Polish Politics and Elections

[Oryxslayer]

Not normally one to post in the International Elections thread, but man is this devastating. How does the world's population love the right so much?

The former Warsaw pact countries has had corrupt and morally despicable pseudo-authoritarian parties of power govern for nearly 100 years now, the traditions and expectations have not changed even as the faces do.

[Oryxslayer]

Pretty shocking that 30-40% of young people are voting for the Neo-Nazis.

Pretty shocking that 20-30% of young Austrians are voting for the Neo-Nazis

12% of young women and 23% of young men voted FPÖ two weeks ago.

In Vorarlberg yesterday, it was 8% of young women voting FPÖ and 27% of young men.

I would say that's a difference compared to Poland, considering we have taken in 150.000 asylum seekers the last few years ... and Poland almost none.

The PiS is not a secular white nationalist party oriented around immigration policy. It is a social-minded Catholic revanchist party that has only begun recently to grow xenophobic and authoritarian. Konfedracja is closer to FPO than PiS amd they got 19 percent among young voters.

But the truth is both countries have a problem with right wing extremism. The difference is that Austrian youth are secular, multicultural, progressive, and economically prosperous. The Polish youth who are like that are in Berlin, Vienna, London, Dublin, Stockholm, etc. So the Polish youth in Poland are split between religiously devout Catholics, typical European leftists, and a huge group of often the stereotypical "economically distressed" sort who never managed to leave the country and want to blame Jews and mythical Islamic hoards for that sad fact. It is that last group who you see supporting neo nazis and parading in nationalist marches.

To further this point, a vote abroad map:



The type of voter who would move to an urban city for work (and contribute to the global trend of left growing in cities, right in rurals) is taking advantage of Schengen.

[Oryxslayer]

https://wbdata.pl/wp-content/uploads/2020/06/pojedynek-Duda-Trzaskowski.png[/img]

See other maps here https://wbdata.pl/mapy-wyborcze-wybory-prezydenta-2020-i-tura/

Looks like Poland is realigning in a way from the usual West vs East alignment to a more "traditional" rural areas vs urban areas alignment or is the map misleading in that way?

Tbf I do think Trzaskowski will win in the rural west in the runoff.

I mean part of the reason the East-West divide is a thing in Poland is thanks to the comparatively high urbanization and historic development of the West. When PO loses by 10 points than the western rurals are going to slide away leaving the committed cities. However, if we look at where Trzaskowski over and underperformed his national result, the East-West divide still is apparent - albeit with a few gaps such as the rural parts of the old Polish Corridor being better for PiS than their neighbors.

[Oryxslayer]

Apologies if this has been asked and answered already, but is there any reason why USA Poles seem to be so much more fash than the rest of the diaspora? I can imagine reasons, but wouldn't want to make assumptions

You see something similar in a lot of South American mail votes - no idea if it holds true here.

The people who end up in the US but retain citizenship are often high skill, high paid residents of the upper-crust who get in via their ability to compliment the US economy. A few other countries are like this, but its usually the US. Neighboring countries (especially in the EU) naturally have an easier time trading workers back and forth similar to osmosis This produces a different sort of resident, often one more working or lower class who got pulled in by opportunity or job listing. Depending on the situation, these types of workers may not even be permanent, they may be saving up money and sending remissions back home for when they can return. Then there is a third category - the distant outpost. The number of voters there will be minuscule, but they will be mainly comprised of embassies, attaches, and college students

[Oryxslayer]

I have been quietly following this contest, and yeah it's a tossup, tilting Duda if you put a gun to my head. However, there is an argument that the polls will be off by enough in one way or another to push the contest away from 50/50. This argument comes down the polls marginally missing the turnout, be this by coronavirus, and the number of absentions since round 1 being higher/lower than expected.

[Oryxslayer]

Nice map as far as turnout is concerned, though making predictions off of turnout is a fools game.


Anyway, I'm just going to make the hellish prediction, someone has to do it after all:

50.1 Duda
49.9 Trzaskowski

[Oryxslayer]

More turnout maps.

