On Elections

How people elect parliaments

Evaluating the justifications of SMD elections

There are a number of well-known justifications for the single-member division (“SMD”) electoral systems such as the plurality (or ‘first-past-the-post’) systems used in the UK, the United States, Canada, India and several other nations, as well as the preferential SMD system used in Australia.

What are the merits of such systems? The three key justifications are maintaining tradition, localness of representation, and the installation of secure governments.

In considering these justifications in the context of Australian electoral law we should ask what does constitutional law have to say about them? We know certainly that the Constitution prescribes both a federal system of governance for the nation, and a system of representative and responsible governance at the federal level (and it would seem at the state level also).

This is the way it’s always been

The first justification for the use of SMDs to elect members to the House of Representatives is the simple notion that this is the way things have always been done, and therefore the way they should continue to be done.

This lack of interest in change is not ‘conservatism’ (in the particular sense of the term meaning a deliberate default position held to protect established and working systems, thereby placing a burden of proof on new alternatives to show that they are superior). It would be a more naturally conservative position to protect and defend the centuries-old ideas of representative and responsible government, rather than merely to protect a number of specific voting system mechanisms that were adopted during the late 19th and early 20th century.

Contrary to popular assumption, the use of single-member-divisions was not an original, or even a particularly early, development in the evolution of the English parliamentary election practice. For most of its history the House of Commons was composed of members elected in constituencies mainly of two members, and in some cases of one, three or four members. The voting method was the block vote.

Australian 19th century colonial systems also did not require single-member divisions.

In the United States the early development of electoral systems often saw multi-member divisions used together with the traditional block voting method. This led to such anomalous results that by the 1840s single-member divisions were proposed, and progressively adopted, as a preferable alternative. The result was systems based on the universal use of SMDs with plurality voting: “SMD-P”, or ‘first-past-the-post’.

By the late 19th century English and Australian opinion had also reached the conclusion that wherever plurality voting was to be used, single-member divisions were fairer. In due course they became the only acceptable option. The unhappy combination of plurality voting in multi-member divisions was finally abandoned in Australia by the 1920s, in Great Britain by 1948, and in the United States in the 1960s. SMD-P was the result left behind by history.

Other democracies that emerged from the British Empire during the 20th century – India, Pakistan, Malaysia and others – took on SMD-P more or less automatically.

As a matter of history SMD voting is not the way things have always been. SMD is a design choice, not a default option. SMD is not the original way in which the House of Commons in the Westminster Parliament was elected.

Local representation

Another justification for using SMDs, despite the severe constraint on choice which they impose, is the idea that we are all meant to be represented by parliamentarians who reside within a defined pool of electors, and thus have the character of ‘local members’.

Clearly there is a deep well of traditional attachment to the notion of local members. Belief in that attachment should be questioned, not only because single-member divisions cause the problems identified elsewhere, but because it’s not clear that it even serves a ‘local’ purpose.

The so-called local divisions for the Australian House of Representatives now have enrolments of around 100,000 electors. In the United States average registration per House of Representatives district was 495,000 in 2012, and in the Indian Lok Sabha in 2009 it was a staggering 1.3 million voters per district. It’s not clear that the work of any person with a small office and many competing responsibilities can have a meaningful association with a population of this size.

Moreover, there is little to link residential localism with the scope of political issues which are dealt with by a national parliament and a national government. Very few such issues have an inherently local nature. The securing of infrastructure assets or services for a local area is almost entirely a state-level concern, and indeed whenever the influence of House members is relevant to securing local infrastructure gains it is usually regarded as an inappropriate use of their influence. The advocacy of local economic interests is perhaps one of the areas of responsibilities which is genuinely shared between state and federal governments, but even there diversions of federal policy and expenditure to favour local employment interests often attracts criticism. In short, the role of a federal MP in securing local advantages is questionable at best.

Perhaps it’s time to make clear that the work of national-level politicians is concerned with very big-picture policy development and other parliamentary functions, and that while they should be answerable for their decisions to various groups within the electorate, we ought not expect that they provide locally direct benefits to them on any day-to-day basis.

Secure governments

A more structurally important justification for using single-member divisions is the view that such electoral systems produce a desirable form of ‘stable’ government – a condition defined as a ruling party or coalition having a secure majority in the House.

Proponents of stable government argue that this way of arranging matters leads to two public goods:

  • executive administrations have time and a clear run at developing and delivering their programs, in particular their election manifesto offerings
  • the system gives the electorate a decision-making architecture in which each election operates as a process for either installing a desired executive, or for removing an unpopular one (or both).

‘Stable’ is a positive-sounding word; who would prefer unstable government? But what is really being sought here is not so much stable government, but secure government. And the security of this form of government is to be obtained by ensuring that at all times the executive has a secure majority in the Parliament. Such a secure position may apparently be artificially contrived by the use of convenient electoral systems – even where doing so works contrary to the mandated achievement of other constitutional goals.

But the proponents of this view make little attempt to reconcile their desired secure, or stable, government with the constitutional notion of responsibility.

Unfortunately for the stable government argument, both limbs of the justification involve a significant conflict with the principle of responsible government. The constitutionally mandated concept of responsible government carries the meaning that the executive must be answerable to the Parliament, and must be so in an ongoing, not a periodic manner. These two principles are deeply etched in the constitutional history of British-based political systems.

