[EM] Down with party poopers and primary poopers

[Kristofer Munsterhjelm]

On 2023-12-19 06:58, Rob Lanphier wrote:
> Hi Kristofer,
> 
> Good to hear some non-American perspective.  The United States' 
> two-party democracy leads us to think that fixing the "duopoly" will fix 
> everything.  There is a certain comfort in the stability with our 
> Coke/Pepsi duopoly in the United States.  You're clearly correct that 
> voters can handle more than two parties (given the multitude of parties 
> in Norway as an example), and I think that it sucks that the top two 
> here in the United States are so safe.  It bothers me that we actually 
> DO have more choices than Coke and Pepsi, and that so many people almost 
> immediately understand that metaphor for "duopoly".  I wish that our 
> collective IQ could be applied to more than beverage shopping (and more 
> generally, that alternatives to "Republican" and "Democrat" in the 
> United States had greater brand recognition than "Coke" or "Pepsi")
> 
> More inline...
> 
> On Mon, Dec 18, 2023 at 6:34 AM Kristofer Munsterhjelm 
> <km_elmet at t-online.de <mailto:km_elmet at t-online.de>> wrote:
> 
>>  One way I tend to think of it is that in a two-party system, the two
>>  parties contain factions that, in a multiparty system, would be
>>  separate parties. Thus, in a two-party system, first the factions negotiate
>>  and then people vote; whereas in a multiparty system, the people vote and
>>  then the factions negotiate.
> 
> 
> That sounds nice.  I'd be surprised if there weren't intraparty factions 
> as well, though.  I have to believe the eight (or so) major parties in 
> your country have at least a little semi-structured infighting.  
> Intraparty factions almost certainly still have to be a thing, even when 
> there are more choices when one leaves a particular party.

Yes. There is definitely a spread of positions within parties, and 
moreso in the parties that have greater support, such as the 
Conservatives and (particularly) Labor.

A more accurate way of putting it would be that larger parties need to 
accommodate a wider range of opinions to stay coherent. Thus you're more 
likely to find significant differences within large parties than smaller 
ones, although the picture is complicated a little by more ideologically 
based smaller parties whose members may have strong opinions about just 
what variant of its ideology is the right one.

You probably can't expose every subtle nuance in a party system without 
having an unreasonable number of parties, but going from two to many 
helps a lot and gives the voters more of a say about which political 
positions deserve power.

>> It seems to me that if there is a possibility for the relative
>> strengths
>> of factions to be decided by the people rather than by party insiders,
>> we should take it. [...] Now, this has to be balanced against the
>> burden of being familiar with more parties. If there were a thousand
>> parties, there would be a big problem because nobody would care to
>> read a thousand party programs.

> I also think that this is a problem for historians and for the 
> electorate being able to learn from history.  Having such a swirling 
> soup of parties makes it difficult for newcomers and outsiders to 
> understand how local politics work.  I suppose it's not any worse than 
> trying to understand what happened to the Rockerfeller Republicans in 
> the United States, though:
> https://en.wikipedia.org/wiki/Rockefeller_Republican 
> <https://en.wikipedia.org/wiki/Rockefeller_Republican>

Yes, though historians would have the benefit of being able to focus 
more thoroughly on their fields. Ordinary voters usually wouldn't 
dedicate so much time to finding out the intricate details of their parties.

I suppose what I'm saying is: if it's too hard for the historians, it's 
too hard for the voters. The historians may cover more time, but the 
voters have less of their own time to spend.

In more closed systems, the historians would anyway try to find out 
what's going on inside the few parties that exist. (The logical extreme 
of this is probably something like China watchers who try to infer the 
dynamics of the CPC based on what is revealed to the public.)

> 
>     But at least in this country, the voters don't seem to have a problem
>     dealing with eight parties.[1] While brand recognition is important,
>     there doesn't seem to be a significant penalty to having eight rather
>     than two.
> 
>     [1] 8.36 parties by
>     https://electowiki.org/wiki/Effective_number_of_parties#Entropy_measure <https://electowiki.org/wiki/Effective_number_of_parties#Entropy_measure>
>     and
>     https://en.wikipedia.org/wiki/2021_Norwegian_parliamentary_election
>     <https://en.wikipedia.org/wiki/2021_Norwegian_parliamentary_election>.
> 
> 
> It's interesting to me that you chose "eight" as the number of parties 
> in Norway when you used the 2021 Norwegian parliamentary election as a 
> reference.

In a multiparty system, some parties may be much larger than others. The 
effective number of parties measure tries to answer the question of how 
many parties exist in terms of "if they were all equally sized, how many 
would there be?". Since some parties have considerably less support than 
others, the ENPP figure will be lower than the number of actual parties, 
but I thought it would be a more appropriate comparison to a two-party 
system, since the two parties already are equally sized.

