[text alternative - opens in a new window]
In Victoria there are a range of mechanisms and points of access which provide for the public to express their points of view and have their opinions considered by legislators and decision makers in Government. While municipal and Parliamentary elections represent one area in which our society expects active participation by all citizens, governments have become increasingly interested in utilising the information and opinions of members of the public between elections as a guide to action and to inform policy development.
This extended interest in community consultation reflects the recognition by government that the introduction and implementation of successful policy initiatives can be assisted when the policy making is recognised as a process of co-production between the State and its citizens (Kramer, 1999:90).
In order to examine how electronic democracy might be employed to increase the involvement and to facilitate the participation of the public in policy making, this Part of the Report focuses on three different areas of political participation, from the highly regulated and formalised, to informal, community based activities that make for a vibrant and resilient community.
These three areas of participation are:
The role of ICTs in the electoral process for Parliamentary and Local Government elections
Means by which the Parliament and Government of Victoria can use ICTs to facilitate formal and informal public consultation processes that have been traditionally undertaken via mail or face-two-face
The role of government in developing the capacity for local communities to act together to resolve local problems or effectively present their concerns and interests to government.
At the heart of the democratic system in Victoria is the election of our representatives who take their position in Parliament and on local councils. Voting is an administrative activity in which care must be taken in the implementation and management of voting methodologies, and it represents a political process of competition between parties and candidates for the good opinion of citizens.
For a large number of Victorians, voting and the electoral process represents the major focus of their political participation.
As the act of selecting individual MPs, and through this process the selection of the Executive of the State, our voting system must be as transparent and reliable as possible. The effective administration of elections provides the duly elected Government with two necessary resources:
The legitimacy to act in the public interest
The legal authority to use public resources and institutions for these purposes.
In addition, some governments use this process to claim particular "mandates" for the implementation of policies discussed during the election period, or as a tacit endorsement of the program of an administration returned to office. While these claims are often somewhat tentative - as Bennett (1996:183-7) observes, the interpretation of electoral outcomes and their political meanings are highly subjective - the electoral process has significant implications for the policy direction of the public sector.
Because of these important determinants, Victoria has long maintained a commitment to the view that the franchise should be offered as broadly as possible, and that the process of voting (or at least maintaining oneself on the electoral roll and participation in the process of voting in some form) should be a minimum responsibility of all adult citizens of the State.
Due to the weight attached to the electoral process in our system of Government, and the use of electoral outcomes to claim significant political legitimacy for the decisions of Government, the process of administering and overseeing this process is explicitly regulated by legislation (the Electoral Act 2002 and the Local Government Act 1989) and an independent Electoral Commissioner, supported by their own bureaucratic organisation.
What is clear is that the electoral process is an administrative and legal process, that:
Makes heavy use of data and information about members of the electorate, candidates, and parties
Involves the communication and aggregation of views (vote casting and vote counting)
Requires an effective communications environment for the citizen participation to be effective: the development of informed participation through communications between candidates and citizens, between citizens and citizens, and between civil society actors (the media, public interest groups, community organisations) and the electorate.
Thus, the Committee notes that in each of these areas ICTs have, or may have, a significant role to play in this process. The Committee notes, however, that issues of electronic and Internet-based voting - which are often considered to be the only major issues of concern in debates regarding electronic democracy - represent one aspect of the overall question of the role of technology in the electoral process, and that any discussion of the use of ICTs must be clearly grounded against two questions (Victorian Electoral Commissioner, Submission No. a27):
What objective does the use of ICTs in this area serve?
Do the advantages justify any costs or risks associated with this change?
The protection of the legitimacy and integrity of the electoral system as the foundation of the rule of law in Victoria must be the central focus of any investigation into potential changes to the system of electoral administration.
Most public submissions received by the Committee expressed overall satisfaction with the electoral process in Victoria and warned the Committee against unnecessary or frivolous changes that would undermine the qualities of the system as it stands. As Stuart Hall stated on the Inquiry discussion list (Post, 29/11/04):
The Committee notes that the use of ICTs in electoral administration need not take an "all-or-nothing approach", and that, if deficiencies in the system exist, the appropriate response would be to introduce changes proportionate to the problems identified. In order to achieve this objective, the Committee has given special consideration to:
Rates of participation
The quality of votes cast
Any issues of systematic fraud.
Given the focus of the Committee on electoral processes as means by which ICTs may be relevant to improving the electoral process in Victoria, the Committee sought to determine specific areas of poor performance, or where existing performance could be improved incrementally (a "continuous improvement" approach).
Voting and maintaining oneself on the Electoral Roll in Victoria are compulsory activities for citizens in Victoria. The rate at which Victorians attend the voting booth is an indicator of the success of the Victorian Electoral Commission (VEC) in administering the compulsory registration and participation requirements of the Electoral Act 2002 and is indicative of the health of our democratic system overall.
As indicated in Table 17 (page over) Victoria has a robust participation rate, which is comparative to the national turnout and significantly higher than nations that do not make voting a compulsory requirement for their citizens. The Committee observes, however, that - for a variety of reasons - nearly seven percent of the Victorian electorate fails to undertake its civic duty to vote for Parliamentary elections and correspondingly fail to have their voice heard in the formal political process.
Period |
Commonwealth |
Victoria Parliamentary |
Victoria, Local Government |
|---|---|---|---|
2001-2003 |
94.9% |
93.2% |
75.2% |
1998-2000 |
95.2% |
93.2% |
73.5% |
1996-1997 |
95.8% |
94.1% |
72.4% |
Sources: VEC, 2004:42; VEC, 2003b:45; ABS, 2004d
The Electoral Commissioner provided evidence to the Committee of the work being undertaken by the VEC to ensure that the Electoral Roll is as complete as possible; a key mechanism in ensuring that (a) voters have the opportunity to cast their ballot and (b) that non-compliance can be investigated and corrective action taken (e.g. the issuing of fines or recognition of exceptional circumstances). This central administrative process includes direct efforts by the VEC (handling application forms, advertising and promotion, etc.) and an increasing number of partnerships with other organisations that can inform it of movements within the Victorian community (Minutes of Evidence, 18/02/05, p. 27) and serve as "triggers" for issuing a change of enrolment form (utilities, educational institutions, the tenants' bond-holding agency, etc.).
Overall, the Committee acknowledges the significant effort of the VEC in looking at ways to triangulate across a wide array of data sources in order to identify potential voters and target them with information and the appropriate compliance forms. This approach, in general terms, appears to be appropriate, and the Committee accepts and shares the view of the Commissioner that:
Clearly this is an area in which continuous vigilance and creativity is required, and the Committee accepts the real limitations of the VEC in locating those who actively resist enrolment.
One area of concern, however, is in the participation rate of people with disabilities.
In recent research conducted by Scope Victoria (2004), the organisation identified that ninety-one percent of people with disabilities voted in the most recent national election, with half of those who failed to vote stating that they were not listed on the Electoral Roll. As disability covers a wide range of impairments, it is not surprising that most people covered by this research (seventy-one percent) voted at a polling booth, but some difficulties with polling places were encountered by one-third of the survey group, which may be a disincentive for participation.
Specific difficulties included:
getting to and accessing the polling booth
no disabled parking
pain in walking or standing for too long
crowds and standing in line
no shade
no ramps, difficulty in negotiating steps
no disabled signs
claustrophobic or room atmospherics problems (lighting). (Scope, 2004:2)
More directly, Mr T. Clark, Manager of Vision Australia, provided evidence to the Committee of disenfranchisement among the blind, particularly where postal voting is the focus for election:
Clearly, the number of people who were unable to cast a vote due to impairment will represent a subset of the 6.6 percent of Australians who are classified as being either severely or profoundly disabled (as per Figure 5 in Part I). However, there does appear to be a discrepancy between the ninety-one percent of disabled people who participate (Scope, 2004), and the general participation rate of 93.2 percent in Victorian State elections. While this percentage is not large overall, the Committee has to consider this in a different manner to the situation of those Victorians whom the Commissioner has identified as wilfully evading the voting process. Any improvements in this area, therefore, would be of benefit to the community overall.
The second area of interest to the Committee was the degree to which non-English speakers and people with limited literacy skills are serviced by our current electoral system. With twenty percent of Victorians speaking a language other than English (LOTE) at home and a changing pattern of language migration to Australia, the State needs to be careful that new immigrants, particularly those with languages not traditionally strongly represented in Victorian society, are being included in our democratic practices and have opportunities to access local political news in their preferred languages.
As indicated in Table 18 Italian and Greek are still strongly spoken in Victoria - recognising the retention of the language into second- and third-generation migrants, but that languages from Asia, the Middle-East, Africa and southern Europe are now increasingly prevalent in our community and represent areas of growth for Victoria in the coming decades.
Source: Victorian Office of Multicultural Affairs, 2003:19
While Victoria attracts the second-largest number of immigrants of the Australian states, the Committee notes that the figure of twenty percent of Victorian LOTE speakers is somewhat deceptive. High levels of English proficiency are found in many of these communities, with the corresponding ability to move between languages depending on need presenting limited barriers to political participation.
Overall, the Committee notes that less than four percent of the Victorian community would consider themselves as having low levels of proficiency in English[16]. Thus, the focus of attention for electoral participation needs to be on some of the most recent arrivals of people from non-English speaking backgrounds. As with any population change, this presents a challenge to Government, as there are often limited resources in the Public Sector for development of translation and interpretation skills[17], by the time second- and third-generation immigrants provide ready access to these skills, the need has significantly abated (with the exception of ageing members of the first generation).
This clearly represents a challenge in the administration of the electoral process, where voter information is often provided largely in English. The traditional approach by electoral authorities to address linguistic issues has been through collaboration with representative groups in these communities, and to support the election with a general LOTE advertising strategy using select media to increase awareness of the electoral process and administrative procedures for enrolling and casting a vote.
On the other hand, the rise of Internet communications and the presence of a vibrant commercial and community media sector does support the political socialisation and informational needs of the diverse Victorian community. The Committee acknowledges that there are a number of broadcasters in this area, including SBS television and local radio, and community broadcasters who provide multilingual services (see Table 19), largely in metropolitan areas where immigrants tend to first settle.
Source: National Ethnic and Multicultural Broadcasters Council, 2005
In addition, the Committee notes that a vibrant and expanding non-English and multicultural press exists (see Table 20) which provides local and international news for Victorians. Many of these services also have some form of online presence, with information provided in English, the specific language of the target audience, or both. The Committee notes the trend towards greater bilingualism in Victoria, rather than a decline in this linguistic diversity.
Source: Margaret Gee's Australian Media Guide, 2005
Informal voting is the casting of a vote by an elector that is not counted. The most common form of informal voting is the provision of a ballot paper that is completed in such a manner as to require the VEC to exclude it from the counting process.
While the prevalence of informal voting is relatively low in Australia (see Table 21, below), the preferential system employed in Victoria is more complex than the first-past-the-post voting systems employed in many other countries, making the voting process somewhat complicated and time consuming. Since the Legislative Council will move to a multimember electorate voting system in 2006, an additional degree of complexity will be introduced into the Victorian system, possibly leading to an increase in informal voting in the short to medium term.
Sources: VEC, 2004:43; VEC, 2003b:46; ABS, 2004d
Even low, informal voting is of concern for two reasons:
Where the informal vote is accidental, the voter has lost his or her opportunity to express a preference (the vote is only counted as "informal")
Where large numbers of informal votes are accidental and the electoral race is close, the informal votes may have altered the outcome of the election, and hypothetically, the composition of Government.
As a general principle, the Government of Victoria and the VEC attempt, where possible, to educate the electorate in the appropriate voting practice, to ensure that the rate of informal voting is minimised. This is undertaken through "civics" education in the Victorian school system and through supporting advertising in and around the voting period. However, it has been noted that in a system in which voting is compulsory, a proportion of voters will deliberately choose to cast an informal vote, either as a general protest against the compulsion of attendance, or because of dissatisfaction with the range of options provided (thus, in areas where optional preferential voting is introduced, the rate of informal voting has declined).
In the table, page over, the rate of occurrence of errors invalidating Victorian ballot papers in the 2001 Federal election (based on research from the Australian Electoral Commission (AEC): Medew, 2003). While it is difficult to assign motive or reason to the various types of error, it would appear that a reasonable proportion of the 3.8 percent of the votes discounted as informal (possibly 50-60 percent, or 1.9 percent of the total vote) may have been the result of human error. Given that at most elections a handful of seats depend on a very small number of votes, the loss of even a percentage point of support to one candidate over another can have significant electoral outcomes.
Source: Medew, 2003
The AEC's research does appear to indicate discernable relationships between informal voting and a number of factors, including:
The complexity of the ballot (the number of candidates), its layout and design
The experience of the elector in voting (normally a function of age)
Language skills of the voter. (Medew, 2003)
This last factor appears to be particularly important, especially where there are large clusters of newly arrived immigrants in a particular electorate. The AEC explains the relationship between limited language skills and informal voting as both technical and cultural:
Secondly many of the voters who are not proficient in the English language arrive from countries where the voting system is far different. It must be remembered few countries practice the alternative vote system. Many use a system where a one or a tick or a cross is all that is necessary to cast a formal vote. There may also be a number of voters who do not comprehend the system, and are afraid of making a mistake and simply return a blank ballot paper. (Medew, 2003)
Thus, the Committee notes that any activity that would reduce the overall informal voting rate, and particularly informal voting in the non-English speaking community, would be a positive advance on the current level of performance in this area.
One strong motivator for considerable change to the Victorian electoral system might be a critical failure of the system, either due to maladministration or systematic fraud (such as mass enrolment of the dead or bogus electors, ballot-box stuffing, miscounting and/or multiple counting, vote purchasing, bribery or coercion of electors or electoral administrators, etc.).
During the course of the Inquiry, the Committee received no credible evidence of systematic electoral fraud being undertaken in the State of Victoria. Given the existing system of voting, the simplest means of electoral fraud would be the casting of more than one vote, either through visiting a number of voting locations or physically voting and casting a postal ballot. While this does occur in Victoria, it appears that the practice is not widespread and is rarely motivated by criminal intent. In the view of Mr S. Tully, Electoral Commissioner:
Overall, therefore, the Committee notes that there are a number of issues in and around the process of voting in Victoria that do require improvement, particularly:
Improvements to the physical environment of voting centres, to ensure maximum venue accessibility
Improving access to the electoral process for people whose written English skills are limited
Means to minimise the casting of accidental informal votes
Maximising the ability of people with vision impairments or mobility problems to cast a valid vote securely and with the right of privacy that serves as protection from coercion or repercussion.
However, the Committee is of the overall opinion that, in considering the processes surrounding the voting processes in Victoria, no critical failure of the system exists that should motivate extensive modification to the current system of casting votes in the State. The Committee considers that a number of improvements could be made to the electoral process via the use of ICTs, and that the VEC's current plans to improve service delivery at physical voting centres is an appropriate response to some of the issues highlighted above.
It should be noted in consideration of the issue of electoral matters that the Committee has resolved to focus on the role of ICTs in facilitating, improving or supporting the electoral process in Victoria. The Committee has deferred making any specific recommendations beyond questions of technological application in this area.
Thus, while there are always many suggestions about means to improve or reshape the electoral system in general terms, the Committee does not make any consideration of, or recommendations into:
Questions of electoral participation (such as extending participation in Parliamentary elections to resident non-citizens, lowering the voting age, non-compulsory voting or enrolment, or the creation of expatriate or "non-territorial electorates"[18])
Electoral systems (such as the introduction of multimember electorates for the Legislative Assembly, optional preferential voting, or first-past-the-post systems)
Boundary changes and similar matters.
It is the opinion of the Committee that these questions are general concerns that are beyond the scope of this Inquiry. The Committee's view is that issues of electoral reform should be undertaken to the greatest possible extent within a technologically neutral context, in the first instance, with consideration of technology as an enabler for the general principles thus determined.
Prior to the detailed discussion of the use of ICTs in specific parts of the Victorian electoral process, the Committee has determined that the conduct of legally binding electoral transactions over telecommunications networks - such as the Internet, landline or wireless telephone service - should be rejected in the immediate term.
