Copy, paste, and obfuscate? Readability and standardisation in integrated reporting

Purpose This study examines how readability and textual similarity evolve and interact in integrated reports, investigating whether narrative accessibility and textual standardisation are distinct characteristics or interconnected features of institutionalised reporting practices. Design/methodology/approach A longitudinal panel of 1,393 firm-year observations from South African listed companies over 2015–2024 is analysed. Readability is measured using Flesch Reading Ease, and textual similarity using cosine similarity between consecutive reports. Fixed-effects panel regressions are supplemented by governance-excluded similarity measures, first-difference models, persistence analysis and an alternative readability measure. Findings Readability declines over time, while textual similarity remains high, with no consistent upward trend. Textual similarity is positively associated with readability, indicating that narrative continuity can coexist with greater accessibility. However, this association becomes insignificant when governance-related disclosures are excluded, suggesting that it is concentrated within a highly institutionalised disclosure domain characterised by recurring reporting requirements and stable narrative structures. Additional analyses confirm the persistence of textual similarity and robustness of the main findings. Originality/value This study extends research that has largely examined readability and textual similarity separately by demonstrating that these characteristics are interrelated and that their association depends on the disclosure domain. Greater readability can coexist with substantial textual standardisation, complicating the interpretation of this as an unambiguous indicator of disclosure quality or transparency. The findings highlight the importance of distinguishing narrative accessibility, textual standardisation and substantive informativeness when evaluating corporate disclosure.