An estimate of upper level substitution bias in the Icelandic CPI

Authors

  • Bjarni V. Halldórsson
  • Oddgeir Á. Ottesen
  • Stefanía H. Stefánsdóttir

DOI:

https://doi.org/10.24122/tve.a.2011.8.1.5

Keywords:

Consumer price index, inflation, substitution bias, mid-year indices, Marshall-Edgeworth index.

Abstract

The household expenditure surveys which are used to calculate the weights of the consumer basket are on average 3¼ years old when used for calculating the consumer price index. The expenditure surveys last for 3 years and it takes 15 months to process the data. The weights of the consumer basket in Iceland are updated yearly. Using historical data we can estimate the upper level substitution bias. We estimate upper level substitution bias in the consumer price index in Iceland as 0.3% per annum. In our data, the inflation estimates are comparable when the consumer basket is updated every fifth year as opposed to yearly.

Author Biographies

  • Bjarni V. Halldórsson
    Reykjavík University
  • Oddgeir Á. Ottesen
    Financial Supervisory Authority, Iceland

Published

2011-06-15

Issue

Section

Peer reviewed articles