Dr Francis Markham
Centre for Aboriginal Economic Policy Research
Australian National University
Reply Paid 83380
CANBERRA ACT 2601
30 April, 2021
Submission on the Interim Report, Indigenous Voice Co-Design
I thank the Co-Design Senior Advisory Group for the opportunity to provide a written submission on the Interim
Report. I write as a non-Indigenous academic, working on Indigenous public policy issues. Although I am part of
the team evaluating the New South Wales Government’s Local Decision Making initiative, my submission does
not reflect the content of that currently unpublished research. Rather, my submission is primarily based on
published academic and government literature.
I write with regards to two points: the protection offered to a First Nations Voice by constitutional enshrinement,
and the proposed articulation of the National and Regional Voices outlined in the Interim Report.
The value of constitutional enshrinement
While I acknowledge that the terms of reference for the co-design process specifically excluded making
recommendations about constitutional recognition, the issue of constitutional enshrinement remains perhaps the
single most contentious issue regarding any proposed Voice. Consequently, a brief comment on it is necessary.
I support constitutional enshrinement for three reasons.
First, the proposed arrangements for constitutional enshrinement have few drawbacks. As Twomey (2015)
pointed out, a constitutionally enshrined Voice would not constrain the powers of parliament or the executive. If
justiciability is a concern, it is possible to include a non-justiciability clause in any alteration to the Constitution.
As such, there is little risk that constitutional enshrinement will diminish the power of parliament.
Second, constitutional enshrinement will enhance the effectiveness of a Voice. Constitutional alteration requires
that a proposition put to a referendum is endorsed by a majority of electors in a majority of states. If such a
referendum were held and were to succeed, it would increase the moral and political authority of the Voice with
the public. This would increase the likelihood that its advice would be heeded by Parliament and Government.
Without a mandate from the Australian people, the risk that the advice of any Indigenous Voice would be
sidelined and ignored is substantial.
Third, constitutional enshrinement of a Voice was a key demand from the Regional Dialogues that led to the
National Constitutional Convention in May 2017. This unprecedented deliberative process involved thousands of
mostly First Nations delegates across Australia (Referendum Council, 2017). Legislating the Voice, rather than
constitutionally enshrining it, would run counter to the voices of Indigenous people expressed through the Uluru
Statement from the Heart. The strong desire of First Nations people for a constitutionally-enshrined Voice is
confirmed by polling published in two Australian Reconciliation Barometer reports (Reconciliation Australia,
2019, 2020). Given the high level of both deliberative and popular support among First Nations people, to
instate a Voice without constitutional enshrinement would be a contradiction: an agreement to listen to First
Nations people, but made without listening to First Nations people about the terms of their representation. To be
blunt, the introduction of a non-constitutionally enshrined Voice would be a somewhat insulting moment of
Popular support for a constitutionally-enshrined First Nations Voice
There is a widely articulated fear that a First Nations Voice to Parliament should not be put to a referendum,
such would be the set back to Indigenous interests should such a referendum fail. A professed disbelief in the
prospects of such a referendum succeeding is indeed one of the reasons that the current Government has
refused to hold such a referendum (Turnbull et al., 2017).
However, public support for a constitutionally-enshrined Voice is already very strong among the Australian
public. In my analysis with William Sanders of public opinion research between 2017 and 2020, we found that
levels of support for a Voice since 2017 among the Australian public have been considerable (Markham &
Sanders, 2020) in their analysis. I have updated that analysis below to include two more recent surveys, the
latest Reconciliation Barometer (Reconciliation Australia, 2020) and the 2021 Australian Constitutional Values
Survey (Deem et al., 2021). As Figure 1 below shows, in most polls, 70–80% of voters with a committed position
on the matter support a constitutionally-enshrined First Nations Voice to Parliament.
Figure 1: Levels of support for a constitutionally-enshrined First Nations Voice to Parliament among
voters with a committed position, 2017–2021.
Sources: Markham and Sanders (2020); Reconciliation Australia (2020); and Deem et al. (2021).
