|
|
|
|
China's Transition, by Andrew J. Nathan
In the most democratic election in the history of any Chinese society, Taiwan voters went to the polls on December 19, 1992, and elected a new Legislative Yuan. This body occupies an important place in the Republic of China's (ROC) constitutional structure, with powers to pass laws, review the budget, approve the nomination of the premier, and interpellate cabinet members. The full Legislative Yuan was last elected in mainland China in 1947, and all elections since then had been for "supplementary" members, never constituting a majority, so that the Yuan operated essentially as a rubber stamp. With its complete reelection in 1992, the Yuan's whole membership for the first time was directly accountable to the residents of Taiwan and the Pescadores, the territory actually administered by the Republic of China.
The ROC's political system was not yet fully democratic. The president was not popularly elected, the electronic media remained government-controlled and biased against the opposition, and the central and local election commissions were stacked with ruling party members--although these commissions seem to have operated fairly in this election. However in 1992, in contrast to earlier Legislative Yuan elections, no one was excluded from candidacy for political reasons (there were age and educational limits), and candidates' freedom of speech was virtually unlimited. There were no charges of interference with the secrecy of the ballot, and the only charge of vote-rigging arose in one district in Hualien County. The ruling Kuomintang (KMT) won 60.5% of the vote and 103 of 161 seats; the opposition Democratic Progressive Party (DPP) won 31.9% of the vote and 50 seats; and minor parties and nonparty candidates won 7.6% of the vote and 8 seats. 1
As Taiwan's political system becomes more democratized, it is appropriate to turn attention from the process of transition to the dynamics of the institutional system that is emerging. 2 The electoral system used to produce the Legislative Yuan is an unusual one in comparative perspective. It belongs to a type which so far as I know is found elsewhere only in Japan, called a "single-vote, multi-member constituency" (SVMM) system. 3 In this system most electoral districts have more than one representative, but each voter has only one vote. Although based on the Japanese model, the ROC electoral system has a higher maximum number of representatives per district--sixteen compared to a maximum of six in Japan where most electoral districts have three to five lower house (Diet) representatives. Japanese and Taiwanese districts have similar sized electorates, averaging about half a million, although in both places the electorates range widely in size from district to district. The ROC's restrictive campaign laws also follow the Japanese example. 4
This type of system differs both from the single-vote, single-member (SVSM) constituency system found in the United States and Britain, in which each district elects the one candidate with the highest number of votes, and from the proportional representation systems found in some parts of Western Europe and elsewhere in which each voter has as many votes as there are seats. The ROC system produces strategic calculations for candidates, parties, voters, and incumbents that are different from those produced by other types of electoral systems, and that are similar but not identical to those found in Japan. These calculations were fully on display for the first time in the 1992 elections. This analysis of the consequences of Taiwan's electoral system is a preliminary one, based on field observation and simple post-election statistics. Its findings are subject to revision on the basis of more sophisticated voting and survey data that will become available later.
ROC Electoral System
In the 1992 elections, 119 legislators were elected from 27 territorial districts. 5 Unlike in Japan, Taiwan's electoral districts conform to administrative lines, being identical with counties or cities except that the two largest cities are each divided into two districts, North and South. Depending on population size, these districts elected from one to sixteen legislators (see table 8.1). In addition, six legislators were elected by aboriginal voters, who are allowed to cast their ballots anywhere in Taiwan on the strength of their
Besides territorial-district and aboriginal representatives, the ROC system provides for two proportional-representation (PR) lists, one for thirty "national at-large delegates," notionally representing the mainland areas of the ROC over which the government claims sovereignty but not control, and the other for six overseas Chinese delegates. Each party gets a number of seats on these two lists proportional to the total vote cast for that party in the territorial districts. Thus KMT won nineteen at-large seats, the DPP eleven; the KMT won four overseas Chinese seats and the DPP two. Functional constituencies used in earlier Legislative Yuan elections were abolished for the 1992 election.
