The role of peers in promoting energy conservation among Chinese university students

China's carbon neutrality goal necessitates promoting pro-environmental behaviors, particularly among university students who represent future leaders. Current emission reduction measures primarily focus on production, neglecting the significant role of consumer behavior. University students, as future leaders, hold a pivotal position in shaping societal energy consumption patterns. This study examines the "peer effect," where an individual's attitudes and behaviors are influenced by their peer group. The research question is: Does a university student's energy-saving behavior improve when their fellow majors demonstrate active energy-conservation behaviors? This is important because peer influence is particularly strong during university years, and in China's education system, students within the same major are often closely grouped, maximizing potential peer influence. While the peer effect has been widely researched, its impact on energy-saving behavior among university students requires further investigation, especially considering the broader context beyond immediate roommates.

The literature review encompasses research on peer effects across diverse fields, including education, wastewater treatment, subjective well-being, and prosocial behavior, highlighting the significant role of social norms. Manski (1993) identifies three factors influencing behavioral convergence: endogenous (peer) effects, exogenous (contextual) effects, and correlated effects. Many studies on peer effects utilize roommate assignments due to their relative randomness; however, this approach only captures a small part of peer influence. This paper aims to address this limitation by focusing on the broader network of peers within the same university major, creating a more comprehensive understanding of peer effects on energy saving behavior. The paper also notes existing research on energy-conservation behavior, focusing on factors like risk perception, motivation, income levels, and the effectiveness of various promotion strategies, including information interventions, gamification, and social media campaigns. While social norms are recognized as important, the specific impact of peers within the same major on university students' energy-saving behavior remains under-researched.

This study uses data from the University Students' Cognition of Energy and Environment Survey (CSCES), a large-scale (2993 effective samples) face-to-face survey conducted at Xiamen University. The survey explored students' energy knowledge, awareness, behavior, and other low-carbon-related aspects. The peer group was defined as students sharing the same education stage, grade, school, and major, leveraging the inherent semi-random assignment of students to majors. A minimum group size of 5 was set, yielding 1617 valid samples. The classical linear-in-means model was employed to estimate the peer effect, where an individual's energy-saving behavior score (sav) is regressed on the average energy-saving behavior score of their peers (peer), along with individual and household characteristics as control variables. The energy-saving behavior score was a composite of five items measuring energy-saving habits. Standard errors were clustered at the major level to account for the correlation among students within the same major. Robustness checks were conducted by using 2SLS estimation method with instrumental variables, varying the minimum group size, conducting placebo tests and controlling for correlated and exogenous effects, and employing alternative estimation methods (ordered probit and ordered logit). Mechanism analysis also investigates the direct (knowledge transfer) and indirect (influence on values and attitudes) components of the peer effect by exploring the influence of peers’ energy-saving behavior on individual’s energy-saving knowledge, willingness, awareness and social responsibility.

The baseline regression results show a statistically significant positive peer effect (p<0.01). A one-standard-deviation increase in peers' average energy-saving behavior leads to a 0.114-standard-deviation increase in individual behavior. Heterogeneity analysis reveals significant peer effects for both male and female students. However, the peer effect is more pronounced among higher-grade students and those with lower levels of depression and more positive environmental values. The effect is also stronger among those who pay attention to energy issues and when there is less difference in energy saving behavior within the peer group. Robustness tests, including instrumental variable estimation, changing the minimum group size, placebo testing, controlling exogenous effects and using alternative estimation methods all yielded significant results, although the magnitude of the coefficient varied across different approaches and settings. The 2SLS estimation shows consistent results with baseline regression. Mechanism analysis reveals that the peer effect operates through both direct (knowledge) and indirect (values, attitudes, social responsibility) channels.

The findings strongly support the presence of a significant positive peer effect on energy-saving behaviors among Chinese university students. This highlights the potential of social influence as a tool for promoting pro-environmental behaviors. The heterogeneity in the peer effect suggests that targeted interventions, focusing on specific student subgroups, can maximize the impact. The mechanism analysis confirms that the peer effect operates through both knowledge dissemination and the shaping of environmental values and attitudes. These findings support the growing literature on social contagion in pro-environmental behavior and suggest that social networks can be an effective complement to market-based mechanisms for promoting energy conservation.

This study demonstrates a significant positive peer effect on the energy-saving behavior of Chinese university students. This effect is heterogeneous across subgroups of students and operates through both knowledge and attitudinal channels. Policy implications include utilizing peer effects in energy conservation campaigns, targeting interventions to receptive groups, and enhancing energy conservation education within universities. Future research could explore the generalizability of findings to other universities and investigate the peer effects of other pro-environmental behaviors.

The study's limitations include its focus on a single university in China, limiting generalizability to other contexts. The study also concentrated on energy-saving behaviors, without extending to other forms of environmentally friendly practices. Future research can address these limitations by replicating the study in diverse settings and investigating the peer influence on broader pro-environmental behaviors.