This study addresses a gap in local government research by advocating for the adoption of computerised content analysis methods to analyse public service accounting texts. Despite the significant volume of textual data produced by local governments – such as financial statements, budget reports, and policy documents – traditional manual content analysis methods have limited researchers’ ability to harness these resources effectively. By employing tools such as the Linguistic Inquiry and Word Count (LIWC-22) software and custom Python code, this study highlights the potential of computerised approaches to enhance the depth, efficiency, and rigour of content analysis in the public sector. Using annual budget reports from eight Irish local governments as empirical examples, the study applies both standardised and customised analytical techniques to assess textual complexity, readability, sentiment and thematic trends. The findings highlight notable differences in text characteristics between urban and rural local governments and reveal temporal shifts influenced by broader policy initiatives, such as climate action planning. Additionally, the study demonstrates the advantages of combining off-the-shelf tools with tailored programming to achieve more comprehensive insights, particularly in capturing nuanced linguistic patterns. This research illuminates the value of computerised content analysis in public sector accounting research. While focusing on budgetary documents, the methods described have broader applicability to other public sector texts, paving the way for innovative research approaches in public governance and administration.