Rethinking referendums

What have we learned from this round of referendums?

The most surprising outcome was that votes for the four ballot items were essentially identical. The polls suggested that many voters saw differences among the four items, and the parties certainly acted like they expected differences. However, the people who voted seem to have almost all voted the same way on all four. I didn’t see this coming, and it forces me to rethink some previous ideas about referendums in Taiwan.

I had thought that the strongest argument for separating referendums from general elections was to create a richer information environment for the referendums. Without candidates to soak up everyone’s energy and attention, voters have more opportunity to learn about the issues involved in the referendums. With more information, they might be less prone to make decisions based on superficial cues. And indeed, there was a pretty robust discussion about the details and wider implications of each referendum. However, with more information, I expected some voters to decide that some items were reasonable while others were not. I did not expect everyone to become highly informed and for half the people to reach one conclusion and the other half to reach exactly the opposite conclusion. One might think that the richer information environment did not matter at all. People voted with their party, just as they would have if the referendums had been combined with a general election.

However, I just cannot believe that all this discussion and all this information did not matter. It may not have mattered in the way I expected, but it must have mattered in some way.

One way it could have mattered was in convincing partisans to vote with their party. The November My-Formosa poll showed that significant numbers of partisans planned to vote against their party’s position on all four items. Another sizable group (though not quite as large) of partisans weren’t sure about how to vote. These voters seem to have either changed their minds and voted with the party or just stayed home. Voters tend to listen to trusted sources. For partisan voters, this often means partisan media or the parties themselves. The information these voters got could certainly have played an important role in aligning voting decisions with their preferred party’s positions.

 % of green identifiers% of blue identifiers
 against or unsure aboutagainst or unsure about
 green positionblue position
   
R17 4NPP19.230.7
R18 pork30.814.2
R19 same day41.534.2
R20 LNG/reef35.545.3

Another possibility is that the campaigns taught voters that what initially looked like a very easy decision was actually very complex. It is easy to imagine many voters two months ago saying, “of course I want to protect the reefs” or “of course I want safe food.” At the time, there wasn’t really a reason to doubt these positions. No one had made a serious argument against them. On pork, for example, the DPP government opened up the pork market last August, but there wasn’t a big information campaign to persuade people that ractopamine is safe. In the referendum campaign, the parties had to face these issues head on and make serious arguments about each topic. These arguments were often hard to follow. For example, in the 4NPP debate, the sides cited numerous technical reports on geology, nuclear waste disposal, energy prices, the status of the construction at different time points, safety inspections, and so on. To make a fully informed decision on the four items, you needed to be an expert in nuclear physics, climate science, oceanography, food science, international trade, political institutions, and a half dozen other fields. It is possible that the rich information environment taught people that these were difficult choices. Without an obvious right answer, I can imagine large numbers of people deciding to just stay home. This seems especially likely for nonpartisans. The partisans could trust that their party’s experts had looked at all facets of the problem and had come to a good conclusion. They could also ignore the difficulty and go out to vote to support their party (or oppose those jerks in the other party). Nonpartisans overwhelmed by the complexity of the choice didn’t have any easy solutions. And if you don’t know what the right answer is, there isn’t much incentive to go vote to express your opinion or to prevent the wrong choice from being made.

I’m not unhappy that all four items failed. Part of this is that I have my own partisan preferences. However, I also have a general skepticism toward referendums. When in doubt, I always vote no. It should be difficult to pass laws, and it should be prohibitively difficult to pass laws when you can’t figure out what you are passing. Moreover, Taiwan’s referendum law stipulates that referendum results can’t be overturned for two years. If you make a mistake, you are stuck with it.

What about the small parties?

The small parties didn’t play much of a role in this referendum. The NPP and TPP both supported some but not all of the items, but their endorsements don’t seem to have had much of an impact at all.

I guess it’s not all that surprising. Referendums are a majoritarian institution. When you need a majority of votes, the small parties are never going to play much more than a supporting role. Still, it’s surprising they didn’t even play a supporting role. They were basically absent.

The most shocking absence wasn’t the TPP or NPP, though. It was the Green Party. Where the hell was the Green Party?!? Two of these referendums were right in their wheelhouse, and pork should also arguably be one of their best issues. How is it possible they weren’t out on the front lines every day screaming about their ideals? Honestly, I don’t know what the environmental position on R20 (reef/LNG). It’s complicated. But this is precisely the kind of thing that an environmentally-centered party should take the lead on. If the Green Party doesn’t have an opinion on environmental questions, what good are they?

What should we expect from future referendums?

