Japan Oct 22 2017
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
March 28, 2024, 09:22:34 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Other Elections - Analysis and Discussion
  International Elections (Moderators: afleitch, Hash)
  Japan Oct 22 2017
« previous next »
Pages: 1 ... 23 24 25 26 27 [28] 29
Author Topic: Japan Oct 22 2017  (Read 40922 times)
Cape Verde
asianzzang
Rookie
**
Posts: 89
Show only this user's posts in this thread
« Reply #675 on: October 31, 2017, 03:57:51 PM »

Fukushima Prefecture

PR Vote Share

LDP+KP 43.0%
HP+CDP 41.4%
JCP+SDP 11.1%
JRP 3.3%
Others 1.2%

Constituency Vote Share

LDP 47.1%
HP+Pro-HP Indep+Pro-CDP Indep 44.0%
JCP+SDP 7.8% (JCP candidate in District 2, 3, 4, 5 & SDP candidate in District 4, 5)
JRP 1.1% (Candidate in District 2)

LDP won District 2, 4, 5. Pro-HP independent won District 3. Pro-CDP independent won District 1.
LDP's high voting share can be explained by good performance in 5th District. HP made a mistake by nominating an MP who had been planning to retire from politics. He wasn't even an MP in the first place but took his seat three months ago when an another MP retired to successfully run for the mayor of Sendai.
LDP won only by 0.8% in 4th District, thanks to a three-way vote split.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #676 on: November 10, 2017, 07:29:09 AM »
« Edited: November 30, 2017, 07:17:37 PM by jaichind »

Sorry if I been away for a while.  I manged to find out how to get the PR vote at the district level by looking at the PR vote by township and then constructing the PR vote by district and have been compiling them so I can do some analysis on how cross voting worked in this election.  

The basic idea is to split the district into different types (in terms of which parties are running) and then for each type compute that a cross-vote free vote share might look like based on the PR result and then compare to actual district results to get a sense of how candidate quality or other factors might be at play.

I will start with with the highest frequency and most decisive type.  Which is LDP vs HP vs JCP.  There are 129 such districts and it was here where the election was won for the LDP.  In such a district you would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occationall HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*JRP + 0.5*NPB
HP = 0.5*NPD+ 0.5*JRP +HP + CDP + 0.5*NPB
JCP = SDP + JCP

The idea here is all the LDP-KP and Right fringe party PR vote (PJK HRP) would vote LDP.  JRP is nominally a HP ally but JRP can be see an Osaka LDP splinter so I expect them to split their vote 50/50 between LDP and HP.  Same for NPD which is nominally a LDP ally but since many NPD voters are ex-Center-Right DP voters I expect the NPD vote to be split between LDP and HP.  Since HP mostly ran ex-DP candidates I would expect HP to win the CDP vote while the JCP candidate should win the Left vote (SDP JCP.)  NPB is a protest party but generally centrist so I expect them to split their vote 50/50 between LDP and HP.

Candidate quality will play a large role,  so I split the 129 seats into 4 types.  67 of them are where HP nominated an ex-DP candidate that has a track record of running in that district.  5 of them are experienced but center-right in their orientation and views.  19 of them are where HP nominated a complete political novice without any sort of real political experience.  38 of them are where HP nominated someone with DP background but without any real experience of running in said district.  



If we go with the 67 where HP nominated an ex-DP candidate that has a track record of running in that district.  The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
33.64%  12.25%  0.31%  0.12%  0.55%  4.13%  21.88%  18.22%  1.61%  7.21%  0.08%
 
Note that NPD PJK and NPB did not run in all districts so their vote share are lower than they they actually go in places where they ran a PR list.

So using my equation one would expect

LDP (plus HRP and minor Right)   48.83%
HP                                             42.36%
JCP                                             8.82%

The real result which was LDP 51 seats HP 16 seats was

LDP (plus HRP and minor Right)   47.50%
HP                                             41.68%
JCP                                            10.82%

Here the candidate quality of HP is high and comparable to the LDP candidate so what the LDP got mostly was as predicted.  HP under-performed because it seems despite good candidate quality some CDP PR voter defected to JCP.  This would be decisive.  Had these defections not taken place HP would have won 6 more seats from LDP to make it 45 vs 22 instead of 51 vs 16.




If we then look at the 5 where HP nominated a candidate with experience in said district but with a Center-Right background.  The PR vote for these 5 districts were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
30.51%  12.42%  0.00%  0.33%  0.39%  3.45%  19.76%  21.83%  1.00%  9.24%  1.08%

So using my equation one would expect

LDP (plus HRP and minor Right)   45.91%
HP                                             43.85%
JCP                                            10.24%

The real result which was LDP 5 seats HP 0 seats was

LDP (plus HRP and minor Right)   49.06%
HP                                             33.29%
JCP                                            17.65%

Here a more center-Right experienced candidate failed to capture LDP PR voters instead bleeding HP/JRP support to LDP and instead lost a bunch of CDP voters to the JCP candidate.  In fact these trends might reinforce each other where JCP was expected to expand beyond the SDP-JCP base into the CDP vote which triggered HP/JRP tactical voting for LDP.




