1. See here for what I think the strongest anti-accommodation position is, though I don’t necessarily subscribe to it. The key point is that, as voters discount marginal shifts by parties, the policy space becomes “lumpy”, requiring parties to move quite far for voters to take a shift seriously. This, however, exacerbates the risk of losing previously aligned voters and irking activists. Put differently: the rise of the radical right and the fact that moderate accommodation isn’t all that effective probably have similar causes.
2. If this is analysis is correct, then we need to understand better what makes people trust parties’ programmatic shifts, eg descriptive representation, group appeals, etc. What are costly signals that they are genuine about this?
3. Finally, there should be more focus on containment, which poses many challenges — as illustrated by debates about the firewall in Germany.
Jacob this post is already partially based on your threads of course ;) and wrt #2, I totally agree. I asked Martin Larsen and he said they are working on something in that direction (experimental), nothing public yet though.
Very cool post! Have been following this/analysing publicly available data.
Two points not raised here about the "Representation Gap" thesis:
(1) Time. The first thing almost anyone thinks in looking at British Election Study data from 2014-2025 is "My goodness an Auth-Left quadrant party would dominate!" (to the point where "Hang the paedos, fund the NHS" is now a meme.
But the *second thing* that should occur to anyone is "Wait - this should have been *even more* the case for the last 70 years. And parties in this space got .1% of the vote max until 1997 (and even then initially by peeling people off from the centre-right in the Auth-Right quadrant)."
All "Representation Gap" arguments struggle with the "what about the last 70 years when the population was - by these measures - waaaaay more Ec Left and Soc Auth? Yes, parties have moved *a bit* - but publics have moved *much, much more* over the period."
Obviously the answer is implicit in the specificity of "70 years" - frustrating as it is to spoil a clean model, there's no getting around the fact that WWII had a huge impact on electoral demand (as well as elite supply) in this quadrant
(2) Space. Voters do not - in general - process politics/parties one issue at a time - certainly not when it's 'nationalised'. Voters appear (looking at distribution of like<party> across political compass space) to perceive parties as having an angular position somewhere around the political compass (hence why every party is not the Central Zentrum Partei).
Parties can rotate around the political compass in a manner that will 'accommodate' the value preferences of voters on one flank (while disappointing those on another). Whether that's strategically beneficial (just in terms of gaining votes *on that flank*, before even thinking about other flanks/voting systems!) is actually quite context dependent (where are other parties/are you perceived as more/less competent than them at the moment).
The point is that your degrees of freedom for accommodation are a bit limited - Danish Social Democrats can roll a bit anti-clockwise (more by emphasising being further to the Economic Left than Soc Auth, but implicitly reducing the distance).
But what you cannnot do is take a bishop's move across the political compass or accommodate* on one just one issue. Like stalling a car the result isn't that you move where you intended to go, but instead everyone downgrades their perception of your competence (e.g. 2024->2025 UK Labour).
* You could do it because you *thought it was effective governing policy* - but not to shift perceived party position on the political compass.
Other issue with "Representation Gap" thesis is that it responds to the "Paradox of Tolerance" with an implicit "you just gotta tolerate everyone and everything".
Whatever you feel about this - and it's worth doing more chewing on the concept than Popper, both in principle and practice - it's an argument that should be made explicitly.
We don't need to euphemise our way around "abstract representation issues".
After WWII, nominally (actual/reconstructing) liberal democracies hang the leaders of the radical right party family from gibbets, proscribed their parties and purged them.
No **** there was a representation gap and a smoking Nuremburg (worst game of Cluedo ever). And one that endured for at least half a century (yes, yes, a bit less in geographically very specific bits of Austria).
A study from Sweden in 2012 is obsolete. What happened was that the mainstream parties (all seven of them) in a naive and ignorant way believed the only goodhearted way was to have open borders in 2014 (a la Merkel).
Even when the systems started breaking down did they dare to suggest that maybe it was time to push the break.
We want politicians who feel like mature adults. When their method doesn't work, they need to admit it and do what is best for the country, not what will win them applauds from the liberal media.
If the politicians in Germany and the UK (and maybe US) had done that earlier, they wouldn't have seen the rise of these alternative parties.
Listen to the voters instead of playing games with vote maximizing.
People like to see everything as either black of white, everything tidily classified in boxes in their brain. The problem is, life is big grey area full of nuances. Dealing with that takes a lot of time, thought and energy. Most people have neither the time nor inclination for that.
I will not lie, I agree 100% to what you write here, but this topic still pisses me off to no end.
To me, the far right is in general is just too dangerous to be kept alive and their existence is detrimental to a liberal democracy. There are just limits to what a state should be allowed to dictate its citizens to do, and the far right need to cross those limits in order to pursue their agenda. At that point, this simply isn't a legitimate dispute about votes anymore.
In Germany there is still a debate going on about if the AfD should be banned or not, and while that topic is of course controversial, I think it should be done. I am sick of this. And if the voters are pissed about it, that's their problem, because there are plenty of options out there that aren't outright in opposition to our constitution.
What's missing is actual follow through, or at least something that can be credibly sold that way. Take Canada, where the Liberals won an election from way behind partly due to promises to listen to voters' concerns on immigration. Granted there were other factors: Trump triggering Canadians' age old anti-American reflex helped given that the Liberals are generally seen as the pro-Canada party, and so did jettisoning Trudeau Jr., who was widely associated with progressives and open border policies
Nice post — thank you!