[Oryxslayer]

More turnout maps.

Who is this map good for?

TBH good and bad for both candidates. Although weak increases in Lower Silesia and Greater Poland are not really good for Trzaskowski.

It also bears stating that in some areas, even strongholds, the Voivodeship is too large of a geography to infer anything from. Rural Mazovia is going to vote very differently than Warsaw, her environs, or Plock. Only fools attempt to predict results by turnout, no matter what contest, no matter who comprises the electorate.

[Oryxslayer]

Duda claims victory.  At least based on exit polls this does not compute

Nah it does perfectly. He who gets out in front in front of the news and starts assuming the mantle of victory will begin to appear like the 'default' winner. The exits show a tight race, so best start building legitimacy now if you end up in front of the courts arguing over what votes matter or who actually won.

[Oryxslayer]

Would the exit poll model in votes abroad or not?

No. It is also missing late votes that were heavily urban. I wouldn't be surprised to see the late exit poll have exactly the opposite numbers.

It's over, stop saying stupid things.

Thank you, random American child.

Seriously. Whenever I follow International elections, I try to do my research and get a lay of the land. Even then though I always defer to locals and avoid provoking arguments when I lack all the facts. A lot of people would do well to follow this example.

[Oryxslayer]

Anyway, final turnout change map would certainly look to favor Trzaskowski, except who knows how much of the northern bias is because of vacationers.

[Oryxslayer]

From EUElects liveblog:

Andrzej Duda won in Canada. He recieved 4720, while Trzaskowski recieved 3119. Despite this fact, Duda won only in Toronto, while Trzaskowski in Vancouver, Montreal and Ottawa.

[Oryxslayer]

From EUElects liveblog:

Andrzej Duda won in Canada. He recieved 4720, while Trzaskowski recieved 3119. Despite this fact, Duda won only in Toronto, while Trzaskowski in Vancouver, Montreal and Ottawa.

I thought the overseas vote was supposed to be pro-Trazskowski?

It is...outside of NA.

[Oryxslayer]

Oh well.

Is it over?

Well, we won't know until the votes are counted. Duda however has much more upside.

[Oryxslayer]

Well, votes are now being counted on the official site.

[Oryxslayer]

Late Late projection for the polish vote is now 51-49 Duda. No diaspora votes included. If Duda loses it will be because of the votes from abroad.

Well, votes are now being counted on the official site.

Can you post the link.  I do not see anything

It's horribly designed, probably purposefully. If you click on a locality all the way down to the precint/poll level, you can see the results for said precinct/poll. Any higher and it's only the counted votes displayed. Someone has to run a scrape program to collect it all, and I'm not sure if there is an relatively fair media station out there who has run such a scrape.

https://wybory.gov.pl/prezydent20200628/pl/frekwencja/2/Koniec/pl

[Oryxslayer]

100% of the vote is here according to reports.

51.03 Duda
48.97 Trzaskowski

High turnout and population growth means that this is the most votes any PiS candidate and non-PiS candidate have ever received for president since the fall of the Soviet Union.

[Oryxslayer]

100% of the vote is here according to reports.

51.03 Duda
48.97 Trzaskowski

High turnout and population growth means that this is the most votes any PiS candidate and non-PiS candidate have ever received for president since the fall of the Soviet Union.

Poland has no population growth ...

Sure she is shrinking/stagnant now, but Poland has added at least 100K people since Lech Wałęsa was elected president, setting the all time high for earned votes. You also have the diaspora votes, which retains (mainly) those who have utilized Schengen for work.

[Oryxslayer]

Polling suggests that the PiS coalition is holding on to a comfortable lead, but taking into account that polls around the world are failing over and over again, these poll numbers should be taken with a massive grain of salt.

This is not exactly directed at you, but more to the whole thread since things are quiet, but this is a good launching point.

It's two weeks until the election.

Polling is fairly stable right now at something like 37% PiS+, 30% PO+, 10% Lewica, 10% Third Way, and 10% Confederation.

