Fri, Jan 19, 2007 page8

• US poll theory does not fit Taiwan

  • By Shih Chih-yu 石之瑜
  • • 
    • 
    • 

  The results of the nation's past few presidential and local elections have been extremely close, so it seems that voters are distributed along two peaks on a graph of their political affiliation, forming a so-called "M pattern."

  But even though the election numbers show an M distribution, we can only infer the cause. Can we go a step further and deduce that the cause of this M distribution is that voters' policy preferences -- namely independence versus anti-independence -- prior to the election also form an M curve?

  The widespread belief that M-distributed voting results are caused by M-distributed political stances is mostly a product of US election theory. However, that US election theory, that is currently drawn upon, should not be applied to Taiwan for two reasons.

  First, the theory supposes that voters have set policy preferences and that candidates establish their platforms in such a way as to win the support of moderate voters.

  Second, it supposes that the scope of policy options available to voters is the same as the scope of policies that the candidates debate. Therefore, in a US-style analysis, if votes are distributed in an M pattern, then the reason is that voters' attitudes are also distributed in an M pattern.

  Taiwan is not the US. Taiwanese voters' stance on independence is often influenced by the candidates. It's therefore very possible that voters will decide their policy preferences after being influenced by the candidates. Thus, what we need to analyse is why a voter likes or dislikes a certain party's candidate, not whether the political views that the voters have independently decided tend toward independence or not. Such is to say that even though voters may be against independence, they won't necessarily vote for an anti-independence candidate, and vice versa. These two suppositions are deserving of some consideration.

  If the only policy issue during an election were Taiwanese independence, the M pattern theory would let us divide voters into three types according to the distribution: those who support independence, those who are against independence and moderate voters who don't support independence but aren't opposed to it either.

  M theory denies the possibility of a fourth kind of voter who could support independence, but who could also be against it. This fourth type of voter cannot exist in US election theory because they are not within the scope of topics discussed by the candidates. It also violates the logic that voters have their set preferences.

  But in Taiwan, surveys have repeatedly shown that there is often a lot of overlap between voters who would be willing to accept Taiwanese independence under the right conditions, and voters who would be willing to accept unification under the right conditions. The overlap could include as many as one-fourth of voters. Taiwanese, therefore, are not a classic model of moderate voters, but instead of "bi-polar voters," simultaneously drawn to both ends.

  Under such circumstances, if a candidate shows enough resolve and determination to give hesitant voters a sense of stability, and if the candidate provides appealing reasons to support independence or unification, voters may form a dependence on the candidate based on their emotional needs. They will naturally begin to fall in sync with the candidate's policies as well.

  However, a candidate might instead follow US election theory and believe that voters already have set policy preferences. Moderate voters may lean slightly toward independence, so in this case the candidate would establish a slightly pro-independence stance. But since that would simply replicate the hesitant feelings of the bipolar voters, they might develop a detachment from the candidate.

  Bipolar voters have no coordinates on the pro-independence versus pro-unification graph because they have been forced into being perceived as ambivalent moderate voters. But at best they are fake moderate voters. If bipolar voters exceed a certain ratio, then its impossible to explain the voting results using the M theory.

  A better way to explain the results would be to say that one party's candidate came across as resolute, while the other's appeared to waver. This would induce many of these apparently moderate voters to temporarily lean towards the more resolute side out of an emotional need.

  For some reason, at each election the Chinese Nationalist Party (KMT) appears to be far out in front, then tends to waver back toward the center to consolidate its support among moderates. But this often ultimately allows the Democratic Progressive Party candidate, who is seen as more resolute, to catch up.

  The KMT, which accepts the M theory, has decided that it needs to go further to befriend pro-independence voters, while maintaining its anti-independence stance. But in contrast, the bipolar voter theory shows the KMT needs to put on a performance for voters to show its resolve, not its wavering.

  Shih Chih-yu is a professor of political science at National Taiwan University.

  Translated by Marc Langer
• Most Popular

  • 1

    Starlux Airlines to apply to join Oneworld alliance

  • 2

    Chinese man captured on a rubber boat

  • 3

    Trump is right about World War III, but wrong about the analysis

  • 4

    Manila pulls ship from Sabina Shoal

  • 5

    Tropical storm to bring rain later this week