Progress is hard to measure

Wellington Regional Council have recently published their Genuine Progress Indicator, which is intended to measure changes in regional well-being. Measuring well-being is very difficult and the technical documentation provided by the WRC shows how hard they have found it to overcome the challenges.

The GPI has been constructed by taking about 100 variables of relevance to well-being, normalising each, and averaging the 100 indices. The Council have declined to weight the aggregation because they recognise that people may disagree over the weighting. They seem to want to avoid arguments over the normative weighting decisions. Unfortunately, weighting everything equally is just as much of a value judgement as any other weighting system. For instance, the council consider the prevalence of smoking to be a negative indicator. Due to the equal weightings, a 1% decrease in smoking in the region would be as good for progress as a 1% increase in incomes, or a 1% decrease in unemployment. With other variables, from access to public transport to dairy farm soil quality, it seems unlikely that many people would agree with weighting them all equally.

There are plenty of other difficulties, too: ensuring comparability of the variables measured and selecting a baseline for normalisation, for instance. What these difficulties illustrate are the importance of value judgments in creating these GPIs, even when the architects try to steer away from making them. Each of us, given the opportunity to choose our own variables and weightings, could come up with a different result for the region’s progress. Because of that it’s hard to take the GPI seriously as a reliable measure of regional progress, except insofar as it is defined by the council’s own preferences.