It is the finding of the Committee that, on the balance of probabilities, the advantages to be gained are mitigated by the current and potential risks associated with security and the excessive cost to the State that risk mitigation would entail. In addition, the Committee considers that the wider community has expressed limited interest in this form of voting, and that the approach may undermine public confidence in the integrity and transparency of the electoral system in Victoria.
The arguments that justify this finding are discussed in this section, in some detail, and the specific implications for electoral activities are discussed in the respective discussions below.
The specific implications of this finding are that:
The Committee does not support Internet-based voting systems or voting over telephone services except in a very small number of exceptional circumstances (current use of these systems for Antarctic voters).
The Committee does not support the transmission of voting data (counts) over the Internet for remote or centralised tabulation.
The Committee does not support the complete enrolment of citizens online at this time.
It is the Committee's view that each of these activities would provide benefits to the electoral process in Victoria and should be considered again when environmental conditions improve.
The basic design of the Internet, as a network of nodes and subnets, provides considerable advantages over other networking technologies and has - in part - driven the massive popularity of this technology across all facets of private and commercial endeavour. The core characteristics of the Internet that have facilitated this growth are:
An open network, with users, nodes and subnets able to participate easily - the system is not "gated"; it can grow and expand rapidly as users and nodes freely join the network
Interoperability of computers using different operating systems and hardware configurations makes participation open to a range of hardware providers, making the system open to the widest possible extent ("agnostic")
The system of packet switching is a robust, fast and reliable means of dealing with communication across a large network in which individual nodes or telecommunications linkages may be of poor quality.
These design fundamentals, however, have also created a system that suffers from considerable security limitations. These limitations are inherent to the design of the fundamental architecture of the Internet, but is exacerbated by the broad range of information literacy exhibited by users and the increasing value of goods and services communicated over the network. Thus:
The communications system, largely based on trust, provides opportunities by which the source of information can be concealed or forged
The popularity of software connected to the Internet is often not a function of its level of security and network integrity, but tends to be driven by other factors
Individual behaviour online can maximise and minimise personal, organisational and network security; the wide range of skill levels of users creates opportunities for miscreants to manipulate and deceive users
The Internet, as an increasingly important part of the global economic system, has attracted criminals proportionate to the opportunity afforded by its inherent security difficulties.
Overall, the Committee notes that the strength (and weakness) of ICT-enabled transactional systems is determined by the least-secure point in the chain of communication. Unfortunately, this tends to be the end-user environment. As Dr R. Clarke of the ANU remarked in his written comments to the Committee:
Thus, the Committee notes that any consideration of the role for transactional activities of high value or importance - and voting is a critical transactional behaviour for the democratic process - must consider the Internet and similar open-networked systems as basically insecure environments.
In making this observation, the Committee notes that it is possible to over-react to these concerns and perceive the threatened environment as one driven by predatory professional criminals. While the best statistical evidence available to the Committee notes that a proportion of computer crime reported in Australia appears to be motivated by "professional" criminal purposes (industrial espionage, fraud), the greater majority comprises attacks motivated by personal vanity (e.g. website defacement), a desire to access the system for personal use (e.g. theft of system resources) or as a target of opportunity (AusCert, 2004:15).
The Committee, therefore, concedes that it us unable to provide a meaningful probability level of risk.
From these general issues, the Committee notes a growing body of security related literature specifically developed on the subject of online electoral systems. The Committee notes three areas of activity:
The work undertaken by the United States Military up until 2004
The Local Government mixed-voting pilots in England in 2002-03
Some work observed in Canada during the delegation to North America.
The first example comes from work undertaken by the United States Military, through the Federal Voter Assistance Program's[19] establishment of the Secure Electronic Registration and Voting Experiment (SERVE).
The SERVE project:
as initiated to provide Internet-based voting in legally binding elections for the United States of America, recognising that that country has six million potential voters who are:
Overseas military personnel who have limited access to mail services
A large group of civilian expatriate
Focused on two critical electoral transactions: voter registration and the casting of votes online
The system was developed by the Pentagon, in conjunction with private providers (e.g. VeriSign as the provider of encryption technology)
Its technical approach to security was the use of 128-bit Secure Sockets Layer (SSL) protocol to employ Public Key Infrastructure (PKI) encryption to guarantee the security and privacy of the interaction between the SERVE servers and the client computers
The development of the system included management processes for internal review of the software developed, and the use of a Security Peer Review Group (SPRG) to oversee the project externally.
Thus, this must be seen to represent one of the most significant attempts by government to develop a wide-scale, Internet-based voting system (multiple ballots and voting and registration rules needed to be accommodated in the system). SERVE was a "live" experiment: it was intended to be active in the 2004 United States Presidential election, collecting legally binding votes. However, due to a well-publicised pessimistic assessment of the security of the system that led to a large amount of negative media coverage, the use of the system was cancelled (Wagner, 2004) for fear of damaging the integrity of the electoral process and the legitimacy of its outcome. The Committee notes that the system may not have been shelved had the political environment in the United States regarding the integrity of the electoral process been different (see The Political Lesson of the United States' Experience); however, following the first election of President George W. Bush in 2000, parts of the United States public were highly sensitive to questions of electoral integrity.
The negative reporting on this system is based on a minority report, released onto the Internet by four members of the SPRG[20] . This report, A Security Analysis of the Secure Electronic Registration and Voting Experiment (SERVE), makes a number of specific criticisms of the SERVE Project:
That the closed nature of software development (proprietary) affords insufficient avenues for inspection and certification of the quality of the software used
Because of this basic design, the system is vulnerable to "insider" attacks by programmers or other staff associated with the project
That the system does not provide the voter with a means (audit trail) to ensure that their vote has been successfully and accurately recorded, and thus the voter has no effective guarantee that their vote was recorded accurately. (Jefferson, Rubin, Simons and Wagner, 2004)
In addition to these specific concerns, the report also provides general criticisms of the overall environment in which Internet-voting systems would be expected to be run. These concerns are more fundamental to the overarching critique of the concept of Internet-voting offered by the analysis. The criticisms are:
The use of the Internet as the delivery mechanism for vote casting opens the system to any number of "general" Internet risks, such as:
Denial of Service (DoS) attacks against the voting system, effectively blocking access to the "virtual polling booth"
Spoofing frauds to lead voters to false voting sites where their vote may be either captured (and discarded - "selective voter disenfranchisement") or their identification details reused (vote theft)
Automated vote buying and selling, such as via online auction systems
Risks that a virus, worm or other malicious active content could be designed to disrupt the voting session, alter votes or record the voter's action
That these attacks could be launched on a large scale, from a foreign legal jurisdiction, by a lone individual or an organised group (such as a foreign government)
That the capacity to estimate the actual probability of the risk is unknown and that, even in the event of a "successful" election, fraud or incidents of attacks against the system may never be identified.
This assessment essentially reiterates the work of the United States National Workshop on Internet Voting, which concluded that:
The Committee notes that some of the risks associated with Internet-based voting systems appear - at this point in time - extraordinarily remote. As Michael Shamos (2004) of Carnegie Mellon University's School of Computer Science observes, excessive perception of risk can border on paranoia:
It is the view of the Committee that attacks against an Internet-voting system are most likely to come from disaffected or misguided individuals and small groups, and that the sophistication of the attacks would likely be low. As Alvarez and Hall (2004) observe, the difficulty of the online environment is often associated with the proliferation of "user friendly" hacking tools that permit:
It is the opinion of the Committee that "dumb" attacks such as: DoS or Distributed Denial of Service (DDoS), phishing emails and website spoofing would be much more likely outcomes than sophisticated worms or "man-in-the-middle" attacks that could substitute voter preferences mid-session without detection.
In discussion with Mr R. Kidd, General Manager Elections of SecureVote, Mr Kidd illustrated to the Committee that while spyware such as keyloggers may be problematic to the security or privacy of some voting systems, the use of mouse movement as a primary means of recording preference is a way in which modern online voting systems evade these basic programs. Other developments include the use of images to replace machine-readable textual elements (such as candidate names), which could be used by a worm to identify a voting field [21].
Mr C. Burton from Everyone Counts (E1C), a provider which uses a separate voting client (not a webpage) to deliver its voting system, also noted countermeasures to many of the attacks highlighted in these reports:
Thus, the Committee observes that the most obvious sources of security risk for the development of Internet-voting systems is focused at:
The design and implementation phase (where appropriate management practices can mitigate or eliminate the prospects of programmers tampering with code)
At the server or network level, through DoS or Domain Name Service (DNS) poisoning
At the desktop level, where the capacity of the individual citizen to secure their own operating environment may be limited. This could be due to a range of factors:
The system is not under their control (it is a corporate network or a public terminal at a library or Internet cafe)
The user has limited knowledge or IT skills
The system is too old to be patched or upgraded against attack
The user has limited financial resources to purchase software to clean and protect their desktop
The operating system (OS).
Even though the Committee does not regard the risk as remotely plausible of a significant attack whereby votes are altered by secretly distributed worms prior to an election[22], the higher likelihood of human factor and "crude assault" risks are just as damaging to the electoral process. While a DoS or DDoS would appear to be the most likely form of assault on the system and could be countered (technically, or though the extension of the voting period), human-factor problems would be much harder to detect (particularly the scenario in which a phishing attack channels voters to a false website).
While these criticisms are strongly worded, the Committee accepted evidence from Mr C. Burton of E1C, who specifically responded to the SERVE report and advised that the Committee should act with caution in adopting it uncritically. Mr Burton stated:
The Committee notes that similar concerns have been raised regarding the earlier critical assessments of Internet-voting concepts by Alvarez and Hall (2004) in their book Point, Click, and Vote: The Future of Internet Voting (2004). The authors note that a number of critical reviews of Internet-based voting systems have been limited in the extensiveness of their studies, that criticisms are regularly generated by a small group of active authors on the subject and that a "conventional bias" against Internet-voting systems has emerged in the United States that is under-supported by tangible evidence.
The basic criticism of proponents of Internet-voting, such as Alvarez and Hall, is that experimentation is needed to determine the likelihood of risks upon which informed risk assessments can be undertaken. Evidence, in this case, is the establishment of operational systems that are run in field conditions and should be subject to "white hat" hacking attempts. Thus, the SERVE minority report remains a hypothetical criticism of a largely untried system.
In considering ICT-enabled voting systems that have been run in real-world, legally binding elections, the Committee has turned to the evidence provided from a number of systematic and ad hoc experiments undertaken in Canada and the United Kingdom. It should be noted that, in each of these cases, failures of security have not been manifest, and thus the experiential evidence to date is far more positive than the analytical assessments of the United States.
In the United Kingdom, the Electoral Commission undertook a wide experiment in early 2003 with a range of new voting methods. The methods employed were used at the local government level as deliberate experiments to ensure that comparative performance and evaluation data were captured. The methods employed included:
All-postal voting
Telephone voting
SMS voting
Electronic voting machines ("kiosk" voting)
Internet voting.
The scope[23] of the experiment was highly impressive:
In order to undertake this process, that government used external technology providers, focusing the management of the process on the development and specification of standards and requirements, and measuring performance against these targets.
The United Kingdom's Electoral Commission noted from this experience that the process of developing these specifications needs considerable experiential learning. It found some areas of the specification setting too detailed and others insufficiently so. These are not problems unique to the electronic voting arena, but reflect general observations of the need to refine and improve contract management and oversight processes when government agencies move from direct provision models to new public management approaches in acquiring "outcomes" rather than systems.
Overall, the view of the United Kingdom's Electoral Commission was that progress had been made through the piloting process, and that the Commission was comfortable with the level of security afforded. In making its final recommendation to government, the Commission considered that a wide range of technical and administrative improvements to the system would be needed, but that - through this specificity - the Commission regarded remote voting systems as effective and manageable (Electoral Commission, 2003:7-8).
In comparing the various methods employed, the Commission was of the opinion that:
Kiosk voting systems were generally preferable to Internet-based voting, due to greater capacity to control the operating environment of the voting system and its user
The design of the PKI approach to vote encryption on the Internet servers was insufficient to safeguard the systems against malicious attacks. However, the Commission did not identify any attacks against the system during the pilot process
Insufficient work was undertaken with regard to alternative systems and standby data centres, and in the event of the failure of a primary site (either through attack or "natural" event, such as power failure) significant voting data could have been lost
Improvements to systems monitoring would be required to identify DoS attacks rapidly in future implementations.
Overall, however, the United Kingdom's Electoral Commission considered the processes of implementation to have been successful, and that the range of suppliers engaged had generally tended to meet the security and management specifications laid down in advance of the process. However, the Committee was cautious about the ability to generalise its findings, for three reasons.
First, the use of a range of approaches allowed the Commission to take some comfort in its security through obscurity. As the Commission stated in its final report:
This approach is not highly regarded by the Committee. As electronic voting processes become more common, and members of the public (electors and non-electors) become familiar with the design and operations of these systems, this protection will quickly vanish. In addition, given that prestige is attached to particularly difficult or bold intrusion attempts by some parts of the hacker community, "cracking" an election system would appear to have greater psychological reward than yet another corporate mail server.
Second, the emphasis of the United Kingdom's approach was - largely - motivated by issues of declining participation. This is not a particular problem in Victoria under compulsory voting, but appears to alter the level of acceptable risk in the United Kingdom, where there is a stronger focus on balancing participation rates and risk. This is clearly different from Victoria, where the emphasis is on improvements to a system that already has high levels of participation (with exceptions, see below).
In the context of these pilots, where participation rates in elections were low, any form of participation was considered to be valuable. However, in the Victorian environment, there is a risk that this approach might have negative impacts on the perception of the value or significance of the process of voting. The University of Melbourne's Centre highlighted this issue for Comparative Constitutional Studies:
Third, the Electoral Commission (2003) of the United Kingdom considers that remote voting for national elections is some time away - possibly six years - but considers ongoing local pilots a means by which to achieve the objectives of reliability, secrecy, security and accessibility. While the Committee notes that the implementation of remote voting in municipal Victoria may be an implementation model for Internet voting in this State, it does not consider it appropriate to maintain different levels of acceptable risk for different levels of government.
It is unlikely in the Victorian context that the public would accept an implementation model that explicitly excludes the practice of Internet-voting for Parliamentary elections due to security risks, but introduces it for Local Government elections. This view was supported by the Submission of the City of Darebin (Submission No. a12, p. 1), City of Manningham (Submission No. a15, p. 2) and the Victorian Local Governance Association in Hearings (Mr A. Davies, Minutes of Evidence, 17/02/05, p. 37).
The Committee observes that online security risks have not been declining, but have been tending to increase over time. This appears to be driven by a number of factors in addition to the issue of basic architectural problems of the Internet, as discussed above.
These factors are:
The development of application software on OS which have basic security flaws or are not maintained properly (patched) to protect against known exploits
Market competition which drives the development and release of software, particularly Internet-based software like browsers, email clients, messaging systems, file sharing and publishing platforms, with a rapidity that often comes at the expense of rigorous security testing
The interoperability of software and systems which may open new vulnerabilities
Consumer use of technologies that have known limitations, such as Wired Equivalent Privacy (WEP) security in wireless routers
The increasing size of the online population and the presence of valuable data and transactions online makes it attractive to criminals
The lower risks of detection and successful prosecution of online crime makes it attractive to criminal experimentation.
While there is evidence that the general levels of information literacy amongst the public has risen over the past decade, and issues of security and privacy are becoming more important to users, the complexity of ICTs places the lay user at a disadvantage in an environment of rapid change. AusCERT states that:
The AusCERT Report goes on to state that, with respect to the capacity for general Internet users to apply mechanisms to verify the identity or bone fides of online identity of people and organisations with whom they interact, the immediate prognosis is not good:
The Committee gave serious consideration to the experiences of the banking sector with regards to these issues, and notes that the sector has experienced ongoing problems with its online banking services and has begun discussing the need for additional levels of security beyond the traditional approach of providing user names (normally customer numbers) and secret passwords. At the time of writing, it appears that the Australian banking industry is actively considering the introduction of "two-factor" authentication approaches to address declining confidence in the traditional password approach (Finextra Research, 2005).
Two-factor authentication improves the security for users by introducing an additional security mechanism ("factor", or "token") to the process. This approach has been introduced by the Bendigo Bank which allows Internet banking clients to purchase one of two types of security token. These tokens work by generating a unique number (based on the device's internal clock) that is entered at the time of logging into the system. Since this number changes constantly, the loss of a particular Personal Identification Number (PIN) will not allow a third party to use the service at a later time. Other examples of this technology include hardware systems, such as USB "keys" with similar characteristics. The specific intention here is to reduce the possibility of intrusion by increasing the level of verification by the user.
