Eunji Seo

This study advances understanding of COz emission dynamics in Malaysia by integrating Structural Decomposition Analysis (SDA) with Machine Learning (ML) techniques to identify key emission drivers and evaluate their stability and predictability during 2010-2020. Using multi-year input-output tables and emissions data, an additive SDA framework decomposes changes in COz emissions into six effects: COz intensity, technology (production structure), household expenditure, investment, government expenditure, and exports. Four ML methods, namely Neural Networks with SHapley Additive exPlanations (SHAP), Lasso Regression, Elastic Net Regression, and Bayesian Regression, are applied to rank driver importance at aggregate and sectoral levels. Results show consistent findings across methods, with COz intensity emerging as the most influential and variable driver, followed by technology and household consumption. In contrast, government expenditure, investment, and exports contribute relatively little to emission changes. Sectoral analysis indicates that emissions fluctuations are primarily driven by electricity, gas and water supply, manufacturing, wholesale and retail trade, and transportation. SHAP results further confirm the significance of intensity improvements in the early decade, demand-driven growth in the middle years, and pandemic-related declines in 2019-2020. The findings provide evidence-based insights for climate policy, highlighting the importance of renewable energy adoption, technological upgrading, carbon taxation, and sustainable consumption in supporting Malaysia's net-zero ambitions and Sustainable Development Goal 13.

Against the backdrop of the transition toward a sustainable society, the green product market has expanded steadily. However, green skepticism poses a significant challenge to promoting sustainable consumption. This study develops a parallel mediation model to examine how green skepticism influences green purchase intention through information seeking and anticipated guilt among Chinese consumers. Survey data were collected from 511 Chinese respondents and analyzed using structural equation modeling (SEM) with IBM SPSS AMOS 30. While prior research typically assumes a direct negative effect of skepticism on green purchase intention, our results show that green skepticism does not exert a significant direct effect on green purchase intention. Instead, green skepticism indirectly undermines green purchase intention by reducing consumers’ motivation to seek green product information and weakening anticipated guilt. These findings challenge the prevailing “skepticism-as-verification” view and suggest that green skepticism may foster information avoidance and moral disengagement rather than deeper cognitive and moral engagement. By identifying these disengagement pathways, this study helps clarify previously mixed findings on the relationship between green skepticism and green purchase intention. It delineates important boundary conditions for the effectiveness of green marketing and policy interventions.

This study examines how Japanese consumers’ awareness of multiple eco-labels, specifically Organic JAS and RSPO, affects their willingness to pay more for eco-labeled cosmetic products. Drawing on the Theory of Planned Behavior (TPB) and Signaling Theory, this study models the interaction between awareness of these two eco-labels and their influence on organic cosmetic buying intentions. Using statistical analysis of survey data from 505 Japanese consumers, the findings reveal that heightened awareness of the Organic JAS label significantly increases the willingness to pay more for eco-labeled cosmetics. However, when awareness of the RSPO label is also present, the positive effect of Organic JAS awareness diminishes, indicating that multiple eco-label signals may interfere with each other. These results underscore the complexity of consumer perceptions regarding eco-labels and suggest that the effectiveness of eco-labels as green marketing tools depends not only on raising awareness but also on understanding how different labels interact to influence consumers’ willingness to pay in the Japanese cosmetics market.

Previous studies have examined changes in software sales using the network effect theory framework based on hardware sales. This study aims to examine the effects of a software series’ length on software sales. We also investigate the moderating effects of how a platform’s life cycle influences the series’ sales. We adopted the double-jeopardy theory and brand loyalty discussions for both products and corporations. To test our hypotheses, we analysed 11,863 video game software samples in the Japanese market, addressing the following research questions: How does the length of a software series influence sales in terms of both the product and the corporate brand? How is the influence of the software series’ length on sales moderated by the stage of the platform’s lifecycle? Our findings show that a longer software series led to higher software sales for both the specific series as well as other software released simultaneously by the company. Moreover, we found that the positive relationship between the video game series’ length and sales of all the firm’s software was stronger when the platform had matured. By examining firm behaviours, such as software series releases and their effects, we confirmed that software series’ length is a core factor stimulating its sales and that of other series produced by a company. We also examined the impact of a platform’s lifecycle on video game sales. Our study contributes to the field by extending the literature on the effect of brand loyalty and the double-jeopardy theory.