Based on Percentage Distribution of Votes by Parties

1. Mean absolute deviation: MAD = Σ | X – √ | / n (also called “index of variation” [Rose and Urwin 1975: 24])

2. Mean squared deviation: MSD = Σ (X – √)² / n

3. Variance: S² = Σ (X – √)² / n – 1

4. Standard deviation: S = √ S² or S = √ Σ (X – √)² / n – 1

5. Lee index: LEE = Σ | X – √ | / 2

6. Variability coefficient: CV = S / √

7. IPR = √ n Σ | X – √ | / (2 (n – 1) Σ X)

Based on Percentages of Parties' Vote Distribution by Constituencies

Cumulative regional inequality index: CRII = 1/2 Σ | votes – voters | (not applicable to turnout).

The formula is divided by 100 to vary between 0 and 1 (Rose and Urwin 1975). When voters' figures are not available, electorate figures are used.

All Constituencies versus Constituencies in Which the Party Is Present Only

For party support, all previous indices have been computed (1) on the total number of constituencies (party not present = 0) and (2) on the constituencies in which the party was present only (party not present = system missing). This distinction does not apply to the CRII or to turnout figures.

Territorial Coverage by Parties

The simplest measure used to assess the spread of parties across regions is the number of constituencies in which a party is present (nominator) as a percentage of the total number of constituencies (denominator).