Research Work Group on CPI measurement.
Higher School of Economics, Perm branch.

The aim of the group was to get a deeper understanding of the Russian CPI methodology, evaluate potential errors and biases

Here you can find:




Hedonic Price Indices
Implemented Microeconomics and Econometrics to show that the Russian CPI could be upward biased by 0,12-0,23% per year due to new goods and quality change effects for personal computers (PC)
We've estimated Hedonic indices for PC that allows to get better estimates of price change for PC: a 11% fall for personal computers in 2005 was estimated and found a 11-23% bias in price index for PC in Russia

Abstract

Economist have noted for decades that Consumer Price Index (CPI) in the developed countries is overstating inflation by 0,5-2,0% per year. A significant part of such a bias is found to be caused by the presence of technology products and differentiated products in CPI. An increasing weight of these products in the Russian CPI may also lead to a substantial upward bias. Nowadays hedonic indices are believed to be the most efficient way to reduce the bias. Following Triplett we have developed the taxonomy of hedonic price indices within the “direct approach”. This first step is essential in constructing a detailed set of hedonic price indices. They can be used in two ways: to estimate the bias in CPI and to elaborate alternative official price indices for IT-products. We've estimated 11% fall of price for personal computers in 2005 with the usage of hedonic price indices, a 11-23% positive bias in price index for PC in Russia was also calculated. We also have found that Russian CPI could be upward biased by 0,12-0,23% per year due to new goods and quality change effects for PC.

B.A. Thesis  (119 pages, 856 kb) Download

Short Article  (20 pages, 335 kb) Download

(Presentation in English) Download

(Presentation) Download

Prices and characteristics Data for 2004 Download

(they were not used in the original paper)

Prices and characteristics Data for 2005 Download

Hedonic Regression Estimates. Download

 

 


Home       |         For Employers       |         Research      |         Personality       |         Awards      |         Contact
Alexander Parkhomenko