Archive for May, 2014

protests and majoritarianism

May 2, 2014

Why are there so many protests?  Of course, there are proximate causes for each incident, but I’m interested in a deeper structural explanation.  I think the real problem is that Taiwan’s government is too majoritarian.  It is perhaps not news that a majority can determine policy here, but I wonder if people realize how the system has shifted over the past two decades.  I think it is fair to say that Taiwan is currently more majoritarian than at any time since Chiang Ching-kuo was president.

What does it mean to say that Taiwan is majoritarian?  In a majoritarian system, minorities have few opportunities to block the majority, or even to negotiate for slight changes.  Moreover, majoritarian systems often create majorities out of mere pluralities.  That is, fragmented groups are pushed together, and then the resulting majority is endowed with tremendous power.  The majority retains this power even when some of the original coalition would rather withdraw.  In short, the danger of majoritarianism is that it can bestow more power on the incumbents than their societal support warrants.  When a majority party has the power to push through policies without regard for minority opinions, the losers have little recourse but to take to the streets.  Working within the system ceases to be a viable option.


There are several sources of majoritarianism.  I’m an electoral systems specialist, so of course I will start with electoral rules.

Taiwan’s president is elected by a simple plurality.  That is, the winning candidate does not need to break 50%; he or she only needs to win more votes than any other candidate.  Perhaps counterintuitively, plurality elections tend to be more majoritarian than majority systems (which typically have a runoff if no one gets 50% in the first round).  In majority systems, small parties have a chance to demonstrate their strength in the first round, and they can use this demonstrated strength as a bargaining chip for concessions from big parties in the second round.  In plurality systems, there is only one round, and voting for a small party can potentially throw the election to the big party candidate from the other camp.  Voters thus tend not to be willing to support small party candidates at all.  The result is that votes are focused on the two major candidates, whether or not voters really prefer smaller parties or less popular candidates.  This can give major parties (and neutral observers) an inflated sense of how much support they have in society.

The legislative electoral system also has strong majoritarian tendencies.  Taiwan’s current mixed member majoritarian (MMM) system has 73 single seat districts, 6 multi-member seats for Aborigines, and 34 party list seats.  Numerically, the single seat districts dominate the system, and they are elected by plurality rule.  Just as with the presidential election, this tends to exaggerate the popularity of the top two contestants.  The party list seats are elected by proportional representation and should be friendlier to smaller parties.  In the 2012 elections, the two big parties combined for 93% of the votes in the single seat districts, but only 79% in the party list tier.  Recall that the list tier elects fewer than 30% of the total seats, so a party with 10% of the list tier votes will only three or four seats.  This meager payoff is a severe disincentive for ambitious politicians to organize a new party.  Even if a politician is willing to try, potential backers will be hesitant to bankroll, volunteer for, or invest too much effort in the new party.  As a result, the party list tier does not do very much to counteract the majoritarian tendencies of the single seat districts.

Why are small legislative parties important?  A small party may start out as a loyal party of the president’s legislative coalition.  However, if the president becomes unpopular, small parties are usually the first legislators to break ranks.  Imagine how much President Ma’s power would be curtailed today if the KMT and PFP had 53 and 15 seats instead of 65 and 3.  With Ma’s unpopularity, the PFP would almost certainly withhold support in the legislature unless the KMT made significant concessions to PFP demands.

The old electoral system used from 1992 to 2004 was far less majoritarian.  The multi-member districts using the single non-transferable vote (SNTV) rule allowed smaller parties to win seats much more easily.  For example in 2004, the PFP won 13.9% of the votes and 15.1% of the seats while the TSU won 7.8% of the votes and 5.3% of the seats.  These semi-proportional outcomes made it worthwhile to try to take advantage of any divisions in public opinion by organizing smaller parties.

Many people have proposed that Taiwan should change to a German-style mixed member proportional (MMP) system.  This would also significantly reduce the majoritarian tendencies engendered by the legislative electoral rules.


Electoral rules are an important source of majoritarianism, but there are others as well.  One source is increasing party cohesiveness in the legislature.  During the democratic era, both the KMT and DPP have steadily become more and more disciplined.  In roll call votes in the Second and Third Terms (1993-1998), legislators defected from their party line over 3% of the time.  In recent years, defections are well under 1%.  More importantly, abstentions have decreased.  In the early 1990s, KMT and DPP legislators abstained from nearly 30% of votes.  Notably, the KMT suffered higher defection and abstention rates than the DPP throughout the 1990s.  Today, this is closer to 10% of all votes.  Relatively poor party discipline meant that the KMT did not have a firm grasp on the legislature after 1992.  When the KMT wanted to push things through, it was not always able to mobilize enough party legislators to do so.  After 1996, when the KMT was reduced to a mere three-seat majority, it effectively lost control of the legislature.  KMT legislators routinely blackmailed their party, threatening not to show up unless the party gave them some sort of payoff.  In other words, individual party members had much more leverage over the party’s eventual decisions, and they could force the party to modify some of its more unpopular proposals.  Today, with much higher levels of party discipline, individual KMT legislators have far less opportunity to force party leaders to listen to their demands.  Power is thus much more concentrated in the hands of the party leadership.

On a related point, the KMT is now less factionalized than it has been in the past.  In the early 1990s, when the KMT still had a relatively large majority in the legislature, it was torn by the mainstream / non-mainstream division.  This division was serious enough that the President Lee had to appoint several premiers from the non-mainstream faction.  Premier Hau and President Lee were more political enemies than allies.  President Lee did not get a firm grasp on power within the KMT until after Hau resigned in early 1993 and the New Party splintered off in the summer of 1993.  Even then, he had to be wary of non-mainstream faction counterattacks.  In contrast to the mainstream / non-mainstream struggles of the early 1990s, the current clash between Ma and Wang has not forced all party members to openly take sides.  Most legislators have simply kept their heads low and tried to stay on good terms with both sides.

