A Replication of “It is Not Merely About the Content: How Rules Are Communicated Matters to Administrative Burden”

ABSTRACT This study replicates prior research to examine how rule communication forms shape administrative burden, while also extending their research by investigating how artificial intelligence (AI) services moderate this relationship. AI services are understood to perform two distinct functions: the labeling function, where chatbots output policy rules directly from databases in response to user queries without further processing, and the optimization function, where chatbots collect, analyze, and distill rule information to present users with simplified, prioritized key points. In Study 1, the original research is replicated in the context of natural disaster relief applications in China. The findings demonstrate that bureaucratic language raises learning and compliance costs, stress, and autonomy loss, while complex information structures also increase learning costs, compliance costs, and autonomy loss. These results confirm that rule communication shapes administrative burden across policies, countries, and languages. However, differences emerge compared to the original study, in which bureaucratic language induced stigma and information complexity had no effect on compliance or autonomy loss, potentially due to contextual factors such as the complexity of Chinese logographic writing. Building on this, the extension experiment in Study 2 demonstrates that AI services reduce administrative burden. Notably, and contrary to expectations, the labeling function reduces learning, compliance, and psychological costs, while the optimization function achieves even greater reductions across all dimensions. Collectively, these findings provide empirical evidence supporting the use of AI to improve rule communication and ease administrative burden.