> By my count, there were ten parties that received seats in 
> the "Storting" (a.k.a. "the supreme legislature of Norway" according to 
> Wikipedia) in 2021.  It would seem based on a cursory glance at the 2021 
> results that the parties can be classified large/medium/small:
> 
>   * The  "Labour", "Conservative", "Centre", and "Progress" parties --
>     these are the four big parties in Norway, and an overwhelming
>     majority of Norwegians at least partially align with one of these
>     parties.  It would seem impossible to really understand the 2021
>     election without understanding these four.
>   * The "Socialist Left", "Red", and "Liberal" parties are the
>     medium-sized parties that earn a modest number of seats.  I'm
>     guessing that each of them must really rely on alliances with at
>     least one of the larger parties to get anything done.
>   * The "Green", "Christian Democratic", and "Patient Focus" parties are
>     the boutique parties that appeal to a very narrow slice of the
>     population.
> 
> As I look back over the past 40 years of elections, my taxonomy above 
> breaks down, since it's based on a single election.  Labour, Progress, 
> Conservative, Socialist Left, Centre, Christian Democratic, and Liberal 
> seem to be the only seven that have a substantial history.  Green, Red, 
> and Patient Focus seem like weird little newcomers, but I guess the Red 
> party is seemingly on the rise, no?  Also, does the modern Liberal party 
> have anything in common with Mowinckel's Liberal party of the 1930s 
> (arguably the Liberal party's heyday)?

Your estimate is roughly correct for the last election. I would say 
Progress is not in the big parties bloc, but that's partly due to its 
history. Without going too much into detail:

The beginning of parliamentary rule was first nonpartisan, but quickly 
the Conservative and Liberal parties formed, broadly representing the 
right and left wing, although the right/conservative was more in terms 
of e.g. allowing the monarch wider powers, and there was not yet an 
organized labor movement.

As history went on, more parties came into existence. This in part 
explains the Liberal party's decline: it started off as the big-tent 
everything from center to the left party, but its domain was encroached 
upon by the other parties establishing themselves. E.g. Labor coming 
into existence as a result of organized workers' struggle. (If the US 
had become multi-party in the late 1800-early 1900s, perhaps Debs' 
Socialist Party would similarly have encroached on the Democratic Party? 
I know too little about US history to say, but it would be a roughly 
analogous development. The Norwegian Labor Party was pretty radical in 
its earlier days too.)

The Centre party and Christian Democrats cater to particular groups: 
farmers and Christians respectively. The very south and some of the west 
of Norway has a considerable higher density of practicing Christians, so 
these are the CDP's strongholds. The Centre party is very pragmatic and 
focuses on increased agricultural subsidies, devolution/increased power 
to the municipalities, etc. Earlier the Centre party aligned itself with 
the right-of-center bloc ("non-socialist bloc"), but it's now aligned 
with the socialist one, trying to be, as the name implies, centrist in 
economic matters and otherwise retain its pragmatic nature. This can be 
seen in municipal elections: in one municipality, they may give support 
to Progress, and in another, to Labor.

The post-2005 coalescing of parties into more durable blocs left the 
Christian Democrats with a serious dilemma. As it holds both 
conservative and liberal Christians, there was a big debate on what 
parties they should align with, or if they should be unaligned and give 
confidence to any bloc on a case by case basis. They first tried to move 
in a liberal direction, but lost support, and some of the most 
conservative Christians tried to form their own party ( 
https://en.wikipedia.org/wiki/Conservative_(Norwegian_political_party) 
), but the attempt fizzled.
Then, as a result of continual whittling, and the wings being evenly 
matcehd yet unable to find a compormise, there was a party vote in 2019. 
The margin was very narrow but the party decided to follow the 
right-of-center coalition. However, this led a considerable fraction of 
the more liberal voters to go vote for somebody else.
In short, their problems were partly due to the decreasing influence of 
religion in Norway and partly due to two wings or factions or equal 
power having too different ideas of what the party should become.

Progress, Socialist Left, and Red are more recent.

Progress started as a right-wing close to libertarian economic party, 
hence the original name "Anders Lange's Party for strong reduction of 
taxes, duties and public interventions". Its leader from 1978 to 2006, 
Carl I. Hagen, oriented it more in a general right-wing 
conservative/populist direction: opposing immigration, focusing on 
national identity and so on. So whereas the Conservative party is more a 
business owners' party, Progress more seeks the support of people who 
think immigration and open borders have gone too far. I guess a very 
rough comparison would be something like: what if the Libertarian party 
joined with tea partiers and southern conservatives and then later 
moderated itself to gain wider support.

Socialist Left is to the left of Labor and basically started as the 
party for socialist voters who didn't approve of or wanted to be 
associated with the USSR. It currently enjoys the support of public 
employees in particular, and is in favor of extended public services.

Red is further to the left, and its origins is as an umbrella group for 
a number of communist organizations to propose candidates. 
(https://en.wikipedia.org/wiki/Red_Electoral_Alliance) It is no longer 
revolutionary, but has specialized itself as a party supporting the 
concerns of the poor and otherwise less well off, particularly in 
economic matters. As Labor has migrated rightwards (starting with New 
Labour/New Public Management type policies in the 2000s; and moreso 
lately with acceptance of more market-based/neoliberal EU systems), this 
has opened up more space for the Socialist Left and Red parties. Some 
voters consider Labor to have become less principled lately, 
particularly under the latest PM, which has also pushed more voters to 
either Centre and Conservative (to the right) or Socialist Left and Red 
(to the left).

