Skip to navigation – Site map

HomeIssues58PoliticsKMT: A Trompe-l’œil Victory


KMT: A Trompe-l’œil Victory

The December 2004 Taiwanese Legislative Elections
Frank Muyard


The recent renewal of Pan-Blue’s parliamentary majorityin an almost exact repeat of the 2001 election resultsmasks more changes than it appears both within the two camps and in the respective parties’ appeal to the electorate. Winners of the elections, the opposition parties keep losing voters. On the Green side, the defeat is mainly due to a mix of tactical errors and political factors. However, the Democratic Progressive Party’s vote share still progresses. The low turnout also shows a rise in the voter disaffection towards Taiwan’s highly partisan political debate.

Top of page

Full text

1The official results of the December 11th 2004 Legislative Yuan (Parliament) elections in Taiwan gave a clear majority to the Pan-Blue camp, made up of the Chinese Nationalist Party or Kuomintang (KMT), the People First Party (Qinmindang, PFP) and the New Party (Xindang, NP). This was to many a surprise as expectations of a Pan-Green victory following the re-election of President Chen Shui-bian in March and numerous opinion polls led pundits to predict a shift of control of Parliament to the greens1. For more seasoned specialists of Taiwan politics such a feat was seen however to be rather difficult to achieve2. The discrepancy between the opinion polls and the final result created a sense of shock and a resounding defeat for the Democratic Progressive Party (Minjindang, DPP) and the TSU (Taiwan Solidarity Union, Tailian), the two parties forming the Pan-Green camp. In the domestic and international media, as well as in Peking, some were quick to read into these results a rejection by the Taiwan electorate of Chen Shui-bian’s Taiwan-centred policies and independence goals as well as a shift in favour of a political rapprochement with China3.

2Although the election’s outcome clearly represents a failure for Chen and the DPP, a closer analysis of the results shows that, compared to the 2001 legislative election, not only both camps include winners and losers but also that the DPP is the only party to see an actual rise in the number of votes received. The Pan-Blue victory is, then, more about merely holding onto a majority than getting a strong new mandate from the voters. The Blue camp’s one seat loss, the redistribution of votes between its three components as well as a very low turnout all indicate that it can hardly claim any progress in political dominance. On the other hand, the persistence of the opposition’s grasp on the legislature is a huge setback for Chen’s second-term plans and will certainly trigger a repositioning and shift in Taiwan’s party strategies with future elections in view.

3This article will first look back at the result and the reasons behind the Pan-Blue victory, then present a more detailed reading of the poll’s figures that leads to a reconsideration of some of the first interpretations of the election results. It will conclude on the poll’s short-term consequences that are already reshaping the country’s political dynamics.

The Pan-Blue victory

4The last opinion poll run by the all-news channel TVBS, gave, on the basis of a 70% turnout, a slight victory to the Pan-Green camp with 106 seats against 105 to the Pan-Blues, the independent candidates holding the balance and the key to the legislature majority4. Such a poll already hinted that the proclaimed majority victory by the ruling party and its smaller partner would not be so easy, especially when the pro-Blue inclinations of most of the independents were taken into account. Indeed, the December 11th election for the Sixth Legislature sent back almost the exact same Parliament as in 2001 (Table 1)5. Pan-Blue won 114 seats, losing one seat, but keeping its majority, while Pan-Green only got 101 seats and the independents 10 for a total 225 seats. The slight increase (+1) by Pan-Green shows the persistence of its progress in legislative elections since 1998, but falls short of its own predictions (between 105 and 120 seats) and above all of the 113 seats landmark that assures control of the legislature.

1. Legislative elections results (by seats)

1. Legislative elections results (by seats)

5Within the Blue camp, the election witnessed a reversal of the 2001 legislative election result where the PFP made big gains at the expense of the KMT, which fell then to its lowest ever level. This time the KMT raised its number of seats by 11 to 79 while the PFP lost 12 of its former 46 legislators for a final 34 seats. The New Party won one seat (Kinmen) in 2001 and kept it in 2004. Three more NP members were actually elected this time although they are counted within the KMT tally because, except for Kinmen, NP candidates officially ran under the KMT banner in order to maximise the number of complementary legislators attributed proportionally to each party’s overall share of the vote 6.

6Still the DPP kept its position as the nation’s biggest party with 89 seats, two extra seats. For the green camp, this result was overshadowed by the loss of one seat by the TSU, which takes it down to 12 legislators, and above all by its continuing minority status in the legislature.

7The percentages received by each party reveal similar trends although with interesting differences: the DPP garnered 35.7% of the votes, a 2.3% growth compared to 2001, which makes it the biggest gainer in this election. The KMT received 32.8% of the votes, apparently a big jump from 2001 (28.6%). But when the NP figures are added, the combined KMT-NP 2004 and 2001 tallies are 32.9% and 31.1%, showing an actual progress of 1.8%, lower than the DPP. For the PFP, the fall is heavy: minus 4.7% on 2001 for a total of only 13.9%, while the TSU saw a repeat of its 2001 score with 7.8% of the votes (+0.03%). The rest went to the independents who received roughly 0.5% more votes than in 20019.6% (Table 2).

8This election thus marked a major victory for the KMT, which for the first time since 1998 progressed in a national election. For the PFP it was a very disappointing result, while bickering and complaining by PFP leaders against its bigger and “bullying” brother started even before election day as its candidates sensed defeat coming. The KMT advances appear indeed to result directly from a shift in the votes and influence back from the PFP: +11 seats for the KMT and –12 for the PFP (+8 and –8 respectively when only directly elected candidates are counted). The outcome constitutes thus a zero-sum game for the opposition, but certainly not for the KMT, which engineered an impressive reversal on 2001. Feelings and expressions of victory were therefore especially strong at the KMT and to a lesser extent at the NP7.

9Most of the KMT and media post-election hype about the opposition’s success must be seriously balanced with the real extent of the Pan-Blue achievement. The KMT success would not have been considered so great if the party and Lien Chan had not been considered finished after their defeat in the presidential election last March. It is thus a timely remainder of the continuing weight of the KMT and its moderate Chinese nationalist orientation in Taiwan’s political field.