The same study found that while support for a First Nations Voice is patterned along partisan lines, the public
statements (or ‘signaling’) of party leaders matters. If the Coalition leadership returns to the idea of a
constitutionally-enshrined Indigenous Voice with support, they are likely to sway another considerable group of
voters towards supporting a Voice. In doing so, it would make the success of a referendum all but assured.
Even without active Coalition support, if a future Labor government puts constitutional enshrinement of a First
Nations Voice to a referendum, it is likely to be successful.
Ensuring representation of regional communities
In the hands of the regions?
Since the advent of the self-determination era in the early 1970s, a tension between competing principles of
representation has animated the design of national bodies to advise the Commonwealth Government on, and
sometimes administer, Indigenous affairs. As the Indigenous Voice Interim Report makes clear, there is a great
deal to be learnt from previous nationally representative bodies, in particular the NACC (1972–1977), the NAC
(1977–1985) and ATSIC (1990–2005). Key among these is a tension between what I describe as localism and
decisionism. The principle of localism is based on the notion that the legitimacy of the institutions of Indigenous
representation relies on responsiveness to local needs, preferences, and priorities. In this view, decisions by
Indigenous people about their future are best made at the community level. The principle of localism is premised
on a model of First Nations decision making which sees ‘the slow emergence of consensus developed over a
period of time during which discussions [take place] at every level from family though larger groups to the
community as a whole…’ (Coombs, 1972, quoted in Rowse, 2000, p. 110). In this view, aggregation of local
opinion upwards is the priority. As a principle of institutional design, this requires facilitation of a slow, iterative
process of information provision, consultation and discussion, feedback and so on. Crucially, this must take
place in local communities. The principle of decisionism, on the other hand, stresses the need for Indigenous
representatives to have the authority to enter negotiations and make decisions on behalf of the people they
represent. Such decisions need be binding if they are to have authority, and unequivocal if they are to be
effective in lobbying or advocacy efforts. The number of parties to such a decision need to be workably small to
facilitate rapid agreement amongst representatives. Alternatively, a method is required for an authoritative
decision to be made in the absence of consensus (e.g. majority vote, executive decision by a chairperson, etc.).
In contrast to the liberal localist approach where the solution to disagreement is conversation, in the decisionist
approach to the solution to disagreement is the agreed vesting of power in an individual or small group with the
authority to decide.
All three previous national, representative First Nations’ bodies were viewed as too decisionist by those who
reviewed their operation. The Hiatt Review of the NACC suggested that ‘the great majority’ of the bodies
constituents knew ‘practically nothing of the formal activities of the NACC’ (Hiatt et al., 1976, p. 45). The
legitimacy of the NACC rested on its ability to ‘develop and maintain close contact with local community
organisations’ (p. 48), an ability that its reviewers found was severely constrained. Coombs, in his review of the
NAC some eight years later echoed this conclusion. According to Coombs, the most frequent criticism put
during consultations was that the NAC was ‘out of touch with Aboriginal communities and their organisations’
(Coombs, 1984, p. 11). Similarly, ATSIC, in its final review before abolition, was criticised for its ‘top down’
model, where ‘few, if any, of its policy positions are initiated from community or regional levels’ (Hannaford et
al., 2003, p. 32).
Put simply, reviews of all previous national bodies have found that they leaned too heavily toward centralised
decision making rather than localism in their decisions. The long-running question about the form of institutional
arrangements that best facilitates an articulation of local First Nations political authority at the national level
remains open. The National Voice proposal provides a starting point to address this question.
The proposed National Indigenous Voice suggests a design once again focused on decisionism. Among the
agreed design principles was the determination to limit the size of the group to facilitate action: ‘Membership of
the National Voice should be restricted to no more than 20 members to ensure the maximum workability and
flexibility of the body’ (Langton et al., 2020, p. 30). Furthermore, it was proposed that the accountability of the
National Voice to the regions need not be strong: ‘linked in some way with the representative structures for
Local and Regional Voices that would be established’ (Langton et al., 2020, p. 30).