Entirely secure government cannot also be responsible government, if the effective answerability of the ministry to the Parliament, and specifically to the House, has been undermined.

And serial government – that is, secure government installed between the holding of periodic elections – is also clearly contrary to the constitutional idea of responsible government, which is premised on the contradictory principle of ongoing, constant accountability.

In short, the notions of the secure government school are actually contrary to principles that are mandated by the design and structure of the Australian Constitution.

Attention should also be given to the kind of political culture which ‘security’ of government creates. Responsible government and diverse, representative parliaments create a climate where minority special interests on each distinct public issue, while they may attract a well-founded number of supporters in Parliament, will generally be rejected by majorities. By contrast secure governments, made possible by artificial and rigid parliamentary majorities, will tend towards a climate in which special interests can engage in political support-trading with governments, and with the major political parties which make up those governments, enabling such minorities to obtain various kinds of benefits or closed policy outcomes. These include rent-taking or taxation advantages for certain industries (or even specific businesses), imposition of the values of specific religious or environmental groups, one-sided interpretations of conflicting interests in policy matters such as intellectual property. Conversely, secure government will tend not to lead to ‘open’ legislative outcomes, such as open laws regarding personal freedoms, open markets, etc.

The distinction that emerges is a political dichotomy not between ‘left’ and ‘right’ versions of economic policy or wealth distribution, but more between ‘open’ and ‘closed’ regimes of government and law, cutting across traditional political topics and interests. We should keep this in mind when we weigh up what overall character of parliament we really want.

The secure government vision and its justification of the use of single-member electoral divisions works contrary to the notion of responsible government. This does not mean that the argument has no place in general political discourse, but it does mean that in attempting in an Australian constitutional dispute to justify any piece of legislation which has an adverse impact on the system of representative and responsible government, the arguments of the ‘stable government’ school will not provide any strength to meet the constitutional requirements articulated in the Australian High Court’s Lange test.

Conflicting principles

Given Australia’s history, it seems counter-intuitive to conclude that the enactment of an SMD electoral system might be contrary to the constitutional powers of the Parliament. In any case, strict versions of equality of voter influence and maximisation of actual representation do not appear to be constitutionally protected, and to the extent that choice is protected it only requires a pretty minimal level of choice.[1] SMDs are thus at least a permissible option for Parliament to adopt. But the political argument for their use is weak, and finds no support in the constitutional structure which is required – that of representative and responsible government.

Consider a counterfactual: If a highly representative voting system (such as STV) was already in place and legislation for a replacement system based on SMDs was enacted, given that the change would threaten to result in manifestly lower outcomes in terms of representative parliaments and responsible government, would such legislation pass the Court’s Lange test? In other words, would Parliament be allowed to legislate to manifestly diminish its own nature and undermine the mandated type of system of democracy?

The SMD electoral system has always been inherently contradictory to the aims of representative and responsible government. How did such a contradiction between the aims of representative and responsible government and the system in use for over a century arise? The answer may be in part be that among the generation of Australians which established the Constitution and enacted the early legislation in Parliament the impact of party discipline in parliaments had yet to fully develop, and the full impact of SMDs on representative and responsible government had not yet become apparent.

Another answer may be that to some of the founders it was apparent, but they had a personal interest in preventing a resolution of the contradiction, and instead sought to sustain the practice of SMD to maintain their own political longevity and influence.

The STV alternative

By contrast with the current system of electing the House, if Australia were to adopt core features of STV – a common quota used nationwide (determined before election day based on past turnouts and results), open (ie: division-less) voting for candidates from all quarters, and free (ie: optional) preferencing – we would have very high levels of choice, very equal influence on election results, and very high (80%+) levels of actual representation. We would probably also put upward pressure back on electoral participation, and certainly we would not be excluding hundreds of thousands of voters from participation by needlessly invalidating their votes.

A common quota used nationwide (as opposed to a variety of division-based quotas of different size), even if applied within states as a form of electoral division, might mean that the precise numbers of members elected to the House would not be strictly fixed, and might vary slightly in response to voter turnout and preferencing trends. That doesn’t seem much of a problem if it is a price paid for near-perfect voter equality of influence.

For the time being, of course, the Australian Constitution does require that the numbers of members elected in every State is pre-fixed, and indeed requires that we must elect members in vote pools demarcated by the State boundaries and by our places of residence. While such requirements endure, a universally simple STV voting system could be slightly bastardised by limiting the numbers elected, if necessary by closing the number of seats won once the required number for a state is reached, even if a high turnout meant that an additional candidate might assemble a quota of votes. Conversely, if not enough candidates win a quota of votes in any given election, the means of filling seats ‘below-quota’ is already an accepted practice in Tasmania and the ACT. Applying the constitutionally fixed state numbers of seats to a national system using a common quota is like using two sieves. It does not necessarily result in better sieving, but the harm to the achievement of electoral goals we seek would not be great.


Malcolm Baalman
April 2014


[1]   The judgement in Langer v Commonwealth (1996) upheld Parliament’s capacity to enact rules for the election of members of Parliament by compulsory preferential voting. The reasoning of several judges indicated that the expression “chosen by the people” is a general description of the required democratic process, and does not confer on individuals a right to any specific form of choice.


This article is based on extracts from my Submission 181 to the Joint Standing Committee on Electoral Matters inquiry into the 2013 election which took place during 2014.

%d bloggers like this: