iiinova — where your digital self keeps evolving

We also find an interesting mechanism: the effect of policy loss is stronger for people who are more highly educated (who are more likely to have more information about what the EU does) 9/

110

This effect is robust across our 6 measures of EU policy outputs as well as with 2 different ways of conceptualising “policy loss”: immediate (the distance to policies in the same year), or cumulative (distance to all policies prior to a given year). 8/

100

At the individual level, pooling <1.1 million respondents in the Eurobarometer surveys, the further EU policies are from an individual’s preferences, the less they like it 7/

100

At the aggregate level, the EU was most popular when it delivered policies that were more centrist (so minimised the "policy loss" for most EU citizens) and became less popular when average policy loss was higher 6/

110

Across all of our measures of EU policies, what the EU has delivered has moved “leftwards” over time, although some measures suggest a move back towards the centre in recent years 5/

200

To test this explanation we measure EU citizens (from almost 50 years of Eurobarometer surveys) and annual EU policy outputs (using 6 different methods) on the same scale 4/

110

What we do in the paper is try to explain why the relationship between political preferences and support for the EU has changed over time, e.g. why people on the right supported European integration most in the 1970s, and why the left support the EU most today. 2/

Simon Hix @simonhix.bsky.social · 05/01/2026

NEW PAPER ALERT! My paper, with @bjornhoyland.bsky.social , on how “policy loss” shapes support for the EU has just been published in @cpsjournal.bsky.social. Check it out! 1/

111936

She's at it again! Another outstanding Substack by @laurenleek.eu, this time looking under the hood of AirBnB listings in London. Amazing analysis, with important policy implications, e.g. this quote open.substack.com/pub/laurenle...

0186

Simon Hix @simonhix.bsky.social · 16/12/2025

Excellent piece on the decline of the “Brussels effect”, by @pmdfoster.bsky.social and @bmoens.bsky.social. This figure, in particular, tells a lot of the story giftarticle.ft.com/giftarticle/...

166

Simon Hix @simonhix.bsky.social · 11/12/2025

Yep. And it looks like our prediction of the trajectory is bang on, eg.

020

Simon Hix @simonhix.bsky.social · 31/10/2025

Scenario 2: assume that some people who vote for parties who wouldn't win seats would switch to other parties (there are lots of ways to do this, and I picked a very simple one) => two three party majority coalitions would be possible: D66-CDA-VVD or D66-CDA-GLPvdA 3/

230

Simon Hix @simonhix.bsky.social · 31/10/2025

Scenario 1: take the actual election result and only allocate seats to the 7 parties who more than 4% or more of the votes => a three party coalition would be possible, between D66, CDA, and VVD 2/

Simon Hix @simonhix.bsky.social · 30/10/2025

Fascinating vote-switching patterns from NL2025 elections, e.g. app.nos.nl/nieuws/tk2025/

171

Simon Hix @simonhix.bsky.social · 29/10/2025

Janan Ganesh puts it well: on.ft.com/48TsW78 Europe and the curse of geography Eg.

2257

Simon Hix @simonhix.bsky.social · 17/10/2025

Caamp in Milan. First date on their European tour. Amazing. Thank you guys.

030

Simon Hix @simonhix.bsky.social · 17/10/2025

What better to way to spend a Friday PM than calculating ENPV in NL? => 1. Dutch voters have no problem coordinating in EP elections, where the effective threshold is 3.1% (100 / 31+1); 2. The party system would still be fragmented in the TK with a 4% threshold, but with fewer parties overall.

4183

Simon Hix @simonhix.bsky.social · 17/10/2025

Marina Hyde on scintillating form about the Royals:

01812

Simon Hix @simonhix.bsky.social · 04/10/2025

Fantastic piece by Simon Schama on the debate between the US Founding Fathers about freedom of speech (and the Alien and Sedition Acts), that includes this wonderful final exhortation to us all: on.ft.com/48b6rKt

2116

Simon Hix @simonhix.bsky.social · 25/06/2025

There is hope in the world 😀

0221

Simon Hix @simonhix.bsky.social · 13/06/2025

Yes, turnout mattered. But, the right coordinated to get their side to abstain if they disagreed with EITHER set of issues. So we don't know whether some might have supported Q1-4 and others Q5. Nb. turnout was not particularly lower in the South or higher in the North, as Pau has pointed out.

000

Simon Hix @simonhix.bsky.social · 13/06/2025

Some places (e.g. Bolzano) was in the right-libertarian corner, others (e.g. Naples) were left-libertarian, while Florence/Bologna/Rome were more libertarian than left, and many smaller rural places where more authoritarian than left. 5/

240

Simon Hix @simonhix.bsky.social · 13/06/2025

Meloni (FdI) and the other governing parties (FI, Lega) told their supporters to abstain. So, the right won and the left lost. But, the %Yes across the provinces suggests a more nuanced story: some were more supportive of labour rights while others were more supportive of citizenship rights. 3/

230

Simon Hix @simonhix.bsky.social · 09/06/2025

And of those who voted in Firenze, almost 75% voted Si! Grazie fiorentini! ❤️ #ForzaViola

062

Simon Hix @simonhix.bsky.social · 04/06/2025

Interesting. And here is the same plot with "Radical Right" added, i.e. Radical Right and Far Right are far more common terms than "Hard Right"

010

Simon Hix @simonhix.bsky.social · 21/05/2025

A fabulous insight in Adam Przeworski’s latest substack. Would love to see if this could be tested somehow.

95810

Then, we look at "European wide swings": whether some party families do better/worse in EP elections (controlling for their size and govt status). We find that greens, anti-Europeans and centre-right parties (!) outperform a second-order prediction, while socialists generally underperform 4/

That said, in line with the classic model, governing parties performed worse than opposition parties in about half of the member states in the 2024 elections: 3/

We look at the classic "second-order national election" model of party performance in the 2024 EP elections compared to all previous EP elections, e.g. this figure 2/

140