The Interim Report provides for two possible options for the membership of the National Voice:
Core model one: Structurally linked through membership, with members drawn from the Local and
Regional Voices level to the National Voice.
Core model two: Directly elected to represent the state, territory or Torres Strait Islands on the National
It is my view that core model one, which emphasises the accountability of National Voice members to local and
regional bodies is more likely to ensure that local communities’ concerns are heard at the national level. A direct
election model is likely to produce a set of parallel institutions (Local and Regional Voices, and a structurally
separate National Voice) that would likely fall into conflict and competition for resources and attention. Members
of the National Voice would face incentives to serve their jurisdictional electoral constituency rather than the
voices of the regional bodies. In contrast, members of a National Voice that are drawn from local and regional
bodies would have an incentive to attend to the concerns of those bodies.
It is at the Local and Regional Voice level that the framers of the Interim Report see community engagement
taking place. In my view, it is also at the Local and Regional Level that community accountability will be
strongest, especially if members of the Regional Voice are required to reside in their regions.
For this reason, it is important that the ‘structural link’ between the National Voice and the Local and Regional
Voices be as strong as possible. As a starting point, the members of the national voice for a state or territory
could be selected by an assembly of the chairpersons of Local and Regional Voices in that jurisdiction. To
further strengthen this link, we suggest that the nominated members of the National Voice be subject to recall by
Local and Regional Voices should their performance prove inadequate. We expect that a National Voice that is
only weakly accountable to Local and Regional Voices will struggle to earn and maintain its legitimacy among
First Nations peoples.
The proposed structure could be made more workable by increasing the number of members of the National
Voice to equal the number of Local and Regional Voices, with each regional body delegating one of their co-
chairs to become a member of the National Voice. The gender balance of the National Voice could be assured
under this arrangement by alternating the gender of the regional delegates across terms. This suggestion is
based on the observation of the Hannaford Review of ATSIC that the multiple layers of delegation between
regions, zones and the ATSIC board were causing problems for local representation (Hannaford et al., 2003).
Distribution of members of the National Voice between jurisdictions
Even under the proposed ‘core model one’, regions may find it difficult to be heard in the National Voice,
particularly in NSW and Qld where 60% of First Nations people live today. It is currently proposed that each
jurisdiction (by which we refer to Australian states, territories and the Torres Strait) provide two delegates to the
National Voice (with a variation suggesting one member each for the ACT and Torres Strait). By providing each
jurisdiction with equal or close to equal numbers of representatives, many regions are likely to feel
unrepresented or under-represented.
The under-representation of regions is substantially worse than in previous models of national representation.
For instance, under ATSIC between 1993 and 2005, 17 ATSIC Commissioners were elected by councillors in
35 regions. However, ATSIC zones and their attendant commissioners were not evenly split between
jurisdictions, and represented a compromise between competing principles of reducing malapportionment,
jurisdiction representation, and reducing zone size. This system worked reasonably well to ensure that each
commissioner represented a similar number of people, except in NSW where Indigenous people were
significantly under-represented (see Table 1). The proposed system would exaggerate that historical under-
representation, with NSW and Qld. becoming further under-represented and Vic., SA, Tas. and the ACT joining
the ranks of the over-represented.
This malapportionment can be justified on the basis that states and territories are responsible for the bulk of the
public policies affecting First Nations people, and each state and territory has distinct laws and histories, thus
necessitating separate representation. Such a justification underlies the make-up of the Australian Senate, with
the original six states each represented by 12 senators, regardless of population size. Such a body might
function quite well, albeit overrepresenting the views and interests of the relatively small Indigenous populations
of Vic, SA, Tas, ACT and the Torres Strait.
