There are two ways of evaluating proportional representation (PR), and they are usually not brought together.
First it can be evaluated as an electoral system that produces a more representative and therefore more democratic legislature, because it awards representatives in proportion to shares of votes. Few outside the ranks of those dedicated supporters appreciate this virtue. A long history of simpler systems has ingrained the idea that someone should win elections and everybody else lose. We ought, however, to be particularly conscious of the virtue of representativeness in recent times. We have had one general election (1996) in which the incoming government secured an overwhelming majority of seats in the House of Representatives with less than a majority of votes (47.3 per cent), and the most recent general election in which the same government was returned with a comfortable majority of seats in the House with less than 40 per cent of the vote (39.7 per cent) and less than its major rival (40.05 per cent). In each case, the Senate, elected by PR, provided a more accurate reflection of the electors' votes. Yet the government, with little fear of ridicule, claimed a mandate arising from the support of the electors and began to call for reform of the unfair Senate.  The psephologists and political scientists may demonstrate the bogus character of this `reform', but the rest of us are at least partly taken in.
Second, proportional representation can be evaluated in terms of its effect of depriving governments of control of proportionally-elected houses, and thereby providing a legislative safeguard. From bitter past experience, we know that governments with the total power conferred by complete control of the legislature tend to become arrogant, overbearing and corrupt, and that an upper house not under government control can provide an antidote to this disorder.
A failure to synthesise these two ways of evaluating proportional representation leads to internally contradictory attitudes. We like someone to win elections and run the country, but we complain when we are the recipients of their high-handedness. We regard proportional representation as a nuisance, but we do not want dictators riding in on 40 per cent of the votes. Because discussion of governance is usually dominated by clichs, this contradiction can be disguised by the simultaneous adherence to contradictory propositions represented by such clichs.
Contradictions and clichs
A very good example of this was provided recently by the Egan affair. The Treasurer in the New South Wales government, Michael Egan, is a member of the Legislative Council and a firm believer in his and his party's right to govern through winning a plurality of votes at an election. When the Council, elected by proportional representation and not under a government majority, sought information about allegedly nefarious government activities,Egan refused to produce it. The Council demanded the information and, when Egan persisted in his refusal, imposed a penalty on him by suspending him from the sittings of the Council. His response was to take the Council to court in an attempt to establish that the Council does not, as a matter of law, have the power to require the production of documents. Although he spent a great deal of the taxpayers' money in legal costs, he singularly failed in this endeavour. Both the NSW Court of Appeal and the High Court found that the Council had the power to act as it did,  and the Court of Appeal found that claims of legal professional privilege and public interest immunity do not protect the government from the exercise of the Council's power. 
One would have thought that Egan would be subjected to severe criticism for spending so much public money in attempting to establish that the public's representatives in parliament do not have the right to information about government activities. One would particularly expect that the Press, in accordance with its usual support for open government and availability of information, would be hostile to Egan's enterprise. On the contrary, at least in the initial stages of his legal battles, the media's attitude was ambiguous. It did not like Egan's secretiveness, but nor did it like the `undemocratic' Council interfering with the `democratically elected government'.
This contradiction was concealed by the use of two dominant clichs: the government must be allowed to govern, and the government must be accountable. These two truisms run through political discussion, academic and profane. When they are used with any meaning, they signify that a government should be allowed to pass all its legislation without tiresome parliamentary processes that are difficult to follow and report, but that governments should not be allowed to keep secrets, particularly where they are up to no good. That there might be some incompatibility between these propositions is usually not detected.
An exemplar of this mode of thought was provided by an editorial in the Sydney Morning Herald at the height of the affair, which said that Egan should hand over the documents and that the Legislative Council should be reformed to stop its interfering with governments.  It did not seem to occur to the editorial writer that `reform' of the Council would mean that the Egans of the world would never hand over any information unless it suited them and that the quality of governance would deteriorate accordingly.
Governments' view of government
This editorial was particularly foolish in that it did not seek to put forward Egan's own defence of his conduct. He and his legal advisers were well aware that an internally contradictory argument might pass muster in a newspaper editorial, but would not go down well in court. His case in court had to be rather more sophisticated. It rested on his own theory of responsible government, which is that governments are responsible only to the lower houses, which they normally control, and not to upper houses. This argument was given very little consideration by the courts, and the Court of Appeal, while expounding the judicial recognition of the principles of responsible government, explicitly held that governments' accountability to parliament does not exclude the Council.
Other audiences, also less susceptible to clichs, required a different treatment of the question. At a parliamentary conference, I asked Egan whether the public had the right to see documents relating to the Fox Film Studios/Showground agreement, which he denied to the Legislative Council. His unusually frank answer was to the effect that the public did not have the right to see the documents, and that the government, having been elected to govern, had the right to determine which documents should be made public. When it was pointed out that the electors had also chosen the Council, he responded that the electors should be confined to choosing a government at election time and should not be allowed to elect a second chamber with the means to call the government to account.  This, though not often stated, is the current theory of the system of government held by our governors. It amounts to a denial of accountability, that is, the right of the public to have the information required for an informed electoral choice.
