Do Institutions Cause Strategic Voting? Evidence from Taiwan
This study demonstrates that Taiwan's 2008 shift from multi-member to single-member districts significantly increased strategic voting, particularly in competitive districts with prior minor-party presence, revealing that such behavioral changes are driven primarily by voters rather than parties.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Voters often face a difficult choice: cast a ballot for the candidate they truly love, or vote for someone they like less but who has a better chance of winning. This dilemma, known as strategic voting, arises when people realize that supporting their favorite candidate might actually help a candidate they dislike even more. Political scientists have long believed that the rules of the game determine how often this happens. Specifically, when an election district is large and elects many representatives, voters feel free to support smaller parties. But when a district shrinks to elect only one person, the pressure to coordinate behind a major contender grows intense. This idea, rooted in the work of mid-twentieth-century scholars, has been tested extensively in countries where the legislature controls the government. However, it remained unclear whether this same logic holds true in systems where the legislature does not directly decide who leads the executive branch.
In 2008, Taiwan offered a rare natural experiment to settle this question. The island nation overhauled its parliamentary system, switching from a complex setup where many representatives were elected from large districts to a simpler system where each district sent only one representative to the legislature. Crucially, while the national parliament changed its rules, the elections for local city and county councils continued under the old, multi-member rules. This created a perfect comparison: two types of elections happening in the same country, with the same voters, but under different rules. Researchers used this unique situation to see if the change in district size alone was enough to make voters abandon their preferred candidates in favor of more viable ones, even in a system where the parliament does not directly control the presidency.
The study focused on a specific way to measure how voters behave: looking at the gap between the candidates who just missed winning and the ones who finished even further behind. In a district where voters are voting sincerely for their true favorites, the candidates who lose by a small margin usually have vote totals very close to those who lose by a larger margin. But when voters act strategically, they abandon the distant losers to support the runner-up, causing the vote gap to widen dramatically. By comparing the voting patterns in the national parliament before and after the 2008 reform against the patterns in the local councils, which never changed their rules, the researchers could isolate the effect of the new system.
The results were clear and immediate. Following the switch to single-member districts, the gap between the second-place loser and the third-place loser in parliamentary elections grew significantly. This shift indicated that voters were indeed deserting their preferred minor-party candidates to back the more competitive major-party contenders. The change was not gradual; it happened right after the reform, suggesting that voters adapted their behavior almost instantly once the new rules were in place. The effect was strongest in areas where the reduction in the number of seats was most drastic and where the races had been closely contested before the reform. In districts where one party had always dominated, the change in voting behavior was minimal, as the outcome was already predictable.
Importantly, the study ruled out the idea that political parties themselves were solely responsible for this shift. If the change were driven only by parties deciding to stop running candidates in certain areas, the voting patterns would look different. Instead, the evidence pointed to individual voters making the choice to switch their support. Survey data confirmed that supporters of smaller parties were more likely to vote for a major party candidate after the reform than they had been before. This suggests that the pressure to vote strategically comes directly from the structure of the election itself, rather than from instructions or coordination by political elites.
The findings also challenged a common assumption about newer democracies. Some theories suggested that voters in younger democracies might not have enough information or experience to vote strategically, leading to more scattered voting patterns. Yet, in Taiwan, voters quickly learned to navigate the new system, behaving much like voters in long-established democracies. The study concluded that the size of the district is a powerful driver of voter behavior, capable of triggering strategic voting even in a semi-presidential system where the legislature does not directly determine the head of government. The change in rules alone was enough to reshape how people cast their ballots, proving that the mechanics of an election can fundamentally alter the psychology of the voter.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.