In theory this looks a lot like the 2019 breakdown just with worse numbers for PiS and better numbers for everyone else. And there wasn't that much of a D'Hondt advantage to PiS then, with most of their seats above the hypothetical fully proportional percentage coming from Confederation. Is this just cause everyone got a decent share of the vote, Polish D'Hondt constituencies are allocated fairly equally (compared to say Spain), or cause PiS voters are hyper-concentrated? And with PiS down, and therefore not likely to finish first in so many constituencies, would they still be the sole beneficiary from allocations? And what about Confederation, would their distribution be fairly uniform or would it be more felt in the PiS stronghold states? (potentially either cause there is more voters to lose there or because its easier to pull away proportional seats in theory if a party is starting from a very high mark)

Essentially, it looks likely from my observers perspective that PiS+ will lose the majority, the three 'willing to cooperate' lists in the opposition probably won't have an alternative majority, but how close should one expect them to be? If this is a incorrect reading of the public polls, please explain.


And then there's the (much less powerful) Senate. Is it right to say the opposition on current polling is clearly favored there? PO has seemingly always done better there than in the Sejm, and in recent times they run a joint slate with the smaller opposition parties. Which would make it PO++ vs PiS+ vs Confederation in most seats, on a map where Duda failed to carry a majority of the seats in 2020.

[Oryxslayer]

The election is in one week, and again I am surprised by the lack of discussion. This is compounded by the fact that polls have seemingly tightened in the final days of the campaign.

So Poland uses D'Hondt allocation of seats across 41 constituencies. D'Hondt as a system has a reputation for rewarding the largest parties and taking these seats from the smaller parties, at least when compared to other PR-style systems. Nowhere is this most apparent than in Spain, with the majority of constituencies having 5 or less seats, de facto meaning that small parties often get zero seats in these districts and its not their fault.

However, the 41 Polish constituencies have awarded their seats roughly equivalently. The constituency nature of this is important, cause some regions have clear comparatively shrunk since allocation, and some grown. There are a few with less, and a few with more, but almost every constancy falls between 9 and 14 seats. What this means is that it isn't too hard for a small party to obtain representation in every constituency, as long their vote totals are large enough.

Mathematically, this benchmark is somewhere around the 10% line nationally. This can be calculated using the 2019 votes shares in the constituencies and applying swings based on polling, be it weighted or simple uniform. Now this isn't uniform, some constituencies will provide 0 seats, others 2, but it averages out to around 41.

This has two important effects. The first is that the small parties aren't getting harmed to much by the system. Winning 41 seats with 10% of the vote is very close to the 46 under pure proportional. Which means there are fewer 'bonus' seats avoidable to be awarded to the big two. The second effect is that it de facto means that if polling is even close to accurate, PiS has lost an absolute majority. Confederation growing ~4% at the expense of PiS means they are getting allocated at least 30 PiS seats, putting the governing party far from the 50% line. But you would expect Confederation to eventually end up supporting a PiS government, so on it's own this has little impact.



This is where we now have to discuss the tightening polls. Now a number of them are seemingly from the selection of polls that in the past are better for the opposition, but even some more PiS friendly ones are showing tightening compared to past numbers. And none are really consistent where this tightening is going, some to PO+ (which would more efficiently lead to opposition gains), and some to Lewica and Third Way. If I had to throw out a guess, I would say this is the result of increasing temporal distance from the German dispute, leading to other issues regaining prominence. But the overall result is similar: there can legitimately be a PO led government with ~10% or higher for the the three smaller parties, and a 5% PiS lead or smaller over over PO+.

This basically works through the peculiarities of D'Hondt, and PiS's poor geographic voter distribution. The easiest way for a big party to potentially pick up a seat in D'Hondt is to finish first in a constituency. Obviously it will come down to how the smaller parties balance out, and how many seats are in a constituency, but eventually there will be a point where the largest party gets allocated a seat cause it is the largest. A 5% margin or less in theory is where PO+ starts finishing first in a lot of western constituencies, since PiS piles up the votes in eastern constituencies where seats are hard to change hands. Obviously this all depends upon how PiS loses votes, be that uniformly or more concentrated such as hypothetically outside of the southeast. Which is why it becomes more possible the tighter the two parties get, no matter if the tightening is caused by smaller party growth or a swing towards PO+.