However, the Committee observes that, outside of the direct cost of providing physical tokens (and accounting for their loss, theft or malfunction), security commentators have noted that there are limitations to this approach and that an increase in use of two-factor authentication will encourage countermeasures from criminals. As Bruce Schneier (2005) observes, these responses are useful in deterring "passive" attempts at password logging, but that in the online environment criminals will develop countermeasures. These are likely to be new attacks that focus on "real time" intrusion:
Given that some ICT-enabled electoral transactions are likely to be focused on short-term windows of participation (in particular, vote casting), the "real-time" attack is clearly the focus for concern.
With regard to the issue of security, it was the view of the Deputy Chief Information Officer, Dr S. Hodgkinson, that the costs of establishing secure voting systems would be prohibitive. With respect to comparative examples in which Internet security issues had been the focus, Dr Hodgkinson observed that:
He added that it was unlikely that the Government of Victoria would be willing to make this investment:
This point was underscored by Dr R. Clarke, who noted that the investment of Australia's banking sector in security has been far in excess of the levels of investment in Internet voting systems development by any government to date, and that this investment is balanced against acceptable levels of risk - the investment of the financial sector is a function of its willingness to date to accept a degree of failure (Meeting, 01/03/05). The point to be highlighted here is that, in the case of ICT-enabled electoral transactions, government must set a much lower level of acceptable risk, with corresponding higher levels of cost (California Internet Voting Taskforce, 2000).
The Committee notes that cost is a clear factor in the degree to which any system has generalised protection against attacks that may be motivated by curiosity or the identification of a target of opportunity by an attacker. Given that the cost of security tokens like those used by the Bendigo Bank ranges from $16.50 to $99.00 per unit, the total cost of providing a token like this alone (excluding the design and implementation of the system to support it) for general electoral purposes in Victoria would be excessive for wide-scale implementation.
In the research undertaken by AusCERT (2004), the organisational investment in security - particularly in the area of maintaining systems that have up-to-date software (patches) and in the general area of human resources - is a clear causal factor in the prevalence of attacks.
Source: AusCERT, 2004:29
While this finding would indicate that an appropriate investment in security would be a serious mitigating factor for the administrators of an electoral system with connection to the Internet or other open network, in terms of:
Ensuring that all security upgrades are implemented immediately upon their release
Having a clear security management protocol
Providing sufficient investment in staff
the Committee notes that this approach is somewhat problematic in an environment in which any change to the configuration of the system (the network, operating system or application software) would require re-certification to ensure transparency, security against vendor tampering and prevention of insider manipulation of the voting system.
Again, this indicates that the costs of security of ICT-enabled transactional systems will be higher than conventional, or even high-risk, online systems, due to the difficulty in maintaining a generally secure environment responsively with the need for any configurations to the system to be recertified at a high level of confidence. This issue was recently highlighted in the Report of the Irish Commission on Electronic Voting (2004), which - while assessing a hardware-based electronic voting system - concluded that serious problems in certification were presented by the rapidity of technological change.
The presence of existing, highly secure government networks that might be used to facilitate ICT-enabled voting around the world (such as the secure Virtual Private Network (VPN) maintained by the Department of Foreign Affairs and Trade[24]) could be considered one way to "piggy back" security upon another organisation or purpose. However, the Committee notes the impracticality of an organisation like the VEC effectively entering into a relationship with another government agency whereby the VEC could exercise an appropriate level of verification of the integrity of its system.
It is the Committee's view that, while the State of Victoria could invest the financial resources required to develop an independent online voting system that would provide extremely high levels of security, the cost of undertaking and, particularly, maintaining such a system in the event of eventual attack would be extraordinarily high. Given that the Committee is not able to envisage the complete replacement of paper-based voting in the immediate term, this cost would be added to the current cost of electoral administration.
As is clearly indicated in the discussion on the policy issues surrounding Internet-based voting platforms (see below), the Committee notes that the cost may provide the level of benefit that has motivated activities in other jurisdictions.
Given the concerns discussed above, the Committee considered the principle of informed consent of members of the public to undertake online transactions in an insecure environment. The Committee, therefore, gave consideration to the degree to which members of the public are aware of the risks of transacting with government or the private sector online (voting, making a legal declaration or application, making a purchase).
While research has been conducted in this area on information technology professionals and managers, business operators and the customers of some industry types (particularly in banking), the Committee notes a lack of quality, broad-scale data collection into perceptions of public risk undertaken by the Victorian or Federal governments. Thus, the Committee notes a number of mixed indicators in this area of public perception of risk and willingness to engage (and accept) in risky behaviour online.
The Federal Department of Communications, Information Technology and the Arts has reported that - as at 2002 - twenty-nine percent of Australian Internet users were cautious about engaging in online transactions because of security risks (National Office for the Information Economy (NOIE), 2003b:32), but that this has not prevented just under fifty percent of Australian Internet users engaging in online financial transactions of some kind, with steady growth over time (NOIE, 2004:19). As of 2005, the number of Internet banking transactions surpassed the number of paper cheques written in Australia (Woolrich, 2005).
Independent advocacy groups have also identified this type of mixed evidence. The Australian Consumers' Association has recently highlighted concerns about decreasing security for Australian Internet users, reporting survey results indicating that fifty-seven percent of Australians have received phishing emails, but that the majority of respondents express confidence in the security offered by institutions like their bank (Innis, 2005). While the Association reported strong use of anti-virus software by survey respondents, the take-up rate of security tokens by Australian online banking customers has been modest, with one in five customers purchasing the added layer of security offered by the Bendigo Bank (Needham, 2005).
Overall, the Committee notes that the technical aspects of Internet security are considerable, and that moves are underway at a number of levels to address the fundamental architectural failings of the Internet and its hardware. Some examples include:
Consideration of ways to develop Internet hardware that is self-defending and includes a degree of intelligence to prevent or limit security vulnerabilities (Waltner, 2005)
The replacement of existing protocols and Internet technologies (such as the development of Internet Protocol version 6 (IPv6) and the work of the Internet 2 consortia)
New attention given by some OS manufacturers to issues of security, possibly leading to better performance in this area
Recognition by application developers of the need for more reliable products, particularly in the development of systems that make a clear distinction between data (message content) and program (executable)
Considerable marketplace interest in developing anti-malware products, including a number of open-source initiatives.
However, overall, the Committee is of the view that the extent of risk in the online environment is not accurately matched by an appropriate assessment of risk by the general community. This view takes a number of dimensions:
The Committee considers that the perception of risk in the general community is lower than actual levels of risk and that this leads to an inadvertently high level of risk-taking behaviour in the online environment
The Committee considers that community education about appropriate security procedures are only now beginning to be systematically developed, and it is unclear how the level of public awareness will change in future years - given the technically complex nature of these issues, the Committee considers that there is only so much individual citizens will do to protect their systems against security risks. These actions are likely to be focused at common irritants (e.g. SPAM, pop-ups, etc.) rather than remote or infrequent risks (voting-system manipulation)
The Committee is concerned that the capacity for a general ICT user to maintain a comprehensive desktop security environment is both a function of technological literacy (which is highly variable throughout the community) and the capacity of the individual to pay for the range of protective software required, or to remove poorly secured "standard" software provided in favour of more secure technology. The Committee does not consider income and education levels as appropriate pre-requisites for the casting of a secure and private vote in the State of Victoria
There have been mixed messages sent to the public by industry and government in this area.
Some of the problems associated with differences between perception and risk were highlighted by the Committee's experience in the United States, where the issue of electronic-voting machine integrity "blind sided" the Californian regulators due to the intensity to which the revelation of problems attracted a strong negative response by the community (which had been using electronic- and mechanical-voting systems for decades) and seriously undermined perceptions of the integrity of the electoral process (Meeting with Mark Kyle, Undersecretary of State, California)[25]. In the Californian case example, this required quick remedial action by the Californian Government, at considerable cost to that state.
This perspective has also been identified in Australia. Mr P. Green, Electoral Commissioner for the Australian Capital Territory has observed that:
Thus, the Committee notes that consideration of the appropriate role of technology in the electoral process must consider both actual risk (be that human factor, or technological factor) and perceived risk. The Committee considers that the systems used in the electoral process must be clear, transparent and effective, but also that the public must share this assessment. The legitimacy of electoral outcomes in Victoria and thus, by extension the legislative process and rule of law, requires public perception of the integrity of the electoral system to match its actual security - neither more or less confident than the system warrants.
The Committee notes that lessons from experiences with electronic- and online-voting systems, while often focused on issues of technology, security and administrative practice, are also largely about the need for political commitment to experimentation with these systems.
Overall, the Committee observed that there is mixed evidence for the potential of the Pentagon voting experiment - as one example - to deliver effective, Internet-based voting. While the SERVE minority report (Jefferson et al., 2004), which is critical of the project, includes a very interesting and useful assessment of the flaws of the approach taken and the underpinning environment of the Internet, the Committee did receive a very detailed briefing from the project team about its responses to most of the criticisms of the approach, such as:
The strength of the 128-bit encryption employed
Positioning of the voting servers close to the DNS root in order to provide resistance against DoS attacks, combined with significant redundancy across geographically diverse sites in case of physical attack on the system
Separation between Internet servers and back-end computers
Inspection of all traffic
Clear separation of management functions between parties
Third-party testing and evaluation.
Given that SERVE was not used during the 2004 Presidential Election in the United States, the hypothetical risks highlighted by the minority SPRG members remain hypothetical. This has positive and negative outcomes with regard to our knowledge.
In addition, the Committee noted in discussions with Mr W. Doherty, the Executive Director of the Verified Voting Foundation, that it appears that the technical issues associated with electronic-voting machines were quickly exploited politically and became a partisan issue - possibly at significant cost to the capacity of the state to have a meaningful public discussion of the problems associated with these technologies. Thus, while the issue of technological failure was largely related to poor planning, contract oversight and organisational capacity to review contractors' work, the issue was exploited in a partisan manner, further eroding public confidence in the electoral system.
Given the range of non-political elections being undertaken online, and the possibility for greater security through some of the technological trends noted above, caution was needed by Government not to over-commit to a process of developing an electronic-voting system while basic architectural conditions are not favourable and where considerable lessons are being learned in less critical environments.
The Committee notes that this view supported by evidence taken by technical experts, but that only the timeframe of technological maturity remains in doubt. Thus while Dr Clarke (Meeting, 01/03/05) considered the possibility of Victorian Internet-based elections "within my lifetime", the submission of E1C observed that:
Internet voting systems have been under development since the 1970s
Legally binding political elections have been undertaken since 1996 (the Reform Party of the USA)
Fifteen percent of private elections employ the technology. (Submission No. a16, p. 4)
It was the view of the E1C organisation that the technology was "maturing", not emerging.
In addition, the Committee noted that one of the problems of SERVE was a lack of political commitment to the process, leading to a considerable fruitless investment of public resources into a project that, at best, may have been avoided by a comprehensive risk assessment prior to systems development. Had the SPRG minority report been released prior to major systems development, the significant cost of developing the system to the United States government might have been avoided.
This finding is not a criticism of the SERVE project team as much as it is a criticism of executive decision making in authorising such extensive development against a critical climate of public opinion. It is the view of the Committee that, while appropriate experimentation does appear to be the most logical way to proceed, including aggressive security testing (such as "cracking challenges"), these systems are most vulnerable to a lack of confidence in the technology by the public. The need to build political support for these practices will be as important as the need for security and systems integrity.
The quality of our democratic system and the governance its produces is based on high voter turnout, coupled with informed participation. While the State of Victoria has laws requiring compulsory voter registration and voting, ensuring that voters are "informed" is an elusive concept and one not readily addressed by government.
Presently, the Victorian Government assists in ensuring informed participation by:
Supporting civics education in the school curricula with collaboration between DET and VEC
General efforts to make public government information widely available (promotion of a rich information landscape)
The provision of rules ensuring access to information from political parties and candidates during elections, either through rules ensuring fair competition for the attention of voters at voting stations or requirements for candidates' statements to be distributed during all postal voting (for Local Government in Victoria).
The Committee recognises the inappropriateness, within a liberal democracy, of "enforcing" informed participation. Traditionally, it has been the responsibility of those seeking public office to make their case for election, and elections are considered to be competitive and open when candidates expend their efforts in persuading and informing the public via a range of media forms and in person.
Thus, while the Committee sees no requirement to change this traditional approach to political education and competition between rivals, the Inquiry gave consideration to areas of the electoral process in which access to information may be limited. These include:
Access to voter information by people with disabilities
Access to voter information by people whose primary language is not English
The capacity for candidates to distribute information online, and for this information to be systematically "discoverable" by voters and media organisations
Specific parts of the community which may need additional information to help them make an informed vote.
It is the general view of the Committee that appropriate steps taken by Government to expand the availability of information upon which the Victorian public can cast an informed vote during State and Local elections is a useful adjunct to the system of private competition between parties and candidates of widely differing skills and resources.
During the course of the Inquiry, and the work of the 54th Parliament, a number of projects were examined that aimed to increase the accessibility of information about individual candidates to the electorate prior to the casting of votes.
These activities included public-sector projects and the work of non-government organisations to fill what they perceived as gaps in the information landscape surrounding elections. This type of work tends to have some similarities with alternative news publishing online, with two exceptions:
The work tends not to be ongoing, but periodic - making the development and maintenance of readership and promotion of the material more difficult over time
The work feeds directly into electoral processes, which tend to be covered by a different set of electoral laws than conventional news media coverage.
In particular, the Committee examined the work of the League of Women Voters' Education Fund (California) and the California Voters' Foundation in the United States. The Committee of the 54th Parliament also examined the establishment of an electoral portal for the Swedish municipality of Ronneby.
In each of these examples, the projects aimed to provide information to voters about their relevant elections, using the WWW as a means to collect, organise and distribute this information. Each served as an interactive "voter handbook", and some of the models also included online debates between candidates, interaction between citizens and candidates, and additional information to support voting (voting location information, enrolment guides and the like).
In each case the project was motivated by perceived limitations in voter access to information about candidates, either in response to low turnout (such as in the United States) or the appropriateness of facilitating direct competition between candidates which have unequal resources (such as between large and small parties, and the involvement of independents).
All of the projects made an explicit decision to allow all candidates to participate, and the Ronneby example is candid about debate surrounding the decision to allow Sweden's far-right political party to participate in this service. In addition, the Committee notes that in Seattle in the United States, the cable access channel provided to the City Council is used to deliver a short video address from each candidate prior to municipal elections (through the "Seattle Channel").
In considering the application of this approach in the Victorian context, the Committee notes that:
The collection and distribution of Candidate Statements (150-word statements) from each candidate in all postal elections was an effort to compensate for the lack of a physical environment in which candidates could campaign (the polling station)
In previous years, the VEC has provided on the WWW copies of the Candidate Statements and how-to-vote cards provided to the organisation by Local Government candidates in these contests
Based on legal advice received by the organisation, the VEC ceased this practice in 2004. (Draper, 2004:10)
With regard to the placement of Candidate Statements online, the Committee received a considerable amount of support for this activity, and that legal barriers to this practice should be overcome.
The Privacy Commissioner, Mr P. Chadwick, stated that:
Mr A. Davies, on behalf of the Victorian Local Governance Association (VLGA), stated that:
Mr R. Spence, Chief Executive of the MAV, stated that there are good reasons for this to occur, and that in the absence of the provision of this material, some councils had considered providing the information themselves, but faced their own legal difficulties:
He went on to add that:
However, the Committee also notes that the Electoral Commissioner expressed a number of reservations about the practice that he felt should be acknowledged by the Government, and emphasised the strong position of independence his position commanded and required. The Commissioner, Mr S. Tully, observed that:
The Committee acknowledges this problem, but considers it worthy of consideration against a number of criteria:
What is the purpose of the provision of the information?
Does it achieve an objective worthy of the cost of administration?
Can the material be managed by the VEC?
With regard to the first question, the Committee considers the provision of this information - in some form - essential to all-postal Local Government elections where the ability to canvass around the polling booth is not possible. This does not, however, necessarily require online publication, and the Candidate Statements are also distributed with voting material. The general principle of more information provision is clearly supported by representative bodies of the Local Government sector.
