Effective Number of Parties

The Effective Number of Parties (ENP) is an index developed by Markku Laakso and Rein Taagepera in their 1979 article "Effective Number of Parties: A Measure with Application to West Europe" (Comparative Political Studies). It addresses a basic problem in comparative politics: simply counting parties overstates fragmentation, because tiny parties contribute little to legislative behavior or government formation. ENP solves this by weighting parties by their share of votes or seats.

The formula is the inverse of the sum of squared shares:

• ENP = 1 / Σ(pᵢ²), where pᵢ is the proportion of votes or seats held by party i.

Two variants are commonly reported:

• ENEP (Effective Number of Electoral Parties) — based on vote shares.
• ENPP (Effective Number of Parliamentary Parties) — based on seat shares.

Intuitively, if two parties each hold 50%, ENP = 2.0. If one party holds 70% and another 30%, ENP ≈ 1.72. A perfectly fragmented legislature with ten equal parties yields ENP = 10.

ENP is central to the study of electoral systems. Maurice Duverger's hypothesis that single-member plurality systems tend toward two-party competition is typically tested using ENPP. Arend Lijphart's Patterns of Democracy (1999, updated 2012) uses ENP as a core variable in distinguishing majoritarian from consensus democracies. Gary Cox's Making Votes Count (1997) uses it to examine strategic voting and district magnitude.

The measure has limitations. It is insensitive to which specific parties hold seats, so ideologically distinct configurations can yield identical scores. Alternatives have been proposed, including Juan Molinar's NP index (1991) and Grigorii Golosov's measure (2010), which adjusts the weighting of the largest party. Despite these critiques, ENP remains the standard fragmentation metric in quantitative electoral research and is routinely reported in datasets such as the Comparative Study of Electoral Systems (CSES) and Michael Gallagher's Election Indices dataset.