The Green party is another example of the increasing support of 
environmentalist parties in Europe in general; and Patient Focus is a 
single-issue party. In a non-party list system, Irene Ojala might have 
been an independent.

(Since Norway is such a large country and the very North is sparsely 
populated, there have been decisions to build hospitals further South 
rather than closer to the few people who live in the very North. This 
led to very long travel times for some, who then strongly objected to 
the decision. The party was formed to advocate for the expansion of 
medical services in the North.)

> Interestingly, the "Christian Democratic" seems to have fallen 
> precipitously from its heyday in the 1980s and 1990s (under Bondevik's 
> leadership).  This Wikipedia article seems to confirm my speculation 
> about intraparty infighting being a thing, even with a lot of parties:
> https://en.wikipedia.org/wiki/Christian_Democratic_Party_(Norway) 
> <https://en.wikipedia.org/wiki/Christian_Democratic_Party_(Norway)>
> 
> That would also be true of the Liberal party:
> https://en.wikipedia.org/wiki/Liberal_Party_(Norway) 
> <https://en.wikipedia.org/wiki/Liberal_Party_(Norway)>
> 
> ...though I suppose the answer to intraparty strife is to FORM A NEW 
> PARTY (or two)!!!!!
> https://en.wikipedia.org/wiki/Liberal_People%27s_Party_(Norway,_1972) 
> <https://en.wikipedia.org/wiki/Liberal_People%27s_Party_(Norway,_1972)>
> https://en.wikipedia.org/wiki/Liberal_People%27s_Party_(Norway) 
> <https://en.wikipedia.org/wiki/Liberal_People%27s_Party_(Norway)>

This shows another way to deal with factions: In the early days of the 
Progress Party, Carl I. Hagen kicked the libertarian wing out of the 
party to reorient the party in its social conservative/populist 
direction. There have since been a number of attempts to try to create 
something either "like Progress but not Progress" or more libertarian; 
none of these have really succeeded.

> I'm not entirely clear how you arrived at "8.36 parties", which seems 
> awfully precise given the fluctuation of seat-earning parties in Norway.

You have a point: this is like the statistics joke "I've measured the 
distance to your house to be 13.8752 miles +/- 10 miles". I just ran the 
numbers through the effective number of parties calculation, and my 
inclusion of decimals might have given an impression of more precision 
than what's really there.

It would be interesting to see the change over time.

>>    There are definitely caveats worth mentioning: it's harder to have a
>>    negotiation phase if the president is directly elected, since
>>    there's no
>>    parliament where this discussion can happen. And we don't vote on minor
>>    positions like city coroner, either. But it's still possible for
>>    presidential systems to be multi-party, and to my knowledge, these do
>>    not have primaries.
> 
> 
> I suspect voters in the United States are way too fond of micromanaging 
> their leadership to allow for taking away voting for city coroner or to 
> vote for a party, and let the party leaders negotiate who gets to be 
> president or prime minister or grand poobah or whatever.  I didn't fully 
> realize just how recent primary elections were anywhere in the world 
> until the 20th century, and that the first time primary elections 
> mattered for the U.S. presidency as in 1912:
> https://en.wikipedia.org/wiki/1912_Republican_National_Convention 
> <https://en.wikipedia.org/wiki/1912_Republican_National_Convention>
> 
> While I'll concede that there are multi-party democracies that haven't 
> found a need for primaries, and I'll concede that the brief history of 
> primary elections here in the United States and elsewhere has been 
> messy, I still think it's a good idea for high-stakes elections.  For 
> high-stakes positions, it seems wise (to me) to allow the general public 
> two big stages of consideration: the primary process and the general 
> election process.  Though many aspects of the current mechanisms we use 
> are awful, I think having  two elections rather than one gives us enough 
> time to collectively think through the decisions we make.  For 
> high-stakes elections, this seems worth it.

I wonder, could this partially be due to FPTP and IRV's first preference 
support? Suppose you have a method that's vulnerable to center squeeze. 
You would want to do multiple elections so that you can get unusually 
strong wings out of the way and focus on something that everybody likes. 
But if you have a method that is better at finding the median voter 
preference, then the presence of wing candidates with strong but limited 
support don't mess with things as much.

If so, we should see less of a need for multiple steps with approval or 
Condorcet than with FPTP or IRV -- although in a sense IRV already 
*simulates* multiple steps. Someone (I don't remember who) called IRV 
"Instant Primary" for this reason.

Then there's another perspective: that having multiple rounds let people 
think between the rounds, and let the candidates readjust their 
arguments, debate etc around what the people are showing to be important.

Consider that IRV countries have two party rule whereas delayed 
countries usually don't, even though top-two seems to be much worse in 
terms of crtierion compliance than IRV. Perhaps this dynamic process 
(that the voters can change their minds as the field thins) helps 
sustain multiple viable choices.

But if that's true, I would probably prefer an actual runoff to closed 
primaries, because it would be better for both rounds to poll everybody 
(hence have a greater chance of finding the actual median voter 
position) than do medians of medians, which could potentially lead to 
their own kind of center squeeze.

-km