The impact of community social capital and trust on levels of accounting conservatism

Abstract

Accounting conservatism reflects the judgment and potential bias allowed per accrual-based accounting by choosing to disclose bad news ahead of holding a higher bar for reporting good news. Financial reporting aggressiveness has been a strong focus of current research; I argue that examining accounting conservatism, representing the other side of the financial reporting quality spectrum, is equally important. Most research has focused on the consequences of accounting conservatism, in contrast to my study, which examines potential antecedents. Essay 1 of my dissertation illuminates the relationship between community social capital and accounting conservatism. Community social capital is proxied by the social norms and networks of the firm's headquarter location and is collected on a county basis. The reputational pressures of social norms may restrain managerial rent-seeking behaviors, thereby resulting in increased accounting conservatism. Social networks facilitate information exchange, thus reducing information asymmetry and possibly reducing the need for accounting conservatism. Hence, the effect of community social capital on accounting conservatism, including the potentially opposing influence of social norms and networks, creates an empirical question and motivates my study. My research extends literature concerning various firm-specific factors associated with accounting conservatism by uniquely examining the influence of the qualitative characteristics of social capital. The results of my study indicate high community social capital is associated with lower levels of conditional accounting conservatism, implying the contracting benefits of high community social capital negate the additional contracting benefits of conditional accounting conservatism. Essay 2 of my dissertation studies the intangible trait of corporate trust and the associated choices about using accounting conservatism. Trust is the reliability and honesty of a relationship. My study measured trust by developing a comprehensive dictionary of trust-related words and phrases via natural language processing known as machine learning. I examined a large sample of firm-year text data from the Q and A section of earnings calls. Corporate demonstration of trust may act as a potential substitute for accounting conservatism, for example, in the context of contracting theory. In contrast, trust may complement corporate accounting conservatism strategies, such as being part of the overall corporate culture and risk aversion theory. Motivated by this potential dichotomy, my study explores the relationship between trust and accounting conservatism. My results indicate higher levels of firm trust are associated with lower levels of accounting conservatism, which supports the tendency of trust to act as a substitute for accounting conservatism leanings. My study adds weight to prior research by examining trust and further delineating its relationship with conditional and unconditional accounting conservatism. My study complements and extends text analysis research. To my knowledge, it is the first study to develop a corporate tone of trust dictionary via machine learning to examine its relationship with accounting conservatism.