Political organisations across Europe talk constantly about participation, yet many of the people they consult come away feeling nothing changed. Online surveys disappear into institutional black holes, town‑hall events are staged for cameras rather than decisions, and party “platforms” function as one‑way broadcast tools. The result is predictable: frustration, disengagement and, eventually, hostility towards the very idea of political consultation. The real problem is not the lack of digital tools but the absence of purposeful design: processes that show participants where their input goes, how it is weighed, and when it shapes outcomes.

Sweden goes to the polls on 13 September 2026, electing all 349 Riksdag seats, 20 regional councils, and 290 municipal councils in a single day. The race is tight. The Social Democrats lead on 32.4% in the most recent polling trend, but the governing Tidö bloc is still in play and the Liberals are hovering dangerously close to the 4% threshold that would eject them from parliament entirely. Every point matters. Every voter counts. We are watching this one closely at Partisan, not because Sweden is the most dramatic election of the cycle, but because it is one of the most interesting laboratories. Swedish political culture combines very high digital literacy with genuine civic seriousness, a media landscape that is trying harder than most to adapt to AI, and a party system structured around eight parties that all need to find their audience simultaneously. The conditions are ripe for experimentation.

In recent election cycles, the traditional political media apparatus have lost its monopoly over political communication. Especially when it comes to how people aged 40-50 and below engage with politicians and opinion leaders. The 24-hour cable news cycle, once the premier battleground for shaping public discourse, has increasingly yielded ground to other, more direct channels of reaching voters. These include personal channels of politicians simultaneously operating on multiple social media platforms, as well as the consistent rise of influencers of all kinds. And podcasts, both in ‘pure audio’ as well as video podcasts, have gained a significant share in the shaping of our social and political discourse, and therefore in how candidates, grassroots organizers, and advocacy campaigns operate and convey messages.

America does not do things quietly. With control of both the House and the Senate on the line, the 2026 midterm cycle is on course to be the most expensive in US history, with political ad spend projected to hit $10.8 billion, up from $8.9 billion in 2022. That is not a typo. Ten point eight billion dollars. For the midterms. But money alone does not tell the story. What is genuinely interesting about the 2026 primary season is not the scale of the spending but the techniques being deployed, many of them for the first time at scale, many of them available to campaigns with budgets measured in thousands rather than millions. The primaries are always the laboratory. What gets tested in Texas and North Carolina in February tends to end up in European campaign playbooks two cycles later. Here is what is actually worth paying attention to.

A year ago, almost nobody outside Colombian legal circles had heard of Abelardo de la Espriella. He was a flamboyant defence lawyer from Barranquilla known for representing some of the country's most controversial clients, for owning clothing and alcohol brands and a Miami restaurant, and for having once recorded music albums. Not exactly the typical presidential profile.

Democratic politics is losing not because its arguments are weaker, but because its attention strategy is irrational. Platforms price political engagement the way they price sneaker ads: optimise for the cheapest clicks from the most likely buyers, not for the hardest minds to reach. That logic pushes civil society and progressive politics into chasing cheap likes inside their own ideological niches, while their opponents pay the real cost of reaching the mainstream. The crisis is therefore less about “losing the debate” than losing a maths problem about where influence actually happens.