Politics does not simply use mass communication. It is increasingly organised through it. Television, YouTube, mobile devices, social media, recommendation systems, advertising, analytics, political institutions and audiences form a coupled environment in which responses to communication become inputs into what happens next. A story attracts attention and is repeated; repetition alters familiarity and expectation; expectation affects response; responses become views, shares, comments, subscriptions, audience figures and votes. Those responses then inform what broadcasters, platforms and political organisations produce next. No participant needs to control or understand the whole for coherent political patterns to emerge from it.
Gullibility is a poor word for one of the most important capacities human beings possess. We are influenceable because we have to be. Language, culture, education, trust and collective organisation depend upon allowing information from other people to alter what we believe, expect and do. No individual can independently verify more than a microscopic fraction of the world upon which everyday judgement depends. Repetition alone can increase the perceived truth of a proposition, including where people possess relevant prior knowledge (Fazio et al., 2015). Propaganda does not exploit some exotic defect found in stupid people. It operates through capacities without which culture itself could not exist.
Sky News Australia provides a useful case through which to examine this larger machinery, but it is a specimen rather than the boundary of the argument. Similar dynamics can operate across left-wing, centrist, state, corporate, activist and platform media wherever political identity, institutional incentives and recurrent communication begin selecting one another. Different factions privilege different enemies, virtues and explanations while using similar mechanisms of compression, repetition, affiliation and antagonism. The useful questions are structural: what forms of communication recur, how do they circulate, what attracts further attention, and what political possibilities become easier to reproduce as those patterns persist? Australia’s communications regulator distinguishes source, content and exposure diversity for precisely the reason that counting outlets alone does not tell us how information and influence are organised (ACMA, 2026).
This also removes the comforting assumption that successful participants necessarily understand the larger systems they reproduce. Political actors, presenters, executives, campaigners, platform designers and audiences can follow locally successful strategies while remaining largely unaware of their aggregate consequences. Money, reach, status and political power demonstrate success within an immediate environment; they do not demonstrate that the larger conditions supporting that success are being sustained. A media organisation can increase engagement while contributing to declining trust. A political movement can consolidate identity while weakening institutional legitimacy. A platform can increase extraction while degrading aspects of the informational environment upon which it depends. The parts can succeed while the whole deteriorates.
Copland, Bruns and Graham (2025) provide unusually direct evidence through which to examine the Australian case. They analysed 20,000 Sky News Australia YouTube videos from March 2020 to January 2021 and traced their dissemination through publicly accessible Facebook spaces, describing the channel’s development into an international component of what they call “network propaganda”. By June 2024, they report, the channel had 4.14 million YouTube subscribers and 4.5 billion views. The important implication is structural rather than partisan. Distribution no longer ends when a broadcaster transmits a message. Audiences share it, supporters endorse it, opponents condemn it, platforms register those responses, aligned actors recirculate it and producers observe what travelled. Communication returns carrying information about communication. The audience becomes part of the machinery selecting what the audience is likely to encounter next.
This is where “the people” enter the machine. Populist communication does not merely discover an already coherent public and give it a microphone. It can participate in producing the identity it claims to represent. Research on populist securitisation describes recurrent constructions of a threatened people and dangerous elites, while comparative experiments indicate that anti-elite identity framing can affect persuasion and mobilisation (Wojczewski, 2020; Bos et al., 2020). None of this requires invented grievances. Housing stress, insecure employment, declining trust, cultural displacement and economic pressure can all be real. Their political meaning remains open. Communication can organise them relationally: this happened because of them; something is being taken from people like us; institutions do not understand ordinary life. Insecurity becomes identity, identity organises attention, attention produces measurable engagement, and engagement tells the communication system which distinctions remain productive.
The statistical character of this process matters. Culture is not a collection of independent messages stored in millions of separate heads. It is a distributed pattern of recurrence in which some descriptions, identities and responses become easier to reproduce than others. People do not need to believe exactly the same thing, encounter the same message at the same moment or respond in the same way. The herd is not a mass of identical minds. It is the larger statistical pattern produced as different minds repeatedly adjust to signals, institutions and one another.
Those adjustments also occur across different timescales. Daily controversies, electoral cycles, longstanding narratives, institutional habits and technological feedback operate at different frequencies. They can reinforce, interrupt or redirect one another. Certain subjects return, certain opponents recur and certain emotional responses become familiar. Complete synchronisation is unnecessary. Partial synchronisation is enough: people remain different while some distributions of attention, expectation and response become aligned. Complex-systems research provides a formal basis for understanding how large-scale coherence can emerge from local interactions among heterogeneous participants without central instruction (Arenas et al., 2008; Dörfler and Bullo, 2014; Rodrigues et al., 2016).
Influence remains probabilistic rather than mechanical. Research cautions against assuming that exposure to online misinformation automatically produces large behavioural effects across whole populations (Budak et al., 2024). This qualification strengthens rather than weakens the systems argument. A communication system does not need to control everyone. It needs only to alter the probabilities from which subsequent attention, interpretation and action emerge. The next communication then enters an environment already altered by what came before.