1. See here for what I think the strongest anti-accommodation position is, though I don’t necessarily subscribe to it. The key point is that, as voters discount marginal shifts by parties, the policy space becomes “lumpy”, requiring parties to move quite far for voters to take a shift seriously. This, however, exacerbates the risk of losing previously aligned voters and irking activists. Put differently: the rise of the radical right and the fact that moderate accommodation isn’t all that effective probably have similar causes.
https://threadreaderapp.com/thread/1982804858206859283.html
2. If this is analysis is correct, then we need to understand better what makes people trust parties’ programmatic shifts, eg descriptive representation, group appeals, etc. What are costly signals that they are genuine about this?
3. Finally, there should be more focus on containment, which poses many challenges — as illustrated by debates about the firewall in Germany.
https://threadreaderapp.com/thread/1993758805000147137.html
https://threadreaderapp.com/thread/1912818177555231055.html
https://journals.sagepub.com/doi/10.1177/00104140261448429
Jacob this post is already partially based on your threads of course ;) and wrt #2, I totally agree. I asked Martin Larsen and he said they are working on something in that direction (experimental), nothing public yet though.
Very cool post! Have been following this/analysing publicly available data.
Two points not raised here about the "Representation Gap" thesis:
(1) Time. The first thing almost anyone thinks in looking at British Election Study data from 2014-2025 is "My goodness an Auth-Left quadrant party would dominate!" (to the point where "Hang the paedos, fund the NHS" is now a meme.
But the *second thing* that should occur to anyone is "Wait - this should have been *even more* the case for the last 70 years. And parties in this space got .1% of the vote max until 1997 (and even then initially by peeling people off from the centre-right in the Auth-Right quadrant)."
All "Representation Gap" arguments struggle with the "what about the last 70 years when the population was - by these measures - waaaaay more Ec Left and Soc Auth? Yes, parties have moved *a bit* - but publics have moved *much, much more* over the period."
Obviously the answer is implicit in the specificity of "70 years" - frustrating as it is to spoil a clean model, there's no getting around the fact that WWII had a huge impact on electoral demand (as well as elite supply) in this quadrant
(2) Space. Voters do not - in general - process politics/parties one issue at a time - certainly not when it's 'nationalised'. Voters appear (looking at distribution of like<party> across political compass space) to perceive parties as having an angular position somewhere around the political compass (hence why every party is not the Central Zentrum Partei).
Parties can rotate around the political compass in a manner that will 'accommodate' the value preferences of voters on one flank (while disappointing those on another). Whether that's strategically beneficial (just in terms of gaining votes *on that flank*, before even thinking about other flanks/voting systems!) is actually quite context dependent (where are other parties/are you perceived as more/less competent than them at the moment).
The point is that your degrees of freedom for accommodation are a bit limited - Danish Social Democrats can roll a bit anti-clockwise (more by emphasising being further to the Economic Left than Soc Auth, but implicitly reducing the distance).
But what you cannnot do is take a bishop's move across the political compass or accommodate* on one just one issue. Like stalling a car the result isn't that you move where you intended to go, but instead everyone downgrades their perception of your competence (e.g. 2024->2025 UK Labour).
* You could do it because you *thought it was effective governing policy* - but not to shift perceived party position on the political compass.
Other issue with "Representation Gap" thesis is that it responds to the "Paradox of Tolerance" with an implicit "you just gotta tolerate everyone and everything".
Whatever you feel about this - and it's worth doing more chewing on the concept than Popper, both in principle and practice - it's an argument that should be made explicitly.
We don't need to euphemise our way around "abstract representation issues".
After WWII, nominally (actual/reconstructing) liberal democracies hang the leaders of the radical right party family from gibbets, proscribed their parties and purged them.
No **** there was a representation gap and a smoking Nuremburg (worst game of Cluedo ever). And one that endured for at least half a century (yes, yes, a bit less in geographically very specific bits of Austria).
A study from Sweden in 2012 is obsolete. What happened was that the mainstream parties (all seven of them) in a naive and ignorant way believed the only goodhearted way was to have open borders in 2014 (a la Merkel).
Even when the systems started breaking down did they dare to suggest that maybe it was time to push the break.
We want politicians who feel like mature adults. When their method doesn't work, they need to admit it and do what is best for the country, not what will win them applauds from the liberal media.
If the politicians in Germany and the UK (and maybe US) had done that earlier, they wouldn't have seen the rise of these alternative parties.
Listen to the voters instead of playing games with vote maximizing.
People like to see everything as either black of white, everything tidily classified in boxes in their brain. The problem is, life is big grey area full of nuances. Dealing with that takes a lot of time, thought and energy. Most people have neither the time nor inclination for that.
I will not lie, I agree 100% to what you write here, but this topic still pisses me off to no end.
To me, the far right is in general is just too dangerous to be kept alive and their existence is detrimental to a liberal democracy. There are just limits to what a state should be allowed to dictate its citizens to do, and the far right need to cross those limits in order to pursue their agenda. At that point, this simply isn't a legitimate dispute about votes anymore.
In Germany there is still a debate going on about if the AfD should be banned or not, and while that topic is of course controversial, I think it should be done. I am sick of this. And if the voters are pissed about it, that's their problem, because there are plenty of options out there that aren't outright in opposition to our constitution.
*Sigh*
>because there are plenty of options out there that aren't outright in opposition to our constitution.
AfD voters don't agree that the AfD is in opposition to your constitution, not that your type much cares about voting so much as power at any cost
What's missing is actual follow through, or at least something that can be credibly sold that way. Take Canada, where the Liberals won an election from way behind partly due to promises to listen to voters' concerns on immigration. Granted there were other factors: Trump triggering Canadians' age old anti-American reflex helped given that the Liberals are generally seen as the pro-Canada party, and so did jettisoning Trudeau Jr., who was widely associated with progressives and open border policies
Playing chess with an entity most interested in chomping on the chess pieces tends to wind up here.