During the life of the current New South Wales parliament, an Electoral Districts Commission must be appointed to re-draw electoral boundaries for the Legislative Assembly. To allow time between the redistribution and the 1999 election, the redistribution should take place during 1997.

After elections, no process focuses the minds of those interested in politics more than the re-drawing of electoral boundaries, and the 1997 redistribution promises to have more political impact than most. The current Legislative Assembly has a finely divided political balance, after the Labor Party narrowly won the 1995 state election with less than half of the state-wide two-party preferred vote¹. In such circumstances, the redrawing of electoral boundaries will be a crucial pre-cursor to the 1999 state election.

Yet the strange fact is that in re-drawing electoral boundaries, the Electoral Districts Commissioners have no scope for considering the political impact of their decisions. They are restricted by a strict set of numerical criteria concerning current and projected enrolments, and several loose principles related to geography and community of interest'. In fact, the Commission will not even consider arguments relating to the political impact of boundaries. Reading submissions to the Commission requires seeing through the veil of community of interest in which the parties cloak their arguments, to see the reality of the political advantage they seek to achieve. That our electoral system pays more attention to the geographical identity of voters ahead of party politics stems from our inheritance of the British system of single member electorates. This equates representation with the geographical location of voters rather than any class or sectional interest they may also possess. It is a view of representation that reaches back to the middle ages, when the early Parliaments at Westminster were called together by the King as representatives of their Counties or Boroughs. After the industrial revolution, the vast disparities in the number of residents in constituencies lead to the various reform acts of the 19th and early 20th century that effectively introduced one person one vote to British politics. However, to this day, the drawing of British electoral boundaries pays far more attention to local government boundaries than is the case in Australia, and far greater differences in enrolment are accepted to maintain community of local interest in representation.

Yet this geographical view of representation pre-dates the development of modern political parties. While boundaries are drawn on the basis of geography and community of interest, clearly the majority of voters support parties based on historical class divisions which bear only limited relationship to geography.

In Europe, very different traditions of representation exist, and the use of proportional representation has equated representation with non-geographic interests, following social cleavages based on class, ethnic and religious divides. That Australia has not abandoned geographic representation for more direct forms is because the existing system has continued to be seen as legitimately translating the will of the people into party representation in Parliament. This acceptance should be compared with New Zealand, where the tendency in the last two decades for the electoral system to thwart the will of the people eventually lead to the replacement of single member electorates by proportional representation.

However, keeping legitimacy in Australia's electoral systems has required considerable tinkering. In the past, the use of different enrolment quotas in rural and urban seats (technically called malapportionment) was commonplace. This was justified at the time by the difficulty of rural communication and travel, as well as the importance of rural industry to the economy. However, as one side of politics (usually the Labor Party) came to see .malapportionment as a tool for keeping governments of the opposite political persuasion in office, its legitimacy was undermined.
In the last two decades, the process of drawing electoral boundaries has been reformed across Australia. With the exception of Western Australia and five seats in Queensland, one vote one value is now the accepted norm. Boundaries are now drawn by independent commissioners and no longer need to be approved by Parliament. Most jurisdictions also specify that Commissioners should pay some heed to expected growth in electoral enrolment.

The point should be made however, that these reforms have dealt only with the issue of equality of electoral enrolment. This is a very different issue to fairness in electoral boundaries. Equality relates to ensuring all votes have the same value, while fairness requires an assessment of whether political parties are able to return members or parliament in proportion to their level of state-wide vote.
With the exception of South Australia, Commissioners are not asked to address whether their boundaries are fair. Boundaries are drawn based on arguments about community of interest, and using strict enrolment criteria. The political impact of boundaries is not assessed.

So there is still scope for achieving electoral advantage through electoral redistributions, though the methods have become more sophisticated. Manipulation is still possible because of the nebulous definition of community of interest. There is often more than one possible solution to re-drawing the boundaries of an electorate, and competing community of interest arguments justifying each. It is in the arguments for the importance of different communities of interest that possible political advantage is revealed.

In the past, New South Wales has often been described as a natural Labor state'. It has been argued that the concentration of Liberal support on Sydney's north shore, and the large number of ultra-safe National Party seats means that the Labor Party appears to be able to gain political advantage even from numerically fair electoral boundaries. (See in particular my previous publication for the NSW Parliamentary Library, "NSW Elections 1984 to 1991: A Comparative Analysis", which deals with the last two NSW redistributions.)

It is within this environment that the 1997 NSW Electoral Districts Commission will function. Its work will be prescribed by a strict set of criteria, as outlined in Section 2 of this publication. Yet the submissions of the political parties, while set out in terms of community of interest, will be carefully designed to maximise political advantage.

After setting out the legislative framework for redistributions in Section 2, the balance of this publication relates to current and projected enrolments for existing electorates. This data is the raw material that will determine the scope of the redistribution. In Section 4, maps are provided showing the electorates projected to be over and under enrolment in April 1999, the date at which legislation specifies all electorates must be within 3% of the projected average enrolment. Section 4 also provides some commentary on how the Commissioners may deal with re-drawing the boundaries.
With two years to the March 1999 election, the day to day political battle will concentrate on the traditional fare of state politics, hospitals, schools and roads, with the addition of the millennial issue of Olympics organisation. Yet during the process of the redistribution, these issues will be mere byplay to the main event of trying to gain political advantage from the new electoral boundaries.

While an official count is not available, it is estimated that the state-wide two-party preferred vote was Labor 48.8%, Coalition, 51.2%.

---

The Author
Antony Green is the Election Analyst for ABC Television, and has worked for the ABC on every Federal, State and Territory election since 1989. He also writes regularly on electoral matters for the Sydney Morning Herald.

Antony studied at Sydney University, obtaining a Bachelor of Science in mathematics and computing, and a Bachelor of Economics with Honours in politics.

Antony is the author of a number of publications for the Parliamentary Library, including "NSW Elections 1984 to 1991: A Comparative Analysis", and "Electing the NSW Legislative Council 1978 to 1995: Past results and Future Prospects". In the current Parliament he has produced three publications based on the results of the 1995 election, "NSW Elections 1995", " 1995 Legislative Assembly Election: Estimated Two-Candidate Preferred Results by Polling Place" and "NSW Legislative Council Elections 1995".

Acknowledgments
NSW Electoral Commissioner, Mr Ian Dickson, and the staff of the Electoral Office, provided great assistance in locating enrolment figures for each month back to 1992. MS Frances Howatt, Australian Electoral Officer for New South Wales also provided assistance with dates and details of roll reviews. I must also thank David McClintock and Thomas Ashelford from the ABC for assistance in creating the coloured maps, and Paul Guilfoyle of the Parliamentary Printing Service for his assistance in e-mailing and printing the maps.
While all enrolment figures have been provided by the Electoral Office, all calculations are the responsibility of the author.