R19 failed, so future referendums will be held in August of odd-numbered years. This creates two very different environments. The 2025 referendums will be held a year and half after the 2024 national elections and over a year before the 2026 local elections. That’s a relatively dead time in the electoral calendar, and the referendums will be the biggest story for a couple of months. In contrast, the 2023 referendums will be held just as the 2024 presidential race is heating up. In August 2023, the two big parties will probably already have nominated their presidential candidates and most of their legislative candidates. Smaller parties will also be making their decisions about how to position themselves. The 2023 referendums are going to be completely subsumed by the upcoming national elections. They will almost certainly be seen as a dress rehearsal for the big clash. There will be very little oxygen for nonpartisan discussion of the various referendum issues.

Taiwan desperately needs to revise its electoral calendar. The easiest thing would be to shorten one presidential term[1] so that the president is inaugurated in early February. All elections and referendums could be held in late November or early December, and there would be a full year between each vote. I’m not holding my breath, though. I think we will have two very different referendum experiences depending on which part of the election cycle it is.

All four items failed, and both parties missed the threshold by about a million votes. One thing we learned from this experience is that you can’t just assume that people will turn out to vote for a referendum. It may simply be easier to mobilize people to vote for a candidate. In 2018, the concurrent general elections meant that the referendum threshold was not a problem, and seven items passed. This includes a few that I have almost no impression of. I learned very little about R7, R8, or R16. My biggest fear was that activists and special interests of all stripes would try to imitate these, putting their pet issue on the ballot and hoping it sails through without much consideration. I’m relieved that R19 didn’t pass, so the threshold will remain an obstacle. It costs money to launch a petition drive, so hopefully only efforts that can withstand public scrutiny and inspire intense, broad public support will bother. Free rides are likely to be more of a problem in the referendums five months before the presidential election because the parties and the general public will be treating them as a dress rehearsal for and bellwether of the upcoming (more important) contest. It will be much easier to get over the threshold, so that is where you might see more efforts by special interests to piggyback on partisan emotions to pass their pet law (which can’t be undone for two years – ka-ching!).

Why do we have referendums?

Ko Wen-je, Hou You-yi, and a few others complained that the two big parties were perverting the referendums by turning them into ordinary party-based election campaigns. They weren’t having rational[2] debates about the issues. Instead, they were just telling people to vote all four yes or all four no.

I don’t have a lot of patience for this complaint. If Ko and Hou sincerely believed this point, they are political simpletons. First, there was a lot of good debate. Maybe they weren’t paying attention, but I heard a lot of people delving into the nitty-gritty of complicated policies. Second, of course the main political parties are going to take a position! These are significant policies, and they are tightly related to many of the parties’ other goals. Ignoring the referendums would be irresponsible.

Third, these referendums were politicized right from the start. The KMT directly sponsored two and enthusiastically supported the other two. They didn’t do that because they were pursuing some apolitical goal. The primary purpose was to inflict political damage on the DPP government. They wanted to give the DPP a black eye and obstruct its policy agenda. The KMT directly or indirectly sponsored all four, so of course they asked voters to vote yes on all four. By the exact same logic, the DPP naturally opposed all four. In a system dominated by two strong parties, this is how almost all referendums will inevitably unfold. If it matters, it will be partisan and politicized. The calm, rational, detached, apolitical debate that Ko, Hou, and others are imagining is just never going to happen.

One of the interesting things in this campaign was that both the KMT and DPP seemed to understand referendums in the same way. Americans often talk about them as a way to go around the parties to make public policies that the parties can’t because of their conflicts of interest. That wasn’t how the KMT and DPP talked about them. Rather, they seemed to understand referendums as a mechanism to restrain a government that has gone too far. Referendums are a roadblock for elected politicians, not a detour around them. As I wrote in a previous post, this is how Sun Yat-sen understood referendums a hundred years ago, and that thinking seems to have trickled down to contemporary politicians. It is significant that three items were attempts to undo a current government policy (and the fourth was an attempt to reverse the longstanding DPP anti-nuclear policy).

Of course, in a democratic polity there exists another mechanism to discourage government overreach: general elections. I generally think that referendums are a lousy way to make public policy. I also think they ask too much of the voters. Representative democracy allows voters to make a fairly simple decision. This set of politicians has a vision for the society based on a set of ideas about what is desirable and possible. That set of politician has a different vision based on a different set of values. Voters don’t need to figure out all the details. They can choose a vision and a set of values and let the elected politicians figure out how to make all the complicated tradeoffs necessary to pursue that vision. Representative democracy works because voters don’t need to understand everything. Referendums demand much more from voters. It might be great in theory, but it isn’t practical to expect each voter to understand the details of every policy. Referendums simply produce decisions by uniformed policymakers.

