TIL about hedonic regression which decomposes the item being researched into its constituent characteristics
In economics, hedonic regression, also sometimes called hedonic demand theory, is a revealed preference method for estimating demand or value of a characteristic of a differentiated good. It decomposes the item being researched into its constituent characteristics and obtains estimates of the contributory value for each. This requires that the composite good (the item being researched and valued) can be reduced to its constituent parts and that those resulting parts are in some way valued by the market. Hedonic models are most commonly estimated using regression analysis, although some more generalized models such as sales adjustment grids are special cases which do not.
1 reply
There was a big online debate about it maybe 15 or 20 years ago and I spent some time looking into the details. It's not a bad idea in principle, but at least in the US and Canada they often take it to crazy extremes in order to hide price inflation.
They might for instance fail to adequately account for the high prices paid by early adopters of something new being representative of something other than the inherent merits of the new thing. Today's cars and vacuum cleaners are indeed better than their 1980s equivalents in many ways, but imo they're not better by nearly so much as to explain the "hedonic" adjustments that have been made.