With regard to the second and third questions, the Committee notes that - having collected the information in the first instance (restricting publication to print distribution only does not limit the data collection and vetting process), the actual cost of distribution via the WWW is very low (and lower still, under the VEC's next-generation CMS than in previous years). In addition, open access to the information allows for access by members of the vision-impaired community through the use of adaptive technologies to access this information conveniently.
Overall, the Committee considers that the practice of publishing Candidate Statements online should be reinstated, with the Minister for Local Government working to address the legal barriers to this activity. This would require adjustment of the Local Government (Elections) Regulations 1995.
The Minister for Local Government should amend the relevant regulations to permit and require the online publication of Candidate Statements, authorised how-to-vote cards, and a candidate photograph by the Victorian Electoral Commission.
In addition, while Part I demonstrated that general adult interest in political issues is relatively high, the provision of additional information through any means cannot be discounted as unimportant. It has been noted that the high usage rate of ICTs by younger voters may mean that this is a vehicle by which to address concerns about their level of engagement with formal political processes. In national data collected for the AEC in 2004, the Committee found areas of concern around the preparedness of younger Australians to cast an informed vote, as per Figure 11.
Source: Print, Saha and Edwards 2004[26]
The Committee notes, however, two issues in reading this evidence:
First, the Print, Saha and Edwards (2004) study provides national data. Victoria should be cautious about assuming these statistics are representative of Victorians
Second, among older Victorians, use of online sources of information for political purposes is not particularly high, and the use of the Internet in this area should not been seen as a panacea.
On the other hand, the provision of this information via the WWW may expand the informational opportunities of Victorians in direct and indirect ways:
Directly:
As an online repository of information for individual members of the public who prefer to receive information in this form
For those who will use accessible technologies to achieve this objective
To support the distribution of campaign information by minor parties or independent candidates which have limited resources to access physical communications (direct mail, leafleting) or commercial (advertising) channels of communication
Indirectly:
As a centralised repository for information "gateways", such as the syndication of the information online and as a reference for commentators such as academics or commercial, public and community journalists.
Thus, it is the opinion of the Committee that Candidate Statements should be collected and published online for all elections under the Electoral Act 2002 and the Local Government Act 2002 in Victoria.
The Local Government Act 1989 should be amended to incorporate the provision of Candidate Statements, authorised how-to-vote cards, and a candidate photograph in all Local Government elections, regardless of voting method (postal, attendance) employed.
The Electoral Act 2002 should be amended to incorporate the provision of Candidate Statements, authorised how-to-vote cards, and a candidate photograph in all Parliamentary elections.
Given these recommendations, the Committee notes that the publication of this material on the WWW should be supported with an effective means by which to locate and promote this material. The work of groups like the League of Women Voters through its Smart Voter website has been to provide a centralised repository for this information which allows voters to search by electorate or postcode. The Committee considers this approach would be useful and could be integrated with the enrolment lookup service already provided by the VEC.
The Victorian Electoral Commission should develop a centralised election portal for Candidates' Statements for future elections.
The provision of this information through the portal should allow for the syndication of this content.
In addition to the recommendations above, the Committee notes that the provision of electoral information to members of the disabled community who do not have ready access to ICTs and the translation of electoral material would be of particular benefit in expanding the participatory opportunities of Victorians who have difficulties with electoral information provided in print or in English.
During the course of the Inquiry, the Committee received evidence about work that is already being undertaken in this area by the Victorian Liberal and Labor Parties, the VEC (in partnership with Vision Australia) and some work at the national level by Blind Citizens Australia (BCA).
In 2002, the two major parties provided "how-to-vote" information at the VEC's four Accessible Voting Centres (AVC) in Braille and large-print formats (Submission No. a3, p. 15). The VEC has noted that its AVCs will be redeveloped for 2006 to include:
Magnifiers
Special lighting
Closed Circuit Television (CCTV)
Specially trained staff. (Submission No. a27, p. 21)
Speaking for BCA, Ms N. Mattiazzo stated that:
The Committee observed the provision of voter information by non-government bodies during its delegation to North America. In California, the California Voters' Foundation had been contracted by the Secretary of State to provide access to candidate information. This relationship saw the non-partisan group source information from candidates for general distribution to the community, utilising its specific areas of expertise (communications) to target the distribution of information and provide guidance to candidates about effective presentation strategies.
The Committee was informed that, while the uptake of these - and similar - services by candidates had been strong, the effective provision of voter information did require a significant investment of effort in order to promote the existence of the service, encourage participation by candidates and provide education for candidates to make best use of the service. Thus, as was the experience of BCA, the simple provision of translation or document conversion and distribution services by community organisations does not appear to be sufficient to automatically generate a high level of uptake, even where the service is free, or nearly so.
In the United Kingdom, the Electoral Commission has also noted the need for specific action in this area, and has taken a "focused" approach through the establishment of a special-purpose, seed funding grants scheme to encourage experimentation in means to increase public awareness and participation. (Electoral Commission, 2002:25)
The provision of accessible information for members of the community who have limited access to conventional printed material in English is regarded by the Committee as a basic democratic necessity in a diverse, multicultural community such as Victoria. The Committee therefore regards the opportunities for partnership with community groups to "link" candidate information with voters who have problems accessing print materials or whose first language is other than English as important to be embraced by the State. The model to support this used in the United States appeared to be the most appropriate way to undertake this activity.
The Victorian Government should provide targeted funding to relevant community bodies to provide access to voter information in a wide range of accessible formats and major non-English linguistic communities for the next State election.
This funding should be provided through the Victorian Electoral Commission, which should be tasked with the process of contract management, oversight and reporting to the Parliament.
The Victorian Electoral Commission should develop and provide a policy manual for the collection and distribution of this information, including information about Victorian electoral law and issues of defamation, and should require equal access to all candidates.
It is the view of the Committee that the development of this approach should be undertaken by the VEC in conjunction with the administration of the recommended information portal, with the strong possibility that the focus of this work will be on the conversion of material stored on the portal.
Given the discussion in the earlier section on A Limited Role for Networked ICTs, this section will not reiterate the Committee's reasons for rejecting adoption of ICT-enabled remote voter enrolment. The Committee notes that the VEC has undertaken a range of activities in this area to provide increased access to the process of enrolment, such as:
The provision of a voter lookup service online which permits quick enrolment confirmation by the elector at any time
The provision of enrolment and similar forms online.[27]
In the opinion of the Committee, this approach is reasonable, and shows that the VEC has been systematically using new technology to improve its services to the Victorian community. The Committee supports this work, and notes that:
This system does permit some remote provision of electoral enrolment forms, through post or facsimile transmission to the VEC
The VEC is working to improve its online offerings, with a proposed significant improvement to its website during the next six to twelve months
The current approach focuses on the collection of signatures or the mark of the applicant, an important piece of verification information that is held by the VEC as a security measure.
The Committee feels that the retention of signatures is appropriate and cannot realistically be accommodated in the online environment at this time for the majority of users. The Committee notes, however, that the increasing prevalence of two technologies place pressure on the status quo:
The increasing prevalence of scanning technology in multi-function printers, which provides for the capacity to email material scanned from paper sources (Davidson Consulting, 2001)
The increasing prevalence of pen-computing (such as though PDAs, tablet laptops or graphical design interface systems for desktop machines) and the implications of this on the remote collection of signatures[28].
Thus, while the Electoral Commission's view that the current Electoral Act 2002 does not permit remote registration by a service such as the Internet, it is the opinion of the Committee that this is only accurate to the extent that the two signatures (applicant and witness) must be provided. The Electoral Act 2002 specifies that (s23):
These provisions are technologically neutral, and thus have facilitated enrolment by facsimile in the past. The prospect of electronically "signed" documents is clearly recognised in section 9 of the Electronic Transactions Act 2000. At the surface this approach would appear to be consistent with the use of facsimile machines and therefore should be investigated as an option for consideration. The Committee notes, however, that while the question of scanning of the form appears relatively straight forward, the issue of collecting data from a pen-computing device remotely may be more complex (though clearly could be accommodated in some implementations of Adobe PDF forms), for example. The Committee considers that the Parliament should consider this approach based on a specific recommendation from the VEC on this matter.
The Victorian Electoral Commission should provide advice to the Parliament and Government prior to the next State election as to the appropriateness of facilitating the delivery of scanned voter registration and postal vote applications via email or similar method.
The Victorian Electoral Commission should provide advice to the Parliament and Government prior to the next State election as to the appropriateness of facilitating the delivery of signed voter registration and postal vote applications using pen-interface technologies.
In addition, the Committee notes, as per the discussion provided in Part II, that there are difficulties with the use of PDFs as the only means of providing enrolment forms online. The Committee notes the evidence of Ms N. Mattiazzo, who stated that from the perspective of a blind citizen:
The Committee notes that the Internet provides a means by which persons with limited vision or mobility could complete the enrolment information online, even if the final act of signing was still required for a print-and-post/fax/mail version of this document. The process would involve the creation of a form online that is accessible to assistive technologies (WCAG compliance) and allows all the relevant information to be entered and then provides a printable version that requires only the addition of signatures or identifying mark.
The Portable Document Format (PDF) version of the existing enrolment form maintained on the Victorian Electoral Commission website should be supplemented with:
An additional version in an accessible format.
An online form allowing users to enter their enrolment details online, but still requiring a printed version to be posted to the Victorian Electoral Commission for verification of the signature.
The Committee notes that an online version of the enrolment form would allow data entered by the user to be captured in a database maintained by VEC and imported into the electoral roll following the arrival of the paper/fax/scanned version of the form. At this time the VEC would also capture and confirm the relevant signature fields.
The advantages of this approach are:
Simplicity of the process for users - print, sign, post
Security through retention of the signature fields
Additional choice to users
Data importation would reduce costs of administration
Data stored in electronic form would limit data entry errors
The possession of the data at the time of form creation would allow the VEC to track the arrival of the final version, and prompt reminders if this version does not arrive in an appropriate timeframe (enhanced service).
Given the discussion in A Limited Role for Networked ICTs, this section will not reiterate the Committee's reasons for rejecting adoption of ICT-enabled (e.g. Internet voting) remote voting systems for Victoria. The Committee does not feel that the introduction of remote voting systems at this time is an appropriate or safe decision.
However, the Committee notes that there have been, and will continue to be, calls for the use of new systems for undertaking voting remotely, and that there is considerable public interest in greater flexibility in the way voting is provided. This is both the view of electoral organisations that have undertaken pilot studies of remote voting (Federal Voter Assistance Program in the United States of America, the United Kingdom Electoral Commission). In the Victorian context, the VEC observes that postal voting has become increasingly popular with voters, as indicated in Figure 12.
Source: VEC, 2003a:29
It is the view of the Committee that there are no general arguments of principle upon which remote voting using ICTs should be rejected in the near term, outside of the problems of security. Elements of an appropriate response, such as a trusted PKI encryption system (e.g. the Federal Government's Gatekeeper PKI project; Smith, 2002:5) may be moving the environment towards resolving the basic problems of security.
The State of Victoria, through acceptance of postal voting, pre-poll voting and all-postal Local Government elections, has already accepted that remote voting and widening voter choice is an appropriate means by which Victorians should be free to cast their ballot.
Thus, given:
The capacity to secure these networks
The capacity of the channel to deliver the appropriate user interface[29]
And reasonableness of cost given demand for new channels
the Committee is of the view that remote voting via ICTs should be introduced.
The key advantages of remote voting systems are that:
They provide access for people who are travelling or who have mobility difficulties
They provide an interface with assistive technologies, such as screen readers or Braille computers, magnifiers or alternative data entry systems
They allow for the distribution of a wide variety of ballot papers in alternative languages, in a means far more cost effective than printing these papers
They allow for speedy (near instantaneous) processing (receipt, verification, counting, processing)
They allow electors to interface with a wider array of informational services while considering and casting their vote.
While the introduction of remote electronic voting systems does not appear to have an impact on voter turnout (based on the comparative data collected in the United Kingdom pilot studies, IDeA, 2000:5; and similar findings from the Canadian experience, Delvinia Interactive, 2004:iii), when provided remote electronic voting options a reasonable proportion of the electorate took up these channels in the UK experience (Norris reports that 24.5% of the electorate took up the option of remote voting using ICTs in the 2003 UK pilots, 2004:19).
Overall, the capacity to provide private voting to members of the vision impaired or non-English speaking Victorian communities would appear to provide the most compelling argument in favour of the adoption of these systems. This key advantage was picked up in the United Kingdom report on the 2003 pilot studies by Scope, a disability advocacy organisation that oversaw a number of the pilots. Scope noted that there was a preference in the disabled community for Internet voting, and this was seen as having benefits over postal systems, particularly with regard to privacy:
The Committee notes, however, that any remote voting system is likely to require continued integration with the postal system in order to deliver voting information and access tokens (such as PIN numbers and/or physical tokens), and thus the limitations of postal networks (security, loss, delay, interception, theft at point of delivery) still apply to Internet voting systems.
Some have argued that the issue of delivery of PINs and other identification information or tokens negates the value of ICT-enabled remote voting. A common example cited is of the Victorian backpacker in a remote location with poor access to news and postal services. The Committee recognises this limitation, but only to a small extent. With the regularisation of Local Government and State elections (with the exception of unforseen events), the capacity for travellers to be aware well in advance of clashes between their travel plans and elections will be reduced.
While the Committee observed that the postal voting system presents some problems in remote voting as currently undertaken (a percentage of loss, the extent of which is not clear at this time; E1C, Submission No. a16; Meeting with Mr R. Kidd, SecureVote), the Committee does not consider Internet-voting as (a) a salve to these problems or (b) necessary to address any specific cases of criminal activity in the Victorian context. The Electoral Commissioner observed that:
On the balance of evidence, the Committee supports the introduction of remote ICT-enabled voting systems at the first available opportunity. In order to ensure that the State Government of Victoria can remain appraised of when this opportunity may arise, the Committee feels that it is appropriate for the VEC to appraise the Parliament on a regular basis as to the status of developments in this area and the wider security environment.
The Committee looks forward to a time when the security environment has changed to the extent that it can recommend pilot projects commence for the State of Victoria.
Parliamentary elections should not be conducted remotely via information and communications technologies in the near future.
This decision should be reviewed by Victorian Electoral Commission following each State election, unless requested to do so earlier by the Minister.
Within 12 months of a positive review by the Victorian Electoral Commission, the Scrutiny of Acts and Regulations Committee table a report in Parliament concerning the desirability of introducing Internet-based voting, either generally or for a limited purpose.
While the Committee has rejected, at this time, the adoption of remote ICT-enabled voting systems, different considerations were given to the use of computer systems in voting centres that are not connected to a publicly accessible network, such as the Internet. These systems are in use around the world, and provide for the casting of the citizen's vote using a computer or similar device.
While there have been a range of issues presented to the Committee regarding security surrounding the implementation of these systems, it is the view of the Committee that electronic voting machines, appropriately designed and configured, can provide a secure voting environment. In addition, as discussed below, there are considerable benefits in making these systems available, not to replace the conventional paper-based voting system, but as a means of providing assistive voting to the disabled and citizens with limited English language skills.
The Electoral Commissioner has stated a clear desire to introduce these systems for trial in future State and Local Government elections, and the Committee considers this an appropriate move forward for Victoria.
The Committee has focused on five reasons the introduction of these systems should be piloted in Victoria:
They provide a means for the casting of private votes by Victorians with no or limited vision, or who may have limited ability to manoeuvre a pen on a conventional ballot paper
They provide a means for the provision of multilingual voting options
The successful implementation of these systems provides one part of the technology needed for remote ICT-enabled voting
They provide safeguards against accidental informal vote casting
They can assist in the tabulation of votes.
Overall, the Committee considers the first three reasons as more significant outcomes in a limited implementation of the systems across Victoria. Importantly, alternative approaches, such as Braille templates and LOTE ballot printing-on-demand were not considered to be practical solutions in the longer term. The use of Braille templates in previous State elections was seen by the Committee as well motivated, but the implementation of these was difficult; will be less feasible with larger Legislative Council ballot papers from 2006; and only of use to that proportion of the vision-impaired community who use Braille (N. Mattiazzo, Minutes of Evidence, 18/02/05, p. 34; T. Clark, Minutes of Evidence, 18/02/05, pp. 41-2). In addition, the distribution of computers and printers to provide ballot papers on demand is seen as problematic (see Difficulties and Limitations).