However, I will suggest two constructive roles for Taiwan’s referendum system. First, in Taiwan’s strange electoral calendar, the 34-month gap between the national elections and local elections is just too long. It’s always difficult to be in the opposition, and it is especially difficult if you don’t have an outlet for your frustrations. Putting the referendums in the middle of that long gap provides that outlet. The opposition can put their energy into referendums that might show the government is out of touch with public opinion and should be restrained. It is important to have an institutional mechanism for opposition to keep it inside the system.

Second, Taiwan’s system is fairly majoritarian, and the party system is grounded in national identity. As you are doubtless aware, the national identity trends have turned pretty decisively against the KMT. There is a danger that the KMT might be turned into a permanent minority, too small to win but too big for a different challenger to emerge. And with this, there is a danger that the DPP will become entrenched in power, unconcerned with electoral pressures because the KMT is unacceptable to most of the population. Japan managed to get through most of the second half of the 20th century with this sort of dominant party system, but an entrenched government with no credible opposition is not generally a recipe for democratic stability. If the KMT insists on remaining unelectable, perhaps referendums can keep the DPP afraid of the voters. One can imagine unhappy voters approving KMT-sponsored referendums in order to show their displeasure and punish the government without facing the risk of actually putting the KMT in power.

All in all, I’d prefer to have elections every 24 months with two (or more) viable parties and no referendums at all.


[1] If it is politically impossible to shorten the current term because Tsai wants her full eight years or the upcoming term because Lai, Chu, Hou, Ko, and others think they will be the president, then change the 2032 inauguration date.

[2] I cringe at the term “rational” in Taiwanese political discourse. It seems to mean “arguments that I agree with.”

6 Responses to “Rethinking referendums”

  1. 兼職歷史研究者 Says:

    “Referendums simply produce decisions by uniformed policymakers.”

    Does it actually mean ‘uninformed’? I cannot imagine uniformed policymakers in Taiwan as yet……

  2. Tony the Taiwanese Turtle Says:

    The green party supported R20 in March but revised their position after the modification of the project in May. They did make some announcement during the last month.

    To be fair, they don’t have the resource or media exposure the two major parties have, but environment NGOs were more active during the debates.

  3. kezza Says:

    I very much agree with having a 24-month separation between elections instead of 34.5/13.5 that we current have, or having a legislature that rotates instead of a full general election (like US Senate or some local councils in England). However, it will probably be more practical to go with shortening one mayoral and council term to start in (e.g.) March/April instead of Christmas and so allowing elections to be held around mid-January. Unless I am mistaken, this only requires amending the Local Government Act, instead of a constitutional amendment for the presidential/LY term.

    • frozengarlic Says:

      Legally, you are correct. However, it would be politically difficult to shorten so many terms, since that would mean that everyone in local government would lose eight or nine months of their term. Ideally, they would have thought about this a decade ago when they extended everyone’s term a full year.

      The other reason for adjusting the presidential term is to shorten the four month gap between the election and inauguration.

      It’s a big mess, but it isn’t high on anyone’s list of priorities.

      • kezza Says:

        Yes, it is a big constitutional mess (I blame CSB and Lin Yi-Hsiung for it).

        Certainly none of the major parties have the appetite for considering the 4-month lame-duck period now (despite cross-party consensus in 2016 to revisit this). Politically I don’t see any president willing to shorten his/her term, so unless we get a constitutional amendment the only other way is to establish a convention along Woodrow Wilson’s unorthodox plan (i.e., appointing/inviting president-elect from opposition party to form the transition cabinet).

        As the LY term is set in stone in 7th amendment to start on 1st Feb for 4-year term, this means the election date of January in year 4N is fixed (unless we have support of both blue and green camp it is extremely unlikely to change). So 24-month mid-term election must occur in January year 4N+2. An 11-month lame-duck period for local authorities would be laughing stock and I don’t see any way to square this politically or legally except by amending LGA or with a constitutional amendment to shorten LY to 2-year term (the latter is basically impossible politically, I don’t see any scenario where legislators will vote to risk their job except under severe voter pressure like 2004 LY election).

        Judging from this week’s episode of LGA amendment, there is simply no hope of any major change being considered in this LY term. The best that we could hope, therefore, is in 2024H1 (certainly no later than 2025H1) to change LGA article 83 for term beginning 2030 (even that is going to be painted as politically motivated to affect 2026 local elections, especially since it is often current legislators that challenge the incumbent mayors), but I’m not holding my breath on this.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: