<p>Leading agencies, such as the Committee on Publication Ethics (COPE), the World Association of Medical Editors (WAME), and the Open Access Scholarly Publishing Association (OASPA), recommend sharing research data. Further, all major publishers also encourage researchers to practice openness and transparency by sharing the data sets. In contrast, many researchers from social science streams are reluctant to share their data sets. We employ multiple methodologies to understand the causes and repercussions of this contrast. Four case studies show no rationale for social science authors not sharing their data sets. Survey results (391 social science Scopus-Q1 authors worldwide) support the hypothesis that the non-availability of data and non-compulsion from journals/editors/reviewers are the major causes for the non-sharing of research data. Interviews with ten senior editors reveal that publishers are reluctant to impose compulsion out of the fear that there would be a reduced supply of articles if data sharing is made compulsory. The study provides reasonable evidence to conclude that those who do not share their research data are most unlikely to have it. To address this serious flaw, we recommend compulsory data sharing to discourage unauthentic research not backed up by underlying data. This paper uses sizable empirical evidence to highlight the lack of authenticity in social science research. It further recommends policy change to make data sharing compulsory to increase the authenticity of social science research.</p>