10Inadvertently, the DPP through its perception of itself as the new representative of the national majority and of 2004 as the start of an era of genuine green power and unstoppable momentum for Taiwan’s self-affirmation also gave its opponent’s victory a stronger impact that would have normally been the case with such a repeat of the 2001 results. To its defence, the pitiful shape of the Blue camp after their failure to regain the presidency, the aggressive street reactions by PFP legislators and supporters in its wake, the refusal of the two Pan-Blue candidates to officially recognise their defeat and their persistence in appealing unsuccessful lawsuits against Chen’s re-election—all behaviour badly perceived by the large majority of the population as opinion polls attested repeatedly throughout the year—, the KMT’s inner dissent, and finally the incapacity of the two parties to finalise their highly publicised merger were also convincing factors for the green camp and part of the media to bet on the disintegration of the blue forces8. Moreover the DPP and Chen Shui-bian knew they would have to prevail in the legislative election to have any hope of breaking the institutional stalemate that characterised the relation between the government and the legislature in the last four years and push forward their policies. The pressing desire to achieve that goal might then have been confused with the capacity to deliver it.

Main factors of the opposition victory

11With the wind of popular support apparently blowing in their direction, the Pan-Green certainly fell into the overconfidence trap. However elections are won on the ground and DPP’s defeat can more directly be attributed to certain choices and errors made in dealing with four main factors of this election: its local nature, the type of strategy and campaign favoured by the parties, the tactics of candidate nomination and vote-allocation, and finally the very low turnout.

The nature of the election

12Legislative elections, particularly in Taiwan—but that would be true for most democracies—, are largely dominated by local questions and are therefore very different from the national-type presidential elections where the electorate must choose the leader who will represent the country and its general political orientation for the next four years from among a few individuals. Consequently, the parties that will best grasp the needs and desire of their local constituents, district by district, are the ones likely to emerge as winners. In Taiwan, classic patron-client relationships have long taken the form of deep-rooted and publicly recognised local factions or local networks of influence that tend to have a huge effect on local and legislative elections in terms of vote allocation and group-voting9. Although the dominance of local factions have been reportedly on the wane, recent surveys of electoral campaigns in the centre and the south show that success of local politicians is still very dependent on their associations with one kind or another—preferably several—of network or faction10. This system has been recognised as constituting the basis of the KMT’s local power for decades and the Nationalist Party, even less wealthy than before, is still at its best when using these political and client-based local organisations to its advantage11.

13This has two consequences. First, the subsequent decision by other parties to play the game of local organisations rather than insist on mere party affiliation and loyalty. For the TSU, born out of the KMT’s “localisation” faction, this goes as natural practice and, in fact, most of its work in the last three years has been to try to transfer the KMT’s local networks support to itself, especially in the south although with limited success. The same goes for the PFP. For the DPP, the stakes are different as a significant part of its programme and image is built on the rejection of local factions and network power-systems criticised as vote-buying machines of a corrupt KMT, and from which it was traditionally excluded12.

2. Legislative elections reusults (by votes and in %)

2. Legislative elections reusults (by votes and in %)

14Reports and first evidence from the election results indicate however that this time the DPP engaged itself into much closer co-operation with local vote-gathering organisations than in 2001, in the south as in the north, leading some observers and DPP strategists to deplore a backward trend away from modern party politics and an increased dependency upon corporatist and local demands that can easily lead to corruption13. Actually, the current polling system and historical strength of local networks in Taiwan make it impossible for the DPP to ignore local faction power and leverage in elections14. In the end, the green camp was however clearly not as efficient as the KMT and the independents in mobilising local networks to support its candidates.

Blue and green camp campaigns

15A striking characteristic of the December 2004 election is the opposite approaches to the poll and campaigning favoured by the two camps: an ideology-oriented, identity-focused, national party and Chen-centred campaign for the DPP and the TSU, while the blue camp chose to privilege a more pragmatic and locally based campaign focused on constituents’ needs and network mobilisation.

16At first, the DPP campaign was actually much less ideological and mixed the promotion of government achievements with promises to do better with a majority victory. The reasons of the shift and the direct intervention of Chen in the day-to-day campaigning are two-fold: the lack of response of the population to DPP’s electioneering up to October and the fear of voter-poaching into the DPP’s pro-independence core electorate by a very vocal and active TSU. From November on, the DPP and Chen Shui-bian ran a near repeat of the March 2004 presidential election campaign15, centred on pro-independence and identity issues, and personalised, on Chen himself—an overwhelming presence outshining the four DPP “kings” originally in charge of the campaign16. It led to an outpouring of pro-Taiwan and anti-China propositions and measures by Chen, the government, and Pan-Green supporters symbolically handing the control of most of the DPP’s campaign agenda to the TSU and Lee Teng-hui’s ideas17.

17Pan-Green main issues were as follows: the strengthening of Taiwanese national identity, upholding national pride—being Taiwanese—, reinforce the nation’s military and political strengths against China (notably through the Parliament’s ratification of a big US arms-package deal), pursue the ongoing school-curriculum and textbook reforms towards a Taiwan-centred ideology, get rid of the last remnants of KMT party-state heritage and influence in the administration and the political field, and last but not least the rewriting of the “Chinese” Constitution into a real “Taiwanese” Constitution adopted through a referendum (but still—for the DPP—keeping the “Republic of China” [ROC], the ROC’s name, emblems and official territory).

3. Number of local electoral district candidates and success ratio (elected candidates/candidates)

3. Number of local electoral district candidates and success ratio (elected candidates/candidates)

18Some of these proposals are still controversial in Taiwan and many people, even without necessarily being opposed to them in theory, did not feel strongly about the timing of their introduction in the political debate and the need to support them in practice. Inevitably these proposals also led to more provocative calls from the green ranks—ike the rejection of Sun Yat-sen as the father of the nation or a shift in administrative exam subjects to only Taiwanese topics—which had then to be officially dismissed by Chen or the government18. Added to the aggressive attacks by Chen on the KMT for its “soft” coup d’état tentative in the wake of its March re-election, and his official call for “name-rectification” (zhengming) of the country’s representative offices abroad, governmental institutions and state-owned enterprises from “China” to “Taiwan”, they generated a strongly polarised electoral debate where conflicts between competing ideologies and nationalisms prevailed over any possibility of compromise or middle-way policies.

