Open Access

In this dissertation, I focus on the link between (internal) corporate governance structures and processes and firms financial reporting quality. Specifically, the dissertation aims to provide insights into the following general research question: What is the effect of different corporate governance stakeholders on the financial reporting quality of a firm? I provide insights into this question through three different articles. While Chapter 2 and 3 of this dissertation are concerned with company owners as one, albeit diverse, group of stakeholders, Chapter 4 focuses on managers and their ability to influence firm outcomes. The paper Do Family Firms Engage in Less Earnings Management? A Meta- Analysis, which is co-authored with Tim Hasso and Dominik Wagner, explores the relationship between family firm status and earnings management and synthesizes and explains previous research findings with the help of meta-analytic methods that are still uncommon in financial accounting research. We find a negative relationship between family firms and earnings management on average across 37 primary studies (and 305 effect sizes in total). Furthermore, we show that the considerable variation in size and direction of primary effect sizes can be explained by researchers choice of study design, earnings management proxy and different institutional settings. The paper Do Institutional Owners Deter Earnings Management? A Meta- Analysis, which is co-authored with Tim Hasso, explores institutional owners as a different set of shareholders and their impact on financial reporting quality. The study enables us to compare the results against the backdrop of the previous chapter and to see different rationales that managers in institutionally-owned companies might have to engage in earnings management. Here, we study 511 effect sizes from a total of 87 primary studies and find that the average effect is slightly negative, meaning institutional owners on average can get more transparent earnings figures from the companies they invest in. Similar to the work we did on family firms, we find considerable heterogeneity between results from primary studies. Specifically, our multivariate meta-regression models can explain 26 percent of the variability in effect sizes, mainly attributable to study design choices, such as how scholars measure the relevant constructs of ownership and earnings management. Conversely, this also indicates that across the 87 primary studies looking at the link between institutional ownership and earnings management, 74 percent of heterogeneity in effect sizes is still not explained. The single-author paper The Fish Rots From the Head Down: Managerial Personality and Financial Accounting Manipulation is concerned with managers - the second main stakeholder in corporate governance - and how managerial personality drives the propensity to engage in fraudulent accounting activities. I use a primary sample of 956 professionals, who work in accounting and finance departments, and ask them to rate their immediate superior on dark triad personality traits, as well as common actions taken by management to obscure and manipulate earnings figures. I find that managers with high ratings for dark triad personality traits engage to a greater extent in fraudulent accounting practices, than managers scoring low on the dark triad scale. Moreover, I can show that traditional risk management mechanisms, like internal audit departments, are only partially effective. Specifically, I find that only internal audit departments that are fully staffed by external personnel can curb the adverse effect of dark triad managers on financial reporting quality. This suggests that managers with dark personalities can take advantage of mixed or entirely in-house internal audit departments. Overall, this dissertation contributes significantly to both literature streams of corporate governance and financial reporting quality. My work can explain a significant degree of heterogeneity in previous findings on the link between different kinds of ownership and earnings management. Further, it stresses that the considerable variation in current findings is not mainly attributable to cross-country differences, as previously suggested, but in no small part attributable to study design features. Both meta-analysis papers are among the first articles that apply advanced metaregression methods in the financial accounting literature. Thus, contrary to previous methods of research synthesis meta-analyses are uniquely able to draw systematic, quantitative and replicate conclusions from a body of academic research. Especially for large, mature literature strands taking stock of all findings on a specific relationship is valuable, as research rarely progresses one paper at a time. Condensing and differentiating cumulative research via meta-analysis is thus a valuable contribution to the field and can be a basis for further theory development and improvements in empirical study designs. Finally, I can provide additional evidence on current research linking executive personality traits and financial reporting practices. Asking practitioners directly about their assessment of managerial personality and the frequency of specific fraudulent actions helped to show the critical role of executive personality and complements recent archival and experimental findings. The survey design also enables me to study so far undetected fraud, which is almost impossible to examine with experimental or archival data and thus a significant contribution to existing literature. Finding links between managerial personality, internal controls and reporting quality contributes to the literature further, as it shows previously unhypothesized relationships that can serve as a basis for future research on the effectiveness of internal audit functions in preventing fraud given different managerial personality traits.