Organised by IMF
This workshop outlines data needs for and practical methods of quality adjustment. The data needs are about price collectors having fairly tight specifications and training about what to do when an item is missing in terms of finding comparable and, if not available, non-comparable replacements and, for the latter, the different data needs and ways of incorporating the replacement. Some methods are relatively straightforward including the indirect ones of imputation and linking; at the other extreme, direct more-complex methods include use of hedonic regressions. Such regressions are also used when there is a rapid turnover in transactions on models with differing characteristics, such as consumer electronics, and can be applied to separately constructed databases taken, for example, on a monthly basis from retailers’ websites on the prices and characteristics of different models on the market. We look at the practice of using such techniques.
Krsinich - quality adjustment in the NZ CPI.doc
Krsinich QA and hedonics in NZ CPI.pptx