Many of the decisions made in 2005-2007 hinged on specific factors or conditions that may not always be present in future redistricting cases. This section explores what might be expected to happen if certain variables are altered. Above all, what would change if redistricting were conducted under unified government? Additionally, how might parties adapt to obtain better outcomes?
21 In D1 and D2, Lin had 10.8% and 9.6%, respectively. Cheng had 11.9% and 11.5% respectively. When the KMT plan won in the lottery, both Lin and Cheng declined to run for re-election.
22 Rather than run in the heavily pro-DPP D3, Liao managed to win a spot on the party list in 2008.
23 Wu eventually did not run for re-election. Instead, he took a position as Vice President of the Examination Yuan in the new Ma Administration.
The non-partisan guidelines set out in the CSERA and in the CEC’s directives clearly constrained decision-makers in the 2005-2007 round. In several cases, the population requirements and the need to respect township boundaries and historical or physical links made it difficult for politicians to come up with alternate plans, especially in the cities and counties with fewer seats. In the future, these constraints will be joined by another: the current legislative districts. Especially in cases in which the number of seats in the city or county does not change, the existing districts will be a powerful focal point. In addition to people being used to them, incumbents will have invested significant time and resources in building up personal votes along those boundaries.
Nevertheless, two factors might erode the power of these constraints. First, administrative reform in 2010 combined Taichung City and County, Tainan City and County, and Kaohsiung City and County, which collectively elect 22 of the 73 (30.1%) district seats. The new, bigger cities have more seats and more possible combinations of administrative districts. For example, consider Taichung. Whereas the old Taichung City and County had three and five seats, respectively, the new Taichung City has eight. Moreover, geography limited the choices in 2005-2007, since Taichung County wrapped around Taichung City (see Figure 2). There were not many possible ways to deal with Taiping, for example, since it is connected to the township to the east by only a few small mountain roads. In the future, Taiping might combined with one or more administrative districts from the former Taichung City. This added flexibility could mean significantly more choices, and that, in turn, could lead to more partisan conflict.
Second, the tradition of keeping groups of townships together might fade. There are no hard institutional rules that country assembly districts or other groups of townships have to stay together. They are held together by cultural ideas of what is natural and by politicians’
mobilization networks. If one party is disadvantaged because the other party has a politician with a potent mobilization network, it might want to break up such a grouping. Over time, parties could chip away at the cultural notions of which townships naturally belong together. This would effectively increase the number of building blocks and choices, which in turn might lead to more partisan conflict.
While the effect of guidelines would not be significantly affected by unified government, this would have a major impact on the composition of the decision-making bodies and their capacity to affect the final plan. In 2005-2007, the two major parties were on relatively even footing. The KMT controlled a few more local governments, but the DPP also held several cities
and counties. This meant that the KMT enjoyed a slight overall advantage in the first stage. The DPP had a small advantage in the CEC. The blue camp ran the legislature, so the KMT should have had an advantage in the third stage. However, roll call votes are costly, so it is difficult to ram things through the legislature, especially with the narrow majorities that are often found in divided government. Control of the legislature turned out not to be very valuable in the third stage, though it was critical for the fourth stage. In the fourth stage, the two parties were on equal footing, and each had a 50-50 chance in the lottery. Overall, divided government meant that neither party was able to dominate the process.
Under unified government, things would be very different. The first two actors, the LECs and the CEC, would continue to have a strong bias in favor of the local executive’s party and a mild bias in favor of the president’s party, respectively. However, the last two stages would be dominated by the governing party. Most importantly, since the speaker and premier would come from the same party, they could easily replace any plan with their party’s most favored plan. In cities and counties governed by the ruling party, the opposition party could simply be ignored.
The ruling party would control all four stages of the process, and so it could unilaterally impose its most favored plan. In cities and counties governed by the opposition party, the opposition party’s best hope would be for the existing focal points to create a consensus or for the size and complexity of the city or county to make coming up with an alternate plan difficult. However, if the governing party were able to settle on an alternate plan, the LEC plan would inevitably be replaced. If more radical plans help spur the other side to make compromises in order to come up with an alternative, opposition-led LECs might strategically propose more moderate plans.
LECs led by the governing party would face no such moderating impulse. In short, under unified government, the most significant restraint on the ruling party would be the need to satisfy the legal guidelines. Within those constraints, the ruling party would be able to impose its preferred plans, regardless of the opposition party’s objections.
The one situation in which an opposition party in unified government might obtain a better outcome than the governing party is if the case were to fall into the LEC-dominant pattern.
In that case, complexity would prevent the governing party from settling on an alternate plan.
Simply put, the LEC’s institutional resources might allow it to resolve problems that individual politicians from the other party could not. However, the other party might be able to overcome these challenges if it were to expend more resources. In 2005-2007, parties largely stayed out of the first two stages, leaving it up to individual legislators to defend their own interests. In the
future, parties might try a more centralized approach in order ensure that their interests are more systematically defended. This might mean that parties get involved during the first stage, perhaps with experts who can design alternate plans and with the political clout to help resolve disputes within the party. If parties do centralize efforts, the LEC-dominant pattern could disappear. In any situation with viable alternatives, both sides would be able to settle on a partisan plan. In this case, some smaller cities and counties would still fall into the consensus pattern, but most seats would be drawn in the partisan conflict pattern.