In five small territorial constituencies that elect only one delegate each, campaign dynamics fit the winner-take-all model familiar to American voters, but the strategic focus of the election falls on the large constituencies that elect multiple delegates. The largest number of seats are won in such districts as Taipei County with sixteen seats, Taipei
City North and South with nine seats each, Taoyuan, Taichung, and Changhua Counties with seven seats each, and Kaohsiung City North and South and Kaohsiung County with six seats each. In these districts not only are there more seats to be won, but also the larger number of votes more heavily influences the distribution of seats from the two PR lists.
Given the system, a candidate in a district with many seats can win with a small percentage of the vote cast, the exact amount depending on how many seats and candidates there are, and on how lopsidedly the top winners win. Mathematically, the number of votes needed for victory in any district by the last winning candidate is the total number of votes, minus the sum of the votes won by the other winning candidates, divided by the total number of remaining candidates, plus one. If there are ten candidates for five seats, and they divide the vote nearly equally, a candidate who gets slightly more than one-tenth of the vote will win. But if the top candidate draws substantially more than one-tenth of the vote, candidates near the bottom of the winners' list will need even less than one-tenth to win. The actual lowest winning votes for each district in 1992 are shown in table 8.1. The minimum winning percentage is even smaller if expressed as a percentage of registered voters. Turnout varies from district to district, averaging 72 percent in the 1992 election. In a total registered electorate of 13,421,170, the smallest number of votes needed to win amounts to as few as 20,000 to 50,000 votes in many districts. 6
Electoral Strategies: Candidates
The electoral system creates an incentive for anyone with a voter base of roughly 20,000 or more to stand for election. 7 The more candidates there are, the fewer votes are needed to win and the greater the incentive for additional candidates to join the race. To discourage frivolous campaigns, the electoral law calls for confiscation of the deposits of candidates who receive less than a certain minimal percentage of the vote. In the 1992 election, 144 candidates lost their deposits, 8 which is a sign of how strong an incentive the system creates for people with small electoral bases to run for a seat.
Once declared, the obvious strategy for each candidate is to solidify a particular electoral base (p'iao-yuan ) of which there are several identifiable types. One is local factions, found mostly in rural areas, although there are some urban factions. According to Ming-tong Chen of National Taiwan University, of 165 Kuomintang members competing in the campaign (including those nominated and unnominated by the party), 72 were supported by local factions. 9 A second type of electoral base is provided by the votes of retired military personnel and their dependents who live in special compounds in major cities; this is supplemented by the votes of serving military personnel loyal to the KMT, many of whom are sent home on leave at election time. Residentially concentrated, with a high turnout rate and a high rate of loyalty to the KMT, the veterans and their families are estimated to cast roughly one million votes--referred to as "iron votes" (t'ieh-p'iao ). Again according to Chen, 56 KMT candidates were supported by the military in this election.
Sub-ethnic constituencies provide a third type of electoral base in some districts. Taipei City, highly urbanized Taipei County, and Kaohsiung City--all districts with large numbers of seats--have heavy populations of migrants from other parts of the island. In Taipei, organized same-place associations (t'ung-hsiang hui ) from Yunlin, Chia-yi, Changhua, and elsewhere, which have memberships in the tens of thousands, supported specific candidates; in Kaohsiung, different candidates appealed to the mainlander vote, the Hakka vote, and the votes of people from Penghu and Tainan. 10 A fourth kind of electoral base consists of the employees and retainers of major enterprise groups (ts'ai-t'uan ), which have large concentrations of voters in certain cities. Candidates with this kind of backing, usually enterprise owners or their scions, are derogatorily referred to as golden cows (chin-niu ). 11
The support of such constituencies is solidified through constituency service and vote buying. Many candidates maintain a network of service centers (fu-wu ch'u) in their districts, through which they maintain contacts with local neighborhood heads and other network leaders. During the 1992 election vote-buying was widely reported, with prices said to range from NT$500 to NT$2,000. 12 But candidates thought to be connected with vote-buying performed less well than expected, according to most press commentators, and a large proportion of voters who accepted money, gifts, or banquet invitations apparently cast their votes independently. This may have reflected both the effectiveness of the anti-vote-buying campaigns mounted by the government and the opposition, and the voters' sense of high stakes in the outcome of the election.
