Are financial conflicts of interest leading to reporting errors in financial and economics academic literature?

Abstract

This thesis explores whether financial conflicts of interest lead to reporting errors in academic literature within the fields of finance and economics. By analyzing 2554 articles and 377962 z-values from 1998 to 2016, the study assessed the impact of conflicts of interest on the accuracy of reported statistical results. The analysis employed quantitative techniques such as the Caliper Test, probit regression, and OLS regression. The Caliper Test revealed a significant over-representation of z- values just above critical thresholds, indicating potential result manipulation. Probit regression showed that, in economics, declaring conflicts of interest significantly decreases the likelihood of reporting significant results, while in finance, it slightly increases. OLS regression indicated that funding negatively influences the proportion of significant results, suggesting higher methodological rigor in funded research. The study emphasized the importance of transparency and concluded that researchers should disclose conflicts of interest, whilst journals need to enforce strict disclosure requirements. Funding bodies should consider implementing policies to minimize biases and unbiased research. In conclusion, financial conflicts of interest significantly affect the integrity of academic research in finance and economics. Promoting transparency and ethical standards is crucial for maintaining the reliability and credibility of published research, supporting better-informed economic and financial policies