If we then look at the 19 where HP nominated a complete political novice without any sort of real political experience.  The PR vote for these 19 districts were

 LDP         KP        NPD      PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
33.69%  12.70%  0.00%  0.18%  0.48%  5.32%  17.72%  19.09%  1.46%  8.81%  0.54%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   49.99%
HP                                             39.74%
JCP                                            10.27%

The real result which was LDP 19 seats HP 0 seats was

LDP (plus HRP and minor Right)   57.16%
HP                                             24.47%
JCP                                            18.27%

Here the fact the HP candidate is a political novice allows the LDP to way over-perform where there are significant JRP and HP PR voters defecting to LDP.  Also the poor candidate quality also means that a much larger bloc of CDP PR voters went over to JCP.  Of course a key point is by going with a candidate with no experience and not someone of DP background the local DP branch most likely did not push for the HP candidate adding to the CDP defection to JCP.   So the Koike plan to replicate her Tokyo Prefecture election  success with political novices completely bombed at the national level.

 


We then look at the 38 district where HP nominated someone with DP background but without any real experience of running in said district.  The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
36.46%  12.98%  0.00%  0.14%  0.54%  4.46%  19.11%  17.67%  1.45%  7.19%  0.00%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   52.35%
HP                                             39.01%
JCP                                             8.64%

The real result which was LDP 38 seats HP 0 seats was

LDP (plus HRP and minor Right)   59.51%
HP                                             27.67%
JCP                                            12.81%

So the LDP over-performed just as much as in the situation where the HP nominated a political novice.  So highlights the need for the main anti-LDP party to have quality candidates that have experience running in the district in insertion.  Since the HP candidate in question does have a DP background the CDP defection rate is lower than in the situation where a novice was running for HP but without a DP background which makes sense.  But it is interesting that the CDP->JCP defection rate here is around 4% but the CDP->JCP defection rate for when HP has a quality candidate is around 2.5%.  So part of it would also be perception of how winnable a race is.  The pro-JCP CDP PR voter would vote HP if they figure the HP candidate could win. If not why not just vote your conscious.


The overall picture here is HP failed to get enough quality candidates to run and where they did the CDP defection lowered there hit rate even in places where they had a chance. Also while the LDP-KP PR vote share was below 46% we have to take into account that PR voters of NPD PJK HRP would default to LDP.  The JRP  PR splitting its vote was expected but disappointing because the whole purpose of HP running as a Conservative party and forming an alliance was to capture the JRP PR vote.

Of course the poor candidate quality problem might be a circular problem.  Because the districts where HP does not have a quality candidate have a greater LDP-KP lean, local DP politicians that might be a quality candidate does not bother wanting to run which in turn means that those that do run end up not being quality candidates.  The resulting LDP blowout reinforces this fact and serve to lower anti-LDP turnout and the circle is complete.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #677 on: November 11, 2017, 06:11:42 PM »
« Edited: November 15, 2017, 04:19:16 PM by jaichind »

The next set of districts to look are the 9 districts where it was LDP vs Ind(HP) vs JCP which is similar in structure to the LDP vs HP vs JCP already covered.   The expected district vote as an equation of the PR vote would be the same as LDP vs HP vs JCP

LDP (LDP plus the occationall HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*JRP + 0.5*NPB
Ind(HP) = 0.5*NPD + 0.5*JRP +HP + CDP + 0.5*NPB
JCP = SDP + JCP

If we look at these 9 districts the PR vote were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
34.09%  10.08%  0.00%  0.08%  0.69%  5.19%  17.41%  23.20%  1.65%  7.61%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   47.53%
Ind(HP)                                     43.20%
JCP                                             9.26%

The real result which was LDP 2 seats Ind(HP) 7 seats was

LDP (plus HRP and minor Right)   39.62%
Ind(HP                                       51.91%
JCP                                              8.47%

Here the superior candidate quality of Ind(HP) was able to capture vote from LDP and KP PR voters.  To be far these 9 Ind(HP) candidates were self-selected.  They all could have joined HP if they wanted to or were prominent leaders of DPJ/DP that HP choose not to take on so HP does not look like just another succession party to DPJ/DP.  All of them were fairly confident of their ability to win ergo they felt a party label was not necessary.  Even in the two seats Ind(HP) lost to LDP it was close.  Again candidate quality shows up as a key factor in victory.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #678 on: November 12, 2017, 05:56:22 PM »
« Edited: November 30, 2017, 07:18:18 PM by jaichind »

The next set of districts to look are the 19 districts where it was LDP vs HP vs CDP.  Here most the HP candidates were novices and more likely to have a Center-Right background while the CDP candidate is the more experienced ex-DP candidate in the said district that HP did not want to take for ideological reasons.   The expected district vote as an equation of the PR vote would be  

LDP (LDP plus the occationall HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*JRP + 0.5*NPB
HP = 0.5*NPD + 0.5*JRP +HP
CDP = CDP + SDP + JCP + 0.5*NPB

Here with a CDP candidate in the fray the CDP PR voter would vote CDP.  The CDP candidate should also capture the SDP and JCP PR voter as well since there is no candidate from the Left in the race.
 Also since the CDP candidate would be, in this case, in a better position to take on LDP, the NPB split would be between LDP and CDP.