The enhanced prospect of access by voters to a private vote, is a recognition of a basic human right and thus the Committee feels this approach is necessary. In supporting the introduction of these systems, BCA noted that:
With regards to the system chosen for the Irish electoral process, the Irish Commission on Electronic Voting provided a positive assessment of the application of the technology[30], stating that:
election results can be calculated and published quickly;
use of the system may secure future reductions in election costs. (Commission on Electronic Voting, 2004)
These appear to be applicable to the Victorian context, and given changes to the electoral process for the Legislative Council from 2006 onwards, the complexity of the ballot paper- and vote-counting process would appear to encourage greater automation. However, the Committee observes that the question of cost should not be a driver, and that short-term costs per vote of investment will be considerably more than conventional methods. As the Electoral Commission of the United Kingdom noted in its 2003 report:
The question of cost clearly relates to scale and the potential for developments to be co-ordinated between levels of Government and across jurisdictions, but also to the number of implementation cycles through which a single technology is put. In its submission to the Committee, SecureVote provided the Committee with a perspective on implementation costs over a longer time frame than has been the experience of either the Australian Capital Territory or the United Kingdom pilots. They observed:
The Committee was impressed with the Electronic Voting And Counting System (eVACS) developed by the Electoral Commission of the Australian Capital Territory. In this system:
Generic PCs were used for the purposes of voting and the cost of acquisition of these machines was relatively low. The PCs were "converted" into voting booths using relatively simple modifications (cardboard shroud, raised numeral stickers on the keypad)
These computers were networked to a central server within the voting environment where the votes were stored. This machine was stored in a highly secured case and provided an independent power supply. The use of a central voting server allowed redundancy to be focused on that machine only, with failures of the individual PCs expected and therefore not preventing voting from continuing
Voters attending the polling place were issued with a barcode token once they were marked off the electoral roll. These voters had the option of voting using the conventional method
The token was used to acquire the electronic ballot paper and confirm the vote cast
The computer provided visual and audio instructions, in a range of languages
The computer notified the user of invalid voting, but would permit it if desired by the user
Voting data was moved physically on two compact disks produced by the voting server and was delivered to the central processing location via two separate means (the data did not travel over the network)
Voting data is stored both as preference data and as keystrokes to allow for cross-verification of the data and voter intention
Hashing was used for the software to ensure software employed was the same as that which had been certified
The software was released as open-source software (OSS, under a General Public Licence) to allow for third-party certification and oversight, errors in the code identified by members of the public were incorporated into the system ("patch") and sat on top of an open-source OS to maximise the transparency of the total codebase
An independent auditing firm audited the software.
In discussions with Elections ACT, the Committee was impressed by the advantage of ensuring that the voting machines remained under the physical control of electoral officials, and that the security environment they faced as a result of excluding the system from an open network produced a relatively simpler kind (ensuring code quality and preventing physical tampering with the machines during the voting, maintaining control over post-voting data). Thus, compared with Internet-voting (or other systems using an open network), the costs of development of the system are far lower, as security is focused on the integrity of the machine and the vote recording and processing code only.
There are a number of limitations associated with the adoption of these machines, which the Committee explored in some detail, both in discussion with Elections ACT, and with government regulators in the United States. In particular, the Committee identified a range of problems associated with electronic voting machines in the United States that have motivated considerable debate surrounding their use and the method of procurement.
These difficulties are well documented, but have resulted in problems such as:
Voting computers fail to record votes correctly
Suspicious e-voting tallies cannot be verified
Touch screens present incomplete ballots to voters
Touch screens register incorrect choice
Tabulation software tallies votes incorrectly
Voting computers break down during election; new election required
Voting computes fail to start up
Smart-card encoders fail to start up
Smart-card encoders fail to operate properly
Daisy-chained machines shut down when one breaks down
Smart-cards encoded incorrectly; voters receive wrong ballots
Ballot too large for touch screen; paper supplements required. (Verified Voting Foundation, 2004)
The Committee notes that each of these areas of concern are based on actual experiences in the administration and implementation of these systems in the United States. In addition, Elections ACT has observed that eVACS has had some hardware difficulties (clock battery failures leading to "dud" machines, "temperamental" barcode readers slowing the process). The issues and implications are discussed in detail below.
Firstly, the Committee notes that the system of electoral administration in the United States has contributed towards the problems experienced in that country. Since elections are administered at the county level, this has led to:
A proliferation of machines and contracts that has created difficulties in certification by the State Government of California
A diffusion of machines down to very local levels of implementation, often where there is insufficient security (voting machines being stored in personal homes, for example); inappropriate storage (machines being subject to damp); rough handling through movement in private vehicles; and limited skills among part-time or volunteer electoral officials (who tend to be older persons).
In the view of the staff of the Office of the Californian Secretary of State, as the regulators, the capacity of the State of California to ensure that the software and hardware provided to county electoral administrators was limited. This was because the State had limited technical resourses to oversee the certification of machines, and because of the wide geographical spread of the systems and limited ability to physically inspect them. This led to issues associated with non-certified code being installed on some machines that was only identified after the election process, leading to public suspicion as to the legitimacy of the process overall.
Second, it is clear that there were fundamental problems with the certification of software used in these machines. Much of this debate focused on one provider, Diebold, for three reasons:
There was a political relationship between the management of the company and the Republican party
The machines had been revealed to be flawed, and some had been involved in the "non-certified" software scandal
The company aggressively attempted to prevent the release of its code and internal memos suggesting the code provided to the State was of poor quality.
As Kohno, et al. (2004) observe in their highly critical review of the source code of one of the United States vendors:
These debates highlight a number of problems:
The need for a certification system which is transparent
The need for a software management process that accommodates the time-critical nature of election preparation
Risks associated with reliance on internal certification
Limited public confidence in systems where the code is proprietary.
Each of these issues were highlighted in the Irish Electronic Voting Commission report (Commission on Electronic Voting, 2004). While this Commission was confident that the specified system would deliver the service and experience of voting, it could not gain - in a timely manner - access to "final" software for the systems, and thus it essentially refused to certify it. The use of the machines, therefore, was rejected.
Given the problems above, it is clear that the implementation of these systems in Victoria would require:
Strong management and oversight by the VEC, which would maintain complete physical control over the voting environment and hardware used
Following the Irish certification issues, there is clear benefit in using established systems whereby the rapidity of software upgrades would be less frequent. There are clearly distinct disadvantages in this area of being at the leading edge of technology development
A clear and transparent software certification process, with more than one checkpoint for certification of the code, and an acceptance (as per the Irish example) to refuse certification, even at the last minute.
In this context, the Committee notes that certification of the code is essentially an information technology version of scrutineering, and thus:
Must be open to the public
Cannot be prevented by issues of patent, copyright or proprietary intellectual property.
On this matter, Dr Clarke (Meeting, 01/03/05) made the following observation:
Dr Clarke went on to state that one of the advantages of developing these systems in open source - as per the eVACS model - is the expansion of the number of parties interested in scrutinising the underlying software.
The value of this approach to external review is highlighted by the report released by the Computer Science Laboratory of the ANU, in conjunction with the Logical and Computation Group of National ICT Australia (Abate, Dawson, Gore, et al., 2003):
Clearly, this type of external oversight must be observed by the electoral administrator, and the Committee notes that this would require a clear and robust process for managing code evaluations outside of the formal certification process.
In addition, the Committee notes that, given that limitations of the code may still not be identified prior to the election process, due caution must demand that an audit trail be provided. It is the view of the Committee that this audit trail would be best provided as a paper version of the voting record - a printed copy.
On this matter, the Committee did receive credible evidence that the need for a paper audit trail for electronic voting is unnecessary from Elections ACT. Their view was that:
The system employed in the ACT provided the voter with a number of opportunities to confirm the vote selection
The use of "counters" on the server provided a double check against loss of the vote
Collection of preference and keystroke data provided triangulated assurance at the data level
The transportation of the voting data using two physical devices that were not moved together provided an additional layer of security
The introduction of printers or similar paper records would be an unnecessarily expense, introduce new areas of systems failure and encourage recounting of the paper ballots by candidates.
In addition, the Committee noted that there may be limited demand for a paper trail from members of the public interested in using electronic voting systems. Certainly, for those Victorians with no or limited English or with severe visual impairment, a paper trail provides them no certainty, if their right to privacy is retained. From evidence presented by Mr C. Burton to the Committee via email, that in the case of his company's use of a confirmation system for Internet voting in the Stratford-on-Avon election, only twenty-three of 3,128 participants utilised their receipting service (Burton, Correspondence, 22/02/05).
The Committee accepts that two-factor vote recording is likely to provide for sufficient security, particularly where the voting system is deployed in a Local Area Network (LAN) configuration. However, the Committee also notes that:
Paper audit trails have been introduced in California following voter backlash against non-physical vote recording systems. This retrofitting has been at considerable cost and required post-hoc adjustments to application software
The paper audit need not present an unnecessary burden in terms of administration, given that the systems are deployed in locations in which paper records are already being stored under strict security.
The Committee notes that the paper audit is not, and should not be, something that the voter takes with them. Instead, it should be a record stored by the administrating officials. Examples might be:
A printer, protected by a transparent barrier that allows the elector to see the record prior to final processing, and then have it deposited in ballot box when satisfied with the accuracy of the record. This record would look as close to possible like a conventional ballot paper
A tape or roll record that records the vote in a continuous role stored by the system.
In addition, while there are concerns about the existence of paper records undermining electronic counting (that is, by providing the paper record, a physical recount is possible), the Committee notes that:
The recount is an essential part of the electoral process and should not be deliberately suppressed
The printed ballot papers would be more effectively processed by ballot-scanning technology (see Automatic Ballot Scanning)
The number of ballots envisaged will not be high in the first instance, as the focus is on the use of these systems as an accessibility tool, rather than as a general voting tool.
On the balance of probabilities, the Committee considers that the use of stand-alone, or locally (closed) networked electronic voting machines, can deliver significant and sustained benefits to members of the Community who have limited or no vision, difficulties with dexterity, or who would benefit from the translation of voting material due to a lack of fluency in English.
Given the limitations of the Braille voting templates used in previous elections, and the low likelihood that these will be able to be employed for the Legislative Council ballot paper for 2006 onwards (because of its estimated size), the Committee considers that action will have to be taken to provide assistive technology for the casting of a private and secure vote by Victorians who are visually impaired or who have a low level of English literacy. The flexibility of electronic systems over the provision of large numbers of paper multilingual ballot papers is clear and this approach seems like an appropriate response to issues of access in our community.
Thus, the Committee is of the view that the introduction of these systems is justified by need. In considering the appropriate configuration for these systems, the Committee would recommend:
A limited trial, based around placement in AVCs and locations favourable to areas of access for LOTE speakers
The use of a LAN model in a closed network, not stand-alone kiosks
The use of electronic counting and tabulation
The provision of a paper trail auditing process
The development of the approach using the OSS development model.
This recommendation has strong similarities with the approach undertaken by the Australian Capital Territory. However, the Committee would prefer a number of differences and recommended different approaches:
The Committee does not consider that it is appropriate at this time to discard the paper trail in favour of electronic record keeping only. This recommendation is based on the strong need to provide the public with confidence in the system developed, using a simulacra of the model of voting with which they have familiarity
The release of the open source code in the Australian Capital Territory example requires improvement in the management of the external review process, and the provision of cleaner, better-documented code in the first instance
The Committee would like to see pre-implementation "white hat" hacking attempts undertaken prior to final certification of the hardware and software.
The Committee acknowledges that this recommendation will be disappointing to groups in the community who would like to see these machines provided universally. In their submission, the BCA proposes the introduction of a system of "kiosk" voting, called Electronic Assisted Voting (EAV). The proposed EAV comprises four elements:
The development of simple electronic voting systems based on a computer, printer and ballot-box system (with the computer acting as an interface between the user and a conventional paper ballot)
The system be developed as stand-alone units (non-networked)
The system be deployed to each polling place
The system has no role in vote counting (Submission No. a20)
While the Committee recognises the principle of universal provision, it notes that limited resources of the VEC and negative experiences in wide-scale implementation would mitigate against a State-wide rollout at this time. The Committee, therefore, considered that a phased deployment of these systems, similar to that undertaken in the Australian Capital Territory, would be an appropriate approach, with extension of the deployment of these systems being undertaken based on:
Successful experiences with development and implementation
Public acceptance and demand.
It is possible that developments in robust hardware design may permit stand-alone units to be employed in the future (indeed, the Indian government has developed a low-cost, portable electronic voting unit it claims as robust, tamper-free and effective; however, this system does not at this time afford a paper receipt, preferential voting, spoken-word interface or the potential for networked redundancy; http://www.eci.gov.in/EVM). The Committee does not regard the recommended approach as preventing development of stand-alone kiosks following successful proof of concept testing of the modified eVACS model.
The Victorian Electoral Commission, in consultation with relevant stakeholder groups, should develop and implement a system of electronic voting machines for local and general elections in Victoria. These machines should:
Permit the casting of a private, unassisted vote for the blind, those Victorians with limited vision, and Victorians with low levels of English literacy
Provide the same voting instructions as appear on the paper ballot in a range of languages other than English
Produce a voter-verifiable paper trail to be retained by electoral officials
Allow for the electronic tabulation of votes cast
Be restricted to a closed local area network under the complete physical control of electoral officials.
The implementation of this electronic voting system should be undertaken with consideration of:
The use of open-source software to provide maximum public oversight of the source code used in these machines
The capacity to collaborate between jurisdictions over the development of hardware and software
The capacity of the Victorian Electoral Commission to pilot the system on a limited basis, and expand provision of these machines based on community need and demand.
The Committee considers that the VEC, in undertaking this work, will note the value of collaboration with Elections ACT in this area, both as a means of reducing implementation costs and as a potential partner with a mature technology that would benefit through collaboration. The modular development of the system would allow different counting methods to be developed against an underlying common architecture, and further permit partnerships from other jurisdiction over time.
While the Committee opposes ICT-enabled remote voting in the medium term, it supports - in principle - this activity due to the significant benefits afforded to a proportion of the Victorian community. In making positive recommendations regarding the adoption of computer-based voting for Victoria, the Committee recognises that structures will be needed in the VPS to ensure that changes in the security and experiential environment surrounding electronic and ICT-enabled voting systems will be required if Victoria is to take advantage of technology as security issues are resolved or mitigated.
The Committee has already noted (above) that there has been significant interest in Internet-based voting systems in the corporate sector and among some unions around Australia, and that a vibrant private-sector industry is developing to support these activities. However, the Committee is concerned to ensure that the VPS also develops its own expertise and knowledge of the technologies and applications for electoral purposes.
The Committee has noted that, in the past, the VEC has undertaken research and participated in experiments overseas. This approach is valuable in improving the Commission's knowledge base, but also in expanding the professional network of the Commission's staff internationally; an excellent management process to establish a community or practice of knowledge and policy transfer.
Given the Committee's interest in the prospects for ICT-enabled voting for Victoria, the Committee considers that a specific allocation of funding to the VEC to continue this work would be beneficial in ensuring that the level of awareness and technical skill in the VEC is maintained and improved, and that, as per Recommendation 52, resources are provided to allow the VEC to provide timely and accurate advice to the Parliament of Victoria if and when the prospect for ICT-enabled voting systems improves to the extent that the Government of Victoria can reconsider its introduction.
The Victorian Electoral Commission should maintain an appraisal of developments in the area of electronic and online voting, and be provided with the resources to send staff to participate or observe the conduct of electronic and online elections internationally and around Australia.
While the issue of electronic or online voting is an area in which the VEC and Committee recognise that difficulties in perception and performance have motivated due caution in the adoption of these systems, the Committee notes that the VEC has for some time adopted electronic vote-counting systems to assist with the administration of the electoral process. The use of these systems provides an alternative means for tabulation, which offer benefits over manual counting in terms of speed and the accuracy of the count.
The Committee has some observations to make regarding:
The current management of electronic vote counting in Victoria
The proposed acquisition of electronic vote-scanning technology for Victoria.