An alternative to the electoral-base strategy is mobilization based on image and ideology. In this strategy candidates reach out through the mass media, handbills, and rallies to try to swing voters with whom they have no network-based connection. In 1992, this type of strategy worked well for two types of candidates. The first was relatively conservative mainlander candidates in Taipei City and Taipei County whose campaigns stressed their educational qualifications, incorruptibility, and willingness to stand up against the party machine and politics as usual. Many of them had been passed up by the ruling party for nomination in favor of factionally supported candidates. The two biggest vote-getters, former environmental administrator Chao Shao-k'ang and ex-Minister of Finance Wang Chien-hsuan, had not been nominated because they were in the cabinet at the time of the nomination process. They then resigned and ran against the party machine with great success (Chao had been the top KMT vote-getter in the 1989 Legislative Yuan elections). The initial impression from street interviews was that Chao, Wang, and other conservative mainlander candidates in Taipei drew their support from the middle class, both mainlander and Taiwanese.
A second kind of image- and ideology-based strategy was used by opposition candidates, stressing their heroic records of political dissidence, past victimization by the regime, willingness to stand up for Taiwanese interests against the mainlander-dominated regime, their commitment to social welfare, women's rights, and Taiwan independence, and their expertise in foreign affairs gained from years of (voluntary or involuntary) residence overseas. Some of these individuals are charismatic, and they attracted large, excited crowds. As part of their strategy, they sometimes engaged in street politics, such as leading groups of supporters to demonstrate and hurl eggs at KMT or government offices. Street interviews suggested that these candidates' base was drawn predominantly from urban workers and rural farmers, and was chiefly male. Several DPP candidates stated that their party is supported by 60 percent of male but only 10 percent of female voters. Despite their lower-class electoral base, DPP campaigners placed about as much stress as do KMT candidates on their educational backgrounds, often including Ph.D.'s earned at overseas institutions.
The traditional constituency-based electoral strategy naturally tends to be preferred by candidates who have a solid electoral base on which to rely. Given the history of the authoritarian regime, these are almost exclusively ruling-party candidates, often incumbents. Their campaign literature stresses sincerity, service, and endorsement by popular leaders like President Lee Teng-hui. They avoid contamination from controversial issues such as Taiwan's international status and often do not bother to display the KMT logo prominently in their campaign material. Their appeal is personal and their goal is to protect their minimum winning vote.
In contrast, candidates using a mobilizational strategy make heavy use of issues like Taiwan independence (for DPP candidates) or corruption (for insurgent KMT candidates) and, in the case of opposition candidates, of the party logo featuring a green silhouette of the island of Taiwan. Instead of seeking to protect a minimum vote base, mobilizational candidates appeal to their supporters to elect them with the highest vote count in their district (tsui-kao-p'iao ), making the argument that this will send a strong message to the government and will strengthen the candidate's ability to represent his or her supporters in the Legislative Yuan. Lacking an organizational base, they have no way to know whether their projected vote count is in the safe range. Public opinion polls provide little useful information as a majority of Taiwan voters report themselves undecided throughout the campaign. By calling for a high vote count, a candidate hopes to overcome the voters' tendency to vote for their second preferred candidate when they believe that their preferred candidate is safely in the lead. Since many of the image-based candidates are insurgents against or loners within their party, they are not concerned that their high-vote strategy may knock a second candidate as do KMT candidates on their educational backgrounds, ofteof their own party out of the race.
