Why Italian politics is "a bit of a mess"
October 17, 2026
Scroll sideways to see the whole chart →
Source: Only Numbers poll by A. Ghisleri, 14–15 October 2026, presented on Realpolitik. Party shares are of voters who expressed an intention; undecided/abstaining respondents shown separately.
Lately I’ve found myself trying to explain, more than once, why I describe the Italian political landscape as “a bit of a mess”. Truth is, that’s an understatement: more than half of Italians who express a voting intention plan to vote for parties with horrible positions.
So, I asked Claude to make me the chart above (sorry if it’s not really mobile-friendly).
You can say it’s my opinion, but my opinion is that if you disagree that dropping support for Ukraine is playing Putin’s game, you are wrong and, worse, not very bright.
The bar shows the latest voting intentions, party by party, arranged from left to right and grouped by coalition; hover to learn how bad things are. 😭
The annotations around it make my point: on the left there are parties with multiple antisemitic stances (yes, really, I’m not saying Pro-Palestine though they like to mask it that way: they hold positions that are actively against Jewish people), on the right parties with multiple racist stances against Muslims (starting from the pretty common around the world “nationals first” position of the current leading party to the really unhinged positions from this moronic retired army general that somehow gained traction outside the current government coalition), and scattered across the whole spectrum an undue Russian influence.
I realize there are probably other horrible (or just plain stupid) positions I haven’t mapped here, and that such a gathering of idiots most likely expresses racism toward other minorities too; but these are the main positions that push more than half the voters in two to shout “Mine! Mine!”.
A disclaimer: the numbers come from a single poll, and polls are polls. But it changes very little if a party is overestimated by one point or underestimated by two; the shape of the problem stays exactly the same.