The Committee makes a number of brief observations about electronic counting currently employed in Victoria. During discussions with the VEC during 2004, the Committee observed that:
The VEC has been using data-entry systems for electronic vote counting for some time
The VEC considers this system as having lower data-entry errors than manual "hand" counting
The VEC administers this process though a component of its elections-management software, a program developed and maintained "in house"
That the data-entry element is undertaken on computers leased by the VEC in or near the place of polling.
The Committee has confidence in the policy rationale for data entry and electronic counting of votes as undertaken by the VEC to date. However, the Committee notes two concerns with the practice as administered currently in Victoria.
Firstly, as per the discussion of certification above, the Committee is not assured that the quality of the vote-entry software or the counting component of the elections-management program has been certified to the extent that it should be. The code maintained by the VEC is not transparent and is internally maintained at this time.
It was the view of the VEC that this is appropriate because the software represents a considerable investment by the organisation, and that it has been subject to ongoing development to improve its performance (Submission No. a16).
Second, the use of short-term leasing of the data-entry locations appear to present some risks to the integrity of the process. While the VEC notes that the computers, while remaining on the LAN, are re-imaged (their hard drives are written over with a "fresh" operating system installation) by the provider prior to the installation of the data-entry software, it is not clear that this would necessarily protect against malware.
Theoretical risks that the Committee identified are that:
It is not clear to the VEC that the image provided by the data centre is necessarily "clean"
It is not clear whether the network separation is provided by a software function or physical break in the connection
The VEC needs to assure itself that there are no secondary storage drives in the LAN machines that remain uncleaned
The VEC needs to consider the hypothetical risk that peripherals connected to the LAN are compromised (such as printer memory).
While there are a number of limitations presented here, the Committee acknowledges that these risks are likely to be remote, and that they are hypothetical only. It would be useful for the VEC to stage an attempted corruption of this system as a test of its integrity.
In the medium term, however, the Committee considers that the protection of this aspect of the electoral process would be further safeguarded by:
Immediate introduction of a system for certification of the vote counting, data-entry, and ballot drawing software used by an independent third party (a software certification firm or similar)
The adoption of a software management and review process similar to that recommended for electronic voting machines in the State of Victoria (OSS with external oversight)
A risk assessment be undertaken into the current procurement practices for short-term data entry terminals.
The Victorian Electoral Commission should be required to provide independent technical certification of its current electronic vote-counting software.
Certification should be required following each modification of the software used.
The Victorian Electoral Commission should be required to ensure that future acquisition or development of electronic vote-counting software be open-source software. This requirement should also specify that:
This source code be released by the Electoral Commission via its website.
A process for the collection, review and incorporation (where appropriate) of comments and bug reports from the public be developed.
The Victorian Electoral Commission as a matter of priority should undertake a risk assessment of the practice of using leased computers for electronic vote counting. This risk assessment should include:
A third-party technical audit.
A review of the administrative practices surrounding the leasing, cleaning, preparation and use of these machines in the counting of votes.
This risk assessment should be tabled before the Parliament at the soonest possible date.
The VEC has expressed a desire to trial the use of automatic ballot-scanning technology in Victoria. This technology uses optical scanners ("digitisers") and character-recognition technology to convert written ballots into digital data[31]. The advantages of these systems are that:
They are fast and can count votes very accurately
They remove the manual, tedious process of data entry from the process of electronic vote counting
They can allow for fast re-processing
They are effective for ballots that are very large
The technology is becoming increasingly common in desktop computing environments and thus the public is familiar with it and its limitations.
The VEC sums up the proposition in its submission:
A similarly positive recommendation of this technology was provided by SecureVote, who noted that the technology underlying the systems are now highly mature (optical character recognition) and that the advantages of speed can be matched with accuracy improvements. The systems are also developed with regard to human oversight, allowing:
A digital image of the ballot to be stored with the data collected
Including "logic tests" to spark review by an operator (e.g. are all the numbers sequential?)
Variable "thresholds" of user review (depending on user requirements) for instances where the software is uncertain of the correct interpretation of the writing.
In addition, the Committee notes that when all Local Government elections in Victoria are brought into alignment, this may place an additional burden on the VEC for the administration of these elections (provided it is the successful tenderer).
Overall, the Committee regards the use of this technology as appropriate, and supports the intention of the VEC to trial this technology. If the technology can, in the Victorian environment, show a higher level of accuracy than current counting systems in a comparative study, then the technology has clear benefits to the administration of the electoral process given the speed by which counting can be undertaking with these systems. The Government of Victoria, therefore, should provide the funding for this technology, based on a business case that can demonstrate the cost effectiveness of the technology as compared with hand- and data-entry counting methods.
In addition, the Committee notes that the acquisition of this technology should see the VEC engage in collaborative efforts with other jurisdictions that may lower the cost of purchase, storage, maintenance and staffing for these systems.
The Victorian Electoral Commission should prepare a business case for the Victorian Government on the use of automatic ballot-scanning technology on a pilot basis. This review should consider:
Models for the implementation of these systems (acquisition, lease, etc.)
Means by which acquisition costs can be shared with other jurisdictions
Costs and benefits of implementation
Effectiveness of the technology and minimum thresholds by which recounts should be required by law (margin-of-error assessments).
It should be noted that the Committee considers the points made in Recommendation 56 relevant to the proposed acquisition of this project.
Of specific interest to the Inquiry was the use of ICTs to expand the capacity of citizens to engage in Government consultation and participation processes, as noted in the Terms of Reference for the Inquiry.
The Government of Victoria, like other governments around Australia and throughout the world, has embraced the notion of community consultation as a means to better ensure that public policy is aligned with the interests of the community - be that the broader general community of Victoria, or specific communities of interest and stakeholder groups - and that information relevant to policy decisions is obtained from those who are often closest to the issues under consideration (Organisation for Economic Co-operation and Development (OECD, 2001:11)).
This view is shared by other jurisdictions across Australia. The Western Australian Government's guide on citizen consultation states that the benefits of investment in consultation and inclusion can be seen to:
Raise the quality of policies
Raise the chances for successful implementation
Reinforce the legitimacy of the decision-making process and its final results
Increase the chance of voluntary compliance
Increase the scope for forms of partnerships with citizens. (OECD, 2002:1)
This view is supported by policy scholars, such as Peter Bridgman (Deputy Director-General of the Department of Education and the Arts) and Vice-Chancellor of The University of Melbourne, Glyn Davis (Bridgman and Davis, 2004:79), who see the growth of consultation as a reflection of shifting community attitudes to government, and changes in government about the role of the public in decision making.
In stating the democratic case for greater use of consultation by government, they argue:
Historically, public involvement in Government decision making has focused on consultation, with participation an expectation of the formal electoral processes only. This has sometimes been referred to as the "Board of Directors model": with the anticipation that the public's role in democratic life was only expressed at elections, where the performance of the government of the day was either approved or rejected in favour of the alternative government presented by the Opposition.
This view, projected from political leaders through their departments and agencies, was reinforced through the response of the public sector, which developed and refined a range of consultation techniques and supporting definitions. These tended to reflect the view that consultation was a process of information collection, with decision making retained by the Executive.
This perspective has been changing over the past four decades, with governments recognising - and citizens increasingly demanding - more active involvement in decision-making processes. The Committee recognises that the increasing expectations of the public for information and involvement in decision-making processes must be provided by Government, but in a planned and managed way that does not favour the few at the expense of the majority.
In this process of change, the use of ICTs has expanded the capacity of the Victorian Government to provide information to the community about policy proposals and plans, and the interactive nature of these technologies has provided a new array of ways by which greater public involvement in decision-making processes can be facilitated.
There is a broad range of definitions used by public sector organisations to define consultation. While each definition is subtly different, they tend to share a number of consensus elements about what consultation is:
Consultation involves two or more parties
It is a process that is structured in some way
Consultation involves the collection and sharing of information or points of view
Its objective is to inform the process of decision making (in that it explicitly occurs before decisions are made).
From these commonalities, there are considerable differences within the definitions used as to other characteristics of consultation. For some, consultation may involve consensus-building (O'Malley, undated), or expressly exclude shared decision making (Scales, 1997:1). For others, consultation represents an ongoing communications process between government and the public, while others see it as a short-term, or "one-off", process of information exchange (KPMG, 2000:81).
These differences are the result of an attempt, wittingly or unwittingly, to conceive of consultation as "an activity" or a "specific process". While this may have been the case a decade ago, the Committee notes that large parts of the VPS have recognised a clear shift to conceive of consultation as an array of techniques and methods with various levels of public engagement and participation.
Bridgman and Davis (2004) illustrate this shift with a "consultation continuum" or typology of methods of consultation that range from activities that require or encourage little or no direct public participation in formulating the final policy decision (information provision) through to activities whereby control over the process of consultation and decision making is handed over to the participants. Within this broad spectrum fit the range of consultation and deliberation approaches used across Government.
The Committee prefers the expression consultation and participation, emphasising the participatory and interactive elements of civic engagement by Government. The lower levels of interactivity consultation invariably "bleed" into the decision-making process in three ways:
They shape community views and attitudes to policy decisions through the provision of (or failure to provide) information
They provide the basis for decision making through the collection of information
They can provide for direct decision making by the citizenry or relevant stakeholders.
On this theme, Professor Stephen Coleman and Dr John Gøtze (2001) provide a three-stage model for thinking about consultation and participation, illustrating differences between one-way (on which subject Part II of this Report is largely concerned), two-way and ongoing information flows between government and the public. These three forms of consultation and participation are:
Consultation and participation, therefore, represents an array of methods and strategies. For any Victorian department or agency undertaking a consultation process, their consultation managers need to ensure that the selection of the appropriate methodology is aligned with:
The nature of, and importance of, the issue under consideration (the priority, scope and scale of the process should match its importance)
The need for information by Government (the extent to which the Government needs to acquire new information to make an effective decision)
The extent to which the establishment of a consensus is required (where levels of dispute over an issue are high) or where a consensus cannot be reached, where the introduction of a open and transparent process for decision making will increase the legitimacy of the outcome
The range of other outcomes that can result from engaging with the public (such as the stimulation or formation of local community groups or representative bodies, increased information provision to the community, or the formation of public-private or public-community partnerships for action).
It has become very clear over the past decade that the inappropriate use of consultation (such as where a decision has already been reached, or in subjects with limited public-interest value) undermines the rationale for the investment of public monies in this activity.
Following this general approach, the International Association for Public Participation provides a spectrum of consultation and participation tools that range from information provision to active participation[32]:
Information provision: fact sheets, web sites, open houses
Consultation: public comment, focus groups, surveys, public meetings
Involving the public: workshops, deliberative polling
Collaboration: citizen advisory committees, consensus-building, participatory decision making
Empowerment: citizens' juries, ballots, delegated decisions. (IAP2, 2000)
These are examples only, with IAP2 listing forty-seven different consultation and participation methodologies in its more detailed literature (2004). While some of these techniques are specific to ICTs, in general each approach is one that can be undertaken on- or offline (though the Committee observes that some techniques are likely to be more valuable in face-to-face environments, particularly if levels of trust are low, and need to be developed and reinforced).
The Committee observes a number of issues in the area of consultation and participation that would benefit from further discussion. These issues focus around the appropriate selection of consultation and participations methodologies, and the need for reflection about the benefits gained from different approaches at either end of the "spectrum of consultation".
First, the introduction of active participation in shared or delegated decision making is an inherently political act. The Committee is not blind to the tensions that shifts towards direct democratic processes bring to the policy process. Parkinson has identified that the citizen's jury model, popular in the United Kingdom, has been:
While, in a review of a similar process undertaken in New South Wales in 2001, Hendriks observed the:
Thus, while the Committee observes the clear benefit of more direct consultation and participation approaches in "cutting through" areas of policy making that are perceived as being "captured" by commercial interests or groups with considerable financial resources or "insider status", the selection of consultation and participation methodologies that provide for deliberative and direct democracy must be undertaken with:
A clear acceptance of the inherently political nature of selecting a deliberative methodology for policy consultation
A recognition of the impact of this approach on established political interests, and the need to secure the relevant political commitment to the process prior to initiation. Deliberative consultation and participation processes cannot be undertaken without the direct and explicit authorisation to do so, by the relevant Minister
Appropriate reflection on the correct approach to use (methodology) and an awareness of the ways in which different deliberative methodologies shape process, and possibly outcome
A refusal to use these processes in a cynical or instrumental manner, engineering desired outcomes through controlling access to the process, or the type of information employed within it (Bishop, et al., 2002).
What is required is a degree of transparency in the way that Government selects and implements its consultation and participation methodologies, and an explicit recognition of some of the underlying motivations about their selection and use. Where the intention of the Government is to select a course of action from an array of information - with public input only serving as one source of input for the decision - then informational and consultation strategies are the most honest approach to the selection of an appropriate tool.
Second, and on a related point to the above discussion, the Committee recognises a problem in seeing consultation and participation as a "sliding scale" of techniques along one axis only (degree of participation and ownership of the process by the community). While the various "continuum" or "spectrum" diagrams of consultation and participation methodologies represent a useful means to consider the type of tool needed to achieve the consultative process by Government, the "ascending scale" model can be misinterpreted to imply a greater normative good ("more democratic") associated with each methodology as one moves from information provision to implementations of direct democracy. See Figure 13, below.
Source: OECD, 2001: 23 (originally adapted from Health Canada)
This is not necessarily the case, and the Committee does not consider it appropriate to over-emphasise devolution of decision making away from Ministers and the Parliament. Methodological legitimacy remains the question that should be considered in any consultation and participation design process, and the role of decision makers is a key aspect of this process.
Thus, while Governments have accepted the value of devolving decisions, either to resolve intractable tensions over policy, or where devolution allows a discrete group within the community to manage its own affairs through co-regulatory models, the Committee notes that this can never be at the expense of the primacy of the Parliament, through its legislative functions and the compulsory electoral process that supports it through mass participation, or Executive Government, as a responsible form of administration.
While it is the clear intention of the Committee to stimulate and encourage the use of more participatory models of consultation, the overuse of these techniques can undermine the democratic processes. The Committee recognises that there is a proportion of the Victorian community which is, and remains, comfortable with their level of political participation contained within the electoral process only, and that the Government of Victoria cannot, and should not, necessarily force more active participation across the community.
This recognises that direct democracy has clear advantages, but if misused, undermines the democratic system through undercutting the responsible decision-making processes of the Westminster system. Thus, while the Committee supports the introduction of direct democracy methodologies in Government consultation and participation processes in Victoria, it considers that these will be undertaken:
Selectively, observing the inherent tensions between direct and representative democratic forms of Government
Reflectively, to build knowledge and practical skills associated with these processes.
Therefore, it is the finding of the Committee that the VPS and Executive needs appropriate information and tools to determine the best form of consultation and participation to apply, and that these tools need to note, weigh and provide guidance on appropriate application of deliberative processes that do not undervalue non-binding consultative processes with respect to devolved decision making.
Illustrating a non-normative classification approach is a useful diagrammatic representation developed by Fife Council in the Scotland. Intended as a guide for policy officers, it shows the respective benefits of different types of consultation depending upon which of the three objectives is required or desired. This type of model, or a further refinement of it, which emphasises:
Benefits and costs
Risks and aspects of best practice
Appropriate or maximum levels of reasonable participation
Example policy areas where the method has been utilised
would be of greater practical value to the VPS (see Consultation
and Participation Best Practice Reference Guide).
Source: Fife Council, 2002:1
Third, even where a department or agency does not pursue direct democracy approaches, the Victorian Government must recognise that policy advice is becoming increasingly open and contestable. This approach has been seen as a public benefit, only countered where there is a need for expediency in decision making. As Professor Iris Marion Young (2000) observes:
At one level, the increasing use of external advisors and consultants recognises that the VPS need not be the sole repository of policy expertise in Government, but this approach can be, and should be, applied more broadly where the Government makes a specific appeal for public participation. The type of inclusion under discussion by Professor Young is a broader approach than simple outsourcing of analysis to specialists and consultants, but one that recognises the value of social difference in determining decisions and making evaluations of policy options.
In this case, we have to recognise that Victoria is an (and is part of a wider) information society, in that:
The implications of this "profound impression" are that:
Government must recognise competing sources of information in policy making
Government must recognise competing capacity of analysis outside of the public sector and traditionally recognised areas of technical expertise
Government must facilitate debate over policy development and decision making through the provision of data, as well as information.
