Current Polling Primary

Since 1998, the DPP started using polling to decide the nomination for the Legislative Yuan elections. But it dictated only 50% of the total outcome. The other 50% of the score would be determined by party member votes. The proportion of the polling has gradually increased to 70% in 2001, and to 100% in 2012.

As only one candidate can be elected in each of the 73 single member districts, electability indeed became the primary criteria for consideration when both parties selected their candidates.

According to its deputy secretary general, when the DPP decides whom it should nominate, it

assesses the winning odds for every potential candidate. Such assessment is normally done by poll.⁴ The KMT also did the same thing and even took one step further—that is, it did not nominate its candidates until the DPP had completed its selection process so that it could send someone to match up against the DPP’s nominee.⁵ In short, for the 2012 Legislative Yuan elections, both parties used poll results to assess candidates’ electability. In the following, we will discuss how the two major parties used opinion polls extensively to gauge the electability of their nominees.

Phases of Candidate Selection: Both the KMT and DPP start their candidate selection procedure with candidate registration. In general, one who seeks for either the KMT or DPP nomination needs to register for nomination selection even he/she is an incumbent. Most of the time, an incumbent would not face any challenger from his/her own party. Yet, even if only one candidate has registered for nomination, the party still assesses his/her electability by a poll result. The KMT’s rule states that one needs to pass 30% threshold of support to become a KMT nominee even if he/she is the only person who seeks for nomination.

If more than two contenders register for nomination selection, the process proceeds into the second phase—negotiation. It is a mechanism that aims to solve conflicts within a party and to smooth the nomination process before heading toward any heated intra-party competition. Yet, negotiation usually takes time. During the negotiation period, parties may again use poll results to see who has a better chance to win. Additionally, the party headquarter may send a senior party officials (normally assigned by the party leader) to handle the negotiation process. In some cases, such negotiation process get through and the party successfully nominates a candidate without further intensifying intra-party competition (or conflict). For example, in 2012, in the 4^th district of New Taipei City, the KMT’s incumbent candidate Lee Hung-jun faced a strong challenger Hsu Bing-kuan, who was a former city mayor of Hsin-chuan area (one of the big cities in New Taipei city metropolitan area). Lee and Hsu had known each other for 30 years and were really close friends. In the beginning of the negotiation process, Lee intended to withdraw himself and let Hsu represent the KMT to join the general election. Yet, the KMT leadership really wanted Lee to stay in the Legislative Yuan. Thus, at the end of the day, Hsu agreed to withdraw from the competition and decided to wait on the sideline for future opportunities.⁶

The above situation, however, may not be the case all the time. It is quite often that multiple contenders refuse to step down and the negotiation process breaks down. For example, in the 2^nd district of New Taipei City, the DPP incumbent Lin Shu-Fen was challenged by Huang Chien-hue, a former Legislative Yuan member who lost to Lin in the 2008 DPP primary. Soon after the deadline for registration, the DPP headquarter sent a senior party leader to initiate the negotiation process.

Yet, the negotiation fell apart quickly as Huang was strongly supported by one of the DPP’s major

factions. It was clear that he would never withdraw from the competition.⁷ In short, although Lin seemed to be a very strong DPP incumbent candidate, a polling primary was still inevitable as some quality contender insisted to challenge her.

Having the negotiation process broken down, both parties’ candidate selection procedures head into the third phrase—polling primary. As mentioned earlier elsewhere, the main purpose of using polling primary instead of closed (i.e. party member) primary was to avoid the domination of primary by the “head (or phantom) party members”. Additionally, both parties believe that polling primary is the most scientific and efficient way to not only resolve internal conflicts but also select the best candidate who is truly electable.

Obviously, the two major parties do not have equal strength in every district. It is predictable that the intra-party competition for nomination would be intense within a party that consistently enjoys a favorable margin in the district. On the other hand, it is not surprising that a party may have difficulty to recruit any quality candidate to run in a district contest that is almost always taken by the other party. It seems that in party’s safe districts, polling primary becomes an effective and acceptable tool for the party to determine who gets nomination; In contrast, as quality candidates do not have strong incentive to run in difficult districts, the party headquarter would directly recruit someone to join the election. Take the DPP’s case as an example. Where the party obtained less than 42.5% of the vote in the last election, the DPP would regard such district as a “difficult” one. Then the party would directly recruit someone to run instead of opening it for competition.

As polling primary became the way for both parties when multiple contenders seek for nomination, how did it work in the 2012 elections? Some details are illustrated in the following sections and are summarized in Table 1:

KMT: potential candidates registered for the polling primary. The major questionnaires used in the polling primary included two parts: 1. intra-party support and 2. inter-party support. In the first part, potential candidates competed for the best supporting rate. This part will consist 15% of the total score. In the second part, each candidate will compete against the DPP nominee to see who has the best winning odds. This part will consist the remaining 85% of the total score.

Candidates would negotiate to decide the dates for the polling primary. On the polling dates, normally two companies would conduct telephone surveys using random sample draw from the district. And in-house sampling was implemented for every survey.

DPP: Potential candidates needed to register for the polling primary. The default questionnaire used in the polling primary asked about inter-party support. But if KMT’s candidate was not available or potential contenders agreed upon a different method (e.g., intra-party support), the DPP headquarter

accepted it. The only thing that the headquarter wanted to control was the primary date. In order to test conteders’ ability to mobilize voters, the headquarter would not announce when the primary would take place until the morning of the primary date. Specifically, the DPP headquarter set up a period for the Legislative Yuan primaries. And in every morning during the primary period, the headquarter randomly selected one district for primary and immediately informed contenders around ten O’clock in the morning. Because DPP’s survey did not adopt in-house sampling, who got to answer the phone became an important factor in determining the results. Thus, once the primary date was announced, it was crucial for the contenders to mobilize their supporters to stay at home after six O’clock to wait for possible phone calls. The DPP believed that such method was the way to test which contender had the best campaign organization that could successfully mobilize his/her supporters.⁸ And each primary was held within a day with three companies conducting telephone surveys. Each one of them needed to complete the 1,200 sample list.

The two parties have adopted a very similar way to conduct polling primary. The most significant difference between the two is the use of in-house sampling. While the DPP cares more about contenders’ ability to mobilize voters, the KMT obviously cares more about representation of the survey sample and tries to avoid possible manipulation.

Table 1: Key Features of Polling Primary (KMT vs. DPP)

KMT Polling Primary DPP Polling Primary Number of Districts 11 13

Conduct Local branch Headquarter

Method Telephone Survey: two companies for three days

Telephone Survey: three companies for one day

Questionnaire Inter-party match up: 85%

Intra-party match-up: 15%

Default:

Inter-party match-up

Date Pre-scheduled Random pick

In-house sampling Yes No

In short, polling primary can be regarded as the current default system. Direct selection and negotiation are NOT default since they only exist when there is no meaningful intra-party competition. The purpose of adopting polling primary is twofold: first, it is designed to find an electable candidate; second, it delegates power to the general public to resolve any intra-party conflict—let the voters decide who should be put on the ballots. And polling primary seems to be the most sufficient way as it is (1) cost-saving, (2) avoid vote-buying, and (3) avoid manipulation from the other party.