Biases in Consumer Price Index Methodology in Pakistan: Suggestions for Improvements

The issues relating to the complexity of the measurement of the Consumer Price Index (CPI) which is regarded as the best and most well known indicator of inflationary trends and without referring to which economic policies cannot be evaluated have long been debated. Any measurement error in CPI may over or understate inflation, which can have serious repercussions on monetary, fiscal and other economic management policies. The report of the Boskin Commission [Boskin, et al. (1996)] has identified the possible sources of bias in the CPI. These biases which this study has also corroborated through a primary survey of selected households relate to commodity and outlet substitution, quality adjustment and new product introduction as well as index calculation in the existing methodologies. In this paper these biases have been evaluated for Pakistan and ways to improve the construction of the Index have been suggested. Other issues in Pakistan relate to selecting a representative product (or good), defining average quality, data collection, weights determination and base year change. The use of the Geometric means index formula and Laspeyre’s Index to reduce the formula bias has been proposed in this study.