some of the races

Today I’m going to write a bit about some of the more interesting district races.  If I have time, I might also say something about the DPP’s party list.  However, since I generally don’t like to react too quickly to news events, I might leave that to a later date.

 

 

Taipei City 4

This is a district with a clear and consistent Blue majority.  In a one on one race, the KMT should always win this district.  However, the Blue camp may or may not end up with one candidate.

The KMT incumbent is Tsai Cheng-yuan 蔡正元, and he wants to run for re-election.  However, he was convicted of embezzlement and, according to the KMT rules, he was ineligible to contest the KMT’s primary.  Tsai never wavered in his determination to seek re-election.  He covered the district with ads trumpeting all his achievements, and it seemed quite likely to me that if the KMT nominated someone else he might still run as an independent.

Two KMT city council members openly considered contesting the nomination.  Wu Shih-cheng 吳世正 eventually dropped out, and Li Yan-hsiu 李彥秀 was the only person to formally register for the nomination.  According to the KMT rules, if there is only one candidate, a telephone survey is still necessary.  That candidate must win at least 40% support to get the nomination.  40% is quite a high threshold.  In several districts, we have seen “I don’t know” or “none of the above” win the race, but that didn’t affect the results with two or more candidates.  In a single candidate race, it was no easy task for Li to get 40% support, especially with Tsai looming in the background.  So Li and Tsai both campaigned hard (even though Tsai wasn’t actually in the contest).  From his point of view, he didn’t want the KMT to nominate anyone.  If they didn’t have a nominee, they might eventually just have left the district open for him as an independent ally.  On the other hand, if the KMT nominated Li, Tsai would be faced with expulsion from the party and a very tough three way race.  In this scenario, the most likely outcome may have been a DPP win.  The green camp should probably get somewhere around 40-45%, and it seems unlikely that either Tsai or Li would have been able to push the other one under 15%.  In the event, Li managed to pass the threshold fairly convincingly, with nearly 50% support.  It seemed we were headed into that three way race.

However, something interesting happened.  The KMT did not immediately announce Li’s nomination (as they did in several other districts right after the survey was completed).  Instead, they waited.  And then the court ruled on the appeal to Tsai’s court case and declared him not guilty.  Suddenly, there were no obstacles to Tsai receiving the KMT nomination.  And immediately, the KMT started to prevaricate, sending out suggestions that they might rerun the primary.  Li wanted no part of this.  From her point of view, she had played by the rules and won, and now she wanted the nomination.  However, the KMT eventually announce that it would hold a new survey with both Li and Tsai.  Again, this was intensely contested, and Tsai eventually won by about 1%.  This time, the KMT immediately announced the results and officially nominated Tsai.  It seems pretty clear that the KMT really wanted Tsai to win the nomination all along.

We haven’t heard much from Li since the contest ended.  There are clearly lots of bad feelings between her and Tsai, but the question is whether she will run against him in the general election.  I think the odds are against it.  She would have to quit the KMT and run against the party nominee in a single seat race, which is no easy task for even the strongest incumbents, much less a city council member.

But wait, there’s more.  After Tsai’s nomination was announced, he was convicted in another case.  This hasn’t inspired the KMT to retract his nomination, but it would give any other blue candidate more ammunition in the general election.

Even if Li decides not to run, there is another possible blue candidate lurking in the wings.  The PFP has announced its intentions to run candidates in several districts as a response the KMT’s blatant efforts to marginalize it.  In 2008, the KMT gave the PFP four spots on its party list and nominated a few PFP figures in the districts.  This year, the KMT has given the PFP almost nothing.  So the PFP is responding by threatening to run candidates, who might draw enough votes to cause the KMT candidates to lose.  One of the PFP’s best politicians is from Taipei 4.  City council member Huang Shan-shan 黃珊珊 is a very respected lawyer.  She also has no ethics questions surrounding her, unlike the KMT nominee.  It is quite easy to imagine her drawing 15-20% of the vote if she runs.  Of course, it is not clear as of now that she will eventually run; the PFP-KMT bargaining game is still playing out.

So we might eventually end up with a straight one on one race, in which case the KMT should win easily.  We also might end up with a three way race, in which the DPP might be able to steal this seat.  Either way, it has already been more dramatic than expected.

 

Taoyuan 5

Taoyuan 5 is not merely a solidly blue district, it is closer to impregnable.  Even in a three way race with the two blue candidates evenly dividing up their votes, the DPP might not have enough votes to win.  So again, the drama is all in the blue camp.

Chu Feng-chih 朱鳳芝 is the incumbent.  She was first elected in 1989 and is now the second most senior person in the entire legislature.  (Speaker Wang Jin-pyng is far ahead of everyone else; he was elected in 1975.)  She came out of the Huang Fu-hsing party branch (the branch for military veterans), and Taoyuan 5 has plenty of retired soldiers, mainlanders, and loyal KMT supporters, so this district should fit Chu very well.  However, much to everyone’s surprise, Chu did not win the party primary.  Instead, the mayor of Pingzhen Township, Chen Wan-de 陳萬得, beat her in the telephone survey.  The KMT duly nominated Chen, and Chu’s only hope to prolong her career seemed to be in convincing the KMT to put her on the party list.

However, Chen was subsequently convicted in a civil case over a financial dispute, and he eventually gave up his nomination.  (One imagines heavy party pressure behind the scenes.)  The KMT has not yet announced a new nominee or even how it will decide on its new nominee.

Chu is back in the picture.  However, she has made it clear that she will accept the KMT’s nomination only if they simply give it to her.  She will not participate in any contest for the nomination.  The KMT seems to want to nominate her, but the fact that she participated in and lost the primary is inconvenient.

Chen also hasn’t completely given up.  He might not be running, but he has chosen that old chestnut of Taiwanese electoral politics: ask voters to prove your innocence by voting for your wife.  In other words, while Chen is too tainted to be presentable, his wife is perfectly clean and innocent.  (Note: please reread the previous sentence with a heavy dose of sarcasm.  Much better.)

No matter how this race turns out, the seat will probably end up in the blue camp’s hands.  In fact, the DPP hasn’t even bothered to nominate anyone yet.

 

Taoyuan 3

Taoyuan 3 has a similar partisan distribution to Taoyuan 5: the blue camp should always win this race, even with two blue candidates.  Inexplicably, the incumbent is a DPP member.  As readers doubtless remember, the DPP won this district in early 2010 in a by-election.  That by-election was, in many ways, the perfect storm.  The KMT nominated an outsider, Apollo Chen 陳學聖， without extensive ties in the district and who had not won the (advisory) surveys conducted prior to the election.  A local politician who had wanted the nomination ran against him and took a fair share of votes.  The DPP nominated a local politician who had extensive local ties but not enough partisan coloration to arouse the loyal blue voters to action.  And in a by-election, the KMT had a very difficult time motivating voters to get out and vote.  With the low turnout, a split KMT vote, and a weak KMT candidate, the DPP candidate somehow squeezed out a victory.  My immediate reaction: enjoy the rest of this term because the DPP will never be able to hold this seat.

A year and a half later, there is an outside chance that the DPP might hold Taoyuan 3.  It is not that the DPP has suddenly become more popular in the district, but that the KMT seems to be repeating all its mistakes from the by-election.  Apollo Chen (and yes, I feel silly every time I write the name Apollo), unbowed by his humiliating defeat in 2010, decided to try again.  His main opponent this time was no mere local politician, but Cheng Chin-ling 鄭金玲, who has served in the Provincial Assembly and legislature since 1994.  Chen managed to beat Cheng in the telephone surveys, but Cheng announced immediately that she would run in the general election, where she will doubtless be a formidable opponent.