In order to appreciate how we have arrived at a situation in which ministers can tell us that we do not have the right to know what they are doing, it is necessary to realise that the concept of accountability, which our commentators support even while supporting Egan's power to undermine it, itself represents a significant constitutional slippage.
The framers of the Australian Constitution adopted a set of institutions that they called responsible government. At that time, this meant that the executive government, the Cabinet, was responsible to the lower house of the legislature in the sense that the executive could be removed from office by that house if that house considered that the executive no longer merited the house's confidence. Even at that time there were dissenting voices who warned that responsible government no longer worked as supposed. Since then, we have become familiar with their thesis in an updated form: the Executive controls the lower house through a disciplined party majority, and the house no longer removes governments or installs new ones, except in times of great political crisis involving splits in the government party, which are now highly unlikely to occur. Responsible government has disappeared, or at least developed into something different.
We now no longer speak of responsible government in that sense. Instead, we settle for something less, called accountability. Governments should be accountable to parliament, that is, obliged to give account of their actions to parliament and through parliament to the public. Governments are then responsible to the electorate at election time.
The problem with this picture is that the system of government has continued to develop, and has moved on again in a way that requires a further reassessment. Governments now expend a large part of their time and energy suppressing parliamentary accountability, seeking to ensure that they are not held accountable by parliament, that old accountability mechanisms do not work and that new ones are not introduced. Just as the party system developed to ensure that governments formed by the majority party are not responsible to parliament, so that governments are never overthrown by parliament, the system has developed further to ensure that governments are not held accountable by parliament, so that they are less likely to be overthrown by the electorate at the next election.
In that context, PR, by denying governments control of upper houses, has prevented the virtually complete suppression of accountability that occurs when governments have that control.
This is demonstrated by a history of the accountability of government to parliament at the federal level.
Accountability measures: history
It is well known that the Senate, over many years, has developed measures to require greater accountability on the part of governments. Some of the more significant measures are:
- As early as 1932, the Senate established a committee to scrutinise delegated legislation, laws made by the executive government, with independent advice and in accordance with criteria related to civil liberties and proper legislative principle. We thereby avoided at the federal level the situation in some other jurisdictions in which delegated legislation has escaped parliamentary scrutiny. In conjunction with the establishment of the committee, the Senate developed legislative measures to ensure parliamentary control of delegated legislation.
- The Senate established in 1970 a comprehensive standing committee system to allow regular inquiries into, and the hearing of public evidence on, matters of public concern, including proposed legislation.
- The Senate established in 1981 the Scrutiny of Bills Committee, which is described below.
- The Senate has increasingly used orders for production of documents to require governments to produce information on matters of public concern. (A particular example may be mentioned: the Senate requires all government departments to place lists of their files on the Internet, to guide people making freedom of information requests.)
- The Senate has frequently amended legislation to include provisions for the appropriate disclosure of information (into this category falls the Freedom of Information Act itself, which was extensively amended in the Senate).
- The Senate adopted in 1989 procedures for the regular referral of bills to committees, so that any bill may be the subject of a public inquiry with opportunity for public comment. (In this connection reference may be made to the fact that the government initially resisted the reference of the goods-and-services tax legislation to a committee, even though, as was pointed out, such a complex legislative change merited close scrutiny and public comment.)
- The Senate conferred on its standing committees the power to examine the annual reports of departments and agencies to determine the adequacy of the reports, and to inquire into the operations of particular departments and agencies at any time.
- The Senate has continued and improved the twice-yearly estimates hearings, which are opportunities for senators to inquire into any operations of government departments and agencies. There is now the ability to have follow-up hearings on particular matters.
- The Senate established deadlines for the receipt of government bills, so that governments cannot introduce large numbers of bills at the end of a period of sittings with the demand that they be passed during that period of sittings. These deadlines are an attempt to remedy the `end-of-session rush' and `sausage-machine legislation'.
- The Senate has taken steps to ensure that governments explain any delay in proclaiming acts of parliament duly passed by the two houses.
- The Senate has amended retrospective taxation legislation to ensure that it is not backdated to vague pronouncements by ministers (retrospectivity is accepted if the backdating is to a clear statement of government legislative intent).
- Many other lesser measures have been taken, such as requiring governments to respond within a limited time to parliamentary committee reports, and placing time limits on answers at question time, so that ministers cannot give 20-minute speeches when they are supposed to be answering questions.
These measures are generically described as accountability measures because they are all founded on the requirement that governments explain what they are doing and why.
It would be difficult to discover anybody who would question the merits of these accountability measures, except ministers currently in power (ministers out of power usually claim to have invented them). Such measures would, I suspect, universally be accepted as meritorious. The significant fact about them, however, is that historically most of them were resisted by the government of day, and accepted only when required by a difficult numbers situation in the Senate. Governments have to be forced to be accountable; they resist accountability and have to be compelled to explain themselves. If this has not always been true, it is certainly true now at least.