19Although DPP leaders and strategists officially defended this new abrasive tone and the take-over by Chen of the day-to-day campaign, more than a few candidates were clearly uncomfortable with it and did not feel very good about its eventual chance of success. Observers alike felt that the strongly ideological rhetoric would prove to be counterproductive19. Successful or not, its final outcome would then clearly be Chen’s personal responsibility.

20In the blue camp, national support was provided to all candidates with big meetings, local visits by party leaders, and a flurry of TV advertisements. But the Pan-Blue parties’ campaign was still essentially locally based, constituency-centred and run by individual candidates and local party chapters. In the centre and the south, the avoidance of any party affiliation went even to the extent of discarding all references to and emblems of the KMT, Lien Chan or the PFP on candidates’ election materials, these partisan associations being seen as detrimental to the local candidates’ electoral prospects. The Pan-Blue campaign however was not devoid of any political and ideological issues. They mainly turned around the conservative and responsible image the KMT has always been keen to develop: maintenance of Taiwan’s Chinese identity and ROC history; avoidance of any provocation and risk of war through increasing “Taiwanisation”; outward rejection of desinicisation; thawing of Taiwan’s political relationship with Peking; reducing the cost of the US arms-package deal considered too expensive for the national budget and militarily unnecessary if cross-Strait relations improve; boosting of Taiwan’s economy through business-oriented measures and closer cross-Strait economic integration.

21But as the highly controversial creation of a “March 19th special Truth investigation committee” by the opposition-controlled Parliament attests, the Pan-Blue’s rhetoric against the ruling party and Chen was as aggressive as its opponents’. In the last weeks of campaigning, political tension increased further when the KMT first requested the hand-over by the DPP of governmental powers to the blue camp in case of the latter’s legislative victory on the basis of political legitimacy and “spirit” of the Constitution, and then threatened to wage chaos in the new legislature if it did not win the polls20. This radicalisation in discourse and tactics by the Nationalist Party surprised many. It was also difficult to understand how continuity in Pan-Blue legislative control could be translated in Constitutional and institutional changes and re-attribution of executive power on the basis of a “new” opposition legitimacy although this was not something implemented on similar past occasions, such as the 2001 legislative election. Whether the Taiwanese constitutional system can be compared to the French and its “cohabitation”, or the US one and its clear division between the three branches of power has yet to be determined as it is clearly a system that takes from both models21. In the meantime, any sort of compromise between the ruling and opposition parties, if any, might more likely take the form of governmental coalitions or case-by-case co-operation in Parliament than a full take-over of the executive responsibilities by the legislative majority.

22As noticed before the poll, the DPP’s campaign strategy totally ran against its previous experience of success in legislative and local elections22. Its eventual failure reinforces the idea that the choice of a presidential-style campaign based on controversial and ideological identity issues was ill-suited to the nature of the election and, therefore constituted one of the main strategic reasons of the Pan-Green defeat. But other factors, essentially tactical, can be seen behind the election result.

Candidates’ numbers and vote-allocation

23On the tactics side, the peculiarity of Taiwan’s single non-transferable vote (SNTV) legislative poll system—in multi-member constituencies—requires the maintenance of a very elaborate organisation by the parties on the local level in order to maximise the number of candidates elected within their estimate vote-pool. First the number of party candidates must be limited to avoid dispersion of the vote between too many candidates resulting in the loss of all or most of them. Secondly, a vote-allocation system has to be implemented to ensure that each candidate receives the minimum expected amount of votes needed for his election. As some candidates (incumbents, local or TV celebrities, party stars) are better known than others, and some are closely related to local networks that can deliver big number of votes, good co-operation between the candidates of a same party or alliance of parties and an efficient electoral organisation are needed to control the transfer of votes to candidates with lesser initial chances of success. Any failure in one of these steps can result in the loss of seats on poll’s day.

24Particularly difficult for the parties is to assure co-operation between their own candidates who are, by virtue of the polling system, competitors for the same share of the vote. Another challenge is to measure and regulate any shift of votes within one party’s own electorate to avoid spontaneous and panic-mode “save-the-weak-candidates” movements that can be go beyond the amount of votes needed and eventually be lethal to “strong” candidates suddenly deprived of too much of their original support.

25The vote-allocation system and limited number of candidates have traditionally been seen as a DPP strength and explain why it so often has a larger share of legislator seats than its share of votes. This was notably the case in the 2001 legislative election, while the blue camp, and especially the KMT, was then criticised for its high number of candidates and poor handling of vote allocation.

4. Legislative elections overall votes and turnout

4. Legislative elections overall votes and turnout

26This time, the three opposition parties co-operated pretty well in limiting the number of their candidates. KMT and NP candidates for local electoral districts numbered only 75 (74 + 1) in December 2004 against 129 (97 + 32) in 2001 while PFP candidates went down to 41 from 61 (Table 3)23. The reduction is significant (-54 candidates) with a Pan-Blue total of 116 candidates (against 170). It reflects a much better appreciation of actual electoral prospects by the three parties and resulted in the end in a successful outcome. In terms of vote-allocation, however, co-operation was difficult and generated numerous complaints by PFP and NP candidates and leaders about the “ungenerous”and “unfair” behaviour of the KMT, which was seen as the biggest potential purveyor of votes24.

