The primary objective of this study is to investigate the opportunities for economic restructuring, resulting in an optimal increase in the energy-environmental efficiency of the global economy. A novel stochastic data envelopment analysis with a perfect object method (SDEA PO) constitutes the methodology of the research. We equip SDEA PO with the projected gradient of the efficiency score. We employ the indicators of the gross domestic product (GDP) and carbon dioxide emissions (CO2) as output and undesirable output, respectively, and population and clean energy consumption as input and undesirable input, respectively. By using the SDEA PO, we obtain a group efficiency score for the global economy; the projected gradient identifies the direction of optimal economic restructuring. The indicator-wise components of the projected gradient determine locally optimal changes in the shares of each economy, serving particular goals. We use a factor analysis technique to aggregate them into one factor vector that determines the multicriteria optimal structural change. The factor vector determines the redistribution of the GDP, clean energy consumption, CO2 emissions, and population, leading to the maximum possible increase in the energy-environmental efficiency. The suggested approach may be used as a tool for decision-making in a variety of two-tier economic systems. •Energy-environmental efficiency of the global economy is examined.•A novel stochastic data envelopment analysis with a perfect object is employed.•Ways to increase energy-environmental efficiency are investigated.•Locally optimal economic restructuring is suggested.•Computer R-program is utilized and provided to support further research.