How we calculate

Methodology

STATE maps representation across many incompatible electoral systems. We show actual results everywhere, but theoretical thresholds and control cost are always system-specific.

Why "votes needed to win" is not globally uniform

A single-member plurality contest in the UK and a nationwide list-PR contest in Israel are fundamentally different units of competition. We never aggregate them naively. STATE always labels the system, the unit type, the method code, and a confidence score so users can interpret every number in its proper context.

Electoral unit types

Single-member constituencies, multi-member districts, regional list districts, and nationwide lists each define "winning" differently. We surface the unit type on every record.

District magnitude

The number of seats a unit elects. STATE treats district magnitude as a first-class field because it determines the meaning of any threshold figure.

Control objectives

We model: majority of seats, executive win, governing control, and all units. Country-level control cost sums the cheapest unit thresholds required to satisfy each objective.

Exact vs estimated

For single-member contests we publish exact thresholds (runner-up + 1 votes). For PR systems we publish marginal last-seat estimates. The method code on every metric tells you which.

National-list systems

Where there is no local breakdown (Israel, Netherlands, etc.) we publish a single national row and flag it with the National-only badge. We do not invent local thresholds that do not exist.

Mixed systems

For MMP and parallel systems we treat local district contests and list-tier allocations separately and visibly. Local FPTP thresholds are exact; list-tier marginal costs are estimated.

Disputed boundaries

Units in disputed jurisdictions are flagged at both the unit and country level. They can be excluded from any control model via the model controls.

Confidence methodology

Each computed metric carries a confidence score (0–1). Exact official results sit at ≥0.95; derived results 0.85–0.95; estimates below 0.85.

Method codes

Exact official

Pulled directly from official primary sources at the unit level.

Derived from official

Computed deterministically from official records (e.g. AV final preferences).

Estimated

Marginal-seat or summary-level estimate; not directly comparable to single-member contests.

National-only

System has no local breakdown; only national figures published.

Non-comparable

Flagged where any comparison across systems would be misleading.