Weighting for what? A comparison of two weighting methods for measuring urban competitiveness

This paper examines how different weighting methods affect the results of urban competitiveness ranking. Using a hierarchical indicator system that considers economic, social, and environmental dimensions of the competitiveness of Chinese cities as a case study, it shows that principal component analysis (PCA) produces rankings of urban competitiveness that are significantly different from the alternative equal weighting (EW) method. Moreover, PCA generates problematic weights for some indicators and assigns a small weight to the environmental performance of cities. The paper investigates these problematic results and argues for the necessity of deliberating more consciously on appropriate weighting methods in the construction of composite indicators in urban competitiveness studies and beyond.

► We compare two weighting methods for measuring cities’ competitiveness.
► The ranking results based on these two methods are significantly different.
► Principal component analysis generates illogical weights for some indicators.
► Principal component analysis reduces the weight of the minority indicators.
► Equal weighting provides more transparent and theoretically plausible rankings.