Sequential Apportionments from Stationary Divisor Methods

Michael A. Jones, Brittany Ohlinger, Jennifer Wilson · 2025

Read Sequential Apportionments from Stationary Divisor Methods by Michael A. Jones, Brittany Ohlinger, Jennifer Wilson (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-20686.txt.

Introduction

Michael A Jones, Brittany Ohlinger, and Jennifer Wilson watch legislatures fill seat by seat. Divisor methods apportion seats by rounding scaled vote shares; because they satisfy house monotonicity, seats can be handed out sequentially as parties earn them. The authors focus on stationary divisor methods, defined by one rounding cutpoint.

With integer votes, the resulting apportionment sequences are periodic. For two parties, the authors characterize exactly which sequences can occur and tie their lexicographic ordering to the cutpoint's value. Pairwise sequences then extend systematically to any number of parties, and the count of distinct sequences is determined. A reframing of size bias emerges: large parties do not simply get more seats, they get their seats earlier. The Adams and D'Hondt extremes turn out to be intimately related.

Catalog record

CREATORMichael A. Jones, Brittany Ohlinger, Jennifer Wilson
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE75 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-20686

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Sequential Apportionments from Stationary Divisor Methods Michael A. Jones, Brittany Ohlinger, Jennifer Wilson Abstract Divisor methods are well known to satisfy house monotonicity, which allows representative seats to be allocated sequentially. We focus on stationary divisor methods defined by a rounding cut point c ∈ [ 0 , 1 ] c\in[0,1] . For such methods with integer-valued votes, the resulting apportionment sequences are periodic. Restricting attention to two-party allocations, we characterize the set of possible sequences and establish a connection between the lexicographical ordering of these sequences and the parameter c c . We then show how sequences for all pairs of parties can be systematically extended to the n n -party setting. Further, we determine the number of distinct sequences in the n n -party problem for all c c . Our approach offers a refined perspective on large-party bias: rather than viewing large parties as simply receiving more seats, we show that they instead obtain their seats earlier in the apportionment sequence. Of particular interest is a new relationship we uncover between the sequences generated by the smallest divisors (Adams) and greatest divisors (d’Hondt or Jefferson) methods. MSC Classification]91B32 91B12 91F10 keywords apportionment sequence, divisor methods, d’Hondt/Jefferson method, Adams method pacs [ † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. 1 Introduction Divisor methods of apportionment are widely used in proportional representation systems to allocate seats to political parties in proportion to their share of the popular vote. Less attention, however, has been given to how these methods determine the sequence in which seats are awarded. This sequence is important in practice as it is used by some coalition governments to determine cabinet positions: while the office of prime minister typically goes to the party with a plurality of votes, other cabinet positions are selected by parties based on the sequence. For example, in Northern Ireland, the greatest divisors method (also known as d’Hondt’s or Jefferson’s method) is used not only to determine the number of seats allocated to each party in the Assembly but also the order in which cabinet posts are selected OGE . Sequential apportionment in this context offers an alternative to the intense negotiation and strategic bargaining that often arise during the establishment of a new cabinet. Using an apportionment method to allocate positions sequentially can help mitigate these conflicts and avoid delays. Just-in-Time (JIT) sequencing problems have the same structure as the sequential allocation of minister positions. In these problems, a number of items must be processed (or produced) to meet a set of demands or users. The items must be processed one at a time, and so the question arises as to what item order satisfies the demands in an equitable way. Although divisor methods have been used to allocate the processing of items, other algorithms have been developed to solve such problems under different optimality conditions, such as minimizing the total deviation or the maximum deviation. Kubiak K provides a comprehensive analysis of JIT sequencing problems, exploring the usefulness of divisor methods for JIT sequencing and analyzing the complexity of algorithms used to solve JIT sequencing problems. Our work has application to sequential apportionment in both fields, as we focus our attention on the possible apportionment sequences from stationary divisor methods, a family parametrized by a rounding cut point c\in[0,1] which includes the smallest divisors method (also known as Adams’ method for which c=0 ), Webster’s or Sainte Laguë’s method ( c=0.5 ), and d’Hondt’s method (used in Northern Ireland, for which c=1 ). Such sequences arise whenever apportionment methods satisfy the property of house monotonicity , which requires that no party’s allocation decreases as the number of total available seats increases. Monotonic apportionment methods give rise naturally to a sequence by considering the allocation of one seat (the first seat), the allocation of the second seat, and so on. Stationary divisor methods have an easily described algorithm for determining this sequence—versions of which are used in awarding minister positions and in JIT sequencing problems. Stationary divisor methods also have a nice property in that the sequences they give rise to are periodic (when vote totals are integers), as proved by Kubiak K in the context of JIT sequencing. Not all divisor methods yield periodic sequences, as evidenced by Hill-Huntington’s method, which is used in the U.S. to determine the numbers of seats each state receives in the U.S. House of Representatives. Periodicity implies that there are only a finite number of possible sequences from stationary methods. In this paper, we identify all possible sequences generated by stationary divisor methods for given party vote totals and provide explicit formulas for the total number of such sequences as functions of party vote shares. This is in contrast to work by Cembrano et al. CCSTV , in which they use a geometric approach to obtain bounds on the number of different possible apportionments (not sequential apportionments) for n parties for the class of stationary divisor methods, as well as for power-mean divisor methods. We also establish necessary and sufficient conditions for determining when a periodic sequence corresponds to a sequential apportionment, as well as methods for identifying the associated stationary divisor method. We combine the periodicity of the apportionment sequences and a notion of weak proportionality for divisor methods to refine the notion of size bias of divisor methods. As identified in Balinski and Young BY , Adams’ method provides one extreme that favors small parties and the d’Hondt method provides the other extreme that favors large parties. However, by applying a lexicographic ordering of apportionment sequences, we reframe this distinction as less about an absolute bias and more about variation in when large and small parties receive their seats. The refinement also leads to a new relationship between both the sequences and apportionments from Adams’ and d’Hondt’s methods. Applying a sequential framework to the apportionment problem is a relatively unstudied field. In addition to the literature analyzing political apportionment cited above, Pukelshein P provides an in-depth study of the use of divisor methods in systems of proportional representation, but does not focus on sequential apportionment. A small number of studies have examined the use or potential use of divisor sequences for the distribution of ministerial positions. O’Leary, Grofman, and Elkit OGE analyze the systems used in the Northern Ireland Assembly, in Danish municipal governments, and in the European parliament to determine committee chairs. Raabe and Linhart RL investigate several models incorporating party preferences and the sequential allocations arising from the d’Hondt, the Sainte Laguë, and Adams’ methods and compare them to actual cabinet allocations in the German parliament. They note that adapting the structure of such sequences may be an advantage to parties given the time pressures they face in selecting cabinet portfolios after an election. Ecker, Meyer, and Müller EMM take a similar modeling approach in analyzing party portfolios for 146 coalition governments in Western and Central Eastern Europe. When cabinet ministry positions are awarded sequentially, Brams and Kaplan BK explain that it may not be rational for parties to naively select their most-preferred minister when available. Further, they

Read the complete text