CDP Candidate quality will play a large role since the HP candidate here tend to be of novice quality so I split the 19 seats into 2 types.  15 of them are where CDP nominated an ex-DP candidate that has a track record of running in that district.  4 of then are where CDP nominated a complete political novice or DP background but without any sort of real political experience in the district in question.



If we look at these 15 districts where the CDP candidate is of high quality the PR vote were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
32.52%  10.03%  0.00%  0.29%  0.35%  3.73%  16.51%  26.30%  1.01%  8.33%  0.94%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   45.52%
HP                                             18.37%
CDP                                           36.10%

The real result which was LDP 7 seats CDP 8 seats HP 0 seats was

LDP (plus HRP and minor Right)   42.42%
HP                                             15.85%
CDP                                           41.64%

Most of these districts  are urban districts where the LDP is weaker.  In seems that the novice but Center-Right HP candidates did lose some votes to LDP but CDP ate into the LDP and KP vote beyond what is expected from what the PR vote would imply.  It seems the CDP campaign did win over from LDP-KP PR voters which went to CDP along with some HP voters given the fact taht the CDP were in a strong position to take on LDP in these districts.  




If we look at these 4 districts where the CDP candidate is inexperienced the PR vote were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
32.62%  9.41%   1.55%  0.23%  0.34%  3.82%  5.93%  25.66%  0.99%  8.82%  0.62%

So using my equation one would expect

LDP (plus HRP and minor Right)   45.60%
HP                                             18.61%
CDP                                           35.78%

The real result which was LDP 4 seats CDP 0 seats HP 0 seats was

LDP (plus HRP and minor Right)   46.15%
HP                                             18.83%
CDP                                           35.02%

Here despite the lower quality HP and CDP candidates, the lack of a JCP candidate mostly kept the result pretty much matching what one would expect from the PR vote which means that marginal HP and JRP PR voters did not migrate to LDP despite the lower quality HP and CDP candidate.  Of course there was not migration toward the CDP candidate which was key for CDP victories in where the CDP candidate was of high quality.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #679 on: November 14, 2017, 08:18:01 AM »
« Edited: November 30, 2017, 07:19:15 PM by jaichind »

The next set of districts to look are the 17 districts where it was LDP vs HP vs CDP (or Ind(CDP)) vs JCP.  Here there does not seem any real prospect of LDP being beaten given there are now 3 opposition candidates running,    Just like the LDP vs HP vs CDP, the HP candidate tends to be inexperienced since the ex-DP candidate of the district most likely went to CDP.  The expected district vote as an equation of the PR vote would be  

LDP (LDP plus the occationall HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*JRP + 0.5*NPB
HP = 0.5*NPD + 0.5*JRP +HP
CDP = CDP + 0.5*NPB
JCP = SDP + JCP

Here the expected vote share are the same as LDP vs HP vs CDP except for the fact that the SDP+JCP PR vote will now go to the JCP candidate.  The CDP candidate here is the most likely to be the most likely to beat LDP ergo I allocated 0.5*NPB to CDP.

Again, we break out these 17 districts into 2 types.  10 of them are where the CDP candidate has experience running in the said district.  While 7 of them are where the CDP candidate is a political novice or lack experience running in said district.



If we look at the 10 districts where the CDP candidate quality is high the PR vote were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
30.75%  11.93%  0.00%  0.43%  0.31%  3.53%  17.49%  23.53%  1.07%  9.70%  1.26%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   45.82%
HP                                             19.25%
CDP                                           24.17%
JCP                                            10.76%

The real result which was LDP 10 seats CDP 0 seats HP 0 seats was

LDP (plus HRP and minor Right)   46.36%
HP                                             17.37%
CDP                                           27.30%
JCP                                              8.88%

There did seems to be some JCP->CDP tactical voting but nowhere enough for CDP to win.  The lower quality of the HP candidate means that it loses some votes for LDP relative to the baseline.  Key trend here is that when we compare this batch of seats vs the LDP vs HP vs CDP, is that LDP tend to over-perform when JCP is in the fray since in both situations the quality of the HP candidate are not great.   It seems that a key trend is that the marginal LDP HP JRP voters are fearful of JCP and would vote LDP vs a viable non-JCP opposition candidate when confronted with the possibility that the JCP candidate could win.  Looking at the different types of LDP vs HP vs CDP as well as LDP vs Ind(HP) vs JCP, the more threatening JCP looks (mostly due to the weakness of non-JCP opposition candidate either due to inexperience or being too right-win to win CDP PR voters) the more likely that marginal LDP JRP and HP voters would vote LDP.



If we look at the 4 districts where the CDP candidate is inexpereinced the PR vote were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
32.77%  11.34%  0.00%  0.22%  0.39%  4.91%  17.30%  23.06%  1.00%  8.33%  0.68%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   47.51%
HP                                             19.75%
CDP                                           23.40%
JCP                                             9.34%

The real result which was LDP 4 seats CDP 0 seats HP 0 seats was

LDP (plus HRP and minor Right)   52.75%
HP                                             17.88%
CDP                                           19.41%
JCP                                              9.96%

Here the poor quality of the CDP candidate means that there is no JCP PR tactical voting for the CDP candidate while CDP and HP loses some marginal JRP HP and even CDP marginal PR voters to LDP.  The combination of a weak non-JCP candidate plus JCP being in the fray gives a large LDP nonperformance.  Although the more Center-Left posture CDP means no large scale defections of CDP PR votes to JCP.
Logged
Kantakouzenos
Rookie
**
Posts: 74


Show only this user's posts in this thread
« Reply #680 on: November 17, 2017, 12:40:29 PM »

This is probably a really stupid question but I can't find an answer to it anywhere. 