[Oryxslayer]

snip

I've been quite busy lately but I do hope to put up some posts about how I think the campaign has gone before election day. For whatever reason, it's always been quite hard to get Atlas people interested in Polish politics, so I rarely look in here.

Oryx, can I ask where you're getting your information on Polish electoral geography? Because it is certainly not the case that PiS has 'poor geographic voter distribution', in fact its distribution is so good that PO could narrowly win the popular vote and still come out with fewer seats. There are a few reasons for this.

First of all, the number of seats per constituency is fixed, so a vote in a constituency with lower turnout technically has greater weight than one in a constituency with higher one. Urban turnout is much higher, and cities favour the opposition. On top of that, votes from abroad all go into the Warsaw I constituency, but it doesn't get any extra seats to account for this, so voters in the capital end up a bit underrepresented.

The other reason is malapportionment - the last time the seat allocation was updated to account for population changes was 2011, and this artificially hands PiS a handful of extra seats.

I think we are approaching it from two different angles. You are looking at it from the perspective of the structural barriers, which do exist. I am looking at it from the perspective of the electoral map. So you will say the opposition winning 50.1% and not winning a majority gives PiS an advantage. I say, given the nature of the opposition's division, the opposition winning 51% and getting a majority is a advantage since PiS are going to come in first in D'Hondt.

For comparison, lets see Spain. The parties that just tried to put Feijoo in power won 46.1% of the vote, to the parties that would put Sanchez into office (everyone else with seats excluding Junts) winning 49%. And The right got 1 more seat cause the PP came in first.

The simple truth is that PiS votes are concentrated in the East, most especially the southeast. In a virtually tied race, lets say the 2020 presidential election, PO wins 22/41 Sejm districts containing 262/460 seats, obvious proportionality makes this complicated, but we'll get to that in a moment. And that's a race they lose 51-49.

PO also won a majority of senate seats in that race. Which is a good hint that the mallaportionment in the Sejm isn't that bad from a partisan perspective. The cities have grown, but so has the religious Southeast. The rural areas have Shrunk, but both the opposition and PiS need rural seats.

In a straight PO vs PiS race, PO wins  more seats with smaller margins, whereas PiS pile on the votes in safe areas.

But that is all two-party stuff. This is not a two party election. I think you are looking just at PO+ versus PiS+, whereas I am looking at it as PiS+ (and Confed kinda) verses PO+, Lewica, and the Third way alliances. PO+ is not going to win the popular vote and I don't think we should expect them to unless they show signs of it.

 That is where the majority of the PiS Sejm advantage comes from that you are saying exists. As I said above, under D'Hondt the big parties get a boost at the expense of others. In 2019, PiS basically swept first place everywhere so they got an absurd bonus. But a closer race sees PiS win less and less, districts and therefore get less.

Here's basically how that works, in a hypothetical pool of 100 votes, and a district with 10 seats. 45 go for party A, 43 for party B, and 12 for party C. This gives 5 seats to party A, 4 to party B, and 1 to party C. Party A won a seat off only 2 votes, whereas every other seat takes at least 8 to get apportioned.

Here's another scenario of 100 voters and 10 seats. Only this time it goes 65-15-15-5 for four parties. 8 seats go to the first party, and one to the second and third. There are obviously a lot of wasted votes for party 2 and 3. But now lets bring that to 60-15-15-10. Now its 7-1-1-1. This is a reduction of the math in the PiS strongholds. PiS mathematically can't get any more seats easily, and driving up base turnout has little to no reward. The other opposition parties are wasting votes. But the smaller, more testimonial parties can make a much more enthusiastic appeal to a smaller section of the electorate only somewhat satisfied by the big two. Very much the case for Confederation. This allows them to with less comparative resources pull away a seat that PiS had to expend much more effort to win.

