Framing Privatization: A Comparative Analysis of Media and Institutional Corpora on the DISCOs in Pakistan

This paper examines how the privatization of Pakistan’s electricity distribution companies (DISCOs) is framed in institutional and media corpus by employing textual analysis using natural language processing (NLP) techniques guided by Framing Theory as the theoretical lens. Two corpora were constructed: one comprising NEPRA’s documents; and the other consisting of English newspaper articles or reports. Textual analysis involved sentiment analysis, Bigram Analysis and structural topic modeling to identify recurring themes and emotional tones. The findings reveal that the sentiments were positive across all corpora, while it was slightly negative in the DISCOs evaluation reports due to poor performance of companies. Moreover, NEPRA’s state of industry reports have, over time, seen an evident shift towards financial efficiency, DISCOs and capacity issues while Performance Evaluation Report of Distribution Companies reveal a reduced focus on safety and service standards. In contrast, the media corpus was pre-filtered using keywords related to the privatization of DISCOs, and the most prevalent topics focused on DISCOs reform. In addition, the bigram analysis of the media corpus shows a crisis-driven discourse, while the NEPRA corpus is characterized by a technical discourse.