Why was this election considered good for Abe if his party ended up losing seats? 
Logged
Oryxslayer
Atlas Icon
*****
Posts: 10,724


Show only this user's posts in this thread
« Reply #681 on: November 17, 2017, 01:07:52 PM »

This is probably a really stupid question but I can't find an answer to it anywhere. 

Why was this election considered good for Abe if his party ended up losing seats? 

Simple. For 90% of the campaign, Abe was expected to lose (much?) more.
Logged
Dereich
Moderators
YaBB God
*****
Posts: 4,892


Show only this user's posts in this thread
« Reply #682 on: November 17, 2017, 01:26:08 PM »

This is probably a really stupid question but I can't find an answer to it anywhere. 

Why was this election considered good for Abe if his party ended up losing seats? 

Simple. For 90% of the campaign, Abe was expected to lose (much?) more.

This, plus Abe ended up hobbling a rising star (Koike) instead of giving her time to assemble a party that could mount a credible challenge to the LDP in the future.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #683 on: November 17, 2017, 01:33:28 PM »

This is probably a really stupid question but I can't find an answer to it anywhere. 

Why was this election considered good for Abe if his party ended up losing seats? 

Well it is a couple of things

1) In July it seems Abe was on the ropes and perhaps on his way out in Sept 2018 if not earlier.  This election put those issues to rest.
2) 2/3 majorities are seen as abnormal in Japan.  Pre-1993 when the system was pseudo-PR it was almost impossible to achieve.  So retaining a 2/3 majority, even if it is with some asterisks is clearly a victory of sorts.  It is like if Bush got 57% of the vote in 1998 after Reagan got 59% in 1984 then it would be seen as a smashing victory given how abnormal what the result in 1984  was.  Of course with FPTP moving to a more FPTP system 2/3 majorities are now much simpler to achieve.  Many people have not caught on to that yet.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #684 on: November 19, 2017, 06:40:47 AM »
« Edited: November 20, 2017, 08:09:43 AM by jaichind »

The next batch of seat to look at is LDP vs JRP vs HP vs JCP.  There are 18 such districts  of them and in such a district you would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occasional HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*NPB
JRP = 0.5*NPD + JRP
HP =  HP + CDP + 0.5*NPB
JCP = SDP + JCP

The idea is the same as LDP vs HP vs JCP except now JRP will win the JRP PR vote instead of being split between LDP and HP.  

HP candidate quality will play a large role just like LDP vs HP vs JCP,  so I split the 18 seats into 4 types.  8 of them are where HP nominated an ex-DP candidate that has a track record of running in that district.  3 of them are experienced but center-right in their orientation and views.  3 of them are where HP nominated a complete political novice without any sort of real political experience.  4 of them are where HP nominated someone with DP background but without any real experience of running in said district.  

If we go with the 8 where HP nominated an ex-DP candidate that has a track record of running in that district.  The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
31.88%  10.73%  0.00%  0.00%  0.45%  8.66%  16.18%  21.71%  1.53%  8.86%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   43.06%
JRP                                             8.66%
HP                                             37.89%
JCP                                            10.40%

The real result which was LDP 6 seats HP 2 seats was

LDP (plus HRP and minor Right)   41.37%
JRP                                             9.95%
HP                                             34.70%
JCP                                            13.85%

Here the candidate quality of HP is high and comparable to the LDP candidate so what the LDP got mostly was as predicted.  HP under-performed because it seems despite good candidate quality some CDP PR voter defected to JCP.  JRP also managed to capture some vote beyond what it PR vote would imply from both LDP and HP.




If we then look at the 3 where HP nominated a candidate with experience in said district but with a Center-Right background.  The PR vote for these 3 districts were

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
29.60%  12.28%  2.55%  0.00%  0.42%  8.34%  15.37%  21.16%  1.07%  9.22%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   43.57%
JRP                                             9.61%
HP                                             36.53%
JCP                                            10.29%

The real result which was LDP 3 seats HP 0 seats was

LDP (plus HRP and minor Right)   44.25%
JRP                                            11.39%
HP                                             26.41%
JCP                                            17.96%

Like its equivalent for LDP vs HP vs JCP, a more center-Right experienced candidate failed to capture LDP PR voters instead bleeding HP support to LDP or JRP and instead lost a bunch of CDP voters to the JCP candidate.  In fact these trends might reinforce each other where JCP was expected to expand beyond the SDP-JCP base into the CDP vote which triggered HP tactical voting for LDP.




