Ongoing advancements in technology have dramatically changed the disclosure media that companies adopt. Such disclosure media have evolved from traditional paper-based ones to the internet as the new platform for disclosing information on companies’ designated websites; however, the new media for disclosures are currently social media platforms. Among the important disclosures that companies make are Corporate Social Responsibility (CSR) disclosures. These are now deemed indispensable due to growing social awareness among societies. While a significant body of the literature is dedicated to studying CSR disclosures and their determinants on the traditional media, studies examining CSR disclosures on social media are quite few. Hence, the aim of this paper is to investigate corporate social media accounts for CSR disclosures, as well as identify their determinants. The sample of the study is comprised of tweets posted on the Twitter accounts belonging to the FTSE 350 constituents. Topic modeling, an unsupervised machine learning approach, is applied to identify CSR disclosure tweets and regressions are run to identify the determinants of CSR disclosure on Twitter. The utilization of topic modeling to reveal CSR content on social media is considered a new approach to mainstream CSR disclosure literature; as such, this paper examines the new phenomenon of CSR disclosures on social media using new research methods. Results show that the popularity of Twitter as a CSR disclosure platform has risen significantly over the past few years. The paper also provides evidence that the presence of women on company boards, especially non-executive women, is positively associated with the extent of CSR disclosure on Twitter. Results highlight the importance of board independence for CSR disclosure in a social media disclosure context, complementing other disclosure platforms. Larger firms are also found to disclose more CSR information on Twitter compared to smaller ones indicating that firm size plays an important role determining a company’s level of CSR disclosure on social media.

技术的不断进步极大地改变了公司采用的披露媒体。这样的公开媒体已经从传统的纸质媒体发展到了互联网，成为了在公司指定网站上公开信息的新平台。但是，用于公开的新媒体目前是社交媒体平台。公司进行的重要披露中包括企业社会责任（CSR）披露。由于社会之间日益增长的社会意识，这些现在被认为是必不可少的。尽管大量文献致力于研究传统媒体上的CSR披露及其决定因素，但研究社交媒体上的CSR披露的研究却很少。因此，本文的目的是调查企业社交媒体帐户中的企业社会责任披露，并确定其决定因素。该研究的样本包括在属于FTSE 350成分股的Twitter帐户上发布的推文。主题建模是一种无监督的机器学习方法，可用于识别CSR披露推文，并进行回归以确定Twitter上CSR披露的决定因素。利用主题模型在社交媒体上揭示企业社会责任内容被认为是主流企业社会责任披露文献的一种新方法。因此，本文使用新的研究方法研究了在社交媒体上披露企业社会责任的新现象。结果表明，在过去的几年中，Twitter作为CSR披露平台的受欢迎程度已显着提高。该文件还提供了证据，表明公司董事会中有女性，特别是非执行女性，与Twitter上的CSR披露程度呈正相关。结果凸显了董事会独立性在社交媒体披露背景下对CSR披露的重要性，是对其他披露平台的补充。与较小的公司相比，还发现较大的公司在Twitter上披露了更多的CSR信息，这表明公司规模在决定公司在社交媒体上披露CSR的水平方面起着重要作用。