In some ways, the campaign rules are unfavorable to mobilizational campaigning. The campaign is limited to ten days, TV news coverage and television advertising are limited, and spending restrictions are severe although widely ignored. But many candidates have long-established public images based on previous public service or activism in the opposition, and many in both parties rely heavily on newspaper advertising and attractive printed handbills. The literacy rate in Taiwan is high, and the novelty of this first fully democratic election seemed to catch the voters' attention. Some rallies drew audiences of close to 10,000. In general, the campaign seemed to mark one stage in a long and perhaps partial shift from faction to image. One sign of this was that, according to one study, more than 80 percent of candidates who ran campaign advertisements on underground cable TV stations were victorious. 13
The system tends to generate intraparty feuds. Candidates who think they are losing have a better chance to save themselves by attacking those of their own party who seem to be winning--in order to peel away party loyalists' votes from them--than by trying to win votes away from candidates of the other party. Last-minute charges against a leading candidate's character by a member of the same party were a feature of campaigns in several districts and by both parties. Since the campaign period is short, there is not enough time for the press to investigate such charges and the targeted candidates have to respond very quickly.
Party Strategies
Under this electoral system, it is rational for a party to nominate the number of candidates that it expects to be able to elect with the number of votes it expects to have, and then to use party organization to allocate its votes among the candidates so that each gets just a little more than the minimum number needed to win. The process of spreading the party vote among a party's candidates so that few votes are wasted on the top candidates is called "vote allocation" (p'ei-p'iao ). Only the KMT has the organizational capability to do this. In each electoral district, KMT candidates sit down with local party officials to decide which candidate will concentrate on which organizations (e.g., the farmer's association, veterans' association, irrigation association) and areas (neighborhoods, villages). Leaders in those organizations and areas are then assigned to deliver quotas of votes to the designated candidate. Party loyalists are discouraged from voting for the candidate they prefer in favor of voting for the candidate that the party organization assigns to them.
The combination of strategic nomination and vote allocation practiced by the KMT is referred to as "organizational warfare," as distinguished from the more mobilizational style used by the opposition DPP. More a congeries of like-minded notables than an institutionalized party, the DPP lacks the capability for organizational warfare. 14 Still, its nomination strategy in the 1992 election reflected the logic of the electoral system. The party put forward only 59 candidates for the 125 territorial constituency seats in the election, nominating in each district only the number it thought that it could elect and avoiding the more adventurist strategy of fielding additional candidates at the risk of dividing the party vote among too many aspirants.. Also, the DPP's most famous senior figures came forth as its nominees. As a result, 63 percent of the DPP's nominees in territorial constituencies (37 of 59) were successful. 15 But given the size of the total vote the party's candidates attracted, some districts might have elected an additional DPP legislator had the party been capable of dividing its vote organizationally.
On the other hand, the KMT's rather similar victor-to-candidate ratio of 58 percent (73 of 125) must be read as signaling a decline in its historically strong party discipline. In many districts, KMT candidates who were denied the party nomination ran anyway, some of them supporters of the KMT's "non-mainstream faction" rather than of the group that controlled the party machinery. When such candidates lost, they carried significant numbers of party votes away with them. Among 43 insurgent KMT candidates, only seven were successful, but some of them were high vote-getters.. 16 These candidacies made it difficult for the party to allocate votes reliably. It was also considered an indication of the decline in voter discipline that the vote totals of candidates supported by KMT-affiliated enterprise groups were lower than expected, even though 27 of 38 such candidates were elected. 17
The KMT faced the dilemma of whether to endorse popular insurgent candidates who ran despite the denial of party nominations, thus rewarding their insurgency but gaining the benefit of their votes for its two PR lists. The party offered insurgent candidate Wu Tzyy (Wu Tzu) endorsement but he refused it; it expelled candidate Ch'en Che-nan from the party but still got the benefit of his vote in its PR columns because he had been a KMT candidate when the ballots were drawn up. Many of the nonnominated candidates came from a group called the New KMT Front, which has called for party reform. According to post-election press commentary, the message to the party organization was that it needs more qualified candidates, and the message to candidates was that they have to count more on themselves.