VI. Conclusion
While politicians doggedly pursued partisan advantage throughout the redistricting process, neither of the two major parties was able to obtain a complete victory. In some cities and counties, one party won. In others, the other party won. In still others, consensus plans were passed. Several key factors combined to produce this result.
One key was that the set of viable plans was quite restricted. In the beginning of the process, the CEC guidelines insisted on big building blocks. Because all but the biggest townships had to be kept whole and groups of smaller townships with historical associations were usually kept together, the number of viable combinations of townships that could meet the population requirements was quite small. Politicians simply could not draw ideal districts with lines that dart in and out to include their best areas and avoid their weakest ones the way that American legislators do. Because of this, the difference between any two alternatives was relatively small.
At the end of the process, the lottery provision also served to eliminate many alternative plans.
Individual legislators did not have the ability to introduce their own plans, so any alternate plan had to have the approval of the party leader. Party leaders do not like to back legislation that splits their party, so they only supported plans that no stakeholders within their party objected to.
Thus, the obstacles to producing viable plans reduced the ability of politicians to pursue partisan gain.
Second, the process empowered formally non-partisan actors in the first two stages. As formally non-partisan actors, the LECs and the CEC, had to publicly justify their actions in non-partisan terms. Moreover, the fact that the first non-partisan body, the LECs, had to win the approval of a second non-partisan body, the CEC, probably encouraged the LECs not to blatantly ignore the non-partisan guidelines in favor of partisan advantage. In the event, the
LECs generally produced plans that looked reasonable to the average observer. There was a striking absence of strangely shaped districts, districts that crossed obvious geographical barriers, or districts that violated other commonly understood conventions. There were clearly partisan motives behind the various designs, but, for the most part, the plans would not have seemed objectionable to an imaginary observer well-versed in local geography, economics, and culture but completely ignorant of politics. If the legislators, whose careers depended directly on the new districts, had had the first opportunity to draw the district, they might not have been so restrained.
Divided government was the third and perhaps most critical key. Because the DPP controlled the executive branch and the KMT controlled the legislature, the final stage effectively became a lottery. This meant that both sides had a fifty percent chance of overturning any plan the other side managed to push through, as long as they could negotiate an alternative plan amongst themselves. Since more extreme partisan advantages increased the incentives on the other side to make compromises to cobble together a viable alternative plan, the possibility of the lottery probably mitigated partisan conflict. Moreover, partisan control of local governments in 2005-2007 was also fairly evenly divided. The KMT controlled more local governments, but the DPP also controlled several populous cities and counties. This meant that the first mover advantage was not universally enjoyed by one party.
The 2005-2007 redistricting process was fairly smooth, especially given the extreme partisan polarization of the period. However, this does not mean that future redistricting processes will not devolve into the sort of partisan warfare that could damage the perceived fairness of the democratic regime.
For one thing, politicians may learn from previous experience how to manipulate the system. This was the experience of Arizona in the 2011 redistricting. A decade earlier, the process had gone relatively smoothly. Rather than build on that tradition of non-partisanship, Arizona politicians figured out how to place closet partisans into non-partisan bodies. As Arizonans learned of these maneuvers, the non-partisan commission lost credibility, and the courts eventually had to step in to replace commission members (Cain 2012, 1833-1834).
Taiwanese politicians have had a decade to reflect on the redistricting process. They will likely come up with some new tricks in the next round.
For another, Taiwan may not have divided government in the next round of redistricting.
If one party controls both the legislature and the executive, the last stage will be quite different.
The governing party will have the ability to unilaterally implement its favored plan. Any plans
favorable to the opposition can be overturned, as can plans that are neutral or merely give the governing party a mild advantage.
This paper has implications for electoral reforms. Based on the experiences of New Zealand, Canada, Australia, and others, it might seem that the most promising avenues for reform are to allow neutral bureaucrats to draw the districts. However, this might not work in Taiwan as well as it has in other countries. Because of Taiwan’s authoritarian heritage as a KMT party-state, the bureaucracy, especially at the more senior levels of the civil service, is often perceived as leaning toward the blue camp. Taiwan simply does not have a century-long tradition of bureaucratic neutrality to draw upon. In New Zealand, the electorate might be willing to believe that the Government Statistician draws boundaries without regard to their partisan impact, but the Taiwanese electorate is less likely to swallow this assumption. For better or worse, Taiwan probably has to rely on people appointed by partisan politicians. At most, it might be possible to weed at the most overtly partisan nominees to the CEC by requiring a supermajority from the legislature for confirmation. On the other hand, giving the CEC stronger influence over the final boundaries, either by making their decisions final or by requiring a supermajority within the legislature to overturn them, might help by empowering the CEC, the least partisan of the major actors.
