Using student subjects in experimental research: a challenge to the practice of using students as a proxy for taxpayers

Abstract

The tax discipline has accepted the practice adopted in the economics and psychology disciplines of using student subjects for experimental research purposes. However, in other social science disciplines (e.g. accounting, management or consumer research), experimental research that uses students as a substitute for another group has been widely criticised as having little external validity. In this research, data are compiled from journal articles over a 20 year period to investigate the extent to which tax experiments use student subjects and the potential implications from this practice. Three research questions are addressed. First, do tax studies using student subjects justify or explain the use of student subjects? Second, are policy recommendations or other extrapolations to the tax-paying population made from the student subject results? Third, is any acknowledgement made of the potential limitations of using student subjects? The findings indicate that approximately half of the tax experiments investigated used student subjects. Of these, approximately 80% provided no justification for their chosen subject sample. A similar proportion did not acknowledge any limitations from the use of a student sample. Around half of the papers using student subjects extrapolated their findings to the general tax-paying population or provided tax policy recommendations. In addition, the research indicates that the tax discipline has not had the same level of robust debate on the use of student subjects in experimental research as that seen in other disciplines. This potentially limits the external validity and utility of many of the tax research experiments undertaken to date.