These points recognise that the Victorian population is better educated than ever before, and that some of the tools needed to analyse large and complex data sets - once the preserve of mainframe computers or large organisational budgets - are now standard desktop applications (spreadsheets, database systems) or available at low cost or as open-source applications (for example, the R project for statistical computing provides a free analysis package similar to commercial offerings).
The prospect of "contestable policy" through the release of Government data is highlighted by the submission of Dr M. Wigan, who observed that:
In addition, from evidence collected in North America, the provision of raw data to local community organisations can be beneficial in stimulating community action on issues of local concern, or in the creation of locally relevant content.
Mr Ari Schwartz of the Center for Democracy and Technology in Washington illustrated this point in his discussion of how local community groups had used government geographic data sets and combined them with pollution data collected by volunteers to place pressure on local administrations to improve water quality.
Similarly, Professor Peter Levine of the Center for Information and Research on Civic Learning and Engagement at Michigan State University talked about the role of schools encouraging multimedia content related to local history, using similar data released by government agencies.
The Committee notes that the State of Victoria has recognised the need for data provision to local groups and activities. The Department for Victorian Communities (DVC) has produced a guide for local organisations, Getting to Know Your Local Community: A Guide to Using Local Data that provides:
Information about how to use data to assist the work of community organisations
A basic statistical primer
An annotated list of available data. (DVC, 2004b)
The Committee regards this work as valuable, and believes it should be continued. The Committee also notes that the DVC has been proactive in drawing together lists of data from outside of the Victorian State Government, making the resource richer to the public.
However, the Committee finds that, just as the Government has responded to criticisms about the over-use of commercial-in-confidence agreements to shelter documents from FoI applications (see Current Performance of the Freedom of Information Act in Part II), care needs to be taken around the commercialisation of Government data sets, where this may preclude oversight and re-analysis.
The primary purpose of the activities of the Government of Victoria is to ensure the public interest, and issues of commercialisation and cost-recovery must be secondary to this value. By nature, re-analysis demands the release of primary data, in its raw form, supported by explanatory information about the way the data was collected.
While some may argue that all data sets should be readily available, the Committee notes that there are instances in which this is not warranted, particularly:
Where the data contains information that directly or indirectly can allow for the identification of individuals (a violation of the Information Privacy Act)
Where the collection of the data needs specific interpretation and clarification in order to be meaningful (essentially where the data needs significant "cleaning" or is of poor quality).
The Committee accepts that not all data collected by the VPS meets high standards of usability or accuracy, and that this is often a function of the relative needs of the organisation at the time of collection (appropriate time and cost investment based on need).
The Committee rejects the proposition, however, that data that can be released to facilitate contestable policy analysis should be withheld due to issues of cost recovery. It is the view of the Committee that this can be used to limit democratic debate, but also reflects three unsound propositions:
That the data is owned by the State of Victoria to the exclusion of its citizens
The State is the only source of analysis and expertise
No mechanism exists to ensure that data can be released for non-commercial purposes and still ensure resale of the data.
While the first and second points reflect a general principle of the collective ownership of the State by its citizens, the rejection of the final point is based on developments in the area of intellectual property licensing. Specifically, the Committee notes the recent development of localised versions of the "Creative Commons" licensing system. This approach allows for the release of intellectual property, such as data, under a flexible licensing system which protects the commercial value of the resource.
This approach of intellectual property, initially conceived and developed in the United States, allows content distributors to:
Determine and publish a Creative Commons licensing agreement, from a template that meets their needs
Provide a range of different aspects to the licence which permit limited free distribution and retain rights to commercial distribution.
This has been localised, and is undergoing further work by a project team led by the Deputy Vice Chancellor, Tom Cochrane and Law School Professor Brian Fitzgerald, of the Queensland University of Technology, and Mr Ian Oi of Blake Dawson Waldron Lawyers.
The Victorian Public Sector should consider adopting approaches to licensing data held that have traditionally been used for commercial purposes on a cost-recovery basis, whereby the release of this data is necessary to provide for contestable policy consultation and upon request. Where possible, alternative licensing schemes, such as a localised version of the Creative Commons licensing scheme, should be applied to data release to allow for public interest use, and should retain the commercial return.
Fourth, the effective use of consultation and participation can also challenge our assumptions about what information or "knowledge" is relevant to the decision-making process. In essence, consideration of consultation techniques and public engagement can open a number of epistemological questions:
What types of evidence are relevant to the process of policy making (what information do we capture)?
Do some data-collection methods have greater value than others (either due to the form of collection, such as qualitative versus quantitative methods, or because of other factors (such as representative sampling))?
This point recognises a clear divergence in the way governments have conceptualised and undertaken the processes of policy making. Following industrialisation, with the growth of the scientific method in the physical and social sciences, there was a tendency in developed nations around the world to view government policy making and program implementation as the task of experts, undertaken using the best scientific methods of the day: clear data collection, classification, analysis and rational selection from an array of alternatives. For many, both outside and inside of government, this scientific and rationalist approach represents the only appropriate way in which public decision making should be undertaken.
While the Committee does not underplay the importance of clear, theory driven and evidence-led processes of policy development and implementation (Scottish Executive, 2003:6), it also recognises that alternative forms and sources of information can provide balance to strictly quantitative and statistical methods. A rigid focus on scientific methods and behaviourist approaches to public policy has sometimes been considered a means by which the public has been shut out of meaningful dialogue with governments, which have favoured a technocratic class of policy advisors.
The shift to consultation and participation, in one respect, is a response to these perceptions, but more importantly is also a means by which government has recognised that different types of knowledge should be available to policy makers (Siedman, 1977:415): experiential knowledge, for example, represents a type of information and policy related knowledge that is only recently becoming recognised as an important input into policy making processes.
In making this positive observation, however, the Committee is cognisant of the difficulties that our expanding understanding of relevant policy-related information can have. These difficulties relate to questions surrounding how policy officers balance different types of knowledge when making recommendations and decisions. The problems of balancing statistical findings (which may be based on broad, but shallow data-collection processes) versus qualitative (textual) case studies will continue to remain a problem across government.
It is the view of the Committee that ICTs will have positive and negative impacts on the resolution of some of these concerns. In negative terms, the Committee notes that ICTs will exacerbate epistemological difficulties through an expansion of the amount of, format of and type of information and knowledge collected by policy makers. While on the positive side, the collection of information in digital format will allow for the increased use of indexing and analysis tools by policy officers as a means to manage information overload.
Thus, the Committee wishes to stress two factors in considering the implementation of consultation and participation methodologies, particularly where new technologies are involved. These factors are that:
There is a risk that policy makers, because of its ease of access, representation, summation, indexing and analysis, will subtly favour consultation and participation collected via ICTs
The Government of Victoria, through professional, systematic and reflective management of consultation and participation processes, needs to be explicit about the equal treatment of information and evidence provided across media forms.
As indicated in the discussion above, the Committee does not regard the use of ICTs in consultation and participation processes as a separate class or type of democratic activity. It is the view of the Committee that the question of how new technology may assist the Government in Victoria in extending its use of consultation and participation methodologies is not one of recommending a single, specific approach for adoption, but noting that consultations processes are able to incorporate an ICT component.
There are three general models for the incorporation of ICTs in consultation and participation:
As a supporting information channel: where information about a consultation process is provided using ICTs, such as a consultation website which includes background information, or the provision of consultation materials via email or other means
ICT support in a physical location: where technology is employed within a physical meeting to support the consultation process. This may be as simple as the provision of audio-visual equipment, or computers for co-authorship of outcomes documents, but also may include the use of networked computers to "scale up" the scope of a consultation process. One example of the latter approach is the use of networked computers and electronic handheld voting devices by the America Speaks group. During the delegation to the United States, the Committee observed how the organisation had conducted "town hall" meetings using these technologies to run interactive meetings with as many as five thousand citizens. In this example, the use of technology was specifically designed to increase the representative nature of the consultative process and was employed in policy areas where "visioning" was undertaken (the redevelopment of the World Trade Centre; Rosegrant, 2003) or significant political consensus required (for a new city plan for Washington aimed at battling civic decline; Potapchuk, 2002)
Online consultation and participation: where the interactive and participative elements of the consultation process are entirely or partially undertaken online. Many examples of this approach exist, but the most frequently cited would be the use of email mailing lists for policy discussions, use of video streaming (webcasting) for "online meetings", or the creation of online equivalents to small group meeting methodologies (such as focus groups or citizen juries).
It should be emphasised here that the Committee prefers to consider the issue of the use of new technologies in consultation and participation processes as more than simply the introduction of online consultation; for example, the use of the Internet to consult with the public. It is clear from the experiences observed by the Committee, and the delegation of the 54th Parliament, that while significant effort has been put into using the Internet in this way, ICTs have a much broader role in civic engagement.
One of the difficulties with the introduction of ICTs in consultation and participation around the world has been an over-emphasis on the use of technology as an end in itself, and under-emphasis on the relationship between technology and clear, demonstrable policy making outcomes (OECD, 2003). This has tended to result in online-consultation processes that were not well planned, leading to failures that caused governments to shy away from further experimentation and innovation.
The Committee notes that governments are not the only victims of excessively high expectations of the benefits offered by new technology. The so-called "dot.com" boom and crash that occurred in the private sector in the late 1990s demonstrates how the impacts of a disruptive technology like the Internet take time to develop and be understood. We are still in the midst of this process of change. As Hill (2003) observed:
It is the view of the Committee that the use of ICTs provides clear benefits to the public, through opening up new channels for participation. In addition, there are examples whereby effective participation can only be delivered through the use of alternative technologies rather than face-to-face meetings or conventional correspondence.
The advantages of using new technologies in consultation and participation processes can be found in:
Areas of access to consultation and participation processes:
Participation by people with limited time or mobility, or who live in remote locations
Non-business hours participation (for asynchronous processes)
Participation by those with a disability
Changing the environment of consultation:
Benefits of informality, where appropriate
Options for anonymity
Reaching members of the community with high information literacy but low levels of participation in formal political processes, such as young people
Exploring new avenues for participation:
The facilitation of citizen-to-citizen interaction and debate
Capacity to increase the numerical level of predication in more complex consultation processes
Introduction of polling and voting practices
Combining processes that focus on individual contributions as well as contributions from peak bodies or community organisations
Applying effective and enhanced information management to consultation:
Provision of greater levels of information, or information in multimedia formats, in a way that allows participants to determine how much or how little information they require
Ability for participants to be provided with information about the results of the process in a direct and expedient manner
The ability to digitally capture information provided, and apply analytical or indexing software to the contributions.
What is clear is that the nature of the various technologies employed will have significant advantages and limitations, depending upon the context of the consultation and participation process being considered.
It is the Committee's view that the State of Victoria (State and Local Government) has already demonstrated an interest in the role of ICTs in the consultation and participation process, and has made a number of positive steps in applying the technology.
It should be noted that, at this time, the emphasis has been on online consultation processes with some emerging work in citizen-to-citizen interaction, and thus the Committee notes a considerable scope for expanding and developing our understanding and experience across a range of approaches to ICT integration in civic participation and participatory democratic approaches.
During the course of the Inquiry, the Committee observed a number of online consultation processes in this State, either first hand, or through evidence provided in hearings and submissions.
The Committee notes a number of relevant activities:
The establishment of the "Have Your Say" pilot online consultation project by the Department of Premier and Cabinet
An online consultation run by the Office of Women's Policy on workplace supports for working mothers
The online consultation registration system by Parks Victoria for development of Box-Ironbark Park management plans
A number of ongoing consultation activities within the Department for Human Services
The Same Sex Relationships Electronic Discussion Group run by the Department of Premier and Cabinet
Incorporation of the capacity for intra-Government and stakeholder consultation through the redevelopment of shared workspaces on the State's Public Sector intranet, "Central Station".
In the Local Government arena, the Committee observed a number of activities to date, including:
Wellington Shire allows citizens to ask questions of its elected representatives via the Internet during webcast council meetings
Darebin City Council has recently concluded, and will continue to deploy, a structured online discussion forum for residents
Moreland City Council and Latrobe City maintain an e-discussion forum for residents to talk about a wide range of local issues
Ballarat City Council solicited email input into its visioning document (Kowalski, 2002:2).
Given the advantages of employing ICTs in consultation and participation processes, it is the Committee's view that the Victorian Government should foster the use of new technologies. This is a corresponding finding to that provided for the Parliament of Victoria, as discussed in Part IV.
Thus, the Committee recommends that all Government and Parliamentary processes be developed with explicit consideration of the role ICTs may have in their conduct. This does not require all Government consultation and participation processes to include ICTs; however, the Committee is keen to ensure that the design of any process of civic engagement in Victoria gives reasoned consideration to the full range of communications and participation tools available. Evidence from Canada has indicated that, with appropriate recognition of the importance of ICTs as an additional channel for government-to-citizen interaction, there is a high level of interest expressed by public managers in the use of these tools (Public Works and Services Canada, 2003:4).
This recommendation, therefore, is an effort to mainstream the use of ICTs in our democratic processes. With increasing numbers of Victorians using ICTs in their day-to-day business and personal dealings, it is inevitable that the public will expect, and the Government will respond to, public demand to access conventional consultations online.
All Parliamentary and Government consultation and participation processes should include, where feasible and appropriate to the target audience, the use of ICTs.
ICT-enabled consultation and participation processes should be developed in conjunction with key stakeholder groups and academic institutions in order to develop relevant frameworks for participation and evaluation.
These processes should be undertaken to provide an evaluation of effectiveness, both qualitative and quantitative.
With regards to online consultation:
Any consultation undertaken without an online component should include in the methodology of the final report of the consultation, a statement outlining the reasons against undertaking an element of the consultation online.
The inclusion of online participation should not be limited to restrict participation in offline forms.
Where possible, online consultation should allow citizen-to-citizen communication, moderated only to prevent incidence of defamation or legal risk to the State or Parliament of Victoria.
In making this recommendation, the Committee wishes to ensure that two caveats are emphasised.
First, the Committee wishes to reiterate that the adoption of ICTs in consultation and participation activities of the Government and Parliament of Victoria should be undertaken with consideration to Recommendation 4, in Part I - that ICT-enabled consultation cannot, nor should be, considered a low-cost means of replacing conventional consultation processes. The use of ICTs in Government consultation and participation processes must be seen as an adjunct, or additional, communication channel only, except in those very rare cases in which the entire population relevant to the consultation processes is online and is willing to engage with Government online.
It is important to reiterate this final point: even where a population may consist of people who all have access to new technologies, this should not be interpreted as meaning that the group necessarily wishes to communicate and participate with Government using these technologies. Democracy, a system of choice, needs to recognise that people have their own preferences about how, when and where they participate.
Second, even where a Government consultation process is not specifically transacted online (e.g. that the consultation does not include an interactive aspect using ICTs), the Committee recognises that the online provision of printed material associated with the consultation processes in advance of any meeting or forum must be considered best practice with respect to allowing those members of the vision-impaired community to access the content using assistive technologies (NILS, Submission No. a3, p. 10). In addition, where the consultation process is likely to include groups whose primary language is not English, this also affords participants time to secure translation or interpretation of the material, if the material is not automatically provided as a translation or translated summation.
In making Recommendation 60, the Committee notes that the Victorian Government and Parliament may interpret this as a requirement for the specific development of an array of new tools. This is not necessarily the case, though areas of investment will be needed.
While there are occasions in which the appropriate consultation methodology necessitates the creation or modification of a specific tool, existing technologies and website functionality can be employed to provide for consultation using ICTs. Thus, any move into online or ICT-enabled consultation and participation by the VPS should be undertaken with reference to the existing capabilities of the department's or agency's technology base, and with reference to technologies that can be shared across Government.
In addition, departments or agencies interested in engaging stakeholders online need to be aware of pre-existing online communities, either those established through projects like the MC2 initiative of VICNET or others that have developed independently of Government (Curtain, 2004). Many professional associations maintain extensive online mailing lists, and there is a range of independent discussion lists and forums that focus on general political and policy discussion (e.g., the Yahoo! Group maintained by On Line Opinion) or are specifically focused on Victoria (for example, insidepolitics.com.au contains a mailing list just for Victorian political discussion).