At present, several concrete reforms are being discussed. One is to increase the number of seats in the legislature. A significant increase in the number of geographic district seats would have two important effects. First, since each city or county would have more seats, drawing lines would become more complex. There would be fewer cases of consensus and more cases in which the politicians could seek partisan advantage. Second, with fewer people in each district, many of the large building blocks, the townships, will be too large, and planners will have to use the smaller building blocks, the li. With smaller building blocks, planners have much more leeway to seek political advantage. Thus, if the number of geographical seats is significantly increased from the current 73, an increase in attempts at partisan manipulation should be expected.
Another proposed reform is to change the system from the current MMM system to a German-style mixed member proportional (MMP) system. In MMP, a party’s seat share is determined by its share of party list votes. This reform would eliminate the incentive for most of the partisan maneuvering discussed in this paper. A party could produce an extreme redistricting plan, but any extra seats it gained in the nominal tier would be offset by losses in the party list
tier.24 MMP would thus reduce redistricting to a mostly non-partisan exercise.
In the current MMM system, redistricting has the potential to incite severe partisan conflict, to produce an unfair field of competition, and even to affect the legitimacy of the democratic system. While these extreme outcomes were largely avoided in 2005-2007, that is no guarantee that future redistricting efforts will go so smoothly. Removing this time bomb from the political process would be beneficial to the future of Taiwan’s democracy.
* * *
Received: 2015.05.08; Revised: 2015.07.20; Accepted: 2015.11.20
24 If overhang seats are awarded, winning an extra district seat may not always be offset by the loss of a list seat.
References
I. References in English
Butler, David, and Bruce Cain. 1992. Congressional Redistricting: Comparative and Theoretical Perspectives. New York: Macmillan.
Cain, Bruce. 2012. “Redistricting Commissions: A Better Political Buffer?” Yale Law Journal 121: 1808-1844.
Cain, Bruce E., and Janet C. Campagna. 1987. “Predicting Partisan Redistricting Disputes.”
Legislative Studies Quarterly 12(2): 265-274.
Courtney, John C. 2008. “From Gerrymanders to Independence: District Boundary Readjustments in Canada.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Cox, Gary W., and Jonathan N. Katz. 2002. Elbridge Gerry’s Salamander: The Electoral Consequences of the Reapportionment Revolution. New York: Cambridge University Press.
Cox, Gary W., and Mathew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press.
---. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. New York: Cambridge University Press.
Denzau, Arthur T., and Douglass C. North. 1994. “Shared Mental Models: Ideologies and Institutions.” Kyklos 47(1): 3-31.
Handley, Lisa. 2008. “A Comparative Survey of Structures and Criteria for Boundary Delimitation.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Jacobson, Gary C. 1990. The Electoral Origins of Divided Government: Competition in U.S.
House Elections, 1946-1988. Boulder, CO: Westview Press.
---. 2001. The Politics of Congressional Elections, 5th ed. New York: Addison Wesley Longman.
Lublin, David. 2008. “Race and Redistricting in the United States: An Overview.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Lujambio, Alonso, and Horacio Vives. 2008. “From Polities to Technicalities: Mexican
Redistricting in Historical Perspective.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Lyons, Michael, and Peter F. Galderisi. 1995. “Incumbency, Reapportionment, and U.S. House Redistricting.” Political Research Quarterly 48(4): 857-871.
McDonald, Michael P. 2004. “A Comparative Analysis of Redistricting Institutions in the United States, 2001-2.” State Politics and Policy Quarterly 4(4): 371-395.
McRobie, Alan. 2008. “An Independent Commission with Political Input: New Zealand’s Electoral Redistribution Practices.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Medew, Rod. 2008. “Redistribution in Autralia: The Importance of One Vote, One Value.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Moriwaki, Toshimasha. 2008. “The Politics of Redistricting in Japan: A Contradiction between Equal Population and Respect for Local Government Boundaries.” In Redistricting in Comparative Perspective, eds. Lisa Handley and Bernard Grofman. New York: Oxford University Press.
Romer, Thomas, and Howard Rosenthal. 1978. “Political Resource Allocation, Controlled Agendas, and the Status Quo.” Public Choice 33(4): 27-43.
Wu, Chung-li. 2008. “A Simple Model for Predicting the Outcome of the 2008 Legislative Yuan Elections in Taiwan.” Issues & Studies 44(4): 1-28.
II. References in Chinese
Teng, Chih-sung, Chin-en Wu, and I-jung Ko. 2010. “Xuan piao kong jian fen bu yu xi ci pian cha: Di liu qi jie li wei xuan ju de kao cha” [Spatial Distribution of Votes and Disproportionality: An Observation of the 6th and 7th Legislative Elections]. Journal of Electoral Studies 17(1): 21-53.
Yu, Ching-hsin. 2008. “Ru he ping gu xuan qu chong hua de zheng zhi xiao guo” [How to Evaluate the Political Effects of Redistricting]. In Ru he ping gu syuan cyu bian cian fang fa lun de tan tao [How to Evaluate Electoral System Change: A Methodological Approach], eds.
Chi Huang and Ching-hsin Yu. Taipei: Wunan.