Exploring corporations’ use of Twitter for direct lobbying

The use of social media for lobbying draws increasing interest. Most studies have focused on the role of social media in outside lobbying (e.g., Kanol & Nat 2017; Vesa et al. 2022) or on interest organizations’ use of social media in general (e.g., Scaramuzzino & Scaramuzzino 2017; Van der Graaf et al. 2016). Less is known about the use of social media for direct lobbying (but see Chalmers & Shotton 2016; Halpin et al. 2021; Widner et al. 2020), where for example Twitter enables a type of “middle-stage lobbying”, in which lobbyists target policymakers directly in front of a limited audience (Figenschou & Fredheim 2020). Another research gap is that studies on the role of social media in lobbying have primarily focused on interest groups, such as NGOs, labor unions, and business associations (but see Van der Graaf et al. 2016). Less is known about the role of social media in the lobby efforts of firms. This is surprising as research has highlighted the important role of individual firms as lobbyist in pluralist systems (e.g., Gray et al. 2004) as well as in corporatist systems (Aizenberg & Hanegraaff 2020). To fill these two gaps, this paper asks: To what extent do corporate representatives use social media for direct lobbying and what factors explain variation in their activity? Our analysis draws on a sample of large corporations operating in Finland, Denmark and the Netherlands. The analysis examines how in-house lobbyists and executive-level employees of large corporations use Twitter for direct lobbying and what factors can account for variation in this. Specifically, we use the Twitter API to identify in-house lobbyists and executives that follow political accounts (those of parliamentarians, ministries, and political parties). To identify direct lobbying on Twitter, we analyze corporate employees’ interactions with political accounts during a selected period: retweets, quote retweets, tweets with mentions, replies, and likes. To distinguish between pure networking and substantial lobbying related to political issues we utilize computational methods (such as naïve Bayes classifier). Our goal is to explain variation in the extent of corporate representatives’ interaction with political accounts – i.e., direct Twitter lobbying – by factors such as revenue, staff size, type of ownership, type of industry and country level factors. In the presentation, we will also discuss some challenges of studying direct Twitter lobbying, such as identifying organizations’ employees in the absence of historical Twitter bios and the challenges in classifying different types of interactions with policymakers. Co-author without ECPR account: Mika Vehka (University of Helsinki).