If we then look at the 3 where HP nominated a complete political novice without any sort of real political experience.  The PR vote for these 3 districts were

 LDP         KP        NPD      PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
33.86%  15.00%  0.00%  0.00%  0.51%  7.58%  17.10%  16.90%  1.25%  7.81%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   49.36%
JRP                                             7.58%
HP                                             34.00%
JCP                                             9.05%

The real result which was LDP 3 seats HP 0 seats was

LDP (plus HRP and minor Right)   54.50%
JRP                                            10.32%
HP                                             22.84%
JCP                                            12.43%

Here the fact the HP candidate is a political novice allows the LDP to way over-perform where there are significant HP PR voters defecting to LDP or JRP.  Also the poor candidate quality also means that a bloc of CDP PR voters went over to JCP but does not seem that much worse than with an experienced HP candidate.  




We then look at the 4 district where HP nominated someone with DP background but without any real experience of running in said district.  The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
31.15%  12.96%  0.00%  0.21%  0.45%  9.04%  19.12%  18.25%  1.37%  7.44%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   44.78%
JRP                                             9.04%
HP                                             37.36%
JCP                                             8.81%

The real result which was LDP 4 seats HP 0 seats was

LDP (plus HRP and minor Right)   46.44%
JRP                                            16.54%
HP                                             25.48%
JCP                                            11.54%

Here the CDP defection to JCP is also similar to when the HP candidate is experienced which shows that in these sorts of seats the defection from HP to JCP by CDP PR voters is mostly driven by ideology but not quality.   On the flip side the large defection of HP PR voters from HP to JRP seems to just has large as in the case as when HP ran an inexperienced candidate so candidate quality is a key driver here.

The main lessons of this set of seats is similar to LDP vs HP vs JCP.  Poor HP candidate quality drives votes toward LDP or JCP while one that has Right wing background tend to drive CDP voters toward JCP without getting anything in return from LDP or JRP PR voters.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #685 on: November 20, 2017, 08:14:42 AM »
« Edited: November 30, 2017, 07:19:27 PM by jaichind »

Similar to but different from LDP vs JRP vs HP vs JCP would be LDP vs JRP vs HP vs CDP.  There are 4 such districts  of them and in such a district you would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occasional HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*NPB
JRP = 0.5*NPD + JRP
HP =  HP
CDP = CDP + SDP + JCP + 0.5*NPB

Here HP is marginalized from both sides. JRP will capture the JRP PR vote and CDP will capture the CDP PR vote.  

The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
31.34%  12.40%  0.00%  0.42%  0.40%  7.61%  13.97%  24.52%  1.28%  8.06%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   44.56%
JRP                                             7.61%
HP                                             13.97%
CDP                                           33.86%

The real result which was LDP 4 seats CDP 0 seats was

LDP (plus HRP and minor Right)   44.67%
JRP                                             8.49%
HP                                             12.05%
CDP                                           34.61%

Even though LDP were destined to win these seats given the 3 way split of the non-JCP opposition, the vote share matched what the PR vote share would imply.  Here without the threat of a JCP candidate most marginal JRP HP PR voters mostly vote along partisan lines and not shifted to LDP to ensure JCP's defeat.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #686 on: November 20, 2017, 08:36:32 AM »
« Edited: December 12, 2017, 09:08:08 AM by jaichind »

Similar to but different from LDP vs HP vs JCP would be LDP vs JRP vs JCP.  There are 9 such districts in Osaka and 2 outside Osaka where JRP is running with support from HP.  You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*NPB
JRP = 0.5*NPD+ JRP +HP + 0.5*CDP + 0.5*NPB
JCP = 0.5*CDP + SDP + JCP

This is similar to LDP vs HP vs JCP except for JRP would capture all the JRP PR vote and the CDP PR vote would be split between JRP and JCP due to the fact that JRP is too Right-wing for the CDP PR voter and yet there is a need for tactical anti-LDP vote by the CDP PR voter.

The PR vote for the 9 seats in Osaka was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
28.03%  14.62%  0.00%  0.00%  0.33%  27.53%  6.33%  13.33%  0.77%  9.05%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   42.99%
JRP                                            40.53%
JCP                                            16.49%

The real result which was LDP 6 seats JRP 3 seats was

LDP (plus HRP and minor Right)   41.88%
JRP                                            40.73%
JCP                                            17.39%

The JRP candidates in Osaka were of high quality and since my model had LDP and JRP winning all the Right leaning PR voters (JRP and HP) anyway there were little reason for JRP to lose any votes assigned to them by my model.  The result was a near match and the district vote shares pretty much matches what my model expected.  Since these races were clearly LDP vs JRP with JRP having strong chance of winning there were little reason for marginal HP or JRP PR voters to defect to LDP to ensure the defeat of the JCP candidate.



The PR vote for the 2 seats outside of Osaka was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
35.87%  13.82%  0.00%  0.00%  0.63%  8.98%  15.22%  16.21%  2.54%  6.72%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   50.32%
JRP                                            32.31%
JCP                                            17.37%

The real result which was LDP 2 seats JRP 0 seats was

LDP (plus HRP and minor Right)   57.01%
JRP                                            28.45%
JCP                                            14.54%

Both districts are fairly noncompetitive and with JRP even with HP support with no real chance to beat LDP and with HP CDP and JRP PR voters defected to LDP as a result.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #687 on: November 23, 2017, 09:31:31 AM »
« Edited: December 12, 2017, 09:08:00 AM by jaichind »

Another bloc of 3 seats are LDP vs JRP vs CDP vs JCP.  