Voter Strategies
The dilemma facing voters in such a system is that while most districts have more than one seat, each voter has only one vote, and that vote cast for a preferred candidate may be wasted if the candidate is widely popular. But if the voter switches to a candidate who seems less likely to win, the risk is that the preferred candidate may lose. To counter voters' temptation to switch in this way, many leading candidates on the last two days of the ten-day campaign advertise that they are in danger of losing, a tactic that looks odd to observers familiar with the American single-winner system in which it makes more sense to claim to be winning, even while warning one's supporters against complacency.
In this context, party-coordinated vote allocation serves the interests of those voters who want the KMT to win the maximum number of seats. Vote allocation assures them that the maximum number of their preferred candidates will win so long as these like-minded voters observe party discipline. Many independent and opposition voters engage in what is called "automatic vote allocation" (tzu-tung p'ei-p'iao ), or family-level allocation of the vote 18 in which members of a family agree among themselves to divide their votes among candidates they like. This family-level division is encouraged by the DPP, which lacks the organizational network to engage in more structured vote allocation. The strategy seems to be useful, judging by the party's ability to distribute votes among its candidates in such large districts as Taipei North and South, Kaohsiung North and South, Pingtung, and Tainan County and City.
On the other hand, some voters may be less concerned with the distribution of seats in the Legislative Yuan than with sending a strong signal to the government by piling up votes on a candidate who they believe carries a certain symbolism. The enormous votes for Chao Shao-k'ang (KMT insurgent) and Lu Hsiu-yi (DPP) in Taipei County, and for Wang Chien-hsuan (KMT insurgent) and Ch'en Shui-pien (DPP) in Taipei North sent such messages. But it is in the nature of the ballot that the message it sends can be both strong and vague. High votes for insurgent KMT candidates, for example, were clearly votes against the so-called "KMT mainstream" of President Lee Teng-hui and Party Secretary-General James C. Y. Soong, but they also seemed to be votes in favor of personal probity, courage, high educational qualifications, and substantial political experience. Beyond this their meaning was open to interpretation.
A number of voters told me that they were voting for the DPP to create a force to check and balance the KMT. The DPP made this more feasible by moderating its platform in comparison to its position in the National Assembly elections in December 1991 in which it performed relatively poorly. The party replaced its 1991 "Taiwan Independence Platform" with an umbrella position of "One China, One Taiwan." Although it foreclosed reunification, this slogan was vague enough to accommodate a variety of positions from Taiwan independence to maintenance of the status quo. The party also developed a series of positions on welfare statism, lower taxes, women's rights, and so forth, that enabled its candidates to escape from the single-issue trap of 1991. Of its 78 candidates, 62 belonged to moderate factions. 19 Voters in 1992 seemed to view the DPP less as a revolutionary force and more as a balancing partner in a stable democratic system. Since this was the third election in which the DPP participated as a party, they may also have been encouraged to support it by the experience of previous elections carried out without excessive violence or post-election retribution from the mainland.
Voters with whom I spoke were also aware that, for the first time, they were electing the entire Legislative Yuan and not merely supplementary members, as in the past. The election outcome would directly affect the conduct of government. Thus empowered, the voters apparently wanted to send a message of caution to the KMT without throwing it out of office. In a similar spirit, some who voted for KMT insurgent candidates may have wanted to balance the KMT mainstream faction within the party without intending to reject its leader, President Lee Teng-hui.
The election also represented another step in the process by which the native Taiwanese majority has been taking over power in the island's political institutions from the previously dominant mainlander elite. Mainlanders, who made up 60 percent of the membership of the former Legislative Yuan, have now been reduced in number to 22 percent, 20 and mainlander politicians campaigned mostly in the Taiwanese language. One result of the election was the appointment of the first Taiwanese premier, former Taiwan Provincial Governor Lien Chan. Some DPP candidates won on strong Taiwanese-identity campaigns, but this was not always a winning card. A number of candidates of the insurgent KMT "wisdom coalition" ran campaigns based on Taiwanese identity, which were not successful, and in Taipei City, twelve of the eighteen successful candidates were mainlanders, even though the mainlander population of Taipei is only 27 percent. 21 The voters showed that they had more than ethnic identity on their minds, and fears of Taiwanese-mainlander polarization were eased.