(In a somewhat surprising way, Chen might be the beneficiary of the district lines drawn by the DPP that weakened the KMT in this district.  Chen is an interloper; he started his career in Taipei City and only moved to Taoyuan after losing his re-election bid to the legislature in 2004.  He threw his lot in with Eric Chu, who was then county executive.  However, when Chu moved on to the vice-premiership and mayor of New Taipei City, Chen stayed behind in Taoyuan.  I’m guessing his core support is a mixture of loyal KMT voters and people involved in the business-construction state.  Cheng, on the other hand, is a military politician.  She, like Chu Feng-chih, draws her support from military veterans, their families, and mainlanders.  Taoyuan 3 is centered on Zhongli City, which has a large military presence.  However, Zhongli is too big to be a single district, so it had to be divided.  The DPP plan cut many of these military areas off and put them in the neighboring Taoyuan 6.  The KMT still has overwhelming support in the remaining areas of Zhongli, but the proportion of military veterans and mainlanders is significantly lower.  If the KMT redistricting plan, which left the military areas in district 3, had been adopted, Cheng almost certainly would have been able to beat Chen for the nomination.  On the other hand, the DPP would never have won the by-election in the first place, so Chen might already be the incumbent.)

I haven’t heard anything about this race in a month or two, so I assume Cheng still plans to run.  She will be a stronger opponent than the local politician who split the KMT vote in the by-election, but I think it is still highly unlikely that the DPP can steal this election again.  First, the turnout rate will be higher.  Second, the stakes will be higher.  In the by-election, only four seats were up for election, and there was no question of the KMT losing power.  That will not be so clear this time.  Third, the concurrent presidential election will polarize the electorate along party lines.  All of this works against the DPP in Taoyuan 3.  On the other hand, the fact that there is even a sliver of hope in this district is a major victory for the DPP.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

KMT telephone primary rules

The KMT recently announced rules for its telephone surveys.  85% of the result will be based on comparisons between each of the KMT candidate and the DPP nominee.  Green camp supporters will be included in these calculations.  The other 15% will be determined by a direct comparison among the KMT contestants.  In this section, green camp supporters’ opinions will not be considered.

This controversy is being reported in the context of  Taipei City 3 (Zhongshan-Songshan), and I’m not sure if it applies to all races, only races in Taipei City, or just this race.  Luo Shu-lei 羅淑雷 expressed satisfaction with these new rules.  She expects to do better among the undecideds and green supporters, and was against any efforts to exclude their opinions.

manipulation and DPP telephone surveys

Right now there is a controversy stewing over the DPP’s telephone polls for the presidential primary.  It seems that some people (and most people are pointing at Tsai Ing-wen’s supporters) are telling their supporters to only support Tsai.  Other people seem to think that this is a betrayal of democratic ideals.  I think there are (at least) two ways to think about this.

But first, let’s look at the rules.  According to the DPP rules, presidential nomination polls must compare the DPP contestants to the KMT candidate, not to each other.  So each respondent is asked for his or her preference on Su vs. Ma, Tsai vs. Ma, and Hsu vs. Ma.  The respondents are not asked which of the three DPP contestants they like the best.  If only one DPP contestant beats Ma, that person wins the nomination.  If more than one beats Ma or if no one beats Ma, then the person with the highest percentage of supporters wins.  If there are two people tied, then the person with the largest advantage (or smallest deficit) over Ma wins.  However, ties are extremely unlikely since they will figure each person’s support out to four decimal places.

Let’s think about this from a DPP supporter’s point of view.  You prefer Su to Ma, and you prefer Tsai to Ma.  Essentially, your response has no impact at all on the result of the nomination contest because you have raised both candidates’ support by an equal amount.  However, what if you strongly prefer one to the other?  Without loss of generality, let’s suppose you strongly prefer Tsai to Su, but you also prefer Su to Ma.  The only way you can help Tsai to win the nomination is to not express support for Su.  Is this unethical?  I don’t think so.  Every election has strategic voting, and this is just another form of strategic voting.  Moreover, these strategic voters don’t have to actually support Ma.  They merely decline to answer the Ma vs. Su question.  This is sufficient to make their preferred outcome more likely.

From the DPP’s point of view, things are a bit murkier.  The DPP rules are set up to privilege moderate swing voters, not their core party supporters.  That is, the DPP made a strategic choice to let swing voters choose their party nominee because that should maximize their chances of winning the general election.  However, their core supporters might not appreciate effectively being told that they don’t matter, and those supporters might respond strategically, as described above, in an effort to play a more decisive role in the nomination process.  So party supporters are effectively subverting the party’s strategic decision to marginalize them.  This does not seem surprising to me.  Why should the people who care most passionately about the DPP not be intensely interested in who it nominates?  I think the DPP rules create incentives that most supporters should not be terribly happy with.

There is something else going on that I find much more interesting.  One DPP member is accusing another of instructing supporters not only to answer that they only support Tsai, but also to misrepresent their age.  Why would you want to misrepresent your age?  This has to do with survey methodology.  Your goal is to infer from the survey respondents what the general population looks like.  Unfortunately, your survey sample (the people you actually interview) rarely looks exactly like the general population.  On some variables, such as age, sex, and geographical distribution, we have very good statistics about what the population looks like.  So we have a pretty good idea if there are too many men or too many senior citizens in the sample.  The usual way to deal with this is to weight the sample.  Suppose people 60 and over are 10% of the total population and you interview 1000 people.  Your sample should have 100 respondents aged 60 and up.  However, suppose you actually only interview 80 such people.  The idea behind weighting is to inflate these 80 people so that they represent the 100 people you should have.  So you multiply each of them by 1.25.  Likewise, if you are supposed to have 200 people aged 30-40 and your sample actually includes 250 such people.  You would multiply each of them by 0.8.  So the devious strategy is to lie about you age and put yourself into one of the chronically underrepresented categories (which if memory serves me correctly are almost always 20-30 and 60+) so that your answer gets inflated, not deflated.

Well, now this is blatant manipulation.  However, there really isn’t much the DPP can do about it.  Once the public learns this trick, the only thing the DPP can do is to stop weighting, which makes the overall results less accurate.  Eventually, I wonder if this (as well as the feeling that strategic voting is immoral or otherwise undesirable) won’t be the eventual catalysts for the DPP (and maybe the KMT) to move away from telephone surveys.

KMT Taichung CC nominations

From a few news stories on the internet, I have put together the results of the KMT’s telephone survey primary for nominations to the Taichung City Council.  I could not find the exact results; all I could find was who won and who lost.  However, one story reporting on the portion of the primary in the old Taichung City listed ranks for the winners (though not for the losers).

Also, a story on the portion in the old Taichung County provided the current jobs of many candidates.  This is interesting to me.  There is a pretty clear hierarchy of positions.  From best to worst, the ranking is township mayor, county assembly (ie: incumbent), speaker of township assembly, regular member of township assembly.

