For Chinese translation, click here. Many thanks to Chen-Yu Chan for the translation.

Last week, the Taiwan Citizen Union (公民組合) announced that it would register as a political party and field candidates in the 2016 legislative elections. Its stated agenda is political reform, beginning with the electoral system. That’s the right place to start, because as currently constituted, Taiwan’s electoral rules make it nearly impossible for small political parties to win more than a couple of the legislature’s 113 seats. Unless the electoral system is changed, the TCU, like hundreds of parties before it, is probably destined to end up as a historical footnote.

The Flaws of the Current System

It wasn’t always this way. Prior to the 2008 election, Taiwan’s legislators were chosen in multi-member districts that allowed small parties to win seats with 10 percent of the vote or less.  When the New Party broke away from the KMT in the mid-1990s, it was able to win 12.8 percent of the seats in its first election, immediately becoming an important player in the Taiwanese party system. Likewise, both the People First Party and Taiwan Solidarity Union made significant gains in the first election they contested, winning 20.4 and 5.7 percent of seats, respectively, in 2001.

That possibility is now much less likely, thanks to a reform package pushed through by DPP-KMT agreement in 2005 over the vociferous objections of the smaller parties. That reform cut the size of the legislature in half and replaced all multi-member districts (複數選區) with winner-take-all single-member districts (單一選區). Predictably, the legislature’s disproportionality greatly increased: in 2008 the KMT won 72 percent of the seats with 54 percent of the district vote, while the DPP won only 24 percent of seats with 38 percent of the vote.The change also reduced other parties to bit players in the legislature, limiting their representation to a handful of seats in the proportional representation (PR) tier (不分區). Worst of all, it greatly exacerbated a structural advantage in favor of one of Taiwan’s two main political camps. By some estimates, the pan-Blue side (KMT+PFP) now can maintain a legislative majority with as little as 45 percent of the popular vote, while the pan-Green side (DPP+TSU) would need to win as much as 55 percent to guarantee control.

Thus, there’s a good case to be made that Taiwan’s electoral system is seriously flawed. But how should it be fixed?

From Diagnosis to Treatment

Electoral systems can affect three key properties of the party system. First, they determine how proportional is the conversion of votes a party wins into seats that party gets in the parliament. Second, they can affect where parties target their political appeals: at moderate voters in the center of the spectrum, or at extreme voters near the ends. Third, they can influence party unity—that is, how likely party legislators are to follow the party line and vote as a bloc. 

The main problem facing Taiwan today is the first: proportionality. The current system is quite majoritarian. Elections to the legislature take place in three separate “tiers” of seats: about two-thirds (73/113) are chosen in plurality single-member districts, an additional 30 percent (34/113) in a PR tier with a five percent threshold (門檻), and about five percent (6/113) from special aborigine constituencies.  The single-member districts are the source of most of the disproportionality, because in these only the first-place candidate gets a seat.

Goal One: Reduce Disproportionality

An obvious improvement, then, would be to reduce the net effect of the single-member districts by increasing the size of the PR tier, ideally to at least half of all seats. An easy way to do this would be to return the size of the legislature to 225 members, and elect most of the new legislators from the PR list. But however it is implemented, the ratio, rather than the size, is what’s important: the greater the share of PR seats, the less disproportional Taiwan’s electoral system will be, and the more likely that parties other than the DPP and KMT will become significant players in the legislature.

An even better solution would be to adopt a version of Germany’s electoral system and make the PR tier compensatory—that is, rather than awarding PR seats separately from district seats, PR seats can be allocated to “compensate” parties that are under-represented relative to vote share in the single-member districts.  The net effect would then be very close to full proportionality. This German-style “mixed-member” system was adopted in New Zealand in the 1990s, and it has dramatically increased the number and diversity of parties in the parliament there: before the reform, the winning party always had enough seats to govern alone, but every New Zealand government since has been a multi-party coalition.

Goal Two: Encourage Moderation

A secondary problem in Taiwan is a long-standing tendency toward extremism and especially incivility in politics. Electoral reform could mitigate this as well. The switch to single-member districts was a step in the right direction on this dimension, because now the winning candidate has to appeal to a broad swath of the electorate in the district, rather than just core partisans or key faction leaders. But this incentive could be strengthened further by introducing ranked-choice voting for these seats.

Under ranked-choice voting, also called alternative vote or instant-runoff voting, voters rank some or all of the names on the ballot, rather than just casting a ballot for one candidate. Last-place candidates are eliminated, and their votes transferred to others according to the voter’s rank, until one candidate ends up with a majority. The strength of this system is that candidates who win tend to be moderate and the “least offensive” of anyone running. For instance, in Australia, where ranked-choice voting has been used for over a century, the main complaint is that elections are boring, because candidates are too nice to one another! Given Taiwan’s tumultuous electoral history, incentives to be civil would be a much-welcome change.

Another feature that keeps extremism in check in Taiwan is the five percent threshold to win PR seats. Some reform groups have called for lowering or eliminating this hurdle. This is a bad idea: it would open the door to party fragmentation and allow much more extreme groups to gain representation in the legislature. Particularly if the PR seat share is increased, this threshold is the main thing preventing even more rabidly pro-unification and pro-independence parties from gaining a foothold in the legislature. Partisan acrimony over the independence-unification issue is already pervasive in Taiwan; it makes no sense to encourage even more.

Goal Three: Don’t Weaken the Parties

The final question to consider is about party unity. Taiwan’s electoral system currently gives parties full say in how their candidates are chosen, rewarded and disciplined. Some reform groups, including the TCU, have called for weakening Taiwanese party leaders’ grip over party nominations, giving legislators greater leeway to cultivate an independent following and vote their conscience in the legislature.

Although this might sound appealing in theory, in practice it would probably be a major step backwards. Taiwanese legislators already have significant independence from their party leadership, and further limiting party control would make Taiwanese parties even less capable of acting with unity of purpose. More worrisome, weak parties are associated around the world with government inefficiency and greater corruption. Rather than improving things, weakening party control over individual members would move Taiwan closer to countries like Italy and the Philippines—places notorious for fragmented party systems, undisciplined parties, and poor public policy outcomes. Intra-party reform is a task best left to the parties themselves.

Crafting a Better Democracy

Taiwan’s democracy is not about to fail, but it is also far from perfect. Better-designed institutions could mitigate the polarization of Taiwan’s political elite, increase public confidence in its elected representatives, and reduce the intense dissatisfaction and alienation that leads to mass protests and confrontations in the streets. Although electoral reform won’t solve all these problems, it can still make a significant difference—and it’s been done in the recent past. Thus, whatever the Taiwan Civic Union’s other arguments, it has at least one thing right: electoral reform is pro-democracy, not pro-blue or pro-green, pro-China or pro-independence. In short, it’s something that Taiwanese of all political stripes should be able to get behind.

(Feature photo of Ketagalan Boulevard and the Presidential Office, by © CEphoto, Uwe Aranas on Wikicommons, CC BY-SA 3.0)


Kharis Templeman

Kharis Templeman is the program manager of Stanford University's Taiwan Democracy Program. He has traveled extensively in Taiwan and the PRC, and was a dissertation research fellow at the Taiwan Foundation for Democracy. He has been looking at party system development and political institutions throughout the world, especially in Taiwan and Asia.
He received a Ph.D. in political science (2012) from the University of Michigan.

Latest posts by Kharis Templeman (see all)