The resistance of governments to accountability measures explains, and is also demonstrated by, the conspicuous lack of those accountability measures in lower houses. Where such measures are adopted, they are rigorously controlled by government; an example being the control exercised by governments over committee systems in lower houses. An enormous amount of mental energy has been expended in recent decades over the subject of parliamentary reform, but the fact is that reforms have taken place only in houses of parliament not under government control.
To take a particular example from the list, the Senate established in 1981 the Scrutiny of Bills Committee to scrutinise legislation, with independent advice, to ascertain conformity with criteria related to civil liberties and proper legislative principle. The government opposed the Senate's establishment of the committee and most government senators voted against it. Some government senators were extremely embarrassed by the government's decision to oppose this new measure of accountability, and some of them crossed the floor to support the establishment of the committee. If a similar issue were to arise now, there would be no embarrassment and probably no dissidents. Governments now regularly oppose all accountability measures, and can rely on their backbenchers' unwavering support in doing so.
Another example is provided by the frequent passage by the Senate of orders for production of documents. These orders require ministers to produce relevant documents about matters of significant public concern and controversy. Most of these orders are complied with, for reasons that I will give later. The significant points are that nowadays senators have to resort to these orders because mere requests for documents can be assumed to be ineffective, and it is assumed that governments will resist the passage of the orders in the Senate and will probably not produce, on various well-worn grounds, at least some of the documents ordered.
Legislative powers and accountability
So proportional representation is favourable to accountability of government. Yet we have that underlying belief that governments must be allowed to govern, that is, must be able to pass their legislation. This leads to a more sophisticated version of the governments must govern/governments must be accountable contradiction. The solution to the dilemma, it is said, is that proportionally elected upper houses should not have any power to amend or reject government legislation, but should concentrate on holding governments accountable through committee inquiries and so forth. This is a well-worn track in the political literature.  Unfortunately, it rests on ignorance about how legislatures actually work. Upper houses have only one hold over governments, their ability to withhold assent from government legislation. This is the only reason for governments complying with accountability measures of upper houses: as a last resort, an upper house with legislative powers could decline to pass government legislation until an accountability obligation is discharged. An upper house without legislative powers could simply be ignored by a government assured of the passage of its legislation. A reviewing house without power over legislation would be ineffective.
This is not a theoretical construct, but a practical, every-day factor in the operations of the Senate. I am frequently asked by senators for advice on what can be done when governments refuse to provide information. I respond by drawing attention to the spectrum of remedies available. At one end of the spectrum is criticism of the government for failure to provide the information; at the other end is refusal to deal with all or some government legislation until the information is produced. The first is usually seen as inadequate, while the other is seen as too drastic in most circumstances. There is then the middle way: initiate a debate in the Senate so that government time that would otherwise be spent on government legislation is expended on the matter in contention. This is the hold that the Senate has over governments: if they do not produce appropriate information when required, their legislative program will be disrupted. Ultimately, there is the threat that legislation will not be dealt with at all. Governments then have to weigh the embarrassment that may be caused by the disclosure of awkward information against the disruption of their legislative program. Usually they decide that the avoidance of embarrassment is not worth the trouble in the Senate. When information is very embarrassing, they opt for the trouble in the Senate. The end result is that much information is made available that would not otherwise be known to the public, while the government may keep its most embarrassing secrets and allow the public to form its own judgment about its motives.
Proportional representation and the Australian solution
A coherent evaluation of government in Australia today must begin with a closer examination of the concept represented by the clich that governments must be allowed to govern. This notion is really based on an avoidance of the foundation problem of government: how to allow governments sufficient control to ensure the continuance of civil society while in turn subjecting them to sufficient control to ensure that their power to preserve does not become a power to destroy.
No system of government provides a foolproof solution to this foundation problem. Perhaps we could come to a realisation that the Australian system provides a solution that may be as good as any. Lower houses elected by single-member constituencies, with an executive dependent on control of those houses, provide relatively stable and strong executive governments, while upper houses possessing legislative powers and elected by proportional representation provide a better reflection of the voters' opinion, a democratic quality control on legislation and a means of ensuring that governments do not entirely avoid public accountability.
Figures supplied by the Australian Electoral Commission. The government case for `reform' appeared in a paper by Senator Helen Coonan, `The Senate—safeguard or handbrake on democracy?', The Sydney Papers, vol. 11, no. 1, Summer 1999. The sources to which she refers in her notes are nearly all hostile to her kind of `reform', a fact not apparent from her paper.
Egan v Willis & Cahill (1996) 40 NSWLR 650; (1998) 158 ALR 527.
Egan v Chadwick and others (1999) NSWCA 176, 10 June 1999.
`Release the Egan papers', Sydney Morning Herald, 2 December 1998, p. 18. For a change in the tone of the press, see `Government must be accountable', Australian, 11 June 1999, p. 10.
This exchange was not recorded, but Egan's theory of government is expounded in the submissions put to the courts in the first case cited in footnote 2.
See Senator Coonan's paper cited in footnote 1. Also R. Smith and L. Watson, eds, Politics in Australia, 2nd edn, St Leonards, NSW, Allen & Unwin, 1993; G. Maddox, Australian Democracy in Theory and Practice, 3rd edn, Melbourne, Longman Australia, 1996.