Finally, we are starting to get into the very bloody portion of the election.  Taipei City actually has more seats than before, while Xinbei City has the same number.  Taichung, Tainan, and Kaohsiung all have a lot of incumbents chasing a small number of seats.  This is not as bad for the DPP, which has historically been underrepresented in these lower-level elections and therefore still has a little bit of space to grow.  However, it is a bloodbath for the KMT.  I count 14 incumbents who did not win nominations in Taichung.  I’m guessing that most of them will quit the KMT and run for re-election as independents

Why would KMT primary losers be less likely to respect survey results than DPP losers?  KMT politicians rely much more heavily on organization, which may or may not include vote buying.  This type of mobilization doesn’t affect the results of telephone surveys very much.  As such, losers can argue that there is a significant gap between these results and the likely results of a general election, when their organizations will come into play.

		spots	name	name	F	Current	Win
	Taichung
1	Dajia	3	姚應龍	Yao Yinglong		incumbent	Y
			林素真	Lin Suzhen	F	Incumbent	Y
			楊永昌	Yang Yongchang		Incumbent
			李鴻榮	Li Hongrong		incumbent	Y
.
2	Qingshui	3*	楊秋雲	Yang Qiuyun	F	incumbent	Y
			蘇麗華	Su Lihua			Y
			顏永滄	Yan Yongcang			Y
			尤璧鈴	You Biling	F	incumbent	?
.
3	Dadu	3*	林汝洲	Lin Ruzhou		Dadu mayor	Y
			吳瓊華	Wu Qionghua	F	Incumbent	Y
			楊忠諺	Yang Zhongyan		incumbent
			謝蒼海	Xie Canghai		Former Wuri mayor
			林士昌	Lin Shichang		incumbent	Y
			何端格	He Duange		incumbent
.
4	Fengyuan	4*	王朝坤	Wang Chaokun		Houli township assembly speaker	Y
			陳本添	Chen Bentian		Incumbent	Y
			張溢城	Zhang Yicheng		Incumbent	Y
			陳清龍	Chen Qinglong		Incumbent
			車淑娟	Che Shujuan	F	incumbent	Y
			劉重迪	Liu Chongdi		Fengyuan township assembly member
.
5	Tanzi	4*	張立傑	Zhang Lijie		Incumbent
			王永通	Wang Yongtong		incumbent
			吳顯森	Wu Xiansen		Daya mayor	Y
			王加佳	Wang Jiajia	F	Incumbent	Y
			羅永珍	Luo Yongzhen	F	Shengang mayor	Y
			何秀香	He Xiuxiang
			賴朝國	Lai Chaoguo		incumbent	Y
.
6	Xitun	3*	嚴榮發	Yan Rongfa
			留峰甫	Liu Fengfu
			黃馨慧	Huang Xinhui	F	Incumbent	Y(1)
			張廖乃倫	Zhang Liao Nailun	F	Incumbent	Y(3)
			吳春夏	Wu Chunxia		Incumbent
			楊正中	Yang Zhengzhong		incumbent	Y(2)
			陳富德	Chen Fude
.
7	Nantun	3*	丁振嘉	Ding Zhenjia		Incumbent	Y(3)
			劉士州	Liu Shizhou		Incumbent	Y(2)
			朱暖英	Zhu Nuanying	F		Y(1)
			陳三井	Chen Sanjing		Incumbent
			黃淑芬	Huang Shufen	F	incumbent
.
8	Beitun	4*	賴順仁	Lai Shunren		Incumbent	Y(4)
			陳成添	Chen Chentian		Incumbent	Y(2)
			唐國泰	Tang Guotai		Incumbent
			林永能	Lin Yongneng		Incumbent
			謝黎芳	Xie Lifang
			沈佑蓮	Shen Youlian	F	incumbent	Y(1)
			吳敏	Wu Min			Y(3)
.
9	North	2	陳天汶	Chen Tianwen		Incumbent	Y(1)
			陳有江	Chen Youjiang		Incumbent	Y(2)
			顏志修	Yan Zhixiu		incumbent
.
10	CW	2	柯貞竹	Ke Zhezhu
			洪嘉鴻	Hong Jiahong		Incumbent	Y(2)
			張宏年	Zhang Hongnian		incumbent	Y(1)
.
11	SE	3*	顏明毅	Yang Mingyi
			賴頤年	Lai Yinian		Incumbent	Y(1)
			林珮涵	Lin Peihan	F	Incumbent	Y(3)
			李中	Li Zhong		incumbent	Y(2)
.
12	Taiping	3*	李麗華	Li Lihua	F	Incumbent	Y
			詹敏豐	Zhan Minfeng		Incumbent	Y
			賴瑞珠	Lai Ruizhu	F	incumbent	Y
.
13	Dali	4*	戴萬福	Dai Wanfu		incumbent	Y
			蔡黃金雀	Cai Huang Jinque	F	Incumbent	Y
			蘇柏興	Su Boxing		Dali township assembly speaker	Y
			陳玉雪	Chen Yuxue	F	Incumbent
			林碧秀	Lin Bixiu	F	incumbent	Y
			曾瑞昌	Zeng Ruichang		Former Wufeng mayor
.
14	Dongshi	2	陳萬通	Chen Wantong		incumbent	Y
			蘇慶雲	Su Qingyun		Incumbent	Y
			冉齡軒	Ran Lingxuan	F	incumbent
.
15	PA
.
16	MA

* includes at least one female

Note: In D2, no survey was held, so apparently You Biling withdrew.

DPP city council nominations

The following are the results of the DPP’s primary process.  As explained in the previous post, the DPP used telephone surveys to determine its nominees in all contested elections.  The results of the telephone surveys are listed in the column labeled “poll.”  Before each district, I note how many seats there are in each district (m), how many people the DPP planned to nominate (nominate), and how many of these nominations are reserved for women.  In the table, I have each candidate’s name in Chinese and English, whether they are female, whether they qualified as a new candidate, whether they are an incumbent, their poll score, and whether they won a nomination.  The DPP considered any candidate as “new” who had never previously won public office (not including the extremely grassroots offices such as township assembly or village head).  New candidates were given a 10% bonus on their survey score.  This bonus is already figured into the numbers reported here.  The data on sex is incomplete.  I only know the gender of incumbents and nominees.  That is, if a candidate did not win a nomination this year and is not already an incumbent, lack of an “F” in the sex column does not necessarily mean that the candidate is a man.

As far as I can tell, only one person won a nomination due to the 10% bonus for new candidates (Qiu Tingwei, Xinbei 3).  Five women won due to the female reserve clause.  Note that this does not necessarily mean that they would not have been able to win without the clause.  For example, Yan Shengguan was the only woman in Taipei 5.  Since the DPP reserved a nomination for a woman in that district, Yan was guaranteed to win and did not need to expend any effort.  In fact, she very nearly won a nomination on the strength of her survey results.

Taipei City 1 (Shilin, Beitou)

m=12               nominate=open           Female=

Name	name	Fem	New	Inc	poll	win
莊瑞雄	Zhuang Ruixiong			Y		Y
吳思瑤	Wu Siyao	F		Y		Y
陳碧峰	Chen Bifeng			Y		Y
藍世聰	Lan Shicong					Y
陳正德	Chen Zhengde					Y
林世宗	Lin Shizong					Y
何志偉	He Zhiwei					Y

Taipei City 2 (Neihu, Nangang)

m=9                 nominate=4                 Female=1

name	name	Fem	New	Inc	poll	win
江志銘	Jiang Zhiming		Y		.1967	Y
王孝維	Wang Xiaowei			Y	.1880	Y
高嘉瑜	Gao Jiayu	F	Y		.1763	Y
李建昌	Li Jianchang			Y	.1530	Y
許嘉恬	Xu Jiatian	F	Y		.1283
林明源	Lin Mingyuan		Y		.0950
陳秀惠	Chen Xiuhui				.0628

Taipei City 3 (Songshan, Xinyi)

m=10               nominate=5                 Female=1

name	name	Fem	New	Inc	poll	win
張茂楠	Zhang Maonan			y	.2882	Y
洪健益	Hong Jianyi			Y	.2315	Y
許淑華	Xu Shuhua	F		y	.1932	Y
許家蓓	Xu Jiabei	F	Y		.1169	Y
陳淑華	Chen Shuhua	F			.1043	Y
葉問	Ye Wen		Y		.0300
陳泰源	Chen Taiyuan		Y		.0360