27On the green side, the number of candidates for electoral districts remained stable (122) with the DPP adding nine candidates for a total of 92 and the TSU cutting nine candidates to 30. Although the “too” high level of DPP candidates was criticised after the poll, as we see, the overall number of green candidates did not change. Besides, as Bruce Jacobs pointed out, the DPP had to nominate more candidates if it wanted to achieve majority control of the legislature25. The problem seems then to reside more in the number of TSU candidates. The TSU has indeed the lowest ratio of elected legislators/candidates of the four main parties (23.3%), expressing obvious over-confidence by the TSU in its goal of doubling its legislators on a narrow Taiwan identity platform. The DPP ratio of success (76.1%) is rather good and close to the KMT. A dysfunctional vote-allocation system between the DPP candidates themselves, as shown by the high number of DPP incumbents who were not re-elected, was also a direct cause of lost seats. Poor co-operation and allocation of votes within and between the DDP and the TSU must therefore more appropriately be considered the culprits in the green camp’s failure to achieve a better result.

28Indeed, the high toll of incumbent casualties was impressive: 37 altogether, including 14 DPP, nine PFP and four TSU for only four KMT (the others six were independents). Well-known political characters and party officials were among the victims26. The renewal of Parliament therefore has many new and some inexperienced legislators especially in the DPP caucus27. This would tend to prove that the parties’ control over their vote-catching system was defective. But it is not true everywhere, even for the DPP who had 100% success in several electoral districts28.

External factors

29Like during the presidential election last March, Peking did its best not to be seen as interfering in the election while, this time, maintaining a string of vocal criticisms and rebuttals of Chen Shui-bian, his government and any step taken by the island’s politicians Chinese leaders saw as independence-oriented29. The PRC also took care of waiting for the final results to enact any measures that could be used by the Green camp to fuel its campaign and boost its electoral chances. However, once the outcome known, Peking did not wait long to announce the coming enactment of an Anti-secession law by China’s rubber-stamp Parliament, and therefore raise tensions again across the Strait30. On the other hand, Peking responded carefully but rather positively to Chen’s new offers of cross-Strait contact and Chinese New Year charters between the two sides. But it waited after the election to speed up the negotiations and agree to the exchange for the first time of non-stop direct flights to serve Taiwanese businessmen in China31. In a sense, the PRC’s tactic was successful as it avoided being a major factor in the election, usually negative to the Pan-Blues. However, the timing of the Anti-Secession Law announcement, following the Chinese Communist Party’s own schedule and institutional process, was clearly felt as overly insensitive to the election outcome and Taiwanese population desire for better cross-Strait relations.

30As for the United States, relations between the Bush administration, especially the State Department, and Chen remained tense, marked by a lack of mutual trust, and a series of official declarations emphatically non-supportive of Taiwan’s government, which as usual had a lot of echoes in the island’s media and political circles32. Although it did not have a clear impact on the election, it showed to everybody that the DPP government needed to amend its relations with Washington and establish a sounder system of consultation and co-operation with its main ally. Neither did the US presidential election have a strong influence as Bush was seen as the better choice for Taiwan by both the government and the opposition. Also no big changes were expected with the nomination of the new Secretary of State, Condoleeza Rice33. Contrary to last year’s presidential pre-campaign, the US factor was therefore not significant although certainly not favouring the DPP. Powell’s remarks in China about Taiwan, obviously expressing deep frustration about the direction taken by Taiwan and the cross-Strait situation, were also badly articulated and clearly oblivious of the political consensus in Taiwan about the country’s status34. For the KMT and other blue parties, even pro-unification factions, and the DPP government, the Republic of China (ROC) is an independent and sovereign country as much as and on a par with the People’s Republic of China (PRC), as attested, among others, by official diplomatic relations the ROC still have with 26 countries and had until 1979 with the United States35. Such remarks probably had an effect opposite to their intended goal, showing how poor Washington’s understanding is about the Taiwan situation and Taiwanese people’s own perception of their country, and in making the government and people more nervous than necessary about the strength of US commitment to Taiwan, and therefore about the kind of future they can expect.

31The need for a better co-operation mechanism is also true for the United States, which can no longer leave its interaction with Taiwan—and strategic interests in the region—so little managed now that the cross-Strait tension and military balance, and the de facto evolution of the terms of the status-quo in China as well as in Taiwan, make war a much more likely scenario than before36. The US Defense Department’s posting of an active duty military officer in Taipei for the first time since 197937, as well as its continuing efforts in helping modernise the Taiwanese forces, and its pressure on Taiwan’s political parties to ratify the acquisition of its latest arms package, all show that Washington is aware of the problem.

32One topic of US-Taiwan relations was nonetheless a subject of intense debate during the election campaign: the US$18 billion arms deal offered by Washington and agreed upon by the Chen government in 2004 after three years of discussion. The government made the approval of this deal by the Parliament a central part of its 2004-2005 programme and budget, and after being rebuffed by the opposition as a main issue of its campaign. Chen and the DPP see the arms deal as essential to cement the informal military and strategic alliance with the United States as much as to improve relations between the two sides by showing the commitment of the island to pay for its future defence against a potential Chinese attack. For that, they prepared a special budget to finance the cost of the deal over a ten-year period. The KMT and the Pan-Blue, big supporters of costly arms deal with the United States when in power, criticise now both the excessive amount and the necessity of some of the weapons to be bought (especially the submarines that nobody really knows by who and where they are going to be built, and which represent a big part of the arms procurement package). The issue has become such a focus of political struggle, that few can distinguish any more the real defence or public expenditure concerns expressed by the opposing camps under their partisan battle38. The arms deal still awaits ratification though Chen promised the Americans to ensure its passage before the end of 2004. Indeed, in spite of meetings between US administration officials and Soong Chu-yu, and his subsequent declarations hinting towards a compromise, the blue camp continues to block the deal in Parliament.

A closer reading of the poll figures

33At a first look, the election results can easily be interpreted as a victory for the Pan-Blue and a reversal of its past (mis)fortunes concomitant to a Pan-Green defeat; and as an impressive continuity in voter preference over a three-year period. This lecture has indeed been largely dominant in opposition party circles and media analyses. Following this line, Chen Shui-bian’s March re-election could then be interpreted as an odd phenomenon only explainable by the “silver bullets” conspiracy theory and other election abnormalities, and the December 11th poll the vindication of Pan-Blue analysis of an unfair and stolen presidential election.