However, co-design group was clearly troubled by the inequalities in the number of representatives in each
jurisdiction, insofar as it has considered the option of providing the two least populous jurisdictions of the Torres
Strait and ACT with only one member. This principle could be extended, devising a fairer system that retained a
total of 18 members, shown in the third column of Table 1.1 This would see five members drawn from NSW, four
from Qld, two each from WA and the NT, and one from the remaining five jurisdictions. Such a system would
ensure that each jurisdiction retained at least one member in the National Voice, while reducing some of the
imbalances in representation introduced in the model proposed in the Interim Report.
Table 1: A comparison of the jurisdictional distribution of ATSIC Commissioners and proposed
National Voice members
ATSIC Commissioners Proposed distribution of National More equal distribution of
(1993-2005) Voice members National Voice members
Jurisdiction n Per cent population n Per cent population n Per cent population
represented represented represented
NSW 3* 10.1%* 2 16.6% 5 6.7%
Vic 1 6.1% 2 3.6% 1 7.2%
Qld ex. Torres Strait 4 6.5% 2 13.4% 4 6.7%
SA 1 5.6% 2 2.6% 1 5.3%
WA 4 3.6% 2 6.3% 2 6.3%
Tas 1 3.8% 2 1.8% 1 3.6%
NT 2 6.2% 2 4.7% 2 4.7%
ACT 2 (or 1) 0.5% (or 0.9%) 1 0.9%
Torres Strait 1 1.5% 2 (or 1) 0.5% (or 0.9%) 1 0.9%
Total 17 5.9% 18 (or 16) 5.6% (or 6.3%) 18 5.6%
Note: The ‘per cent population represented’ figure refers to the average percentage of the Indigenous population of Australia that a
commissioner or member from each jurisdiction represent. * The ACT is included in the NSW calculations for ATSIC as it formed part of the
Queanbeyan Regional Council from 1993. ATSIC population totals are calculated from the ABS’s 2001 Estimated Residential Population
figures, while the National Voice population totals use 2016 Estimated Residential Population figures.
This apportionment is made on the basis of ‘Dean’s method’, which was recently adopted to apportion Senate seats to the ACT and NT in
the Electoral Amendment (Territory Representation) Act 2020.
Coombs, H. C. (1984). The role of the National Aboriginal Conference: Report to Hon. Clyde Holding Minister
for Aboriginal Affairs. Department of Aboriginal Affairs.
Deem, J., Brown, A., & Bird, S. (2021, April 9). Most Australians support First Nations Voice to parliament:
Survey. The Conversation. http://theconversation.com/most-australians-support-first-nations-voice…-
Hannaford, J., Huggins, J., & Collins, B. (2003). In the hands of the regions: A new ATSIC. Department of
Immigration, Multicultural and Indigenous Affairs.
Hiatt, L., Luther, M., O’Donoghue, L., & Stanley, J. (1976). National Aboriginal Consultative Committee: Report
of Committee of Inquiry, November, 1976.
Langton, M., Calma, T., Odegaard, D., Griggs, R., Buckskin, P., & Hope, L. (2020). Indigenous Voice Co-design
Process—Interim Report to the Australian Government (p. 239).
Markham, F., & Sanders, W. (2020). Support for a constitutionally enshrined First Nations Voice to Parliament:
Evidence from opinion research since 2017 (Working Paper No. 138/2020). Centre for Aboriginal
Economic Policy Research, ANU. https://doi.org/10.25911/5fb398ee9c47d
Reconciliation Australia. (2019). 2018 Australian Reconciliation Barometer. Reconciliation Australia.
Reconciliation Australia. (2020). 2020 Australian Reconciliation Barometer. Reconciliation Australia.
Referendum Council. (2017). Final Report of the Referendum Council. Referendum Council.
Rowse, T. (2000). Obliged to be difficult: Nugget Coombs’ legacy in indigenous affairs. Cambridge University
Turnbull, M., Brandis, G., & Scullion, N. (2017). Response to Referendum Council’s report on Constitutional
Recognition [Joint Media Release]. Prime Minister’s Office.
Twomey, A. (2015). An Indigenous advisory body: Addressing the concerns about justiciability and
Parliamentary Sovereignty. Indigenous Law Bulletin, 8(19), 6–9.