1504.0 - Methodological News, Jun 2008  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 13/06/2008   
   Page tools: Print Print Page Print all pages in this productPrint All

Hedonic price index construction for laptop computers

Hedonic regressions for personal computers (PCs), in conjunction with a traditional matched model approach, are currently used in the Producer Price Indexes (PPI). Hedonic regression is a method for predicting the price of a good through the characteristics that make up that good. In the area of PCs, for example, sales information on a range of computers is available, together with information on each PC's characteristics such as RAM size, speed, hard drive size, etc. Through standard regression techniques, one can model the relationship between the price of the good and these characteristics.

The use of hedonic regression in a price index is motivated primarily by a good having a high rate of quality change. This means that the number of one-to-one matches in successive months is reduced, and so traditional matched model approaches become less effective. The bias that is inherent through this reduction in matches can be somewhat reduced by estimating the prices of those goods that are not matched; this is why the hedonic regression is used.

A review was recently conducted by the Prices Branch recommending that the timing was appropriate to investigate the possibility of extending the hedonics regression to laptop computers, as laptop sales have been very strong in the last couple of years. On top of this possibility for a laptop hedonic price index, the desktop hedonic regression also requires a review to test its robustness and applicability in the current environment. The Analytical Services Branch (ASB) has started to assist in this review and in investigating the feasibility of introducing a hedonic price index for laptops. For laptops, it is expected that the form of the regression equation will be similar to that for desktop computers, with slight changes in the explanatory variables included.

For more information, please contact Steve Lane on (02) 6252 7833 or Charity Liaw on (02) 6252 5578.