本次讲座以民国重要历史人物王伯群为例，探讨档案研究与人物研究的深度互动。主讲者将散见于两岸三地的书信、公文、日记等第一手档案，还原这位集民主革命先驱、交通部部长、大夏大学创办者于一身的复杂历史形象。讲座将深入解析如何通过档案挖掘，突破单一历史评价，展现人物在时代变革中的多重面相与内心世界。特别关注王伯群在民主共和、抗日救亡、高等教育、交通建设等领域的贡献，揭示档案如何让被遮蔽的历史重新彰显。通过这一案例，不仅展现历史研究的求真意趣，更思考档案在文化传承与集体记忆建构中的当代价值。

Using the full text of official annual reports from municipal governments and local firms in China, we develop a state-of-the-art machine learning model to predict one-year-ahead earnings changes. Our approach significantly expands the set of predictors beyond financial data by incorporating measures of government-business nexus and key macroeconomic variables. By employing ensemble learning, our model achieves an AUC of 78.37%, demonstrating substantial improvement over conventional machine learning techniques. Our results validate the critical predictive value of both the ensemble method and the newly incorporated measures of government alignment and macroeconomic factors.

This paper investigates the impact of peer-to-peer lending on a firm’s workplace safety. By utilizing the temporal and spatial variation in Lending Club’s entry in US, with a difference-in-differences approach, we find that peer-to-peer lending is associated with a decrease in death rates in working place. We demonstrate that our findings are more significant in firms with higher unemployment rates, higher institutional ownership, a higher proportion of ethnic minorities, higher market concentration, and fewer labor unions. Our study contributes to the literature by revealing the social welfare of P2P lending platforms in enhancing worker safety and health.

This paper examines whether and how the introduction of peer-to-peer (P2P) lending platforms affects opioid-related mortality, one of the most severe public health crises in the United States with over 150 daily overdose deaths. The theoretical relationship is ambiguous: while improved credit access may alleviate economic stress and reduce opioid intake motivated by financial distress, it could potentially enable opioid prescribing by providing additional financial resources. Exploiting the staggered introduction of Lending Club across U.S. states as a quasi-natural experiment, we find that P2P lending entry leads to a reduction of 17 opioid-related deaths per million population. The effect operates through improved psychological well-being, as evidenced by reduced work intensity, decreased mental health-related mortality, and lower suicide rates. Cross-sectional analyses reveal stronger effects in areas with limited banking presence, higher household leverage, and lower sunlight exposure, consistent with P2P lending being particularly beneficial where financial constraints and psychological vulnerability are more severe.

This study investigates the impact of Uber, a gig economy platform, on mental health. Utilizing a difference-in-differences approach, we analyze county-level data to assess the effects of Uber's entry on psychotropic drug-related mortality rates. Our findings reveal a significant decrease in mortality rates post-Uber entry, suggesting a positive influence on mental health. The impact is more pronounced in areas with lower financial inclusivity, higher unemployment, greater household debt and higher proportion of ethnic minorities, indicating that Uber's flexible income opportunities may alleviate financial stress, provide employment opportunities and improve psychological well-being. This research provides new insights into the role of the sharing economy in public health and mental health policy-making.