The concept of well-being has emerged as a promising alternative to GDP for assessing countries’ progress, yet its multidimensional nature poses diverse measurement challenges. Among these challenges lies the treatment of negative externalities, representing socio-economic and environmental costs that substantially impact countries’ performances in well-being production. This study aims to investigate the implications of different methodological approaches in managing bad outputs within an efficiency analysis encompassing the 38 OECD countries. Four distinct approaches, well-established in existing literature, are initially implemented using the DEA methodology and subsequently through Bootstrap DEA to quantify biases arising from limited data usage. The findings carry a dual perspective. Methodologically, the approaches differ in the emphasis they place on discrimination power – when bad outputs are treated as inputs – versus accuracy – when bad outputs are transformed before analysis. Empirically, the results suggest that high absolute well-being levels might not necessarily align with efficient production, suggesting potential improvement even in the wealthiest countries by using national resources more efficiently.
Tackling negative externalities in well-being efficiency analysis: accounting for socio-economic and environmental costs
Gabriella De Bernardo
;Gennaro Punzo;Rosalia Castellano
2026-01-01
Abstract
The concept of well-being has emerged as a promising alternative to GDP for assessing countries’ progress, yet its multidimensional nature poses diverse measurement challenges. Among these challenges lies the treatment of negative externalities, representing socio-economic and environmental costs that substantially impact countries’ performances in well-being production. This study aims to investigate the implications of different methodological approaches in managing bad outputs within an efficiency analysis encompassing the 38 OECD countries. Four distinct approaches, well-established in existing literature, are initially implemented using the DEA methodology and subsequently through Bootstrap DEA to quantify biases arising from limited data usage. The findings carry a dual perspective. Methodologically, the approaches differ in the emphasis they place on discrimination power – when bad outputs are treated as inputs – versus accuracy – when bad outputs are transformed before analysis. Empirically, the results suggest that high absolute well-being levels might not necessarily align with efficient production, suggesting potential improvement even in the wealthiest countries by using national resources more efficiently.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