Incumbent Strategies
The newly elected legislators have relatively short three-year terms, so incumbents will need to cultivate their local constituencies continuously. Those who ran constituency-based campaigns will work hard to channel central government project funds and social services into their local bases. Those who ran image- and ideology-based campaigns will compete for media attention and, in the case of opposition legislators, engage in street politics to show that they are serving the values of the constituencies that elected them. Each type of candidate may also try to diversify his or her base, with constituency-based candidates trying to polish their images for honesty or issue commitment, and issue-based candidates trying to provide local services, help constituents with government bureaucracies, and attend more weddings in their districts.
The Legislative Yuan is expected to play a larger role in policy-making than it did in the past. As the only directly elected organ of ROC government, it is the one with the most popular legitimacy. Legislators are now more answerable to their constituents than to party leaders, and with the disappearance of the rubber-stamp majority of elderly mainland-elected legislators, the Yuan's bargaining power vis-Ã-vis the cabinet will be at a new high. In a political culture that values consensus, the DPP's 31 percent of the seats (50 of 161) entitles it to be consulted extensively by the government. Some DPP members even believe it entitles the party to seats in the cabinet.
Although the ruling KMT has 64 percent of the seats (103 of 161), the poor performance of "organizational warfare" in this election weakened the ability of the party to control its members' votes in the legislature. On issues ranging from mainland and foreign policy to social welfare and administration of the six-year infrastructure program, the administration will reach out for support in the Legislative Yuan where it can find it, which will not always be along party lines. On a number of these issues, the interests of the "mainstream" (pro-Lee Teng-hui) KMT faction converge more closely with those of the moderate or "Formosa" faction of the DPP than with the "non-mainstream" minority faction within the KMT. Legislators seeking to prove their worth to their constituents will exercise the interpellation power more aggressively than it has been used in the past. Party discipline is likely to be relatively ineffectual in overcoming the priority of local interests in legislative politics. In short, the election produced what can be called a locally penetrated center. 22
Although the candidate who ran on the slogan "Taiwan priority," Wu Tzyy, failed to be elected, his slogan reflects the direction in which the Legislative Yuan will probably push government in the coming years. Since the voters did not vote on strict ethnic lines, Taiwan priority does not necessarily mean Taiwanese as opposed to mainlander priority. But the interests of the residents of the island, Taiwanese and mainlander alike, will dominate the legislators' concerns. One obvious consequence is that the chances for reunification on the terms currently being offered by the PRC are dimmer than ever, and the ROC's "flexible diplomacy" is likely to become even more flexible in the search for the international respectability that Taiwan's residents crave. 23
Winning seats did not require candidates to build coalitions across the electorate to "aggregate interests." Local factions, sub-ethnic constituencies, enterprise groups, and issue- and ideology-based groups were all able to place their representatives in the Legislative Yuan. From now on, the Yuan is likely to reflect the shifting levels of ideological and ethnic polarization among voters. The 1992 election seemed to reveal a moderate electorate, but one containing some polarized groups. Voters crossed ethnic lines to vote for candidates with images of personal probity, and they rewarded the DPP for moderating its party platform on independence. But some mainlander candidates came to power chiefly with mainlander votes, and some groups of voters supported fiery DPP politicians who have devoted their careers to pushing aggressively for political change and who openly favor Taiwan independence.
Now that the government has ceased to use the criminal law to censor political discourse, the electorate is the only arbiter of what can be discussed and how. Since the electoral system makes it possible for different groups of voters to reward different types of politicians, we can expect Taiwan's politics to continue to be as lively as ever. Politicians of all shades of opinion will compete for the public ear, and while the danger of polarization is always present, powerful forces are pushing the electorate toward an aggregate preference for moderation. The most important of these forces are Taiwan's growing wealth and its vulnerability to military and political pressure from the mainland.
