Association for Public Policy Analysis & Management

Achieving UN Sustainable Development Goals is the common vision of the global community. However, the complexity of the relationship among development goals determines that the 17 SDGs are hard to be realized one by one or simultaneously. It is necessary for researchers to lay emphasis on the studies of the trade-off and synergy relationship among Goals. As the existing literature pays little attention to the health goals, this paper attempts to take SDG3 as the central Goal and supervised variable, based on regional Sustainable Development Goal database from Tsinghua University in China, to reveal the relationship between SDG3 and other development goals or indicators. Key findings through feature selection algorithm in machine learning, panel regression model and spearman correlation analysis are as follows. SDG3 has one optimal Goal subset having 7 Goals and one indicator subset containing 17 indicators. SDG3 has a complementary relationship with most of the development goals and indicators and a negative non-conflict covariant relationship with fewer Goals and indicators. Among them, Goal 1 (No Poverty), Goal 2 (Zero Hunger), Goal 5 (Gender Equality), Goal 6 (Clean Water and Sanitation), Goal 9 (Industry, Innovation and Infrastructure), Goal 10 (Reduced Inequality) are positively connected with SDG3. There is a negative covariant relationship between Goal 11 (Sustainable Cities and Communities) and SDG3. This paper mainly contributes to provide a new methodology for the data reduction of Sustainable Development Goals and indicators; meanwhile enrich the understanding of the relationship between different Goals.