This PhD thesis by Peter Hein van Mulligen examines the hedonic method, one specific method for explicit quality adjustment. The methods are based on the assumption that characteristics and not finished products are important to users, producers and end prices.
To measure inflation, one of the most important economic indicators, prices indices are used. To construct price indices, prices of more or less identical products are compared through time. Such price comparisons are also done between countries to determine relative incomes or production levels of countries, as exchange rates are unsuitable for these purposes. However, because products are often not ‘more or less’ identical, quality bias occurs.
The thesis examines the quality issue that emerges when prices are compared through time, and across space. Traditional measures only deal with this quality issue implicitly, if they examine it at all.
The hedonic method is applied to two types of durable goods: computers and cars. For computers, various quality-adjusted price indices are calculated for all computer sales in the Netherlands from 1999 to 2001. For cars quality-adjusted international price comparisons are conducted for the car industry in a number of important car manufacturing countries.
The conclusion of the thesis is that other methods than the hedonic methods offer a better prospect to minimise quality bias in price indices, but that for international price comparisons the hedonic method is more suitable to remedy the quality issue.
Appendix with all results of the regressions on scanner data for computers