Taipei City 4 (Zhongshan, Datong)

m=10               nominate=5                 Female=1

name	name	Fem	New	Inc	poll	win
簡余宴	Jian Yuyan	F		Y	.3188	Y
李文英	Li Wenying	F		Y	.1540	Y
王世堅	Wang Shijian				.1534	Y
黃向群	Huang Xiangqun			Y	.1282	Y
梁文傑	Liang Wenjie		Y		.0971	Y
蔡易餘	Cai Yiyu		Y		.0949
朱政騏	Zhu Zhengqi		Y		.0330
許界元	Xu Jieyuan		Y		.0208

Taipei City 5 (Wanhua, Zhongzheng)

m=8                 nominate=4                 Female=1

name	name	Fem	New	Inc	poll	win
劉耀仁	Liu Yaoren			Y	.2362	Y
童仲彥	Tong Zhongyan		Y		.2267	Y
周威佑	Zhou Weiyou			Y	.1933	Y
陳嘉銘	Chen Jiaming			Y	.1628
顏聖冠	Yan Shengguan	F		Y	.1528	Y
周永鴻	Zhou Yonghong		Y		.0281

Note: Yan Shengguan won due to the female reserve clause.

Taipei City 6 (Da’an, Wenshan)

m=13               nominate=5                 Female=1

name	name	Fem	New	Inc	poll	win
周柏雅	Zhou Boya			Y	.3372	Y
徐佳青	Xu Jiaqing	F		Y	.2272	Y
李慶鋒	Li Qingfeng			Y	.1258	Y
阮昭雄	Ruan Zhaoxiong		Y		.1154	Y
柯景昇	Ke Jingsheng				.0792	Y
江蓋世	Jiang Gaishi				.0690
林一方	Lin Yifang		Y		.0463

Xinbei City 1 (Danshui, Sanzhi, Shimen, Bali)

m=3                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
呂子昌	Lu Zichang			Y		Y

Xinbei City 2 (Xinzhuang, Linkou, Taishan, Wugu)

m=10               nominate=5                 Female=1

name	name	Fem	New	Inc	poll	win
張晉婷	Zhang Jinting	F	Y		.1806	Y
何淑峰	He Shufeng	F		Y	.1767	Y
賴秋媚	Lai Qiumei	F	Y		.1633	Y
陳科名	Chen Keming			Y	.1525	Y
陳文治	Chen Wenzhi			Y	.1476	Y
張嘉玲	Zhang Jialing		Y		.1270
陳明卿	Chen Mingqing		Y		.0524

Xinbei City 3 (Sanchong, Luzhou)

m=9                 nominate=6                 Female=1

name	name	Fem	New	Inc	poll	win
李余典	Li Yudian				.1317	Y
李倩萍	Li Qianping	F		Y	.1246	Y
李坤城	Li Kuncheng		Y		.1233	Y
陳啟能	Chen Qi’neng			Y	.0957	Y
林克欣	Lin Kexin		Y		.0896	*
邱婷蔚	Qiu Tingwei	F	Y		.0836	Y
鄭金隆	Zheng Jinlong			Y	.0785	Y
李世東	Li Shidong		Y		.0779
林秋貴	Lin Qiugui		Y		.0712
黃秀玉	Huang Xiuyu		Y		.0643
李友親	Li Youqin			Y	.0595

Note: Li Kexin’s was denied the nomination due to suspicions of cheating.  The last nomination was awarded to Zheng Jinlong.

Note: Qiu Tingwei originally won due to the 10% bonus for new candidates. After the Li Kexin case was settled, she won her spot over Li Shidong, also a new candidate.

Xinbei City 4 (Banqiao)

m=9                 nominate=5                 Female=1

name	name	Fem	New	Inc	poll	win
王淑慧	Wang Shuhui	F			.2234	Y
張宏陸	Zhang Honglu		Y		.1607	Y
林水山	Lin Shuishan		Y		.0998	Y
黃俊哲	Huang Junzhe			Y	.0972	Y
李婉鈺	Li Wanyu	F	Y		.0952	Y
王月明	Wang Yueming	F		Y	.0818
盧輝煌	Lu Huihuang		Y		.0617
石一佑	Shi Yiyou		Y		.0478
黃炳煌	Huang Binghuang		Y		.0473
詹加鴻	Zhan Jiahong		Y		.0421
許弘業	Xu Hongye		Y		.0267
廖林麗玲	Liao Lin Liling		Y		.0162

Xinbei City 5 (Zhonghe)

m=7                 nominate=2                 Female=0

name	Name	Fem	New	Inc	poll	win
林秀惠	Lin Xiuhui	F		Y		Y
張瑞山	Zhang Ruishan			Y		Y
江永昌	Jiang Yongchang			Y		*

Note: Jiang is technically only “permitted” (報備) to run rather than nominated.  Legally, this makes no difference.  The DPP is probably doing this to circumvent rules about how long you must be a party member before getting a nomination.

Xinbei City 6 (Yonghe)

m=4                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
廖筱清	Liao Xiaoqing	F		Y		Y
許昭興	Xu Zhaoxing	F				Y

Xinbei City 7 (Sanxia, Yingge, Tucheng, Shulin)

m=10               nominate=6                 Female=1

name	name	Fem	New	Inc	poll	win
吳琪銘	Wu Qiming			Y	.1976	Y
陳世榮	Chen Shirong				.1708	Y
林銘仁	Lin Mingren			Y	.1555	Y
歐金獅	Ou Jinshi			Y	.1313	Y
彭成龍	Peng Chenglong		Y		.1241	Y
曾進益	Zeng Jinyi		Y		.1188
高敏慧	Gao Minhui	F		Y	.1021	Y

Note: Gao Minhui won due to the female reserve clause.

Xinbei City 8 (Xindian, Shenkeng, Shiding, Pinglin, Wulai)

m=5                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
陳永福	Chen Yongfu			Y	.6638	Y
李新芳	Li Xinfang	F		Y	.1767	Y
吳春美	Wu Chunmei		Y		.1595

Xinbei City 9 (Ruifang, Pingxi, Shuangxi, Gongliao)

m=1                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
顏世雄	Yan Shixiong					Y

Xinbei City 10 (Xizhi, Jinshan, Wanli)

m=4                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
周雅玲	Zhou Yaling	F		Y		Y
沈發惠	Shen Fahui					Y

Taichung City 1 (Dajia, Daan, Waipu)

m=3                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
吳敏濟	Wu Minji		Y		.4699	Y
易錦隆	Yi Jinlong		Y		.4160
陳献宗	Chen Xianzong		Y		.1141

Taichung City 2 (Qingshui, Shalu, Wuqi)

m=5                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
王立任	Wang Liren					Y
楊典忠	Yang Dianzhong					Y

Taichung City 3 (Wuri, Dadu, Longjing)

m=5                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
陳世凱	Chen Shikai					Y
劉淑蘭	Liu Shulan	F		Y		Y

Taichung City 4 (Fengyuan, Houli)

m=5                 nominate=3                 Female=

name	name	Fem	New	Inc	poll	win
吳富亭	Wu Futing			Y		Y
翁美春	Weng Meichun	F		Y		Y
謝志忠	Xie Zhizhong			Y		Y

Taichung City 5 (Tanzi, Daya, Shengang)

m=6                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
林竹旺	Lin Zhuwang			Y		Y
許水彬	Xu Shuibin			Y		Y
廖述鎮	Liao Shuzhen			Y		Y