So yes, do I think the opposition needs more than 50% for a majority? Yes. And the polls that do suggest they could get one tend towards 51-52%. How much above that line though is fairly small for D'Hondt when you are losing first place by several percent points. That line can also come down if PiS losses are not uniform, and their bedrock voters in the eastern constituencies, where PO+ has a harder time pulling away seats, remain the firmest. It also comes down if Third Way get a bit above 10%, cause they seemingly have a temporarily efficient coalition. For only 1 or more 2%, they temporarily outperform the perfectly proportional hypothetical seat allotment.

[Oryxslayer]

https://twitter.com/EuropeElects/status/1711358435051065710?t=3VpV0vc3oE9fZxeXIP...

This is a good public presentation of the math that I have done in the background. A PiS leaning outlet, compared to the averages, gives us a 6% lead over PO and only 9% for all the smaller tickets. But this is only a 8 seat majority for PiS+Confed. PO+ and/or the minor parties - including Confed here since they take for PiS - doing a little bit better rapidly drops that total. PiS falls below PO+ in more Western constituencies, transferring D'Hondt 'bonus' seats  awarded in allocation to the opposition. The minor parties doing a bit better gives them representation approximatly everywhere, including in the places PO+ can't make a targeted appeal towards. That's why I say 35-30-10-10-10 is the divider, with PiS or PO governments becoming increasingly likely depending on the deviations from that line and in what direction.

A more divided constellation of parties getting a majority in D'Hondt is hard when they are also losing first place,  and yet somehow the PiS voter concentration makes it possible.

[Oryxslayer]

SNIP

You come off as confrontational. If I did previously, then I am sorry. So I am just going to TL:DR what I am trying to convey:

Empirical studies based on other, more popular concepts of disproportionality show that the D'Hondt method is one of the least proportional among the proportional representation methods. The D'Hondt favours large parties and coalitions over small parties due to strategic voting. In comparison, the Sainte-Laguë method, reduces the disproportional bias towards large parties and it generally has a more equal seats-to-votes ratio for different sized parties.

If we were to reverse the 2019 election using the calculator (arguably a much neater version of my stats tables, but without regional weights) that you linked for example, PO+ suddenly starts to have a much more efficient vote total. The seat counts do not perfectly reverse, I would say largely thanks to how voters abroad are handled (which is stupid and benefits PiS, but you either play by the rules of the game, no matter how hypothetically gerrymandered and oppressive, or you boycott and delegitimize the contest) but also cause PO's "lane" has a third party competing for allocations, so the west can never be as favorable to PO as the east to PiS in seat allotments.

All this fire and fury has been to try and say pure D'hondt is not a proportional system from my perspective, getting more and more unproportional the number of tickets you add to system and the larger a single party becomes (and the smaller the number of seats per consituency, but thats not a issue here). It is quite literally FPTP but with more than one seat, or more 'posts,' per district. And every such system has some clickbait newsgrabber of how a mathematically precisely distributed coalition can end up with absurd representation when compared to votes won. Division is punished harshly and unity greatly. And so I am trying to explain, perhaps in a roundabout way, how those harmed by the system can still make themselves be the winners.

[Oryxslayer]

9 p.m. IPSOS EXIT POLL

PIS 36.8% - 200 seats
KO 31.6% - 163 seats
TD 13% - 55 seats
Lewica 8.6% - 30 seats
KON 6.2% - 12 seats
BES 2.4% - 0 seats

KO + TD + Lewica MAJORITY

Unless this is really wrong (which history says it won't hopefully)...that shouldn't be safe from fluctuations of a percent or two. Confederation falling again below the effective threshold in a bunch of constituencies screws not just them, but also PiS who now needs their votes. It also saves Lewica who could be losing a lot more if Confederation didn't plummet.

If things do change, its going to be because of how and if seats are apportioned to Lewica and Confederation in various constituencies, lets keep an eye on that during the court.

[Oryxslayer]

Did either of the exits bother on the four referendums? Though i can probably guess the results in at least 2.