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*NPB
JRP = 0.5*NPD + JRP +HP + 0.5*NPB
CDP = CDP
JCP = SDP + JCP

All 3 seats are in Osaka so one would expect JRP to win the HP PR vote with CDP and JCP just winning the CDP PR and SDP + JCP PR votes respectively.

The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
28.06%  15.42%  0.00%  0.00%  0.31%  27.84%  5.72%  13.54%  0.65%  8.47%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   43.78%
JRP                                            33.56%
CDP                                           13.54%
JCP                                             9.11%

The real result which was LDP 3 seats JRP 0 seats was

LDP (plus HRP and minor Right)   42.08%
JRP                                            35.74%
CDP                                           13.51%
JCP                                             8.66%

JRP slightly over-performed what the PR vote would imply but otherwise the vote shares came out as expected.  Since these seats was basically LDP vs JRP and HP expected to go to JRP there was no tactical voting by marginal JRP or HP PR votes for LDP despite the fact that the JCP is on the ballot.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #688 on: November 23, 2017, 09:39:53 AM »
« Edited: December 12, 2017, 09:07:51 AM by jaichind »

Another bloc of 4 seats are LDP vs JRP vs CDP.   To be fair one of the CDP candidate is an Ind(CDP) candidate that is fairly high quality that I counted as CDP.

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + HRP + 0.5*NPB
JRP = 0.5*NPD + JRP +HP + 0.5*NPB
CDP = CDP + SDP + JCP

3 out of the 4 seats are in Osaka so one would expect JRP to win the HP PR vote and CDP to win CDP + SDP + JCP PR votes.

The PR vote was

 LDP         KP        NPD     PJK      HRP      JRP        HP         CDP       SDP      JCP      NPB
28.25%  14.31%  0.00%  0.00%  0.36%  21.49%  8.08%  18.57%  1.24%  7.69%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   42.93%
JRP                                            29.57%
CDP                                           27.51%

The real result which was LDP 1 seats CDP 3 seats was

LDP (plus HRP and minor Right)   37.95%
JRP                                            26.41%
CDP                                           35.63%

Without JCP running and CDP being a viable choice as a result there were clear tactical voting by LDP PR voters for CDP to stop JRP as well as by JRP-HP PR voters for CDP to stop LDP.  This result shows how brittle LDP is once an opposition candidate is of high quality and presents a choice for marginal LDP JRP and HP PR voters that ensures that JCP will not win no matter what.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #689 on: November 23, 2017, 09:47:36 AM »

I will be on a business trip to Asia for a couple of weeks which would include a stop of around 4 days in Japan.  Hopefully the 福岡(Fukuoka) prefecture election commission website is accessible to me from Japan so I can download the last of the PR vote info and complete my data analysis.  Right now the 福岡(Fukuoka) election commission website seems to block non-Japanese IP addressed.  I am too lazy to setup a VPN to get around this since I am visiting Japan soon anyway.

Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #690 on: November 30, 2017, 07:11:23 PM »
« Edited: December 12, 2017, 09:09:41 AM by jaichind »

We should now look at the 9 seats that the KP are running.  8 of them can be grouped into 3 KP vs JCP,  3 KP vs CDP, 2 KP vs CDP vs JCP.

For the 3 KP vs JCP you would expect the vote share to be (as an equation of PR vote)

KP = LDP + KP + NPD + PJK + HRP + NPB + JRP + HP
JCP = CDP + SDP + JCP

The idea here is that JCP wins the Center-Left vote and KP should win everything else given JRP and HP de facto support of KP.

The PR vote for KP vs JCP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP        JCP      NPB
26.74%  18.25%  0.00%  0.29%  0.33%  13.55%  10.93%  16.05%  0.94%  11.97%  0.95%
 
So using my equation one would expect

KP (plus HRP and minor Right)     71.04%
JCP                                            29.96%

The real result which was KP winning all 3 seats are

KP (plus HRP and minor Right)     63.07%
JCP                                            36.39%


 

For the 3 KP vs CDP you would expect the vote share to be (as an equation of PR vote)

KP = LDP + KP + 0.5*NPD + PJK + HRP + NPB + JRP
CDP = 0.5*NPD + HP + CDP + SDP + JCP

The idea here is that CDP will be able to pull in HP support but JRP still go with KP.  HP's backing of KP is implicit and most if not all HP PR voters will vote CDP while JRP backing of KP is explicit.

The PR vote for KP vs CDP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP        JCP      NPB
26.06%  17.85%  3.06%  0.00%  0.44%  16.45%  7.52%  19.30%  0.97%  8.34%  0.00%
 
So using my equation one would expect

KP (plus HRP and minor Right)     62.34%
CDP                                           37.66%

The real result which was KP winning all 3 seats are

KP (plus HRP and minor Right)     54.89%
CDP                                           45.11%
 


For the 2 KP vs CDP vs JCP you would expect the vote share to be (as an equation of PR vote)

KP = LDP + KP + 0.5*NPD+ PJK + HRP + NPB + JRP
CDP = 0.5*NPD + HP + CDP
JCP =  SDP + JCP

Here KP keeps the Center-Right votes while CDP and JCP splits the Center-Left vote.  

