Unveiling global narratives of restoration policy: Big data insights into competing framings and implications

Abstract

Restoration has become a key environmental policy goal of the contemporary era. Yet, what restoration means and how it is pursued remains an object of debate. This study examines the nature of restoration discourses on Twitter - a large, open, and global record of public discussions around contemporary restoration matters. We apply machine learning-powered text analysis of about 350,000 geolocated tweets spanning 2015-2022, focusing on four main restoration terms - landscape restoration; forest and landscape restoration; ecological restoration; and ecosystem restoration. Findings reveal a wide diversity of environmental policies framed through the language of restoration, underscoring its public appeal and use by different institutions from global to national and subnational scales. Restoration discourses foster both ecological and human-centered framings, with the former being more prominent. Other distinct discourses convey promotional efforts, momentum building, political engagement by proponent actors, and what restoration should deliver. Only a few discourses feature quick fixes such as tree planting, potentially implying that contemporary restoration interventions are more diverse than headline-grabbing targets to plant trees. There is little discussion of rural livelihoods, tenure rights, or tradeoffs between environmental objectives and local needs. Although the discourses vary across the restoration terms, we find some shared discourses as well as unique ones. We underscore how restoration discourses carry different worldviews with implications for the purported socio-ecological benefits of restoration. Our work shows how data-driven analysis of social media can shed light on the rhetoric of restoration policy agendas and their nuances among a broad spectrum of social and policy actors.

Keywords

Discourse analysis; Environmental values; Global environmental agendas; Topic modelling; Machine Learning; Twitter

Published in

Geoforum
2025, volume: 161, article number: 104241
Publisher: PERGAMON-ELSEVIER SCIENCE LTD

SLU Authors

• Fischer, Harry

  • Department of Urban and Rural Development, Swedish University of Agricultural Sciences
    

UKÄ Subject classification

Environmental Studies in Social Sciences

Publication identifier

• DOI: https://doi.org/10.1016/j.geoforum.2025.104241

Permanent link to this page (URI)

https://res.slu.se/id/publ/141226