Seat Allocation in Parliaments.

 

A key institute in old and modern democracies is the parliament: a collection of persons that have been elected to represent the people, and whose main task is to control legislative power – we will refer to these elected persons as members of parliament or MPs. Almost always, members of parliament are grouped into different parties, and members of the same party share a basic point of view on how society should be organized. To exercise their right and duty of controlling power, there is a physical location where the members of parliament meet, discuss, vote, gossip, and do all the things that members of parliament are supposed to do – in the sequel, we will use the term ”parliament” to denote this physical location. In this note, we are especially interested in distributing MPs over the seats in the parliament so as to optimize the functioning of the parliament. Of course, this objective can be concretized in different ways – we intend to inventarize different, not necessarily conflicting ways of achieving this objective. It is also interesting to observe that the particular physical lay-out of the seats in the parliament matters for this objective; most parliaments have a (half-) circular shape where sets of seats are separated by corridors. We will close this note by giving a list of some parliaments that we find interesting, give the current seat allocation, and give our “optimal” seat allocation.

 

Let us list a number of properties present in almost all seat allocations in parliaments all over the world:

(i)                MPs from the same party are grouped. Thus, neighboring seats (to be defined later) are as much as possible allocated to MPs of the same party. This facilitates intra-party discussion.

(ii)              Not all seats are equally important. Seats in the front have more visibility, and allow direct access to the debating spot (a term we use for the physical position from where the government representative, or other MPs can be interrupted and questioned). Usually, large parties (i.e., parties with more MPs than others) occupy one or more front seats. Seats in the back may be considered less important, while seats with access to corridor may be preferred over seats between other seats. Notice that there are parliaments such that each seat has access to a corridor.

(iii)             Western democracies have parties that are often labelled somewhere in a “left-right” spectrum. In many parliaments, this left-right positioning is reflected in the allocation of seats. In a way, MPs from different parties that are considered to hold similar views, are allocated to neighboring positions in the parliament.

 

 

A way to precisely model the interconnection between different seats in a parliament is to define the concept of neighboring seats. We use the following definitions. Two distinct seats are neighboring when direct contact between two MPs in these seats is possible. With direct contact we understand that whispering and drawing attention (without alerting others) is possible. In practice, this means two seats that are positioned next to each other on the same row, or that are positioned such that one seat is directly behind another seat, are called neighboring. Of course, in some cases deciding whether two seats are neighboring can be a debatable choice. Further, seats are called half-neighboring when the seats are on opposite sides of a corridor. Indeed, MPs on opposite sides of a corridor can still easily make contact. We realize that much more fine-grained ways exist to define the concept of neighboring – we use this as a first proxy.

 

 

Our model is an integer programming model that maximizes a particular objective function representing intra-party communication. In this objective, one point is gained when a pair of neighboring seats is allocated to members of the same party; half a point is gained when a pair of half-neighboring seats is awarded to members of the same party. The model has constraints that ensures that large parties receive front seats, and depending on the physical lay-out of the particular parliament, the model has constraints to ensure that a left-right allocation is adhered to. If desired, many other constraints can be added; we mention here: (i) particular seats that, for historical or nostalgical reasons should be allocated to a particular MP of some party, or (ii) parties whose members hold wildly different views, and in order to avoid conflicts should not receive (half-) neighboring seats.

 

A crucial advantage of a model-based proposal for a seat allocation is its neutrality; once there is agreement over the basic principles that should guide the seat allocation, the outcome of the corresponding model is should be perceived as unbiased.

 

 

 

 

The Netherlands, Second Chamber – Facts

Number of Seats: 150

The Layout (including neighboring seats)

Number of Parties with MPs: 12

Parties and number of seats:

(33 20 19 19 14 14 9 6 5 4 3 3 2)

 

Our solution, value

Current solution, value

Remarks

 

 

The Netherlands, First Chamber – Facts

Number of Seats: 75

The Layout (including neighboring seats)

Number of Parties with MPs: 12

Parties and number of seats:

(13 12 10 9 9 8 4 3 2 2 2 1)

 

Belgium, Kamer der Volksvertegenwoordigers – Facts

Number of seats: 150 (notice however)

 

Our solution, value

Current solution, value