What’s in a Number? The Implications of Thailand’s Ballot Design
by Allen Hicken
Figure 1. Thailand's Ballot Design for the Upcoming Elections on March 14th
“What number will you choose?” shouts the speaker. “Three!!” the crowd cries in return. “What number?” “Three!!” “What?” “Three!!”
This particular call and response came during a rally in Kamphaeng Phet province, but similar scenes have been playing out all over Thailand in the last several weeks as candidates, campaign workers, and party leaders hit the campaign trail in search of votes. The speaker went on to explain that if voters forgot the number they could hold up the OK sign and count their fingers. Or, if they had been a boy scout, they could use the scout salute as a reminder.
All political campaigns involve a certain amount of voter education, whether it is reminding people of the party’s policies, the good things the candidate has done (or promises to do), or the misdeeds and missteps of their opponent. And in fact, speakers devoted time to each of these things during campaign events. However, they also spent a significant amount of time educating voters about how to properly cast their votes—a critical job given the significant changes to the election ballot this time around. This article discusses three features of Thailand’s ballot as well as some of the implications of these features.
While much of the scholarly work on elections understandably focuses on electoral rules and electoral administration, ballot design—the way in which ballots are presented to voters—can have measurable effects on voter behavior. Where a candidate or party appears on the ballot shapes the number of votes they receive (Koppell and Steen 2004), for example. The symbols, photographs, layout, and color can also serve as powerful cues for voters (Reynolds and Steenbergen).
One of the main areas of research in this area is how ballot design affects the difficulty of the cognitive task voters face. For example, consider the ballot used in the 2007 Philippines election and the ballot used in the most recent election, both presented below. By requiring voters to write in the names of their preferred candidates for each office, with no prompt on the ballot to remind them of who the candidates were, the 2007 ballot placed a much greater burden on voters compared to the 2022 ballot.
Figures 2 & 3. Philippines' Ballots from 2007 & 2022 general elections
Source 1: https://en.wikipedia.org/wiki/File:Sample-ballot.jpg
Source 2: Source: ABS-CBN News
How does the ballot design for Thailand’s 2023 election (see Figure 1 above) score in terms of the burden it places on voters? First, for this election Thailand opted to reintroduce two separate ballots to fill the two tiers of its mixed member system—a party list ballot where voters select a party to fill 100 seats, and a constituency ballot where voter select their preferred local MP to fill 400 seats. The two ballot system has been used in past Thai elections, 2001, 2005, 2007, and 2011, but for the 2019 election officials elected to use a single/fused ballot, where voters cast a single vote for their most preferred party, and that vote counted towards both a party’s vote total and towards the party’s constituency candidate. The two ballot system makes voting more complicated for voters, but it also gives them more choices—i.e. if they choose they can split their votes between two different parties, one for the constituency election and one for the party list. By itself the reintroduction of the two ballot system increases voter’s cognitive load to a minor degree, but the two other changes to the ballot design have much greater implications for the burden on voters—the lack of information on the candidate ballot and the inconsistent numbering scheme.
Compare Thailand’s candidate ballot to the 2022 Philippines ballot. The former lacks any cues for voters apart from the candidate number. No reminder about the candidate names. No information about which party candidates belong to. Adding to voters’ burdens is the decision by the Electoral Commission to delink the party and candidate numbers. In 2011, using a similar ballot system, parties were assigned a single number and that number applied to both the party list ballot and to each of the parties’ candidates on the candidate ballot. Thus, a voter who wanted to support the Pheu Thai Party could choose #1 on each ballot, regardless of where they lived. For 2023 the party list and constituency ballots are no longer linked, meaning voters who want to support a given party have to remember two separate numbers. Adding another layer of complexity, while a party’s number is the same on every party list ballot, the number assigned to the party’s constituency candidates varies across constituencies.
This new ballot design has clearly increased the burden on voters, candidates, and parties, but how might these changes affect candidate and voter behavior? Research in other countries tells us that where ballots lack partisan cues or make it harder for voters to cast a straight-line party vote, voters are more likely to split their votes across parties (Calvo et al. 2009). The lack of number linkage across ballots also increases the incentives for candidates to cultivate a personal vote (linked to their unique number) and pay comparatively less attention to encouraging voters to also support the party. Indeed, this has been the case in the campaign events we have observed this election. The party’s number usually receives some attention, but the focus is on making sure the voters know the candidate’s number. The Move Forward Party even directed its candidates to only use their candidate number without mentioning the party number, hoping the party and name and symbol on the party list ballot would be enough of a cue for voters.
Some candidates and party officials we have spoken to despair at the inability of voters to recall the correct two numbers, and many expect that a significant number of voters will end up splitting their votes. If such vote splitting is the result of strategic behavior by voters in line with their interests this is fine. However, the added burdens this ballot design places on voters raises the chance the voters will inadvertently split their voters, thus undermining their ability to effectively voice their political preferences.
Koppell, Jonathan GS, and Jennifer A. Steen. "The effects of ballot position on election outcomes." The Journal of Politics 66, no. 1 (2004): 267-281.
Reynolds, Andrew, and Marco Steenbergen. "How the world votes: the political consequences of ballot design, innovation and manipulation." Electoral Studies 25, no. 3 (2006): 570-598.
Calvo, Ernesto, Marcelo Escolar, and Julia Pomares. "Ballot design and split ticket voting in multiparty systems: Experimental evidence on information effects and vote choice." Electoral Studies 28, no. 2 (2009): 218-231.