34For the DPP, different legislative and presidential election outcomes are however explained by tactical and technical errors in its December poll campaign organisation, compounded by over-confidence within the green camp and vote-fatigue in the electorate. Moreover, the DPP’s progress both in terms of percentage of the vote and seats would tend to demonstrate that its programme still managed to convince more voters and therefore that the election loss cannot be attributed to its pro-independence stance nor express a disaffection of the electorate

35Actually, a closer look at the result shows that  although the Pan-Blue victory is without any doubt a real political success, its electoral basis is rather fragile; and that there is a clear consolidation of DPP strength in the electorate that could lead to better outcomes for the greens in the future. Further, voters tended to favour middle-of-the-road programmes, “centrist” positions, and perhaps did not vote for lack of such choices in this election.

36The first indication that the Pan-Blue election result is paradoxically more a setback than progress compared to 2001 is shown in the percentage of votes received by the two camps. Whereas overall the blue camp reached 49.74% of the valid votes in the last legislative election, this time the opposition parties only got 46.85% of the vote, a decrease of almost 3% (Table 2). In the meantime, voter support for the green camp increased by 2.4% to 43.5%. The difference between the two camps therefore narrowed from 8.6% to 3.3%, a result that cannot in any case be considered a success for the opposition.

37Second point, the turnout factor. At 59.16%, it is the worst turnout of any national-level election since the beginning of free and competitive polls in the 1990. Turnouts for legislative elections are generally lower than for presidential elections, so nobody will be surprised that the December elections did not match the 80.28%, 82.69% and 76.04% respectively for the 2004, 2000 and 1996 direct presidential elections. However compared with the three previous legislative polls since 1995, this turnout was considerably lower. In the 1995 legislative election, turnout was 67.65%, in 1998 it was 68.09%, and in 2001 66.16% (Table 4). Abstention in December 2004 was thus 7% higher than the time before (and an average 8% higher than in the last decade). Combined with a 2.3% increase in the share of the eligible voters compared to the whole population (itself rising) since 2001, this means that a considerable smaller number of people actually went to vote. While 735,671 more people were eligible to vote than in 2001, the December 2004 election witnessed 672,706 less people casting their vote39. Taiwanese political parties hence clearly failed to impress the voters and make them vote: if we add the diminution in turnout to the rise in the eligible population, more than 1.4 million old and new voters decided not to vote this time.

38But this overall loss was not proportionally transferred to each of the different parties this time. The parties’ actual wins and losses in absolute numbers rather than in percentage give us a better image of their power of attraction for the population. The DPP emerges here as the only party that saw rise in voter numbers: an increase, albeit small, of 23,689 votes (Table 2). For the KMT, one needs to add to its 2001 tally the NP votes since they concluded an electoral alliance this time (and add NP’s Kinmen candidate votes in 2004). In 2001 the total of KMT and NP votes climbed to 3,218,991, while in December 2004, the sum of their voters only reached 3,202,218, resulting in a loss of 16,773 votes. For the PFP, the reduction was much more significant: a fall of 567,223 votes, or close to 30% of its 2001 vote. The TSU, while maintaining its vote share in percentage, did register a loss as well, as it gathered 44,848 less votes than in 2001. The independent candidates’ number of votes hardly budged (-5,341) although their percentage of the vote rose 0.5%40.

39The 2.89% loss of the Pan-Blue in vote percentages translates then in absolute figures into 584,000 less votes, while the green camp saw its support wane only by 21,159 votes. In other words, the Pan-Blue vote diminution represents 84% of the overall decrease in the number of votes cast (or of the 7% increase in abstention), while the green camp loss constituted only 3.2% of it. On the basis of these numbers, the Pan-Blue legislative victory cannot be presented as a big success in its extending its influence or attraction on the Taiwanese electorate neither as giving it a strong new mandate―and needless to say even less over-riding the Chen Shui-bian mandate received in March with 11.8% more votes than in 2000 (or almost 1.5 million more votes). Furthermore, it shows that the gain of 11 seats by the KMT at the expense of the PFP did not correspond in any gain in vote numbers for the KMT.

40When compared to other parties’ results, one can hardly say that the DPP campaign failed as it is the only one to have gained more voters than in the previous election. The DPP’s biggest failure consists in not having attracted more people to the polling stations. This is especially true as higher turnouts are generally considered to favour the green camp. The blue camp’s core electorate is indeed seen to be more reliable and ready to vote to express its opposition to the current government policies and regain power. If this were the case, the continuous lowering of Pan-Blue vote numbers and vote-share would mean that its electoral basis is fast diminishing (584,000 votes in three years).

41The DPP is therefore clearly consolidating its political dominance and electoral basis both against the blues and the TSU. As we noticed, tactics were a strong factor in the green camp’s setback. Indeed, many DPP or TSU candidate defeats were by very small margins of one or a few thousands votes, and often with two candidates of the same party or one from each party defeated. Although in local districts, candidate affiliations do not always recoup party affiliation and then cannot ensure a perfect transfer of votes between candidates. Better co-operation in the candidate nomination process and in vote-allocation between the DPP and the TSU could have ended very differently on election day. Add to that a few more percent in turnout, and the green camp might have wrested victory from the others. From these figures, one can even see that pre-election polls were not so wrong after all as they were based on a much higher turnout.

42Chen’s identity issues-focused campaign seemed to have two opposite results: to consolidate the DPP’s electoral basis and voter fidelity; and to scare away or discourage non-party affiliated and “centrist” voters to vote for the Green camp. On the basis of the election outcome, this could be seen as a tactical error. But on a longer-term perspective, it might have strong strategic value in helping to achieve two things. First, by reaffirming its pro-independence programme and identity, the DPP has managed to generate a growth of core-support for the party and influence in the population while eliminating the menace of the TSU over its vote-pool, especially in the south. Voters who chose the DPP in December would probably not abandon it in future polls unless it makes huge mistakes in governance or changes its political identity. They either saw themselves in the DPP rhetoric or at least actively favoured it against both Pan-Blue and TSU programmes41.

