The digital economy is a key driver of industrial upgrading and regional growth. Focusing on Gansu Province—an under-represented, less-developed region in northwest China—this study constructs a multidimensional digital economy index (DEI) for 2009–2023 under a unified normalisation and weighting scheme. Two complementary MCDA approaches—entropy-weighted TOPSIS and SESP-SPOTIS—are implemented on the same 0–1 normalised indicators. Robustness is assessed using COMSAM sensitivity analysis and is benchmarked against a PCA reference. The empirical analysis then estimates log-elasticity models linking modern logistics production (MLP) and the DEI to the provincial GDP and sectoral value added, with inferences based on White heteroskedasticity–robust standard errors and bootstrap confidence intervals. Results show a steady rise in the DEI with a temporary dip in 2021 and recovery thereafter. MLP is positively and significantly associated with GDP and value added in the primary, secondary, and tertiary sectors. The DEI is positively and significantly associated with GDP, the primary sector, and the tertiary sector, but its effect is not statistically significant for the secondary sector, indicating a manufacturing digitalisation gap relative to services. Cross-method agreement and narrow sensitivity bands support the stability of these findings. Policy implications include continued investment in digital infrastructure and accessibility, targeted acceleration of manufacturing digitalisation, and the development of a “digital agriculture–smart logistics–green development” pathway to foster high-quality, sustainable regional growth.

In recent years, universities have increasingly been treated as everyday laboratories for greener lifestyles, yet the link between what students say about the environment and what they actually buy is still quite uncertain. This study applies the Theory of Planned Behaviour to examine how Subjective Norms and Perceived Behavioural Control are connected to Green Consumption Attitude, Green Buying Intention and Green Purchase Behaviour among university students in China. Based on 181 valid questionnaires and analyses conducted with PROCESS Model 6, the findings suggest that both Subjective Norms and Perceived Behavioural Control are linked to more positive attitudes toward green consumption and higher levels of purchase intention. Green Buying Intention is found to be the most stable predictor of behaviour, whereas Green Consumption Attitude alone neither shows a significant direct effect nor serves as an effective single mediator. In most cases, the two predictors influence behaviour through a stepwise route in which changes in attitude are followed by shifts in intention and, only after that, by actual purchasing. This sequential pattern underlines the importance of intention in explaining why supportive environmental views do not always lead to consistent green actions, and it provides a basis for universities, public agencies and businesses to design more targeted measures that make green choices easier and more attractive for students.