Conclusion
The 1992 Legislative Yuan election solidified a tacit pact of peaceful coexistence and peaceful competition between a former authoritarian ruling party and a former anti-system opposition movement. Opposition leaders who were in prison or exile just a few years ago took seats in the nation's highest elected body. Despite the bias of the electronic media, the dominance by KMT officials of central and local election commissions, stringent campaign limits, and widespread charges of vote-buying, the electoral system was sufficiently acceptable to both parties so that each participated in it with full energy and each accepted its verdict. The ruling party hailed the election as a step toward two-party democracy, acknowledging the possibility of a future peaceful transfer of power. The DPP stated its intent to "fulfill the responsibilities of an opposition party in a modern democratic state." 24 In short, after years of struggle between regime and opposition, the opposition has been granted legitimacy within the system and has accepted the legitimacy of the system. In retrospect, it was a remarkably peaceful transition, despite some large street demonstrations, political arrests, and the breaking of some desks and chairs in the Legislative Yuan over the preceding few years.
Not only is the ROC electoral system unusual in comparative perspective, but its relationship to the party system reverses a common historical pattern. Most party systems were shaped through history by their countries' electoral systems, among other factors. In Taiwan, the party system was formed in the absence of national-level elections, although Taiwan has been holding hard-fought local elections for years. Now for the first time the electoral system has started to affect the party system. 25 Its initial impact appears to be to promote internal fragmentation in the ruling party, while encouraging a fragmented opposition party to remain so. Other forces are also at work, so the outcome is by no means predetermined.
Indeed, the electoral system itself is subject to change at the parties' initiative. A relatively minor reform that has been discussed would allow each voter two votes, one for a local candidate and the second for the two proportional-representation lists. Such a change would probably favor the opposition because it would allow voters beholden to ruling-party candidates to cast a second, balancing vote for the opposition. It would also create an incentive for each party to pay more attention to its platform and to its image as an organization. A bolder reform would be to shift to the single vote, single member system, with one legislator per district and smaller districts. Such a change would offer potential gains and risks to each party. Although the current system allowed an opposition with less than one-third of the vote to establish a strong foothold in the Legislative Yuan, it also enabled the ruling party, with its strong local bases, to maintain a majority position that will be difficult to shake. For the opposition to move from a minority to a majority under the current system, it would have to recruit scores of additional attractive candidates and painstakingly build local party machines that could compete with local pro-KMT factions. If just one seat per district were to be filled, the DPP could field its strongest candidates in one-on-one battles with KMT nominees, and it might be able to break out of what now looks likely to be protracted minority status. But under such a system, the KMT would be able to turn its 60 percent electoral majority into a near-100 percent control of the Legislative Yuan if it could get more than 50 percent of the vote in each district.
The complexity of the system makes it difficult to say unambiguously whose interests it favors. It also creates numerous vested interests among currently elected politicians with their small electoral bases, and the constituencies they serve. So far, neither party has decided to support change in the system, and it may be difficult to construct a consensus for change in the foreseeable future.
Note 1: Lien-he pao , December 20, 1992, p. 1. Party totals include self-nominated candidates without party endorsements. Back.
Note 2: An account of the early stages of the transition is in Yangsun Chou and Andrew J. Nathan, "Democratizing Transition in Taiwan," Asian Survey 27:3 (March 1987), pp. 277-299; reprinted as chapter 8 in Andrew J. Nathan, China's Crisis: Dilemmas of Reform and Prospects for Democracy (New York: Columbia University Press, 1990). Back.