Taichung City 6 (Xitun)

m=5                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
陳淑華	Chen Shuhua	F		Y		Y
張廖萬堅	Zhang Liao Wanjian			Y		Y

Taichung City 7 (Nantun)

m=4                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
張耀中	Zhang Yaozhong			Y		Y
何文海	He Wenhai					Y

Taichung City 8 (Betun)

m=6                 nominate=3                 Female=1

name	name	Fem	New	Inc	poll	win
曾朝榮	Zeng Chaorong			Y		Y
王岳彬	Wang Yuebin			Y		Y
蔡雅玲	Cai Yaling	F		Y		Y

Taichung City 9 (North)

m=3                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
范淞育	Fan Songyu			Y	.4698	Y
賴佳微	Lai Jiawei	F		Y	.4169	Y
游金隆	You Jinlong		Y		.1133

Taichung City 10 (Central, West)

m=3                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
黃國書	Huang Guoshu			Y	.6160	Y
江正吉	Jiang Zhengji				.2500	Y
王世勛	Wang Shixun				.1340

Taichung City 11 (East, South)

m=4                 nominate=3                 Female=1

name	name	Fem	New	Inc	poll	win
何敏誠	He Mincheng			Y	.3719	Y
鄭功進	Zheng Gongjin			Y	.2270	Y
陳福文	Chen Fuwen			Y	.2145
邱素貞	Qiu Suzhen	Y		Y	.1866	Y

Note: Qiu Suzhen won due to the female reserve clause.

Taichung City 12 (Taiping)

m=4                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
何明杰	He Mingjie			Y		Y
黃秀珠	Huang Xiuzhu	F		Y		Y

Taichung City 13 (Dali, Wufeng)

m=6                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
何欣純	He Xinchun	F		Y	.4443	Y
劉錦和	Liu Jinhe				.2832	Y
李天生	Li Tiansheng			Y	.2087	Y
林明正	Lin Mingzheng				.0638

Taichung City 14 (Dongshi, Shigang, Xinshe, Heping)

m=2                 nominate=1                 Female=0

name	name	Fem	New	Inc	Poll	win
蔡成圭	Cai Chengui			Y		Y

Tainan City 1 (Baihe, Houbi, Dongshan)

m=2                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
賴美惠	Lai Meihui	F		Y		Y

Tainan City 2 (Yanshui, Xinying, Liuying)

m=4                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
李退之	Li Tuizhi		Y		.3159	Y
賴惠員	Lai Huiyuan	F	Y		.3020	Y
趙昆原	Zhao Kunyuan			Y	.2842
陳芝伶	Chen Zhiling		Y		.0980

Tainan City 3 (Beimen, Xuejia, Jiangjun)

m=2                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
侯澄財	Hou Chengcai			Y	.7491	Y
郭再欽	Guo Zaiqin		Y		.2509

Tainan City 4 (Xiaying, Liujia, Madou, Guantian)

m=4                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
陳文賢	Chen Wenxian			Y	.5251	Y
楊麗玉	Yang Liyu	F		Y	.2994	Y
邱健吾	Qiu Jianwu		Y		.1755

Tainan City 5 (Jiali, Xigang, Qigu)

m=3                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
蔡蘇秋金	Su Cai Qiujin	F				Y
陳朝來	Chen Chaolai			Y		Y
蔡秋蘭	Cai Qiulan	F		Y		Y

Tainan City 6 (Shanhua, Anding)

m=2                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
樑順發				Y		Y

Tainan City 7 (Danei, Shanshang, Xinhua)

m=2                 nominate=1                 Female=0

Name	Name	Fem	New	Inc	poll	win
林志聰	Lin Zhicong		Y		.6735	Y
王俊仁	Wang Junren		Y		.3265

Tainan City 8 (Yujing, Nanxi, Nanhua, Zuozhen)

m=1                 nominate=1                 Female=0

name	name	Fem	New	Inc	poll	win
王俊潭	Wang Juntan					Y

Tainan City 9 (Xinshi, Yongkang)

m=7                 nominate=3                 Female=1

name	Name	Fem	New	Inc	poll	win
林宜瑾	Lin Yijin	F		Y	.3109	Y
陳秋萍	Chen Qiuping	F		Y	.2588	Y
郭國文	Guo Guowen		Y		.2526	Y
蘇泓文	Su Hongwen		Y		.1019
李國璧	Li Guobi		Y		.0560
李建志	Li Jianzhi		Y		.0198

Tainan City 10 (Annan)

m=5                 nominate=4                 Female=1

name	name	Fem	New	Inc	poll	win
郭信良	Guo Xinliang			Y	.3559	Y
王錦德	Wang Jinde		Y		.1998	Y
郭清華	Guo Qinghua			Y	.1616	Y
涂韶芳	Tu Shaofang	F		Y	.0934	Y
劉益昌	Liu Yichang			Y	.0883
唐儀靜	Tang Yijing	F		Y	.0656
黃永田	Huang Yongtian		Y		.0353

Tainan City 11 (North)

m=4                 nominate=2                 Female=1

name	Name	Fem	New	Inc	poll	win
陳怡珍	Chen Yizhen	F	Y		.4633	Y
唐碧娥	Tang Bi’e	F			.2696	Y
陳宗彥	Chen Zongyan			Y	.2141
吳杰	Wu Jie		Y		.0530

Tainan City 12 (Central-West)

m=2                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
邱莉莉	Qiu Lili	F		Y		Y
林俊憲	Lin Junxian			Y

Note: Lin Junxian agreed to withdraw after the telephone surveys were held.

Tainan City 13 (Anping)

m=2                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
李文正	Li Wenzheng			Y		Y
翁朝正	Weng Chaozheng					Y

Tainan City 14 (East)

m=6                 nominate=3                 Female=1

name	Name	Fem	New	Inc	poll	win
王定宇	Wang Dingyu			Y	.4709	Y
蔡旺詮	Cai Wangquan			Y	.2149	Y
陸美祈	Lu Meiqi	F	Y		.1696	Y
郭朝武	Guo Chaowu				.0890
蔡麗青	Cai Liqing		Y		.0557

Tainan City 15 (South)

m=4                 nominate=2                 Female=1

name	name	Fem	New	Inc	poll	win
莊玉珠	Zhuang Yuzhu	F		Y	.4949	Y
陳進益	Chen Jinyi			Y	.3015	Y
杜媽政	Du Mazheng		Y		.2036

Tainan City 16 (Rende, Guiren, Guanmiao, Longqi)

m=5                 nominate=3                 Female=1

name	name	Fem	New	Inc	poll	win
王雅雲	Wang Yayun	F		Y		Y
陳文清	Chen Wenqing					Y
劉正昌	Liu Zhengchang			Y		Y

Kaohsiung City 1

(Taoyuan, Namaxia, Jiaxian, Liugui, Shanlin, Neimen, Qishan, Meinong, Maolin)

m=3                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
林富寶	Lin Fubao			Y	.4686	Y
蕭育穎	Xiao Yuying	F		Y	.2993	Y
鍾盛有	Zhong Shengyou			Y	.2321

Kaohsiung City 2 (Jiading, Hu’nei, Luzhu, Alian, Tianliao)

m=4                 nominate=3                 Female=1

name	name	Fem	New	Inc	poll	win
張文瑞	Zhang Wenrui			Y	.3099	Y
陳明澤	Chen Mingze			Y	.2922	Y
黃炎森	Huang Yansen				.2333
葉香	Ye Xiang	F			.0906	Y
鄭顯達	Zheng Xianda		Y		.0739

Note: Ye Xian won due to the female reserve clause.