43At the same time, the disappointing result of the election should slow down DPP radicals as they realise that they are still not the national majority. It can be used by DPP leaders to steer a more moderate policy with the next elections in view―national assembly, city and county chiefs (2005), Taipei and Kaohsiung mayoralty (2006), presidential and legislative (2008). This would help them win over more centrist voters by good governance and compromise with the opposition while holding firm on ROC sovereignty, independence and its democratic system, and steadily reinforce the population’s Taiwanese “consciousness”.

44To conclude, in the elections, DPP support rose more than for any other party, both in vote percentage and vote numbers. The Pan-Blue voter base weakened considerably compared to previous elections showing a diminishing influence of the KMT and PFP in the population as a whole. All parties failed to draw the electorate to the voting-booth, especially the centrist voters, with stronger negative consequences for the DPP. The electorate tended to reject radical parties in both camps (PFP and TSU) in favour of (relatively) moderate ones. The KMT under Lien Chan succeeded in reversing its decline but the number of votes did not progress. Retail politics, local factions and local networks are still important components of success in the legislative elections. Finally, tactics (nomination, vote-allocation) were a major factor in determining the final outcome.

Consequences of the poll and recomposition of Taiwan’s political field

45The election shows, once again, that opinion polls should not be trusted too much, especially in Taiwan where reliability is marred by partisanship and scientifically unsound methods.

46For the Pan-Blue, new rifts in the alliance and not so clear skies for the KMT are now to be expected. A party continuously losing voters, even when economic conditions (2001) or the external political (2004) environment are bad for its opponent, is facing a real challenge in ensuring any future political dominance. The KMT’s success in this election was mainly built on the return of the “lost children” of the PFP and the NP to the old party, and on a very efficient mobilisation of its core base of local supporters. It vindicates the ideological return to a pre-Lee Teng-hui and orthodox KMT engineered by Lien Chan during the last four years in order to win back the true-blue voters who deserted him for Soong and the PFP in 2000 and 200142. The KMT now faces the challenge of internal reforms in a somewhat better position and the old leaders will have a chance to leave with dignity. Lien Chan should retire as KMT chairman in August and the party will have to find a new leader43. But, if the true-blue voters are not going to leave the party or the Alliance, as the poll’s figures indicate, they do not constitute a sufficient majority to ensure wins in further elections, especially in times of higher turnouts. The Blue camp’s appeal appears lower than ever as it was unable to keep more than half-a-million of its own 2001 voters, not to say to attract more centrist voters. This can be linked to the dissatisfaction of the electorate, especially the young and middle-class voters, towards the Pan-Blue leaders for their reaction to last March’s election, their political obstruction in Parliament, their inability to reform and ensure a generational transfer in their leadership, and finally for their programme, which is not more attractive than the green’s. That includes the pro-China orientation and calls by the blue parties for an economic integration with China that scares many in the lower middle-class who are unable to invest in China and see hardly any benefits for them in it.

47In terms of immediate consequences, this election ended any possibility of a quick merger between the KMT and the PFP, which already decided that it would be better alone than swallowed by its big ally44. Beaten by their rival’s electoral machine, the strong anti-KMT and grass-roots factions of the PFP, represented by the party vice-president Chang Chao-hsiung and Soong himself are now even less inclined to relinquish the fruits of their work in their last four years and their personal fame and influence for a marginal position in a newly merged blue party keeping the KMT’s name and centralised organisation.

48In the opposite camp, the defeat led the DPP’s leaders to thoroughly rethink their tactics and methods for the coming years. The goals are for Chen to avoid becoming a lame duck for its entire second-term, and for the party to find a way to implement some programmes and policies through co-operation with Parliament. Party and President therefore changed their tone and announced that to respect the popular will expressed in the election they will seek compromise and co-operation with all parties including opening government portfolios to opposition politicians or forming a coalition cabinet45. The idea of collaborating with the blue camp, especially with the PFP, was then repeatedly affirmed by DPP legislators, government officials and Chen’s close collaborators46. The result was a “10-point consensus” signed by Chen and Soong to lower partisan hostilities and foster smoother legislative co-operation.

49This says a lot about the nature of Taiwan politics, much less ideological and strictly partisan, and more pragmatic and power-focused than Western analysts would believe based on the experience of classic European party politics. But this was also made possible by two things. First the imperative to deny the “legitimacy” argument branded by the KMT to take over governmental power in exhibiting a image of consensus for the DPP and Chen and therefore to show that the President is still the source of executive power and ultimate political decisions in Taiwan’s constitutional system. Secondly, the evolution and manoeuvring of the PFP in the wake of its crushing election setback and end of its merger with the KMT.

50The PFP’s new strategy focuses on re-centring the party at the middle of the Taiwan political spectrum. While keeping on its true-blue mainlander voter base, anti-independence ideology and objective of political reconciliation with China, Soong is rebuilding the image of its party as a bridge and a balance between the two camps to get the country out of the political deadlock created by the DPP/KMT opposition in the legislature. The PFP is now officially ready for co-operation with any party in order to achieve progress in matters of national and people’s interests. DPP’s overtures have already legitimised the new image of the PFP. However both parties’ core supporters reacted rather negatively to the “consensus” and the PFP runs the risk of losing more true-blue voters to an evermore old-guard looking KMT.

51An official PFP-DPP coalition, as the latter publicly wished (maybe more by political calculation than by real intent), would therefore not be possible now. But with the new and rather open-minded Hsieh Chang-ting government installed, other avenues of co-operation could be worked out with the PFP or the KMT, dangerously over-passed by its partner in a “centrist” image.

52Political co-operation in Taiwan usually needs long and repeated meetings and discussions that do not ensure any eventual outcome, and depends always on each party’s short-term strategy. Whether the Taiwan political environment changed enough after the December election to bring a new type of co-operative spirit between the government and the legislature will only be attested by 2005 political evolution. In any case, Soong and Chen showed after the election they still are the most acute and imaginative Taiwanese politicians, and in a sense true representative of modern politics compared to the classic and institutionally-groomed Lien Chan and other KMT leaders.

