Strategic voting

Strategic or tactical voting refers to the practice of casting a vote based not solely on sincere personal preference but on expectations of how other voters will vote and how different voting strategies may influence the final outcome. The aim is to maximise a voter’s satisfaction with the electoral result, even if this involves supporting a candidate who is not the voter’s genuine first choice. Strategic voting is a recurrent feature of electoral politics across systems, and its prevalence is deeply shaped by the structure and incentives embedded within each voting method.
Despite efforts to design fair and strategy-resistant systems, Gibbard’s theorem demonstrates that no deterministic single-winner voting rule can ensure that sincere voting is always a voter’s best strategy. Nonetheless, under more restrictive assumptions—such as single-peaked preferences or dichotomous approval structures—certain voting rules can be strategy-proof or significantly more resistant to manipulation.

Conditions influencing strategic behaviour

Strategic voting is closely tied to the characteristics of the voting system. Systems using large electoral districts, including list-based proportional representation without thresholds, are typically more resistant to manipulation, as each vote has limited marginal effect. In contrast, systems with biased apportionment, electoral thresholds, or small district magnitudes, such as those used in many versions of the single transferable vote (STV), create opportunities and incentives for strategic coordination. In Ireland, for instance, STV often leads to extensive vote-management operations by parties aiming to distribute support efficiently across multiple candidates.
Under certain structured preference assumptions, however, strategy-proof guarantees do emerge. With single-peaked preferences, the median voter rule is resistant to manipulation, while with dichotomous preferences, approval voting and score voting can be strategy-proof in theory.

Common types of strategic voting

A rich academic literature identifies several forms of strategic behaviour that voters may adopt depending on the rules in play.

Compromising

Also known in various contexts as the “lesser-evil” strategy, compromising occurs when a voter ranks or rates a less-preferred but more viable candidate above or alongside their true favourite in order to block the election of an even less desirable candidate. This is closely linked to the favourite-betrayal criterion: whether a voting method ever incentivises a voter to rank someone above their sincere favourite.

• Most affected: plurality voting, two-round systems, instant-runoff voting (IRV), Borda count, approval voting, and score voting.
• Immune: Coombs’ method and antiplurality voting.
• Political effect: enables and reinforces Duverger’s law, favouring two-party competition as minor candidates are abandoned in favour of viable contenders.

Burying

Burying refers to the strategic ranking of a strong rival candidate artificially low in the ordering to increase the chances of one’s preferred candidate. This form of insincere demotion can, in extreme conditions, lead to a “race to the bottom,” where strategic distortions become so widespread that the eventual winner is someone few voters genuinely favour.

• Most affected: Borda count, antiplurality voting, approval voting, score voting, and many Condorcet methods.
• Immune: IRV and plurality voting.

Pushover or Party Raiding

In this tactic, voters artificially elevate a weak rival candidate—not to elect them but to eliminate a stronger competitor, particularly in sequential elimination rules or open primaries. Party raiding is most associated with attempts by one party’s supporters to manipulate the nomination process of another.

• Most affected: multi-round procedures such as IRV, partisan primaries, and two-round systems.
• Immune: plurality voting and rated methods such as score and approval voting.

Compression

Also called truncation or levelling, compression occurs when voters deliberately refrain from distinguishing between two candidates, giving them equal rankings or ratings. This can obscure genuine preferences and exaggerate or diminish perceived support.

• Most affected: rated voting methods including score voting and approval voting (when used with broad approval strategies).
• Immune: random ballot systems and methods requiring strict rankings.

Frequency and susceptibility

All systems are technically open to manipulation in some circumstances, but the severity, difficulty, and frequency of successful strategic voting vary significantly. Research finds that:

• Range voting (score voting with broad scales) may be especially vulnerable at its worst.
• Majority judgment displays comparatively strong resistance.
• Condorcet systems vary widely depending on tiebreak and cycle-resolution rules.
• Electoral list systems demonstrate low manipulation rates unless thresholds create pressure for tactical consolidation.

Advanced simulations of elections—such as those using Monte Carlo methods applied to preference polls—continue to reveal substantial variation in strategic vulnerability across systems.

Mitigation through system design

One proposed approach to reducing voter-driven strategy is the creation of declared-strategy voting (DSV) methods. These systems incorporate optimisation logic directly into the voting rule, effectively “strategising on behalf of the voter.” DSV adaptations exist for plurality, approval, and score voting, attempting to reduce incentives for independent manipulation.
Switching from highly manipulable systems to those with stronger theoretical or empirical resistance can also reduce strategic distortion. However, no method eliminates strategy entirely.

Coordination mechanisms

Strategic voting requires awareness or anticipation of how others will vote. This generally demands some form of coordination, commonly supplied by:

• Opinion polls, which signal which candidates are viable and help voters avoid splitting support across similar choices.
• Historical voting patterns, which shape expectations of which parties or candidates have realistic prospects.
• Campaign-endorsed tactical voting, where parties or civic groups explicitly encourage strategic support to defeat a common opponent.

In proportional systems, coordination often centres on passing or avoiding electoral thresholds. The 2018 Swedish election, for example, saw supporters of smaller parties evaluate whether their preferred party could realistically surpass the threshold required for representation.

Organised tactical voting

While individual voters often act strategically in isolation, organised tactical voting by political parties is less common. A notable historical example is the Gladstone–MacDonald Pact in the 1906 United Kingdom general election, in which the Liberal Party and the emerging Labour Party coordinated candidacies to avoid splitting the anti-Conservative vote. Such agreements aim to maximise shared electoral goals by managing competition within similar ideological blocs.