John Stuart Mill and proportional representation. A misunderstanding

ABSTRACT

Mill scholars usually classify as ‘proportional’ the theory of political representation sustained by John Stuart Mill. However, Mill did not use the term ‘proportional’ in his texts, and ‘proportional representation’ as understood today has little to do with what Mill really proposed. A hypothesis is offered to explain the origin of the ‘proportionalist’ interpretation that was accepted at the beginning of the 20st century. It is concluded that specialists in Mill should not use the word ‘proportional’ when referring to his conception of political representation. In addition, political scientists should not quote Mill as the founder of proportional representation.

KEYWORDS:

• Proportional representation
• John Stuart Mill
• political participation
• electoral systems
• majority government

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Nagel (2001, 12, 216) distinguishes between ‘proportional representation’ as applied to specific electoral systems and ‘proportionality’, the representative ideal to which those electoral systems aim. Mill enthusiastically advocated an electoral system, the Hare system. Since then, almost every political science or political philosophy contribution on the topic of political representation published throughout the last century (from Humphreys 1911:198 to Ryan 2014, 11) considers him a supporter of the ideal of ‘proportionality’. Even Wikipedia affirms, in its entry on ‘proportional representation’, that ‘the case for proportional representation was made by John Stuart Mill in his 1861 essay Considerations on Representative Government’, (accessed 7 July 2019).

2. Colomer (2007, 269) mentions several incorrect translation examples: Hoag and Hallett (1926), Iain Mclean (1991) and Colomer (2001). However, the misinterpretation is even older (Sterne 1871, 50). In 1969, Pitkin was already conscious of the problem (Pitkin 1969, 263).

3. All page references to the writings of Mill are to the Collected Works of John Stuart Mill, 33 volumes, John M. Robson, General Editor, Toronto, University of Toronto Press, 1963–91. I include the corresponding book or letter in Italics.

4. This happens even with the presentation of his work. For instance, the conference originally titled ‘The Redistribution of Seats/Conference at the Reform League Rooms’ is titled in Mill’s Collected Works as ‘Proportional Representation and Redistribution’ (C.W. XXVII, 239). However, the expression ‘Proportional Representation’ does not appear in the text. It is in the mind of the editors, but not in Mill’s work and, as I intend to show, was not in Mill’s thought either.

5. This opinion has changed today. According to a 2005 survey conducted among 170 electoral systems scholars, STV was the second most preferred option out of nine main categories of electoral systems. The first one was mixed-member proportional systems (Bowler, Farrel and Pettit, 2005, 7).

6. According to Hare’s system, voters indicate on their ballot papers their orders of preference for each candidate. Any candidate who receives the quota with the first preferences is elected. If a candidate receives more than the established quota, his surplus is transferred to the candidates who come out with the second choice on the surplus ballots. This process continues until 658 candidates reach the quota and are elected. This peculiar transfer mechanism achieved more notoriety than any other aspect of the proposal. Both systems share such a mechanism, and this is probably the reason for the confusion arising between the Hare system and the later STV. The main difference between them is that the Hare system permitted electors to rank order for up to as many as 658 candidates, being the whole country a sole district. Modern STV, however, uses geographic districts, usually between three and seven representatives. Versions of STV are today used for national elections in several countries: in Ireland and Malta for lower house elections, in Northern Ireland for local, European and Assembly elections, in Australia and India for their upper houses, in Australia for some regional elections and in Scotland and New Zealand for local government elections (Mitchell 2014, 247).

7. Both men were mistaken here. Bagehot was right to assume that ranking a huge number of personal candidates is beyond the cognitive capacity of electors. The Australian Senate offers an empirical confirmation: the electoral system may require voters to rank thirty or more individual candidates, but in 1984 the Group Vote Ticket (also known as’ above the line voting’) was included. Since then, over 90% of electors rely on tickets previously prepared by their political parties (Economou 2016, 117).

8. Mill is probably the best exponent of this process, since he successively defended two of the four systems: the Cumulative Vote, in his Thoughts about Parliamentary Reform, and the Hare system, in his Considerations (Hart 1992; Parsons 2009). Indeed, at the end of chapter 7 of Considerations, he included under the ‘Personal representation’ label all the contemporaneous alternatives to the Originating system. The problem is that those proposals are very different from each other. Mill cited the existing systems in Switzerland (Party List), in the USA (Cumulative List) and in Australia (the Hare System). Mill stressed that ‘the scheme is rapidly making its way’. Too much rapidly, indeed. Early on, the expression ‘proportional’ prevailed over the term ‘personal’. In 1878, for example, the Spanish edition of the Considerations already translated ‘personal representation’ into ‘proportional representation’ (Mill 1878). The same happened in several languages. See, for example, Petit Catéchisme de la Representation Proportionnelle, (1891), Belgium (anonymous).

9. Thompson (1976, 116) assigns Mill an intermediate position in the mandate-independence controversy. That is undoubtedly right, but here I do not want to situate Mill on that scale, but to illuminate why he is surprisingly situated on one of its ends.

10. Specialists in Mill are aware that there is a difficulty here. Urbinati, for example, suggests that Mill intended an ‘elitist use of proportionality’ (2000, 760). In my opinion, what is problematic is the customary use of the term ‘proportionality’ still applied to Mill and Hare (Ryan 2007, 165; Thompson 2007, 175; Riley 2007, 226; Kumar 2013, 138; Ryan 2014, 11). Maybe the use of the term ‘quotism’ would result in a more accurate clarification of Mill’s political thought.

11. In 1910, the ‘Royal commission appointed to enquire into electoral systems’ declared that STV ‘was not originally invented as a system of proportional representation’ (1910, 35); in 1984, Dummett affirmed – correctly, in my understanding – that ‘STV is obviously a majority procedure’ (1984, 280); in 2014, Farrell and Katz wrote that ‘in a technical sense it is a misnomer to identify STV as a proportional system at all’ (2014, 13).

Additional information

Funding

This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) under Grant DER2015-69217-C2-1-R.

Notes on contributors

Jorge Urdánoz

Jorge Urdánoz is Professor of Political Science in the Universidad Nacional de Eduación a Distancia (UNED, Madrid) and professor of Philosophy of Law in the Universidad Pública de Navarra (UPNA, Pamplona). He has been Visiting Scholar in Columbia University (sponsor Nadia Urbinati) and in the New York University (sponsor Steven Brams). His main research topics are political representation, electoral systems and theory of democracy. He has published several books and articles.