be detected in the regression analysis. 6. Generally, pricing equations will be a function of cost and demand shift variables, except when the demand function is iso‐elastic and moves proportionally (see Bloch, Citation1992; Olive, Citation2002). 7. It should be noted that this test is only indicative, as it does not allow for the cross‐industry constraints and the short‐run dynamics implied by the empirical model. 8. An ARDL(3, 3) model is initially chosen in order to capture lags that could occur for up to one year. 9. As the series used in this study are in index form, not all of the regression estimates for Equations (Equation8), (Equation9) and (Equation10) are easy to interpret when series are in levels. Characterizing an index as an unknown number multiplied by its true value, parameter estimates will generally be the true estimates multiplied by unknown constants. The exceptions are the speed of price adjustment estimates for Equation (Equation10) which will be the true estimates only when industry prices (and, therefore, lagged prices and price differences) are multiplied by the same unknown number. In contrast, if the time series variables are transformed into natural logs, the slope coefficient estimates can be directly understood as the unknown numbers fall into the estimates of the constant coefficients (that is, the θ d1). Given the similarity of the results shown in Table 2, using series in levels does not seem to be a serious problem in the estimation of Equation (Equation10). 10. Kremers et al. (Citation1992) indicate that the speed of adjustment in the single equation has an asymptotic distribution that ranges from normal to Dickey‐Fuller under the null of no cointegration. The exact nature of the distribution is determined by the short‐run dynamics and the signal‐to‐noise ratio. In the absence of an exact analytical distribution for a panel under the null, the t‐distribution is assumed. However, that Student's t ratio might over‐reject the null is a further reason to describe the results for the production lag as weak. 11. This median lag formula is appropriate for the partial adjustment model, so the values derived here are approximations of the true median lag. 12. Pesaran and Shin (Citation1999) indicate that valid inferences on the long‐run parameters can be made using standard normal asymptotic theory when the variables are I(1) and the model is of the single equation ARDL type.
Год издания: 2008
Авторы: Michael Olive
Издательство: Taylor & Francis
Источник: International Review of Applied Economics
Ключевые слова: Firm Innovation and Growth, Economic Growth and Productivity, Global trade and economics
Другие ссылки: International Review of Applied Economics (HTML)
CiteSeer X (The Pennsylvania State University) (PDF)
CiteSeer X (The Pennsylvania State University) (HTML)