The PR vote for KP vs CDP vs JCP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP        JCP      NPB
27.09%  17.32%  0.00%  0.00%  0.47%  20.35%  8.36%  15.05%  0.85%  10.51%  0.00%
 
So using my equation one would expect

KP (plus HRP and minor Right)     65.23%
CDP                                           23.41%
JCP                                            11.36%

The real result which was KP winning all 2 seats are

KP (plus HRP and minor Right)     54.73%
CDP                                           24.88%
JCP                                            20.39%

One can tell by looking at these 8 seats that the social stigma of KP is even worse than JCP.  There seems to be large defection from LDP and JRP PR voters away from KP and even toward JCP.  The KP  vs JCP match-up is telling that that even JCP over-performed fundamentals even as JCP has it own social branding issues.
Logged
Lachi
lok1999
Sr. Member
****
Posts: 3,347
Australia


Political Matrix
E: -1.06, S: -3.02

Show only this user's posts in this thread
« Reply #691 on: November 30, 2017, 07:45:27 PM »

I have a question.

Why was there a CDP -> Kibo defection?
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #692 on: November 30, 2017, 08:05:52 PM »

I have a question.

Why was there a CDP -> Kibo defection?

None of the blocs of seats I looked at has CDP PR ->HP cross voting at an aggregate level.  There are cases where that took place in individual seats and usually it involves a very weak CDP candidate whch leads to some tactical voting by CDP PR votes for HP which on average are fairly weak as well in seats where HP has to face CDP.   This is the case because the "quality" ex-DP candidate in a HP vs CDP seat usually went with CDP and not HP.  Where a quality ex-DP candidate went with HP, CDP would not bother with running a candidate.  CDP was formed to support ex-DP candidates that HP refused to pick up.
Logged
Lachi
lok1999
Sr. Member
****
Posts: 3,347
Australia


Political Matrix
E: -1.06, S: -3.02

Show only this user's posts in this thread
« Reply #693 on: December 01, 2017, 06:36:34 PM »

I have a question.

Why was there a CDP -> Kibo defection?

None of the blocs of seats I looked at has CDP PR ->HP cross voting at an aggregate level.  There are cases where that took place in individual seats and usually it involves a very weak CDP candidate whch leads to some tactical voting by CDP PR votes for HP which on average are fairly weak as well in seats where HP has to face CDP.   This is the case because the "quality" ex-DP candidate in a HP vs CDP seat usually went with CDP and not HP.  Where a quality ex-DP candidate went with HP, CDP would not bother with running a candidate.  CDP was formed to support ex-DP candidates that HP refused to pick up.
No, not pre election, there has been a post election defection from the CDP to Kibo.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #694 on: December 02, 2017, 09:51:05 PM »

I have a question.

Why was there a CDP -> Kibo defection?

None of the blocs of seats I looked at has CDP PR ->HP cross voting at an aggregate level.  There are cases where that took place in individual seats and usually it involves a very weak CDP candidate whch leads to some tactical voting by CDP PR votes for HP which on average are fairly weak as well in seats where HP has to face CDP.   This is the case because the "quality" ex-DP candidate in a HP vs CDP seat usually went with CDP and not HP.  Where a quality ex-DP candidate went with HP, CDP would not bother with running a candidate.  CDP was formed to support ex-DP candidates that HP refused to pick up.
No, not pre election, there has been a post election defection from the CDP to Kibo.

I do not see it.  So far the only defection to HP is ex-DP leader 前原 誠司(Maehara Seiji) that ran as a pro-HP independent (IND(HP)) and then joined HP after the election.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #695 on: December 10, 2017, 09:44:31 PM »
« Edited: December 12, 2017, 09:12:34 AM by jaichind »

Now we have a bloc of seats which are 1-on-1 against LDP.  First we have LDP vs JCP where pretty much all of them are in LDP strongholds and all the other opposition parties will just let JCP go and run against LDP.  There are 11 of them in all.

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + NPD + PJK + JRP + HP
JCP = CDP + SDP + JCP

The ideal here is JCP gets the Center-Left and Left vote and LDP wins everything else.

The PR vote for LDP vs JCP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
36.92%  15.46%  1.74%  0.12%  0.77%  4.38%  14.95%  16.50%  1.84%  7.32%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   74.34%
JCP                                            25.66%

The real result which was LDP winning all 11 seats are

LDP  (plus HRP and minor Right)  75.93%
JCP                                            24.07%

So other than a minor defect of the CDP PR vote over the LDP it pretty much worked out as as the PR section would imply.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #696 on: December 10, 2017, 09:48:58 PM »
« Edited: December 12, 2017, 09:12:00 AM by jaichind »

Then we have LDP vs SDP where just like LDP vs JCP they are in lean LDP strongholds and all the other opposition parties will just let SDP go and run against LDP with explicit backing from JCP.  There are 5 of them in all.

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + NPD + PJK + JRP + 0.5*HP
SDP = 0.5*HP + CDP + SDP + JCP

The ideal here is SDP gets the Center-Left and Left vote as well as the anti-LDP part of the HP vote and LDP wins everything else.