Kaohsiung City 3 (Yong’an, Gangshan, Yanchao, Mituo, Ziguan, Qiaotou)

m=5                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
陳政聞	Chen Zhengwen		Y		.4049	Y
翁瑞珠	Weng Ruizhu	F	Y		.3214	Y
謝志富	Xie Zhifu			Y	.2737

Kaohsiung City 4 (Nanzi, Zuoying)

m=8                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
林瑩蓉	Lin Yingrong	F		Y	.4852	Y
黃昭星	Huang Zhaoxing			Y	.2413	Y
張豐藤	Zhang Fengteng		Y		.1711	Y
許仁圖	Xu Rentu		Y		.1023

Kaohsiung City 5 (Dashe, Renwu, Niaosong, Dashu)

m=4                 nominate=3                 Female=1

name	name	Fem	New	Inc	poll	win
張勝富	Zhang Shengfu				.3524	Y
錢聖武	Qian Shengwu			Y	.2990	Y
曾欽宏	Zeng Qinhong		Y		.2411
林芳如	Lin Fangru	F	Y		.1075	Y

Note: Lin Fangru won due to the female reserve clause.

Kaohsiung City 6 (Yancheng, Gushan, Qijin)

m=4                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
李喬如	Li Qiaoru	F		Y		Y
連文堅	Lian Wenjian			Y		Y

Kaohsiung City 7 (Sanmin)

m=8                 nominate=4                 Female=1

name	name	Fem	New	Inc	poll	win
洪平朗	Hong Pinglang			Y	.1865	Y
康裕成	Kang Yucheng	F		Y	.1746	Y
林進興	Lin Jinxing				.1702
林武忠	Lin Wuzhong			Y	.1593	Y
黃淑美	Huang Shumei	F		Y	.1209	Y
潘金英	Pan Jinying		Y		.1166
李帝慶	Li Diqing		Y		.0431
劉彥賢	Liu Yanxian				.0288

Note: Lin Jinxing agreed to withdraw after the telephone surveys were held.

Kaohsiung City 8 (Qianjin, Xinxing, Lingya)

m=6                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
郭建盟	Guo Jianmeng				.2794	Y
周玲妏	Zhou Lingwen	F		Y	.2185	Y
蕭永達	Xiao Yongda			Y	.1876	Y
李文良	Li Wenliang			Y	.1817
湯東穎	Tang Dongying		Y		.1328

Kaohsiung City 9 (Fengshan)

m=8                 nominate=3                 Female=0

name	name	Fem	New	Inc	poll	win
張漢忠	Zhang Hanzhong			Y		Y
陳慧文	Chen Huiwen	F		Y		Y
顏曉菁	Yan Xiaojing	F		Y		Y

Kaohsiung City 10 (Qianzhen, Xiaogang)

m=8                 nominate=4                 Female=1

name	name	Fem	New	Inc	poll	win
吳銘賜	Wu Mingsi			Y		Y
鄭光峰	Zheng Guangfeng			Y		Y
李宛蓉	Li Wanrong	F		Y		Y
陳信瑜	Chen Xinyu	F		Y		Y

Kaohsiung City 11 (Daliao, Linyuan)

m=4                 nominate=2                 Female=0

name	name	Fem	New	Inc	poll	win
韓賜村	Han Sicun				.4111	Y
蔡昌達	Cai Changda			Y	.3536	Y
洪村銘	Hong Cunming		Y		.2354

DPP City Council telephone survey procedures

The DPP is determining its nominations for this year’s city council elections by telephone surveys.  This is, to my knowledge, unique worldwide.  Here in Taiwan, the KMT sometimes claims to also use surveys, but they have not institutionalized their procedures to the extent that the DPP has, and the KMT always reserves (and occasionally employs) the right to ignore the surveys and do something else.  As such, I view the KMT’s surveys as simply playing an advisory role.  Lots of parties around the world do this.  In contrast, the DPP takes its survey results as binding.  If you win the survey, you win the nomination.

This is interesting in and of itself, but there is more.  The electoral system the DPP uses in its surveys is quite esoteric.  Taiwan’s multimember districts use the SNTV system, but the DPP surveys use something just a little different.  In fact, I’m not quite sure exactly how to classify this system.

Here are the important rules.  There are two conditions, and each uses a different set of rules.  If there is only one nomination or if there are three or fewer candidates, survey respondents will only be allowed to express support for one candidate.  (OK, this part is easy to understand: it’s just a simple survey and plurality wins.)  If there are four or more candidate contesting two or more nominations the following rules will be used.

1. Each respondent is asked for his first and second preference.  The first preference is given 2 points, and the second preference is given 1 point.
2. If, when asked for his first preference, the respondent insists that he cannot differentiate between his two top choices, each will be given 1.5 points.  If the respondent cannot differentiate among three or more choices, his answer is coded as “don’t know” and no one is given any points.
3. If, when asked for his second preference, the respondent cannot give a clear answer, refuses to choose another candidate, or indicates that he only supports his first choice, the first preference is given 3 points.

The DPP does three separate surveys with at least a sample size of 1068 for each race.  The results of each survey are averaged and calculated out to the fourth decimal place to get the final result.  (Note: in previous elections, the DPP required a sample size of 3000.  They also have used a filter question in the past to disqualify respondents who support the KMT.)

The source for this is this document:

http://www.dpp.org.tw/upload/history/20100604113242_link.pdf

I am looking primarily at pp 20-25, 43-49.

So each respondent has three votes that he can cast in the following ways:

1. A: 3
2. A: 2, B: 1
3. A: 1.5, B: 1.5

Since it is impossible to cast any number other than three total votes, we could normalize the three votes to one vote, which is what the DPP does when they report the results.

This fits somewhere in the family of limited votes.  To review, the limited vote has districts with m seats, where m≧1, and each voter casts somewhere between 1 and m-1 votes.  The top m vote-getters win the seat.  Standard plurality elections are part of this family.  M=1, and each voter casts one vote.  SNTV is also part of this family.  M>1, and each voter casts exactly one vote.  However, here we have a case of voters casting more than one vote, and they can either spread their support among two candidates or concentrate it on one candidate.

I think this counts as limited vote with cumulation, except that cumulation is required, not simply allowed.  You are not allowed to give one vote to three separate candidates.

Research on limited vote systems is very sparse (except for the SMP and SNTV variants).  I need to go and look up the article, but I think I remember reading about a limited vote system for the London city council in the Victorian era.  The upshot was that everyone cumulated their votes on the local candidate, so the election essentially turned into SNTV.  That is what candidates in Taiwan seem to be trying to do.  Every banner, ad, and billboard asks people to express exclusive support for them.  However, I’m dying to know if this is actually what happens.  Are there successful candidates who get large shares of their support from second preferences?  Is there any pattern to whether a voter splits his support or not?

Another interesting question about these surveys has to do with sampling error.  In short, there is no concession made to sampling error.  If A gets a higher score than B, then A beats B no matter how small the difference is.  This, of course, is just the type of thing that drives statisticians crazy.  To simplify, the sampling error for a survey is roughly 1/Ön, where n is the sample size.  Each survey has a sample size of n=1068, so the error for each survey is .03, or 3%.  So when you do one survey, you get one answer.  If you do the same survey 1000 more times, you’ll get 1000 more answers, all slightly different.  However, 950 of those answers should be within 3% of the actual value in the population (which is what we are trying to measure).

