Taipei 3 update

The KMT announced the results of its preliminary poll in Taipei 3. (For background, see my previous post.) Incumbent Lo Shu-lei羅淑蕾 did not win the poll by more than 5%, so there will be a full-blown primary in this district between Lo, Wang Hung-wei王鴻薇, and Chiang Wan-an蔣萬安.

Lo immediately cried foul, saying there were irregularities in the polling process. According to her, several of her supporters who answered phone calls complained that they were not allowed to answer the survey. When a man answered, the interviewer asked to speak to a woman. When a woman answered, the interviewer asked to speak to a man. Sounds bad, eh? Actually, not at all. This is called “in-house sampling,” and it is a standard part of the KMT’s survey protocols. The basic idea is that while the telephone number is randomized, the person answering the telephone may not be. In order to ensure randomization, the interviewer first asks how many people live at the residence and then, according to a predetermined table, asks to speak to the second oldest female, the youngest male, or whoever else the table demands. With in-house sampling, it is quite normal that the person answering the phone is not sampled. Either Lo simply has no idea how this process works (note: they usually brief the candidates on the process, and the candidates are allowed to have observers watch the process), or she is simply making up a fake complaint. I think it is the latter.

(Aside: The DPP does not use in-house sampling. The KMT’s philosophy is that the polls should reflect public opinion as accurately as possible. The DPP’s philosophy is that polling primaries are partially a contest of mobilization capacity, so they encourage candidates to mobilize supporters to go home and sit by the phone, especially if opinions inside their household are divided. Generally speaking, in-house sampling is theoretically superior, but, in practice, it doesn’t affect results all that much. In this case, since Lo reportedly won by 4.2%, it is conceivable — though certainly not obvious — that in-house could have driven her lead under 5%.)

Lo hasn’t lost yet; why is she crying foul? More than most candidates, Lo needed to win in the first round. Over the next six weeks, her three biggest advantages will all fade. First, now that she has been shown to be vulnerable, her organizational base will start to crack. She has built fairly close ties with all the neighborhood heads, going to all their local events and representing their demands to the bureaucracy. This organizational base was one of her main advantages. However, now that she is not an overwhelming favorite to get the nomination, some of these neighborhood heads will reconsider their affiliation. Many of these loyal KMT footsoldiers probably were never that happy with her in the first place, due to her tendencies to openly criticize party leaders. If Chiang looks like he will be the local legislator for the next four years, many will decide that the smart thing is to change sides, and that will happily allow them to support a more reliable party member. Second, Chiang Wan-an somehow succeeded in garnering a substantial level of support in only 17 days. Most voters only know one thing about him, that he is John Chiang’s son and Chiang Kai-shek’s great grandson. Some voters may have been uncomfortable supporting someone whose last name was his primary (only?) appeal. Now he has about six weeks to introduce himself to voters and flesh out that picture. If he does this wisely, he should be the favorite to win the nomination. It is instructive that almost none of the media reports mentioned Wang. If Chiang can frame the election as a contest between a loyal KMT member who seeks to better the country by improving the party from within versus a loose cannon who doesn’t hesitate to damage the party’s reputation for the sake of making a splash on TV, he will win the nomination easily. However, if Chiang fails to define himself appropriately to the media, this could easily become a contest between Lo and Wang. Lo’s third advantage in the first round was that Wang’s support was heavily concentrated in a small part of the district. Only about a third (the parts in Songshan District) of the legislative district falls within Wang’s city council district. In the other two-thirds (Zhongshan District), Wang is relatively unknown and has little organizational support. Wang now has another six weeks to work on getting votes in Zhongshan. A fourth advantage for Lo could also disappear in a hurry. Right now, the anti-Lo vote is roughly split between Chiang and Wang. If a neutral media source, such a TVBS, were to publish a poll of this race that showed either Chiang or Wang as significantly ahead of the other, this three-way race could quickly become a two-way race. Lo would almost certainly lose that contest.

As of today, I no longer think that Lo is the favorite. As I see it, Chiang is now most likely to win the nomination and the seat. It will be interesting to see if Lo or Wang turn to negative campaigning to try to take him out. Because he is so unknown, he is the perfect target for mudslinging. With an established politician, you know enough about him or her that it is nearly impossible for any single new bit of information to fundamentally change the way you think about him or her. With an unknown, every bit of information can completely rewrite the story. If I were in the Lo, Wang, or DPP camp, I’d be thinking about how to redefine Chiang Wan-an as the 2016 version of Sean Lien. If I were in the Chiang camp, I’d be trying to build up a more robust image of him as someone who has not relied on the family name, who has lived a modest lifestyle, and who is earnest and down-to-earth.

In the end, Lo has no one to blame but herself. She simply could not discipline herself. Time after time, she lambasted President Ma and other KMT leaders on TV and in print, and she seemed to love being in the limelight and willing to say anything to attract attention. Many people have observed that she sounded more like an opposition politician than one from the KMT. It should not be all that surprising to her if the KMT voters who dominate her district want a legislator who represents the values, ideas, and interests of the KMT.