Author
Frances Krsinich
Hedonics is the use of regression modelling to measure price change, where price-determining characteristics are included in the regression model as explanatory variables, so that the price change can be determined independent of the effect of these characteristics. For price measurement from samples with changing characteristics over time, this technique is being used to an increasing extent internationally. Statistics New Zealand uses a hedonic estimation method to measure the price change of used cars in the consumers price index, and has recently used hedonic methods in some benchmarking exercises to assess the quality of non-hedonic estimation methods for the measurement of both housing rentals and house prices. Use of Hedonics at Statistics New Zealand provides details of these uses, and raises some points for consideration:
- The possibilities of using survey data rather than catalogue data. We were fortunate to have existing survey data, but consideration could be given to developing and collecting survey data for use with hedonic estimations where appropriate.
- Where individual-level characteristics are not available, but small aggregates are, the use of a 'semi-hedonic' approach using dummy variables for the small aggregates (such as suburbs, or model numbers) is likely to be a more effective and efficient approach than a stratification approach based on groupings of the same small aggregate variables.