53The political battle’s horizon is now set on a series of minor and local elections in 2005 and 2006, appetizers and rehearsals for the real prize of the 2008 presidential and legislative elections. As the analysis of the December poll figures shows, the blue camp is not in a very good shape. It would need to devise strong new programmes and strategies to have a chance to wrestle the presidency from the DPP or even repeat its legislative success if future turnouts improve. It cannot count only on Ma Ying-jeou’s charisma and the KMT’s traditional support networks as shown by their continuing loss of votes in the few last years.

54In the DPP camp, the necessity of compromise to achieve more in the government will certainly be an asset in building a more moderate image. The new tandem of Su Tseng-chang, who replaced Chen Shui-bian as DPP president, and Prime minister Hsieh Chang-ting could then well become the next DPP presidential ticket. For many in the party, that would constitute a “dream team”. Respectively former Kaohsiung mayor and Taipei County chief, both have won consecutive electoral successes in a tough political environment and are known for their moderation on top of real charisma and competence.

55For the TSU, its role as the radical wing of the green camp should persist. Both Lee Teng-hui and DPP independence advocates considered it important to keep alive and vocal their long-term objective of formal disentanglement of Taiwan from the ROC constitutional framework. Signs of openness to co-operation with the opposition are nonetheless also appearing, clearly with the goal of presenting a more responsible image for the party. Still, the TSU should remain an appreciated but marginal addition to the DPP’s chances of further success as the ruling party has proved to be the only political force making repeated gains in the last series of elections.

56  Besides this, the political atmosphere and success opportunities in Taiwan will also much depend on the evolution of cross-Strait relations. Appeasing signs followed the groundbreaking exchange of direct flights between Taiwan and China for the Lunar New Year. But tensions rose again with the March 8th enactment by the Chinese National People’s Congress of its “Anti-secession” law even if most Taiwanese desire better political and economic relations with the mainland. Once again, US reactions will be crucial to manage the cross-Strait situation and indicate where the strategic Washington-Taipei-Peking triangle might be heading.

57The next Legislative Yuan election in 2008 should also present a rather different picture as it should be held the same day as the Presidential poll, the legislature’s seats would be cut by half, and a brand new polling system would be implemented based on single-member constituencies and dual votes. Constitutional amendments voted in August 2004 must however still be ratified by the last convening of National Assembly this year after its mid-May 2005 election. It will be the first electoral test of the year and will show how the Taiwanese people react to the political parties’ repositioning.

