Redistribution methodology

Determining boundaries

In drafting suggestions, I seek to comply with the criteria set out in subsection 66(3)(b) of the Commonwealth Electoral Act 1918, taking into account:

1. community of interests, including economic, social and regional interests;
2. means of communication and travel;
4. the physical features and area; and
5. the boundaries of existing divisions within the State.

Where possible, I aim to use clear and recognisable boundaries to ensure that electors can readily identify the division in which they live. I will split SA1 areas where a geographic feature makes for a more logical division boundary. In cases where split SA1 elector data is not available, I make a reasoned judgement as to the likely distribution of electors. The Redistribution Committee will, of course, have access to more precise data to ensure compliance with numerical requirements, However, in these situations, I always aim to ensure that either division could afford to contain the entire SA1 population within the numerical threshold.

In pursuit of improved community of interests representation, I make use of the full numerical threshold under subsection 66(3)(a), recognising that this threshold exists for a reason. I believe the reason for an allowance is to facilitate better alignment with community of interests. I therefore may draw divisions close to the permissible upper or lower enrolment threshold to achieve the best outcome.

My approach generally follows a cascading process, beginning with an anchor division – usually located in a corner of the state or an area with limited capacity to expand. From this starting point, I work across the state, adjusting neighbouring divisions in a sequential and logical pattern. Smaller states or territories may not require this approach and I'll alter the approach depending on how many divisions I need to modify. For instance, in the 2025 South Australian redistribution, a statewide shift was not needed. Most divisions were within, or at least very close to, threshold. The divisions that did require signficant changes, serendipitously were next to one another and could accommodate a relatiely minor adjustment.

Boundaries of existing divisions

I use the boundaries of the existing divisions as building blocks in my suggestions. However, in doing so, I note that subsection 66(3A) of the Electoral Act, the existing boundaries must be considered as subordinate to the other criteria in subsection 66(3). Therefore, existing boundaries must always yield where they divide a community that could otherwise be united, provided numerical tolerances permit such a change.

Community of interests

Community of interests is a broad and often contested concept. In practice, it refers to groups of electors who share access to similar services and amenities, participate in the same social and cultural networks, share demographic characteristics, or identify with a common place or identity. It is rare for an electoral division to represent just one single, homogeneous community. Almost all divisions will contain multiple, interconnected communities that function together.

Local government areas can be a useful indicator of communities of interests, particularly in rural and regional areas, where several townships often share services and economic activity. However, in metropolitan areas, local government tends to be less reliable as an indicator of community cohesion. Boundaries may reflect outdated administrative arrangements, or recent cost-cutting amalgamations, that do not correspond with contemporary lived communities. Accordingly, I am more willing to divide local government areas in urban contexts where other factors and boundaries result in more coherent representation of community.

Suburb or locality boundaries, while sometimes arbitrarily or inconsistently defined, have been used by previous Committees and are commonly understood and familiar to electors. They can be useful in both metropolitan and regional areas, particularly where no other clear geographic boundary exists in the area.

Communication and travel

Means of communication and travel are fundamental to determining effective and representative boundaries. Communities connected by major roads, public transport corridors, or shared infrastructure are more likely to function cohesively. Where separate communities must be included with the same division I advocate for this to occur along transport and communication corridors where possible.

While major roads, highways or railways are used as boundaries to a great extent, I note care must be taken to ensure that a boundary is not drawn down a road or railway that results in inadvertently splitting otherwise united communities. For instance, in rural areas, sometimes a railway station or major road will become a community hub with electors and residents from either side accessing the facilities. In these types of cases the railway line should never be used as the boundary, at least not through the part of the division where the station or hub is located.

Crucially, divisions should never be drawn in a way that makes them only contiguous on a map. A boundary that results in two parts being physically separated by impassable terrain, such as forests, ridges, national parks, un-spanned waterways or undeveloped bushland, may satisfy technical contiguity, but fails the practical test of means of communication and travel. Electors in these areas may feel disconnected, not only from one another, but from their elected representative. This is especially true when the divide is grossly uneven. A division with a small pocket of electors disconnected from the bulk of the division will never feel represented. This risk should be avoided wherever possible, even at the expense of greater numerical neatness.

Physical features and area

Natural geographic features can help define logical and effective electoral divisions. Rivers, ridgelines, reserves and open space corridors can serve as both practical boundaries and symbolic dividers of community identity, particularly where they also limit communication or development.

Where such features serve as real-world barriers to movement, access or interaction, they should be used to inform electoral boundaries. However, much like roads and railways, it’s equally important to avoid using these features where the act as shared connectors, such as riverfront precincts, lakeside parks or transport corridors along valley floors.

Political fairness

My suggestions are written in a manner that is wholly impartial. I do not consider the political implications of any boundary changes, nor do I attempt to analyse voting patterns or booth-level results, or try to balance the number of members between parties. In my view this is consistent with the intent and structure of section 66 of the Electoral Act, which deliberately omits any reference to political outcomes or representation.

Fairness in this context means neutrality. Boundaries should be drawn only with regard to the criteria spelt out in the Electoral Act and the needs of electors, not to the political consequences for parties or candidates. Electoral boundaries should reflect people and communities, not polls and politics.

In this role, I am impartial, as reqiured in my employment as a member of the Australian Public Service.

Drawing the line

Ultimately every redistribution must draw boundaries, and those boundaries will necessarily divide some communities. It is an unavoidable consequence of the task. There will always be electors who feel they belong more with those just across the street or creek than with others further away in their own division.

When determining boundaries, I seek to minimise those cases but recognise they cannot be eliminated entirely. My goal is always to ensure that, when a line must be drawn it is done so logically, transparently, and with the greatest possible regard to the criteria of subsection 66(3).

I believe my submissions to the various redistribtuion committees over the years have achieved this, and I hope to continue providing input and analysis to this vital electoral process into the future.