Stochastic multi-attribute acceptability analysis (SMAA): an application to the ranking of Italian regions

We consider the issue of ranking regions with respect to a range of economic and social variables. Departing from the current practice of aggregating different dimensions via a composite index, usually based on an arithmetic mean, we instead use stochastic multi-attribute acceptability analysis (SMAA). SMAA considers the ‘whole space’ of weights for the considered dimensions. The methodology is applied to the ranking of Italian regions, showing that although the north–south divide is definitely wider than the one measured simply in terms of gross domestic product. There are southern regions that perform generally better than those belonging to their broad region: a kind of ‘northern regions within the southern broad region’. This result poses interesting questions about the uneven development of Italian regions.本研究探讨基于多维度经济与社会变量开展区域排序的议题。现有研究通常通过综合指数整合不同维度，此类综合指数多基于算术平均构建，而本研究摒弃这一范式，转而采用随机多属性可接受性分析（stochastic multi-attribute acceptability analysis，SMAA）。该方法将所考量维度的全部权重空间纳入分析范畴。本研究将此方法应用于意大利各区域的排序分析，结果显示：尽管仅以国内生产总值（gross domestic product，GDP）为衡量标准，意大利的南北差距已颇为显著，但实际差距幅度明显更大。部分南部区域的整体表现优于其所属大区的平均水平，堪称“南部大区内的北部区域”。这一研究结果为意大利区域非均衡发展的相关议题带来了富有启发性的探讨空间。