The Methodology
First, 47 city-specific measures were divided into the 12 broad but diversecategories (see charts). Many of these had to be normalised for the number ofsec a&b households. Next, each of these were re-normalised. Twelve intermediateindices were then derived by giving equal weights to each of the measures for each of the12 categories. The final index was created for each city by taking the average of each ofthese indices. The 12-category indices and the one overall index were then rescaled sothat they had a minimum value of 0 and maximum of 100. A higher value indicates betterperformance.