Top of page


1 Cf. Kathrin Hille, “Taiwan ruling party casts KMT into the shade”, Financial Times, October 18th 2004; AFP, “Pro-independence parties tipped to make gains in Taiwan vote”, December 8th 2004.
2 Cf. Shelley Rigger, in “March to December: Taiwan’s year of elections”, CNAPS, Brookings Institution, November 17th 2004.
3 Keith Bradsher, “In Taiwan Ballot, Ties With Beijing Seem to Be a Winner”, New York Times, December 12th 2004. This to the dismay of most international academics, cf. Huang Tai-lin, “US scholars cite need to fight foreign media’s ‘spin’”, Taipei Times, December 13th 2004.
4 Cf. AFP, op. cit.
5 On the December 2001 legislative election, see Lin Jih-wen, “Taiwan’s 2001 Elections and their Political Impact”, China Perspectives, No. 39, January-February 2002, pp. 53-61.
6 For the specifics of Taiwan’s legislative elections, see GIO, “Background information on the 2004 Legislative Yuan election”,
7 The election result was quite good for the NP compared to its crushing defeat in 2001, and legitimises its alliance with the KMT. Its candidates elected under the KMT banner received very high support, ensuring some lasting influence for the tiny party among the Pan-Blue electorate and leaders, especially in Taipei and its suburbs. The three new NP legislators are Lai Shyh-bao (Taipei South, who got the highest number of votes in the whole city, 70,773), Fei Hung-tai (Taipei North) and Lei Ching (Taipei County 3rd district, the second highest national vote-getter with 72,467 votes). Altogether the NP had eight candidates, seven running with the KMT, plus incumbent Wu Cheng-tien in Kinmen. For district-by-district results, cf. Lianhe bao, December 12th 2004, p. A11.
8 See Lawrence Chung, “Kuomintang in crisis”, Straits Times, August 8th 2004.
9 For a general analysis of local faction influence in Taiwanese elections, see Mikael Mattlin, “Nested pyramid structures: Political parties in Taiwanese elections”, China Quarterly, No. 180, December 2004.
10 See Bruce Jacobs, “Faction, blood and party”, Taipei Times, December 5th 2004. And Fiorella Allio, “Pour une étude anthropologique des élections taiwanaises”, conference given at the CEFC Taipei Office, December 16th 2004. Mattlin and Allio’s researches show that local factions (difang paixi), local networks (jiceng mailuo) and even local “organisations” (difang zuzhi) have to be differentiated.
11 For an historical review of KMT power-system, see Yun-han Chu and Jih-wen Lin, “Political development in 20th Century Taiwan: State-building, regime transformation and the construction of national identities”, China Quarterly, No. 165, March 2001.
12 For KMT and the DPP’s different approaches to the question, see Christian Göbel, “Beheading the Hydra: Combating Political Corruption and Organised Crime”, China Perspectives, No. 56, November-December 2004, pp. 14-25.
13 Cf. B. Jacobs and F. AlIio, op. cit. Not everybody is so pessimistic, for instance Michael H.H. Hsiao who sees in the election’s result a continuing trend towards a two-party political system, and lauds the influence of  “civil society”’s social groups and organizations on the Taiwanese political field. Cf. Intervention at INPR Conference on “The implications of 2004 legislative election in Taiwan”, Taipei, December 12th 2004.
14 See Shelley Rigger, “The Democratic Progressive Party in 2000: Obstacles and opportunities”, China Quarterly, No. 168, December 2001, pp. 944-959.
15 For an analysis of March 2004 presidential election and campaigns, see Frank Muyard. “Taiwan: The Birth of a Nation?”, China Perspectives, No. 53, May-June 2004, pp. 33-48.
16 Su Tseng-chang, Hsieh Chang-ting, Yu Hsi-kun and Chang Chun-hsiung.
17 With the symbolically important differences: change of Taiwan’s official name, total rejection of the ROC’s (and Chinese) heritage and name, full-Taiwanese new constitution and its rapid implementation, all parts of the TSU’s programme of fast-track independence but not supported by the DPP programme.
18 “Chen reaffirms Dr. Sun history to halt distortions”, China Post, November 14th 2004.
19 Cf. Lawrence Eyton, “Taiwan elections curiouser and curiouser”, Asia Times Online, December 9th 2004. For the DPP’s past electoral experience, see also S. Rigger, op. cit.
20 “Mixed reaction to Lien's call to form gov't”, China Post, December 7th 2004.
21 For an analysis of the Constitution and the potential for Taiwanese “cohabitation”, see Da-chi Liao and Herlin Chien’s article in this issue of China Perspectives.
22 Cf. L. Eyton, op. cit.
23 To have the total number of candidates from each party, one must add the nation-wide and overseas Chinese lists of candidates to the sum of the local electoral districts candidates: KMT-NP: 104 (103+1) in 2004 against 186 (144+42) in 2001. PFP: 65 against 89. DPP: 129 against 111. TSU: 40 against 55. Pan-Blue total: 169 in 2004, 275 in 2001. Pan-Green: 169 against 166. These extra candidates do not directly compete in the election and are elected in proportion of the overall result of their party in the electoral districts direct vote.
24 Cf. Caroline Hong, “Pan-Blue allies upset with KMT”, Taipei Times, December 8th 2004.
25 Bruce Jacobs, “Interpretation and implication of Taiwan’s legislative elections”, Taiwan Perspective e-paper, No. 55, January 10th 2005.
26 Besides Chen Hsue-sheng (KMT), there are DPP’s Tuan Yi-kang, Shen Fu-hsiung, Chen Mao-nan, Chou Ching-yu, TSU’s Chen Chien-ming and Hsu Teng-kun, from the PFP Chin Huei-chu, Pang Chien-kuo, Hsieh Chang-chieh, and in the independent ranks Walis Peilin and Su Ying-kuei. Lianhe bao, December 12th 2004, p. A6.
27 The election also saw the return of 11 former legislators: seven KMT, two DPP, one PFP and one independent. Cf. Zhongguo Shibao (China Times), December 12th 2004, p. A9.
28 In Taipei City North (5/5), Tainan County (5/5 plus a green-leaning independent), Tainan City (3/3), Kaohsiung County (5/5), Kaohsiung City 1st district (3/3), Chia-I County (3/3), Chia-I City (1/1), Nantou County (2/2), Taichung City (3/3), Ilan County (2/2), Keelung City (1/1).
29 For a good summary, see David Brown, “Campaign fallout”, Comparative Connections, Pacific Forum CSIS, vol. 6, No. 4, January 2005.
30 Cf. Willy Lam, “Beijing’s post-election policy”, China Brief, vol. 5, No. 1, January 4th 2005.
31 Cf. “Landmark deal on first direct flights in 56 years”, China Post, January 16th 2005.
32 Cf. D. Brown, op. cit.
33 Cf. Chris Cockel, “Rice likely to remain tough toward Taipei, Beijing : scholars”, China Post, Novmber 18th 2004.
34 Cf. Joseph Kahn, “Warnings by Powell to Taiwan Provoke a Diplomatic Dispute”, New York Times, October 28th 2004.
35 Cf. Hsu Yung-ming, “‘Identity problem’ is not political”, Taipei Times, November 22nd 2004.
36 See Arthur Waldron, “Which way forward for Taiwan?”, International Assessment and Strategy Center Online, November 3rd 2004.
37 Cf. AFP, “US to post military officers to Taiwan mission: Jane”, December 19th 2004.
38 “Arms deal to face opposition from 'Pan-Blue' in Legislature again”, China Post, November 9th 2004.
39 And 610,496 valid votes. The number of invalid votes decreased by almost a half (78,940 against 141,150 in 2001, a reduction of 62,210). In March 2004 presidential election, 13,251,719 votes (of them 12,914,422 valid) were cast for an eligible population of 16,507,179. Chen Shui-bian and Lu Hsiu-lien received 6,471,970 votes for only 6,442,452 votes for Lien Chan and Soong Chu-yu.
40 The lower number of invalid votes explains the rest of the “lost votes” in 2004. See previous note.
41 Reasons explaining why DPP voters did not switch to the TSU also include a perception of Lee Teng-hui’s party as too radical and uneasiness in voting for former members and officials of a party, the KMT, long associated with corruption and inner dealings.
42 The KMT even celebrated in February the return to the party of 18 former heavyweights, including Hau Pei-tsun and Lin Yang-kang. Cf. “18 ex-KMT heavyweights rejoin opposition party”, China Post, February 5th 2005.
43 Caroline Hong, “Lien makes August date”, Taipei Times, December 18th 2004.
44 “‘Door is shut’ on Pan-Blue merger, PFP Chairman says”, Taipei Times, December 13th 2004.
45 “Chen tries a new strategy”, Taipei Times, Dec 19th 2004.
46 Lee Wen-chung, “A coalition between DPP, PFP is plausible”, Taipei Times, January 12th 2005.
Top of page

List of illustrations

Title 1. Legislative elections results (by seats)
File image/jpeg, 108k
Title 2. Legislative elections reusults (by votes and in %)
File image/jpeg, 64k
Title 3. Number of local electoral district candidates and success ratio (elected candidates/candidates)
File image/jpeg, 88k
Title 4. Legislative elections overall votes and turnout
File image/jpeg, 72k
Top of page


Electronic reference

Frank Muyard, “KMT: A Trompe-l’œil Victory”China Perspectives [Online], 58 | march - april 2005, Online since 01 April 2008, connection on 01 December 2023. URL:; DOI:

Top of page

About the author

Frank Muyard

By this author

Top of page


The text and other elements (illustrations, imported files) are “All rights reserved”, unless otherwise stated.

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search