The buried lede here is that someone had to use Facebook to make an official statement about elections.

In a Facebook address Tuesday watched by tens of thousands of people, Venancio Mondlane again demanded a recount of the October 9 vote which the electoral commission said was won by the Frelimo party in power for almost half a century.

“We lost 50 people shot by the authorities who were supposed to protect these people,” said Mondlane, referring to a police crackdown on waves of protests he called against the election.

The conflict stems from Mozambique’s post-colonial history, one of my longest and most focused research areas, as I wrote on this blog in 2006:

It begs the question what Mozambique would have looked like if someone hadn’t assassinated Mondlane (February 3, 1969). Killing a powerful liberal-but-left American university professor of history, a respected leader within FRELIMO, ended his moderating influence over a freedom movement. FRELIMO was operating more peacefully under Mondlane as he and immediate colleagues left out rigid dogma or hierarchy; they openly invited interplay of conflicting views and positions. His assassination by the US regressed freedom and propelled turmoil.

Basically FRELIMO (Mozambique Liberation Front) ruled since independence from Portugal’s fascist dictatorship in 1975. The party transformed from the fight for liberation into (arguably due to American assassination and subterfuge) political dominance, maintaining power through a combination of legitimate support and contested electoral processes.

Several key aspects stand out in terms of today’s news:

• The use of Facebook for official opposition communications reflects both the weakness of traditional media access for opposition voices and the growing importance of social media in African political discourse. This echoes patterns seen in other African nations where social media becomes a crucial platform for opposition voices when traditional media is state-controlled.
• The allegations of 50 deaths in election-related violence, if verified, would represent one of the more serious instances of electoral violence in recent Mozambican history. However, election violence has been a recurring issue in Mozambique, particularly during local elections.
• The demand for vote recounts is a common opposition strategy in contested African elections, seen previously in Kenya, Zimbabwe, and other nations where ruling parties maintain long-term control.

How much of this violence is attributable to Facebook owning the platform in Mozambique for public discourse? I’m not saying Facebook necessarily was causal in the violence, rather that it’s more a symptom of broader issues in Mozambican democracy. The platform’s unnatural high-exit barriers and undemocratic privatization of infrastructure reflects the lack of alternative spaces for political opposition to communicate with supporters. And also that Facebook may be causal.

When Facebook becomes the de facto platform for political communication, it intentionally creates a dangerous anti-democratic dependency where a private foreign company effectively controls access to political discourse. Facebook’s algorithms and content moderation policies tend to amplify political tensions and shape how opposition movements organize and communicate. This isn’t unique to Mozambique, given we saw similar dynamics in Myanmar, Ethiopia and… the United States.

Facebook causing violence, like Tesla causing chemical spills, isn’t the core issue here as bad as it may seem, however. Facebook’s dominance was a symptom of institutional weaknesses that has setup a transition into becoming the cause (perhaps similar to how FRELIMO went from liberation to domination). A foreign-state service monopoly mindset of American businessmen likely exacerbates this. The key issue is in fact the privatization of what should be public democratic infrastructure.

Just as FRELIMO’s transition from liberation movement to ruling party was enabled by control of state resources, Facebook’s transition from communication platform to political infrastructure was enabled by network effects and data monopolies.

The critical difference lies in accountability. While FRELIMO must at least maintain some veneer of democratic legitimacy in a government role within Mozambique, Facebook faces no such local (or even international) constraints. This creates an unprecedented situation where crucial democratic infrastructure is controlled by an entity with no democratic accountability to the population it serves.

The Myanmar and Ethiopia parallels also obscure how Mozambique represents something distinct and more like the United States. We are witnessing a case where Facebooks’s role in political communication was normalized before its potential for amplifying violence was fully understood. This makes it an important case study in how privatized democratic infrastructure becomes dangerously entrenched even in the absence of acute crises.

Arguably the “soft” path to platform dependency might actually be more dangerous than the more visible crises in Myanmar or Ethiopia.