AbstractsPolitical Science

Measuring media coverage of electricity public policy in New Zealand for the period 2006-2007: a new approach

by Jeanette Elley

Institution: AUT University
Year: 0
Keywords: Content analysis; Media studies; Electricity policy; Text analysis; Renewable energy; Media influence; Muliaga; Climate change; Peak oil; Fossil fuels; Political economy; Critical analysis; New Zealand Herald; New Zealand electricity; Natural language processing; Computational anaphora
Record ID: 1305282
Full text PDF: http://hdl.handle.net/10292/7409


This study measured media coverage relating to electricity public policy using an original content analysis technique, simultaneously responding to calls for more research on the media-policy nexus, and for more precise content analysis methods. The project involved the collection of more than 6000 electricity articles online from nzherald for the period 2006-2007 (about 5% of all articles). Articles were systematically analysed for keywords and a categorisation framework was iteratively compiled, in order to facilitate automated content analysis. The goal was to identify trends relating to electricity policy, and broader trends relating to the underlying paradigm within the media, by analysing phrase repetition frequencies. A unique prototype text analysis tool was developed, incorporating customised key indicators which distinguished between explicit (actual), literal (bonus), and implicit (hidden) coverage. Utilising aspects of natural language processing (NLP) the automated tool also handled exceptions and nick names, enabling the recognition of extended “coverage” of a search target over multiple sentences, even in the absence of proper names. Issues such as climate change, peak oil, fossil fuels vs. renewables, carbon tax vs. emission trading, sustainability vs. growth, SOE privatisation, and the Muliaga disconnection were examined, key players identified, and ranked lists compiled to indicate relative coverage in each case. Phrase recognition using an unvalidated prototype tool encountered certain limitations of performance and interpretation but nonetheless allowed an extent and precision unavailable in similar manually coded studies. Major findings were the dominance of the financial sector (especially the stockmarket), governments, and fossil fuels. Climate change received significantly increased coverage in the second year of the study, as did the concept of sustainability. However, biofuels and energy efficiency received limited coverage. Furthermore, the peak oil issue, alternative energy sources (e.g. marine and solar power), and distributed generation received very little coverage at all, reflecting a missed opportunity on the part of the media to contribute to a more resilient society in the face of looming environmental threats.