The electoral calendar also leads to higher levels of majoritarianism.  In the 1990s, there were major elections nearly every year.  1991: National Assembly; 1992: legislature; 1993: county governments; 1994: Taiwan governor and municipal mayors; 1995: legislature; 1996: president and National Assembly; 1997: county governments; 1998 municipal mayors.  Because there was always an election on the horizon, governments had to continually worry about public opinion.  In fact, there was a lot of grousing that there were too many elections.  Many in the KMT complained that the government was unable to make good long-term policy because it was always worried about short-term public opinion.  The current electoral calendar is completely different.  We now only have two elections in any given four year period.  All the national elections are combined into one big election day, and all the local elections are combined into a second big election day.  Moreover, the local elections come relatively late, roughly 34 months into the presidential cycle.  Because of this, presidents can simply ignore public opinion for the first half of their term.  After all, there will probably be time for it to bounce back.  And even if the president’s party does poorly in local elections, these do not affect the balance of power in the national government.  There is no way for the public to restrain the president’s party in the legislature during the term.  They must wait a full four years, at which time the cycle starts anew.  Policymakers have thus attained that 1990s wish: they can make decisions without worrying about the impact on public opinion.  Unfortunately, while policymakers like to think that they hold a unique wisdom and make good decisions, unpopular policies are often unpopular for a reason.  One of the primary aims of democracy is precisely to prevent policymakers from ramming through policies without considering their adverse effects on large groups of people.

It probably goes without saying that divided government diffuses power.  When one party holds the presidency and another controls the legislature, unhappy people on both sides of the divide have a friendly institution willing to listen to their complaints.


In other countries, there are other mechanisms to spread power more broadly.  Many countries have bicameral legislatures.  A second house elected by a different method makes it harder for a small plurality to control the government so thoroughly.  My colleague Jih-wen Lin has shown that government coalitions in Japan tend to include extra parties precisely because the governing coalition wants to maintain power in the upper house.

Some countries have federal systems.  In a federal system, strong regional governments can resist unpopular policy initiatives from the national government.  In unitary systems like Taiwan, local governments do not have the right to pass laws, and they often do not have independent fiscal powers.  This severely limits their capacity to serve as a check on the national majority party.

An independent court system can check majority powers.  In most liberal democracies, minorities can turn to the courts to protect their interests.  Unfortunately, the judicial system is one of the weakest parts of Taiwan’s government.  The judiciary rarely dares stick its nose into controversial issues until they have festered for so long that society has already reached a consensus.  For example, an independent judiciary almost certainly would have declared the Assembly and Parade Act unconstitutional long ago.  There is good reason for the judiciary’s reticence: it enjoys very low levels of public confidence.  There is widespread belief that the judiciary, especially at the top levels, follows instructions from the KMT.  At the very least, few people believe the judiciary is neutral.  My colleague Chung-li Wu is currently doing interesting research on systematic political biases in judicial outcomes.

Regional cleavages can disperse power.  If different groups have distinct regional power bases, a multi-party system can result.  In India, for example, regional parties dominate many states, and the national parliament has two, large, fractious coalitions.  In Taiwan, the KMT and DPP are the two major parties in almost every corner of the country (except Jinmen and Matsu).

Supermajoritarian rules can also restrain majorities.  For example, in the US Senate, the filibuster rule allows minorities to block (or severely slow down) proposals if they can muster 40% of the votes.  Since minorities can usually get 40% of the votes, majorities have a strong incentive to modify their proposals to make them more acceptable to the minority.  In Taiwan, a simple majority is required to win votes.  However, there is some degree of consensualism in other legislative stages.  For example, inter-party negotiations require the assent of all party caucuses.  Further, in recent years, opposition parties have taken to occupying the podium to stop votes from being held, much in the same way that minorities in the US Senate uses the filibuster (or the mere threat of the filibuster) to prevent votes from being held until the majority compromises a bit.  However, tactics such as occupying the podium are not recognized in the formal legislative rules and are not generally seen as legitimate opposition tactics.


All in all, power is highly concentrated in today’s Taiwan.  This is the result of several factors, some of which have been slowly building for decades and some of which suddenly snapped into place a few years ago.  The aggregate effect is that the KMT party leadership currently has more actual power than at any time since Chiang Ching-kuo was alive.  This is great for the winners.  However, with unassailable power, the KMT leadership has very little incentive to worry about public opinion.  It can push its favored policies without regard for public backlash.  Even worse, because it does not have to worry about public opinion, there is no incentive to do the hard work of selling policies to the public and building a solid consensus to support them.  We have seen this quite vividly with the cross straits services trade agreement.  Rather than trying from the outset to build a pact that could win broad support and then going out and energetically building that support, the government has preferred to keep the pact out of the public eye as much as possible.  This may save the KMT a lot of effort, but it is not necessary healthy for the system.  From opponents’ point of view, there are almost no points of access from which they can try to modify or block government policies.  The opposition parties in the legislature are impotent, the ruling party in the legislature follows orders from above, the courts are useless, local governments only have limited power, and there is usually no upcoming election to throw their energies into.  There simply is no way to work within the system to stop policies that they do not like.  All they can do is go out onto the street.  When even street protests seem useless, even more radical steps are on the horizon.

If we are indeed in a “constitutional moment” when it is possible to change the basic architecture of the system, one of the top priorities should be to shift the system in a more consensual direction.  The goal should be to encourage minorities to express their opposition within the normal political institutions, and this requires giving them enough power that they can achieve moderate successes.  Minorities should not have absolute vetoes, but neither should they be entirely powerless.  At present, the system has become so majoritarian that it is threatening to rip itself apart.