Rank-index method

From HandWiki

In apportionment theory, rank-index methods[1]:Sec.8 are a set of apportionment methods that generalize the divisor method. These have also been called Huntington methods,[2] since they generalize an idea by Edward Vermilye Huntington.

Input and output

Like all apportionment methods, the inputs of any rank-index method are:

  • A positive integer [math]\displaystyle{ h }[/math] representing the total number of items to allocate. It is also called the house size.
  • A positive integer [math]\displaystyle{ n }[/math] representing the number of agents to which items should be allocated. For example, these can be federal states or political parties.
  • A vector of fractions [math]\displaystyle{ (t_1,\ldots,t_n) }[/math] with [math]\displaystyle{ \sum_{i=1}^n t_i = 1 }[/math], representing entitlements - [math]\displaystyle{ t_i }[/math] represents the entitlement of agent [math]\displaystyle{ i }[/math], that is, the fraction of items to which [math]\displaystyle{ i }[/math] is entitled (out of the total of [math]\displaystyle{ h }[/math]).

Its output is a vector of integers [math]\displaystyle{ a_1,\ldots,a_n }[/math] with [math]\displaystyle{ \sum_{i=1}^n a_i = h }[/math], called an apportionment of [math]\displaystyle{ h }[/math], where [math]\displaystyle{ a_i }[/math] is the number of items allocated to agent i.

Iterative procedure

Every rank-index method is parametrized by a rank-index function [math]\displaystyle{ r(t,a) }[/math], which is increasing in the entitlement [math]\displaystyle{ t }[/math] and decreasing in the current allocation [math]\displaystyle{ a }[/math]. The apportionment is computed iteratively as follows:

  • Initially, set [math]\displaystyle{ a_i }[/math] to 0 for all parties.
  • At each iteration, allocate one item to an agent for whom [math]\displaystyle{ r(t_i,a_i) }[/math] is maximum (break ties arbitrarily).
  • Stop after [math]\displaystyle{ h }[/math] iterations.

Divisor methods are a special case of rank-index methods: a divisor method with divisor function [math]\displaystyle{ d(a) }[/math] is equivalent to a rank-index method with rank-index function [math]\displaystyle{ r(t,a) = t/d(a) }[/math].

Min-max formulation

Every rank-index method can be defined using a min-max inequality: a is an allocation for the rank-index method with function r, if-and-only-if:[1]:Thm.8.1

[math]\displaystyle{ \min_{i: a_i \gt 0} r(t_i, a_i-1) \geq \max_{i} r(t_i, a_i) }[/math].

Properties

Every rank-index method is house-monotone. This means that, when [math]\displaystyle{ h }[/math] increases, the allocation of each agent weakly increases. This immediately follows from the iterative procedure.

Every rank-index method is uniform. This means that, we take some subset of the agents [math]\displaystyle{ 1,\ldots,k }[/math], and apply the same method to their combined allocation, then the result is exactly the vector [math]\displaystyle{ (a_1,\ldots,a_k) }[/math]. In other words: every part of a fair allocation is fair too. This immediately follows from the min-max inequality.

Moreover:

  • Every apportionment method that is uniform, symmetric and balanced must be a rank-index method.[1]:Thm.8.3
  • Every apportionment method that is uniform, house-monotone and balanced must be a rank-index method.[2]


References

Template:References list



  1. 1.0 1.1 1.2 Balinski, Michel L.; Young, H. Peyton (1982). Fair Representation: Meeting the Ideal of One Man, One Vote. New Haven: Yale University Press. ISBN 0-300-02724-9. https://archive.org/details/fairrepresentati00bali. 
  2. 2.0 2.1 Balinski, M. L.; Young, H. P. (1977-12-01). "On Huntington Methods of Apportionment" (in en-US). SIAM Journal on Applied Mathematics 33 (4): 607–618. doi:10.1137/0133043. ISSN 0036-1399. https://epubs.siam.org/doi/pdf/10.1137/0133043.