Antagonism can become part of this recursive machinery. Anger attracts attention; attention produces circulation; circulation provides information about what attracts attention. Opposition does not necessarily interrupt the process. People who despise a commentator can watch, quote, condemn and recirculate that commentator. Hostile political groups can become increasingly coupled because each provides the other with evidence of threat and material for another cycle. Agreement is unnecessary because coordination can occur through opposition as readily as assent. The enemy changes while the structure of enmity persists.
This matters politically because autocratic, plutocratic and kleptocratic forms of power become easier to reproduce under particular informational conditions. Autocratic tendencies benefit when institutional constraints are experienced primarily as illegitimate obstruction and opponents as enemies. Plutocratic power becomes less visible when recurrent cultural conflict overwhelms attention to concentrated economic power. Kleptocratic behaviour becomes harder to constrain when political identity displaces consistent standards of accountability. This is a structural argument about informational conditions, not an allegation that any particular media organisation or participant consciously intends such outcomes.
A functioning propaganda ecology consequently requires neither conspiracy nor central command. Platforms can seek engagement, commercial organisations audiences, political organisations support, commentators successful formats and audiences explanation, entertainment or belonging. Each action can make sense locally while their interaction produces something nobody intended. The deeper problem is recursive selection: communication alters the conditions under which subsequent communication occurs. Repeated across populations and timescales, those altered conditions can become cultural structure. Culture then affects which political identities, explanations and performances are easiest to reproduce.
This is what I mean by the Herd Reich. The phrase deliberately corrupts the contemporary neologism “Nerd Reich”, shifting attention from technological elites alone towards the statistical organisation of populations through recurrent communication. It is not a claim that audiences constitute a totalitarian mass, nor that a particular media organisation is equivalent to a historical fascist regime. The point is that technological power, political ambition, concentrated wealth, institutional incentives and mass behaviour can become coupled without any one element controlling the whole. Populations can become statistically organised before they become politically organised.
The Australian example matters because it makes a much larger transformation visible. Treating one outlet as an isolated villain would reproduce the same compression being criticised. Contemporary communication systems can measure reaction, identify material associated with that reaction and feed related material back into culture at extraordinary speed. What is selected need not inherently be true or false, left or right. It needs to recur. The danger appears when recurrence makes collective interpretation progressively harder: distrust without discrimination, grievance without causal depth, identity without permeability, certainty without correction, enemies without end.
A society need not collapse for this to matter. It continually reconstructs itself from the altered probabilities left by previous cycles. Eventually the important statistic is not how many people believed any particular story. It is what kinds of people, institutions and political futures the communicative environment has made progressively easier to reproduce.
References
Arenas, A., Díaz-Guilera, A., Kurths, J., Moreno, Y. and Zhou, C. (2008) ‘Synchronization in complex networks’, Physics Reports, 469(3), pp. 93–153. doi:10.1016/j.physrep.2008.09.002.
Australian Communications and Media Authority (ACMA) (2026) Media Diversity Measurement Framework. Canberra: ACMA.
Bos, L., Schemer, C., Corbu, N., Hameleers, M., Andreadis, I., Schulz, A., Schmuck, D., Reinemann, C. and Fawzi, N. (2020) ‘The effects of populism as a social identity frame on persuasion and mobilisation: Evidence from a 15-country experiment’, European Journal of Political Research, 59(1), pp. 3–24. doi:10.1111/1475-6765.12334.
Budak, C., Nyhan, B., Rothschild, D.M., Thorson, E. and Watts, D.J. (2024) ‘Misunderstanding the harms of online misinformation’, Nature, 630, pp. 45–53. doi:10.1038/s41586-024-07417-w.
Copland, S., Bruns, A. and Graham, T. (2025) ‘Sky News Australia as network propaganda: how a niche cable channel became an international right-wing propaganda machine’, Media International Australia, published online 24 December. doi:10.1177/1329878X251406899.
Dörfler, F. and Bullo, F. (2014) ‘Synchronization in complex networks of phase oscillators: A survey’, Automatica, 50(6), pp. 1539–1564. doi:10.1016/j.automatica.2014.04.012.
Fazio, L.K., Brashier, N.M., Payne, B.K. and Marsh, E.J. (2015) ‘Knowledge does not protect against illusory truth’, Journal of Experimental Psychology: General, 144(5), pp. 993–1002. doi:10.1037/xge0000098.
Rodrigues, F.A., Peron, T.K.D.M., Ji, P. and Kurths, J. (2016) ‘The Kuramoto model in complex networks’, Physics Reports, 610, pp. 1–98. doi:10.1016/j.physrep.2015.10.008.
Wojczewski, T. (2020) ‘Enemies of the people: Populism and the politics of (in)security’, European Journal of International Security, 5(1), pp. 5–24. doi:10.1017/eis.2019.23.
