Taxpayers’ Non-Compliance in the Digital Era

<p dir="ltr">Governments continue to face challenges in securing high levels of tax compliance, despite tax revenue being central to public finance. Incremental technological developments, such as data matching and electronic lodgement (myTax) in Australia, have reshaped compliance enforcement and prompted renewed attention to the drivers of taxpayer non-compliance. This thesis develops and tests the Taxpayer Propensity to Offend (TPtO) model to explain intentional tax non-compliance in the digital era, comparing behaviours between those lodging via myTax and those lodging via tax agents. </p><p dir="ltr">The thesis employed a sequential explanatory mixed-methods design across three studies. Study One analysed longitudinal ATO sample-file data (2003-04 to 2019-20) to compare demographic and deduction behaviours across lodgement methods, confirming systematic differences between myTax and tax agent users. Study Two introduced a vignette-based survey (n = 310) to test 14 hypotheses of the TPtO model using Structural Equation Modelling. The findings revealed that perceived severity is the most consistent deterrent, while other drivers – fear of apprehension and perceived fairness – differ significantly by lodgement method. The model also required refinement, as tax morality was highly correlated with fairness and lacked discriminant validity. Study Three conducted 20 in-depth interviews, analysed using grounded theory, which validated and enriched the survey results. Participants viewed compliance as shaped not only by deterrence but also by perceptions of fairness and trust and confirmed that tax morality is embedded within fairness rather than a standalone driver. The refined TPtO model therefore integrates six constructs – perceived severity, perceived certainty, perceived control, fear of apprehension, perceived fairness, and propensity to offend – while recognising lodgement method as a moderating variable. Theoretically, this advances compliance research by bridging cognitive, emotional, and persuasive dimensions within a digitalised context. Methodologically, it demonstrates the utility of vignette-based surveys complemented by qualitative validation. Practically, the findings provide evidence for policymakers and the ATO on tailoring compliance strategies: technological deterrence is most effective for myTax users, while fairness-based messaging resonates more with tax agent clients. </p><p dir="ltr">Limitations include the focus on individual Australian taxpayers (excluding entities), the Australia-specific context which may limit international generalisability, and the online survey design which may under-represent less technologically literate taxpayers. Nonetheless, the results offer valuable insights for tax authorities in developed countries on tailoring strategies to diverse lodgement methods to mitigate non-compliance. Future research could extend the TPtO model to businesses and other jurisdictions, and track taxpayers’ behaviour over time as digital enforcement matures.</p>