Note 3: Cf. David Butler, "Electoral Systems," in Butler et al., eds., Democracy at the Polls: A Comparative Study of Competitive National Elections (Washington, D.C.: American Enterprise Institute, 1981), pp. 11-19; Rein Taagepera and Matthew Soberg Shugart, Seats and Votes: The Effects and Determinants of Electoral Systems (New Haven: Yale University Press, 1989), p. 28. Secretary Hsu Kui-lin of the Central Election Commission told us that this system dates back to the 1947 elections on the mainland, and was modeled on the Japanese system. Back.
Note 4: Japan had 130 electoral districts electing 512 lower house representatives, compared to 27 ROC districts electing 119 delegates. See Gerald L. Curtis, The Japanese Way of Politics (New York: Columbia University Press, 1988), pp. 165-175. Japan later abandoned this system. Back.
Note 5: Territorial districts are those listed in table 8.1, which excludes the aboriginal, at-large, and overseas Chinese constituencies. Except as otherwise noted, the entire discussion below concerns these territorial constituencies. Back.
Note 6: This is leaving aside certain small districts such as P'eng-hu, Lienchiang, and Chin-men counties where even smaller vote counts are sufficient to win because of small population size. Registered voter and turnout data are courtesy of the Central Election Commission. Back.
Note 7: The minimum electoral base needed to encourage a candidate to run depends on the size of the district, the number of seats to be filled, and the projected size of the vote for the top vote-getters. Back.
Note 8: China Post , December 21, 1992, p. 15. Deposits are lost when the candidate receives less than 10% of the quotient resulting from dividing the total number of eligible voters in the candidate's constituency by the total number of the officials to be elected therein. For example, if there are five candidates in a district, a candidate will lose his deposit if his vote falls below 2% of the total of registered voters. The more candidates, the smaller the percentage of the vote required to avoid forfeit. Back.
Note 9: Briefing at Institute for National Policy Research, December 14, 1992, later updated by Chen in personal communication. See Joseph Bosco, "Taiwan Factions: Guanxi , Patronage, and the State in Local Politics," Ethnology , 31:2 (1992), pp. 157-183. Back.
Note 10: Chung-kuo shih-pao chou-k'an (American edition), no. 52 (December 27, 1992-January 2, 1993), pp. 38-39. Back.
Note 11: A news magazine article claims that the Yi-kuan-tao, a religious sect, is also a KMT electoral base, casting its votes in an organized way for specific KMT candidates (ibid., pp. 40-41). Back.
Note 12: China News , December 15, 1992, p. 1. The New Taiwan dollar (NT) is exchanged at approximately 25 to US$1. Back.
Note 13: Ibid., December 21, 1992, p. 2. Back.
Note 14: The DPP's 50 legislators come from 14 identifiable factions, according to The China News , December 21, 1992, p. 3. Back.
Note 15: In this calculation (and for the KMT below) only candidates with official party nominations were counted. Back.
Note 16: Chung-kuo shih-pao , December 20, 1992, p. 6. Back.
Note 17: Tzu-li wan-pao , December 20, 1992., p. 6. Back.
Note 18: Tzu-li wan-pao , December 20, 1992, p. 2. Back.
Note 19: This number includes all the officially nominated DPP candidates, in territorial, aboriginal, at-large, and overseas constituencies (Lien-he pao , December 20, 1992, p. 2). Back.
Note 20: Kau Ying-mao in Chung-kuo shih-pao , December 21, 1992, p. 10. Back.
Note 21: Chung-kuo shih-pao , December 20, 1992, p. 7. Back.
Note 22: Thanks to Ming-tong Chen for this term. Back.
Note 23: These trends have been clear for some years, as argued by Andrew J. Nathan, "The Effect of Taiwan's Political Reform on Taiwan-Mainland Relations," in Tun-jen Cheng and Stephan Haggard, eds., Political Change in Taiwan (Boulder, CO: Lynne Reinner, 1992), pp. 207-219; reprinted as chapter 9 in Nathan, China's Crisis. Back.
Note 24: Tzu-li tsao-pao , December 20, 1992, p. 2. Back.
Note 25: I am grateful to Gerald L. Curtis for suggesting this point. Back.
Email this citation