Remember, however, that in these multimember districts, 10% or 15% support can often win the last spot.  If one person gets 13% (or more accurately, somewhere between 10% and 16%) while another gets 12% (really 9%-15%), you simply can’t be sure that the former is really stronger than the latter.  Doing three surveys helps a bit, but the answer isn’t three times better.  The error is roughly 1/Ö3204, or 1.77%.  There are going to be lots of instances in which the difference between the last winner and the first loser is more than twice that, which is roughly what it would take to say that the former is more popular than the latter in a statistical sense.  I haven’t looked at it, but it wouldn’t surprise me if the last winner and first loser were not statistically significantly different in a majority of races.  However, the DPP makes no allowance for this.  If you get a lower value, you lose.

(Note: In previous elections, the DPP required each survey to have a sample size of at least n=3000.  1/Ö9000 is 1.05%.  Better, but still not necessarily conclusive.  And very expensive.)

Of course, you now expect me to conclude that the DPP’s system is unreasonable.  Nope.  The purpose of a primary system is not to choose the most popular candidates.  It is to persuade the losers to stand down so that the party can succeed in the general election.  Think about this from the loser’s perspective.  There was a fair criterion that was well understood in advance.  The loser might not be less popular than the winner, but there is very little reason to believe that he is more popular.  Moreover, in the general election, the party nominees will enjoy the benefits of the party label.  If the loser ran as an independent, he would probably lose much of his support (since most of his supporters are also DPP supporters and would want to vote for a DPP candidate).  In other words, if you couldn’t win the primary, chances are pretty dim that you might be able to win the general election, even if you think you really had more support in the general population than the winner and just got an unlucky sample.  If the losers stand down, then this system works.

(Yes, I will be paying attention to what happens next.  This is one of the reasons that I’m so excited to get data on people contesting the nominations.)

Tainan Mayoral Race Set

In the past week, both parties have finalized their nominations for the Tainan mayoral race.  The DPP has nominated Lai Qingde 賴清德, while the KMT will nominate Guo Tiancai 郭添財.  The early polls (and all other indicators) make Lai a prohibitive favorite to win the general election in November.

The DPP released the results of its nomination surveys last Friday.  Here are the results:

	Lai Qingde	Xu Tiancai	Su Huanzhi	Li Junyi	Ye Yijin
	賴清德	許添財	蘇煥智	李俊毅	葉宜津
Pollster
Shanshui	46.27	29.84	19.60	2.62	1.66
Guanchajia	43.52	33.17	18.63	2.78	1.90
Jingzhan	43.05	34.80	18.70	1.65	1.80
Average	44.28	32.60	18.98	2.35	1.79

As you can see, Lai won by a comfortable 12% margin, well above the margin of error for a single poll.  Moreover, the three polls yielded nearly the same results.  While the Shanshui poll was a bit more favorable to Lai, the difference was only about 3%, not an unreasonable result for a random sample.  Like the Kaohsiung surveys, the DPP got a clear and clean result here: Lai was clearly the winner.

It’s probably a good thing that Lai did win by so much, because comments from Xu’s camp lead me to believe that he was looking for an excuse not to accept the results.  On the one hand, he doubted the fairness of the polls because he didn’t think he should have lost by that much.  (Had he lost by less, you can bet that he would have thought he should have won; politicians are like that.)  On the other hand, he doubted the fairness of the polls because he wondered if there was some sort of sampling problem.  His argument was roughly as follows.  He was particularly strong in two townships in Tainan County, Dongshan 東山鄉 and Madou 麻豆鎮, and his support organizations (後援會) there reported that none of them had received a telephone call.  He concluded that there must have been something wrong with the sampling.

Ok, lets do a little math to see how likely that is.  In the 2008 presidential election, there were 1432399 eligible voters in Tainan County and City.  Of those, 55353 lived in Dongshan and Madou.  How many of those could be in Xu’s support organization?  Let’s assume (optimistically) that 80% of the eligible voters are politically active enough to potentially participate.  Then assume (again, optimistically) that 65% of those lean green, and 50% of the green leaning voters supported Xu.  Finally (and most unreasonably), let’s assume every single one of those people is in Xu’s support organization.  That’s 14392 people, or roughly 1% of the total electorate.  (In fact, I’d be shocked if his support organization in those two towns were even one tenth that size.)  Assuming the sample sizes were 1000 for each of the surveys, or 3000 total, about 30 of those people should have been sampled.  If (and here we assume perfect reporting from every member of the organization back to the top) none of them were in fact sampled, we might be a bit suspicious.  On the other hand, if it turns out that Xu’s support organization is only one twentieth that size, roughly 700 people, then only 1.5 of them should have been sampled.  There are lots of sets of random samples that would not include any of these people.  In short, Xu’s complaint is only reasonable under the most heroic of assumptions.  Frankly, if Xu Tiancai, a supposed expert in financial matters, doesn’t understand basic notions of probability, (I probably shouldn’t finish this sentence).

In the day or two after the results were announce, Xu succumbed to better judgment and indicated he would not split the party by running as an independent.

Why did Xu and Su lose?  I have heard two good explanations.  The first is that both Xu and Su were two-term incumbents, and voters wanted a change.  Lai certainly hammered this point throughout his campaign.  If he didn’t think it was working, he probably would have stopped saying it.  However, I’m not completely convinced that this was the real key.  Voters in Taichung City seem to have few qualms about giving Jason Hu (another two-term incumbent) four more years.  One might argue that Hu is a KMT member, and DPP voters take notions of rotation of power more seriously.  Ok, but Taipei County voters didn’t seem too worried about sending Su Chenchang back into power which would have made it three of the past four years for him in that office.  I think this probably helped Lai, but I doubt it was decisive.

The other explanation is that this primary (as well as the Kaohsiung primary) was a rejection of Chen Shuibian.  Of all the candidates in the Tainan race, only Lai did not personally go and visit Chen in prison.  In fact, Chen publicly complained that Lai had not visited him.  Chen’s favored candidate was Xu, who was a member of Chen’s Justice Faction going back to the 1992 election.  In fact, Chen’s support was a central theme of Xu’s campaign (though perhaps not as important as Xu’s record as mayor).  It may be true that Lai’s victory really represents a desire to move away from the Chen era (and Lai is much closer to party chair Cai Yingwen).  However, remember that this was a multi-candidate election.  If we are going to interpret Lai’s support as a rejection of Chen, we must also remember that 55% of the Tainan electorate supported candidates who actively curried favor with Chen.  So let’s not get carried away speculating on the national implications of this primary.

I don’t know much about Lai Qingde, and voters usually know more about their executives than their legislators.  I have a feeling this was really more of a rejection of Xu and Su than a victory for Lai.    My guess is that both of the above factors mattered, but that dissatisfaction with the performance of the two incumbents was probably the decisive factor.  Of course, I have no evidence for this line of speculation.

Today’s newspaper brings reports of the KMT side of the race.  After weeks of trying desperately to find someone else, the KMT finally admitted that it was stuck with the declared candidates (just as in Kaohsiung).  Four waves of surveys all showed that Guo Tiancai 郭添財was slightly more popular than Li Quanjiao 李全教, by margins of between 1% and 3%.  Now, 1-3% is not significantly different in a single poll, but if you find those differences again and again over several polls, eventually you can determine that one is, in fact, higher.

Li has accepted the decision and expressed support for Guo, though let’s not imagine he was swayed by my argument about repeated polls.  He would probably be fighting a bit harder if this were a winnable race.  However, the “prize” is to waste a lot of money and energy in what is almost certain to be a humiliating defeat.

The UDN has an interesting article on the decision between Guo and Li.  (Well, interesting to me: it has the smallest headline and the least space on the entire page.)  Guo, it seems, was the safe choice.  Li had much more grassroots support and much better organization.  However, Li also has a court case pending and is considered much more controversial and attackable.  Guo, in contrast, is a former education bureaucrat who is now serving as vice-president of a technical college.  He hasn’t cozied up to grassroots power brokers; in fact, they complain that he is a cold fish.  However, this race is really about setting the table for 2012, and it wouldn’t do Ma Yingjiu any good to win a few more votes this year while sullying the party image.