The PR vote for LDP vs SDP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
39.89%  13.20%  0.00%  0.00%  0.77%  4.02%  15.55%  13.65%  7.79%  5.12%  0.00%
 
So using my equation one would expect

LDP (plus HRP and minor Right)   65.66%
SDP                                           34.34%

The real result which was LDP winning all 5 seats are

LDP  (plus HRP and minor Right)  67.65%
SDP                                           32.35%

So other than a minor defect of the HP PR vote over the LDP it pretty much worked out as as the PR section would imply.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #697 on: December 12, 2017, 09:15:34 AM »
« Edited: December 12, 2017, 11:12:11 PM by jaichind »

We then have the 3 seats which is LDP vs HP.  

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + 0.5*JRP
HP = 0.5*NPD + 0.5*JRP+ HP + CDP + SDP + JCP

The idea here is the HP should be able to all PR votes HP and Left of HP and split JRP and NDP vote with LDP.

The PR vote for LDP vs HP was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
37.73%  11.84%  0.00%  0.30%  0.71%  3.17%  22.26%  15.87%  2.07%  6.05%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   52.17%
HP                                             47.83%

The real result which was LDP 3 HP 0 seats are

LDP  (plus HRP and minor Right)  54.95%
HP                                             45.05%

It seems HP under-performed but that most likely is a function of the fact there are only 3 seats in this bloc and 1 of them is in 岩手(Iwate) where the HP candidate has a LP background and there seems to be some anti-Ozawa tactical voting by anti-Ozawa but Center-Left PR voters.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #698 on: December 15, 2017, 07:48:15 AM »

We also have 12 seats which are LDP vs CDP.  8 out of the 12 are with an experienced CDP candidate in said district and 4 of them are with an inexperienced CDP candidate in said district.

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + JRP
CDP = 0.5*NPD + HP + CDP + SDP + JCP

The idea here with CDP's Center-Left image the JRP PR voter would go LDP but NPD PR voter would still split their vote.  HP base which is mostly ex-DP would go with CDP.

If we go with the 8 districts where the CDP has an experienced candidate the PR vote was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
32.23%  11.13%  4.47%  0.00%  0.63%  2.99%  13.01%  25.93%  2.41%  7.20%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   49.21%
CDP                                           50.79%

The real result which was LDP 3 CDP 5 seats are

LDP  (plus HRP and minor Right)  49.94%
CDP                                           50.06%



If we go with the 4 districts where the CDP has an inexperienced candidate the PR vote was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
36.30%  13.73%  2.18%  0.00%  0.64%  3.59%  12.38%  23.11%  1.47%  6.60%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   55.34%
CDP                                           44.66%

The real result which was LDP 4 CDP 0 seats are

LDP  (plus HRP and minor Right)  58.99%
CDP                                           41.01%

So for districts where the CDP candidate quality was high LDP and CDP preformed mostly at par as implied by the PR vote.  For district with inexperienced CDP candidate there is some defection of most likely are HP PR voters over to LDP.  This is expected and fairly logical.
Logged
jaichind
Atlas Star
*****
Posts: 27,151
United States


Political Matrix
E: 9.03, S: -5.39

Show only this user's posts in this thread
« Reply #699 on: December 19, 2017, 12:08:32 PM »

We also have 18 seats which are LDP vs IND(OPPN).  15 out of the 18 are with an experienced IND(OPPN) (mostly ex-DP) candidate in said district and 3 of them are with an inexperienced IND(OPPN) candidate in said district.

You would expect the vote share to be (as an equation of PR vote)

LDP (LDP plus the occational HRP or minor Right in candidate) = LDP + KP + 0.5*NPD + PJK + 0.5*JRP
IND(OPPN) = 0.5*NPD + 0.5*JRP + HP + CDP + SDP + JCP

The idea here is that LDP will get the LDP ally and Right wing parties while IND(OPPN) will get all opposition party support with the two sides splitting the NPD and JRP PR votes.

If we go with the 15 districts where the IND(OPPN) has an experienced candidate the PR vote was

 LDP         KP        NPD     PJK      HRP       JRP        HP         CDP       SDP      JCP      NPB
32.69%  12.33%  0.46%  0.41%  0.71%  3.87%  16.45%  21.82%  3.18%  8.09%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   48.29%
IND(OPPN)                                 51.71%

The real result which was LDP 3 IND(OPPN) 12 seats are

LDP  (plus HRP and minor Right)  46.60%
IND(OPPN)                                 53.40%


If we go with the 3 districts where the IND(OPPN) has an inexperienced candidate the PR vote was

 LDP         KP        NPD     PJK      HRP       JRP        HP       CDP       SDP      JCP      NPB
37.72%  8.94%  0.00%  0.00%  0.74%  3.58%  16.73%  22.62%  2.44%  7.23%  0.00%

So using my equation one would expect

LDP (plus HRP and minor Right)   49.18%
IND(OPPN)                                 50.82%

The real result which was LDP 3 IND(OPPN) 0 seats are

LDP  (plus HRP and minor Right)  57.34%
IND(OPPN)                                 42.66%


So just like LDP vs CDP, experienced IND(OPPN) candidates over-performed what the PR vote would imply while inexperienced ones significantly under-performed.   
Logged
Pages: 1 ... 23 24 25 26 27 [28] 29  
« previous next »
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.102 seconds with 12 queries.