So, how is the horse race looking?  A UDN poll from last Friday says Lai 57, Guo 15, undecided 27.  (The same poll had Lai 59, Li 15, undecided 26.)  I don’t think Lai will win by a 4-1 margin, but a 2-1 victory is not out of the question.

Chen Ju wins Kaohsiung nomination

The DPP announced the results of its first telephone surveys today.  Chen Ju beat Yang Qiuxing by a margin of 59-41 to win the DPP’s nomination.  More precisely, the results are as follows:

	Chen Ju	Yang Qiuxing
DPP survey center	0.5988	0.4012
Guanchajia	0.5739	0.4261
Jingzhan	0.5950	0.4050
Average	0.5892	0.4108

Reporting survey results to the hundred of one percent is a bit ridiculous since the sampling error is roughly three percent each way, but we’ll ignore that this time.  The good news for the DPP is that all three organizations came up with the same result, so there is little doubt about Chen Ju’s clear victory.

Yang Qiuxing responded by saying he would accept the results, so the contest is effectively over.  In this sense, the system has worked very well.

To Americans (like me), primaries seem natural.  How else could you pick a candidate?  In fact, in most of the world’s democracies, candidate selection is a purely internal party matter.  Twenty years ago in Taiwan, the general electorate had no influence over candidate selection.  In a sense, the “democratization” of nominations is a failure of party politics.  (By “democratization,” I simply mean the enlarging of the electorate.)  Parties could not make their nominations stick for two reasons.  First, the voters didn’t have strong enough party identifications to vote for whoever the parties put forward.  To use a term from the old one-party American South, there weren’t enough “yellow dog Democrats,” people who would even vote for a yellow dog as long as it was the Democrat nominee.  Second, the losing candidates (seeing that voters might still vote for them in the general election) often refused to withdraw.  Instead, they often ran as independents.  Even if they didn’t win (and a significant number did), they might cause the party nominee to lose by splitting the party vote.

The DPP was the leader in institutionalizing the nomination process.  In the 1980s and early 1990s, the DPP central executive committee decided nominations.  In 1992 (I think), they tried a two tier nomination process, with party cadres deciding part (30 or 50%, I’m not exactly sure) and the general party membership deciding the rest.  This was a disaster.  The cadres formed voting alliances along factional lines instead of choosing candidates on their individual merits.  The general party membership was even worse: faction leaders registered huge numbers of new party members.  They often paid their membership fees and listed their contact info at the same address.  In other words, the party had no access to these members.  The party image suffered from all the dirty stories caused by this nomination procedure, and losers had no reason to withdraw since the primaries weren’t really a test of general popularity.

My favorite episode was the 1996 presidential nomination.  The DPP ran a two stage primary.  In the first stage party members and cadres voted, narrowing the field from four to two.  (Lin Yixiong and You Qing were eliminated; Peng Mingmin and Xu Xinliang went on to the next round.)  Then Peng and Xu went on the road for a month.  Each night, they held a debate in a different county or city.  After the debate, the audience voted.  The voting was really fun.  Each person was given a specially minted coin, which they could deposit in either the Peng slot or the Xu slot in a specially designed vending machine.  At the end of the voting, the coins for each candidate were counted up.   The DPP was trying to expand its primary electorate so that (a) mobilization wouldn’t be the determining factor and (b) the nominee would get a mandate from a primary electorate that was like the general electorate.   They succeeded on the first goal.  Xu mobilized every night, and almost every night his supporters were outnumbered by  Peng’s self-mobilizing supporters.  However, Peng’s supporters were mostly the die-hard Taiwan independence fundamentalists, so the nomination was essentially decided by a small, radical slice of the total electorate.

In the late 1990s, the DPP started experimenting with telephone surveys, and it quickly became apparent that this system accomplished nearly everything they wanted.  They don’t corrupt the party membership, it’s hard to manipulate a survey (when done by a neutral organization), and the sample can mirror the general electorate.  Best of all, they give a “clear” result that losers find very hard to defy.  (I will probably discuss at a later date how ludicrous it is to use surveys in multimember districts, and even in single seat races the winner does not always win by a statistically significant margin, but these are points that seem lost on candidates and voters.)  Since then, telephone surveys have been used more and more by all parties.  Nowadays, they are generally (but not always) the determining factor, not simply one element of a complex nomination process.

All of this is a roundabout way of saying that the Kaohsiung primary has been a rousing success for the DPP.  There were two very strong candidates, either of seems capable of winning the general election.  With only a bit of rancor, they have conducted an intense contest that resulted in a clear victory for one and a concession of defeat by the other.  Given that the KMT’s best chance to win the election seems to be a split in the DPP, this is no small achievement.

DPP primary poll procedure

The DPP is using telephone surveys to decide its nominations for mayors and city council elections.  I will discuss the results in other posts.  Here, I just want to document the process.

There are six survey organizations that might be used: the DPP survey center, Shih-hsin University survey center,T Taiwan Real Survey, Shanshui Survey, Jingzhan Consulting, and Guanchajia Marketing (中央黨部民調中心、全國公信力民意調查股份有限公司、世新大學民意調查研究中心、山水民意研究股份有限公司、精湛民意調查顧問股份有限公司、及觀察家行銷研究有限公司).

The surveys for Kaohsiung and Tainan mayor will take place from May 3-7.  On the morning of May 3, the DPP will draw lots, choosing three of the organizations to survey each race.  They will also decide which survey is held on May 3-4 and which is held on May 5-6.

In the event, the Kaohsiung survey was first, and the Tainan survey was next.  They also picked the order for the city council surveys: (1) Kaohsiung, (2) Taichung, (3) Tainan, (4) Taipei and Xinbei.  (I’m not sure why Taipei and Xinbei go at the same time.  This seems strange to me.)  These surveys will begin on May 17.  Each morning, they will randomly select one district for the relevant city.

KMT telephone survey results

The KMT finished its telephone surveys for the Taipei City,  Tainan City, and Kaohsiung City mayoral nominations last week.  They did not release the numbers, but here are the general results.

Hao Longbin 郝龍斌 won in Taipei City and will be nominated soon.  This surprises no one.

In Tainan City Guo Tiancai 郭添財 beat Li Quanjiao 李全教, but only by about 3%.  Xie Longjie 謝龍介 was a distant third.  Since this is so close, the KMT will try to negotiate a compromise between the two leaders.  If the negotiations yield no results, it may try another round of surveys with just these two candidates.

In Kaohsiung, Huang Zhaoshun 黃昭順 won.  Lin Yishi 林益世 trailed by about 5%, and Su Yinggui 蘇盈貴 was third.  The KMT will try negotiations or a second round among the top three.

So Tainan and Kaohsiung remain unresolved.

To me, the most interesting comment was from one of the losers in Kaohsiung.  Hou Caifeng 侯彩鳳 remarked that the results were basically what she expected.  After all, Huang is the only candidate who is actively campaigning.  Which makes me ask: Why?  Why aren’t any of the others campaigning?  The DPP candidates are trying like hell, and have been doing so for several months.  In contrast, two of the KMT candidates only announced they would run a week or two before the surveys, and Lai Fengwei 賴峰偉 even announced his candidacy while he was in Penghu.  Of course there are close ties between Penghu and Kaohsiung, but he couldn’t even be bothered to travel to Kaohsiung?  Why didn’t the KMT candidates (other than Huang) exert any effort?