Any consultation approach should, as a matter of course, identify key stakeholders and representative organisations with insight or relevance to the policy issue under consideration. The proliferation of email list systems in the community means that the Victorian Government is able to form partnerships with community organisations to use their resources and subscription base as part of the consultation process, or as a passive source for information gathering.
The advantages of these practical approaches to toolset development have been recognised by some governments and civil society groups engaged with technology. Speaking from the perspective of a policy practitioner in government, Ms Elizabeth Richard (2000) of the Canadian Federal Government observed that governments and Members of parliament need to be more active in utilising the existing areas of policy discussion and debate, even if this is simply for environmental scanning purposes:
While from the corresponding position outside of government, Mr Michael Weiksner and Mr Scott Reents, founders of the e-the-people discussion board, explained to the Committee's delegation to North America how greater involvement in their civil society conversation space would be a good way to support and encourage these types of "online agora" civic spaces. It was their view that the involvement of public officials and policy makers would be beneficial to both parties, increasing the "pull" of these sites to members of the public, and providing the government and decision makers a ready source of interlocutors.
More specifically, an approach to online consultation that saw the incorporation of existing third-party organisations as a source of expertise and/or participants, has more direct benefits in improving consultation outcomes. This view was made strongly by Dr Liss Jeffrey of the byDesign eLab group during the Committee's visit to Canada. Dr Jeffrey, whose organisation had provided expertise in the delivery of the Canadian Foreign Policy Dialogue, observed that departments or agencies interested in online consultation should consider the following questions:
Are there organisations in the community that currently maintain a list or system that could be used for online consultation by government?
Are there organisations in the community that possess special expertise with regard to the moderation of consultations with the specific target group (e.g. an organisation with specific cultural experience, or with credibility to act as an impartial broker in highly charged areas of political debate)?
Does a highly structured process need to be undertaken at all? Can government solicit and receive the type of information it needs through engagement with an existing online community?
Clearly these are questions of "fit". Where significantly representative and vibrant online communities already exist, incorporation of a government-sponsored consultation process in these groups would provide benefits in:
Reducing the costs and time associated with the recruitment of participants
Reducing the costs and time associated with the development or modification of tool sets for the conduct of online consultations
Faster access to information: the need to establish and formalise ground rules for participation of the participants is negated - these groups will have their own cultural norms and rules
Higher levels of trust through engagement of the community in its own online "space"
Stimulation of non-government civic activity.
Online consultations should be undertaken using the best available technology and techniques available, and the design of these systems should be undertaken in conjunction with key stakeholder groups. Where appropriate, these consultation and participation processes should be based within community organisations.
In order for ICTs to find their appropriate place in the array of public engagement tools available to policy makers, the Victorian Parliament and Government cannot shy away from promoting these activities. While some governments have taken bold moves into online consultation, establishing whole-of-government portals and consultations tools (such as seen in the United Kingdom, and Queensland), the Committee notes that many governments have experimented very tentatively with the interactive power of ICTs.
It is the view of the Committee that experimentation is important, but that timidity on the part of the Victorian Government will not assist in developing an effective understanding of how these tools and processes can assist governments to produce better policy outcomes for citizens using ICTs. Stephen Clift (2004), a well-known advocate of online consultation, has been critical of the way in which online consultation has been implemented in some jurisdictions. His view is that some governments:
Clearly, there are risks in implementing ICTs in consultation and participation processes, which include:
Failure to produce meaningful outcomes (lack of informative response received given the effort expended by the department, agency or parliament)
Malicious attacks or failure of the technology to provide reliable outcomes
Deliberate misuse by unrepresentative elements of the community, to promote specific political purposes, or to "spoil" the outcome
Lack of interest by the public.
Yet, the Committee contends that these risks are inherent in all consultation processes. Given a degree of "hype" often associated with new technologies, the implementation of ICTs in consultation and participation activities does need to be realistic, with a clear statement of the expected benefits as well as the limits of the process.
In evidence presented on behalf of the City of Darebin on its experience with online consultation, Mr P. Shanahan observed that:
Considerable investment had been required in establishing the forum, particularly around the moderation of messages and recruitment of participants
Careful consultation between the list managers and policy officers was required to ensure that valuable information could be solicited from participants
The purpose of the list needed to be closely aligned with decision-making processes in order for participants to see a benefit from their participation. (Submission No. a12, pp. 5-7)
Indeed, the Committee's experience with its own discussion list was that considerable investment was required to maintain participation in the list, and when staff were not actively encouraging participation, it tended to fall away.
Thus, it is the view of the Committee that the process of stimulating the use of ICTs in citizen engagement will need a demonstrable political commitment in the medium term. While the recommendations above aim to promote online engagement methods across the whole of the public sector, as a "standard" tool considered in the development of any consultation process, the Committee notes that the introduction of a number of demonstration projects should be undertaken to:
Increase awareness across the VPS of the Government's commitment to extending citizen participation using ICTs
Engage the community and raise its expectations of Government to include the appropriate use of ICTs in consultation and participation
Begin a process of general awareness raising as to the capacity of ICT-enabled consultation to enhance public participation, but not provide a panacea to the inevitable reality that politics and policy making is about balancing the views of sometimes-conflicting parties, and that tensions are an inevitable part of the political process.
The State Government of Victoria should identify and develop a number of policy areas where demonstration online-consultation projects can be conducted.
Given the focus of Recommendation 60 on stimulating the use of ICTs for consultation and participation processes across Government, and the desire of the Committee - through Recommendation 62 - to encourage visible political commitment and action in this area, the Committee considered three additional questions:
To what extent has the VPS the capacity to effectively develop and implement the use of ICTs in consultation and participation practices?
What structural mechanisms will be required to develop and sustain a vibrant and innovative community of practice in and around the area of ICT-enabled citizen engagement?
Are there eternal drivers that present risks in allowing the current process of learning to continue?
The Committee examined whether activities such as online consultation require a specific skill set, and best-practice models to develop and sustain these skills. During its delegation to Canada, the Committee sought to gather specific evidence on this issue, since this country's national government has taken a number of policy decisions on the development of a structural response to the skills issue. The Committee was also particularly attentive to the evidence collected by the delegation to the United Kingdom of the 54th Parliament.
The Committee is cognisant of an array of skills within the VPS. These include:
The experience gathered during projects conducted to date, as noted in the discussion above
Expertise in the area of online consultation and moderation practices has been developed by VICNET for past projects, and the non-profit organisation Infoxchange Australia is currently providing moderation skills for the DHS
The Inquiry received submissions from two private-sector organisations (Big Pulses and JustSys) which had developed specific online consultation and participation administration and moderation skills (Submission Nos. a8 and a9)
Online Dispute Resolution (ODR) training has been provided to the staff of Consumer Affairs Victoria by the International Conflict Resolution Centre of The University of Melbourne, which offers a one-day course on this subject (Submission No. a24, p. 2).
While the Committee considers these activities to be a positive step towards the development of a community of practice in and around the area of ICT-enabled consultation and participation, the Committee did receive a number of submissions that noted the particular difficulties of some Government organisations in developing and sustaining a minimum skill set required to introduce and maintain these services effectively.
The Cities of Port Phillip (Submission No. a22, pp. 4-5) and Darebin (Submission No. a12, p. 9) and the VLGA (Minutes of Evidence, 17/02/05, p. 236) were very clear in their statements about the need for additional assistance from the State Government in the development and maintenance of skills for online moderation and consultation purposes. In making these statements, none of the organisations stated a preference for a single, specific training or education package; emphasising instead the need for areas of expertise and experimentation to be identified and linkages formed between, and across, levels of Government.
The Committee gave particular attention to the risk of defamation arising from participation in online discussion forums. While the Committee does not wish to over-emphasise this aspect of risk - given the comparative infrequency that the publication of content online has resulted in legal action when compared with the explosive growth of participatory publishing - the Committee notes that the risk of defamation does act as a strong driver for clear policy processes surrounding the use of ICTs in consultation and participation, as well as the development of an appropriately informed and skilled set of moderators for this process.
It is the view of the Committee that the general issue of defamation remains largely outside of the terms of reference of the Inquiry. While it notes a number of issues associated with the laws regarding defamation in Victoria, the primary concern of the Committee with regard to defamation is not to challenge the existing intention of the legislation, but to concern itself where defamation law:
Has direct relevance to the matters included in the Terms of Reference
Requires specific action on the part of the Victorian Government in the administration of electronic democracy initiatives.
In Victoria, the specific prohibition for defamation ("defamatory libel") is currently included in the Wrongs Act 1958, though a stand-alone act is currently being developed.
The Committee observes and notes that the Federal Attorney-General has proposed a national law covering defamation, and is asserting the capacity of the Commonwealth to assume responsibility for defamation based on a combination of powers granted under the Constitution of Australia (Corporations, Telecommunications and Territories; Hull, 2004).
The status of this approach remains unclear at the time of writing (Gibbs, 2004), and the States and Territories have, as at the November 2004 meeting of the Standing Committee of Attorneys-General (SCAG), adopted a uniform piece of model legislation for adoption by each State and Territory.
The outcome of this decision will be the introduction of uniform, consistent legislation across Australia. However, it is not clear whether the Federal Government will press ahead with its proposal - a step that would introduce new complexity (Grattan, 2005) to the administration of this law. Thus, while South Australia has already introduced this legislation (The Advertiser, 2005) and it is accepted that all the remaining jurisdictions will follow suit by January of 2006, there is potential for the benefits of a uniform code for defamation to be undermined through the introduction of a Federal law. This may give rise to:
Legal challenge and uncertainty
Different treatment for some matters, where the Federal jurisdiction does not apply.
The model laws provide a number of certainties, but is still situated within the common law (for a full draft of the model legislation [opens in a new window). The law:
Recognises truth as a defence (there is no need to add a public-interest test; truth alone is an adequate defence against defamation), this is a national acceptance of the legal practice that Victorians have enjoyed for many years
No longer provides the capacity for the dead to be defamed (preventing suits from family members)
Does not provide for corporations to sue for defamation (as was the case in New South Wales)
Requires that actions must be initiated within one year (placing a cap on the time period, but also possibly hastening litigation)
Provides a cap for damages (AAP, 2004).
Given the developments in this area, this discussion focuses on:
The applicability of concerns regarding defamation to ICT-enabled consultation and participation processes
Cross-jurisdictional issues
Action to be taken by the Victorian Government in the introduction and management of electronic democracy.
The Committee is of the opinion that issues of defamation should not be used as a barrier to progressing the implementation of ICTs in public participation and consultation in Victoria[33]. The preferred approach of the Committee would be for the Federal Government to acknowledge the action taken by the States and Territories in normalising the law and to indicate that no further action will be taken by the Federal Government that would re-instate uncertainty.
One of the common misconceptions about the Internet and ICTs in a general sense, is that ICTs represent some form of pure, ungoverned space, with infinite personal freedoms. While the Internet was very loosely regulated in its early period of growth, and governments have found considerable difficulties in applying local laws to what is an international network of communication and communities (Chen, 2004b), the Internet is not, and should not be, beyond the scope of the law.
With specific regard to consultation and participation via ICTs, the issue of censorship and editorial control over published content is an area in which the Committee recognises there lies considerable tensions between those who desire complete freedom of expression and absence of regulation online (either because of a negative perception of the motivation of governments or due to a philosophical view that a free interchange online can lead to a form of mutualism and enlightened civic discourse) and those who view aspects of communication online as undermining the social norms of behaviour (the distribution of mistruths, attacks on character and the use of anonymity for this purpose).
The first perspective often perceives the encouragement of ICT use in civic participation processes by Government as an activity by means of which the State becomes the provider of the technology and "space" in the first instance, and then moves into the role of passive observer of the debate it fosters. This view is captured by the submission of the VLGA. The VLGA, as an organisation that focuses on fostering and sustaining local democracy, stated that:
This point of view focuses on concerns that moderation may, at its most benign, act as a "brake" on the interchange of ideas between citizens or, at its more insidious, represent an act of censorship by the department or agency overseeing the process. This can be a particularly crippling form of censorship which, unlike other forms of censorship, where the absence of the message is often noted and can thus be "overseen" (e.g. the work of the Australian Film and Literature Classification Board in the regulation of films often prompts considerable public debate around decisions made by the Board), the control of online messages may be difficult to observe externally (Penfold, 2001).
It is the view of the Committee that, while this concern is valid and should be addressed through professional moderation undertaken with a clear statement of policy (see below), moderation is a necessary element of online consultation processes of Government, because of the legal risks associated with defamation and the implications for the State and Parliament of Victoria, as publishers of this content.
Unlike some of the online communities established in the United States of America, where the First Amendment of that country's constitution favours the right of free speech over the protection of individual reputation, the law in Victoria and the model Bill maintains a strict prohibition on defamatory libel. Libel is the publication of a falsehood that damages or undermines another's reputation. The laws against libel have been designed to protect individuals against personal loss through malicious attacks upon their reputation. A common definition is that cited in the case of Parmiter v. Coupland (1840), which defines defamation as a communication:
The action taken through litigation focuses on a recovery of damages associated with the loss of reputation and esteem of those associated with the plaintiff, although some jurisdictions provide for mitigation through retractions or apologies and the possibility of imprisonment.
It can be argued that, without oversight of the content of messages, no legal liability for the transmission of defamation material exists. This argument, that the owner of the service is not acting as a "publisher", but is merely an "innocent disseminator", was upheld in the case of Laurence Godfrey v. Demon Internet Limited (1999) to the extent that the service provider becomes a publisher if informed of the presence of defamatory content and does not act to remove it (Akdeniz, 1999). Locally, the Melbourne PC Users Group settled a similar case by Godfrey in 1998 - a serial litigant - based on disparaging remarks against him published on one of their email discussion lists.
The Committee notes, however, that this arrangement cannot be made for Government-sponsored online consultation and participation processes, as the content of the discussion must be read by an officer or appointed facilitator of the hosting organisation in order to compile the views of participants.
The Committee is of the opinion that there is no demonstrable need for the State Government of Victoria to provide and service general discussion lists or online communities for Victorian citizens beyond those it currently maintains through the MC2 service. This view is based on the recognition that, as in other jurisdictions around the world, a range of general discussion forums has been developed outside of Government, often with great creativity and innovation. The prevalence of these existing services and the availability of relatively simple tools to establish them, counter-indicate the need for Government action in this area. In addition, the Committee reiterates and underscores the view that consultation and participation processes run by the Parliament and Government of Victoria should be tied to specific policy outcomes, and thus need to be relatively instrumental in meeting specific objectives.
While this view could be considered as dismissing the issue of defamation, the Committee notes that the issue of jurisdiction and defamation has been of particular interest in the community in recent years, largely due to the impact of new technology on what remains, at this time, State-based laws. As Handford (1983) observes, the introduction of mass media has highlighted the fact that each Australian jurisdiction maintains its own set of laws covering defamation, and that the differences between these laws present particular problems for, particularly, the publishers of printed matter distributed nationally and the integration of television and radio into nationally syndicated networks (1983:454).
While this legal environment has been recognised by commercial and public media companies as a source of considerable difficulty in the preparation of material for distribution through increasingly integrated distribution systems, the development of the Internet as an international medium of public and private communication in which non-professional publishers are active has added new difficulties to the regulation of defamation. In particular, issues exist where:
The publisher of online content is subject to litigation in a foreign jurisdiction, possibly one of which they are unaware of the laws associated with defamation
General members of the public find themselves subject to defamation.
This problem was recently highlighted and to some extent resolved, in Dow Jones & Company Inc. v. Gutnick (2002). The basis of this case was an appeal made to the High Court of Australia against a Victorian Supreme Court decision that a case of defamation against the Dow Jones Company could be heard in Victoria. Dow Jones was seeking to have the decision overturned, arguing that the relevant jurisdiction was New Jersey in the United States, where the material was loaded onto the company's webserver.
In this case, the High Court determined that, with regard to the publication of material online, Australian State jurisdictions applied where it could be shown that the material published was accessed (downloaded) from that jurisdiction. In making its finding, the High Court considered that defamation, as a legal protection against damages received to one's reputation, by nature must occur at the point of consumption, and that Dow Jones, in arguing the publication occurred where its Internet hardware was located, was not presenting a reasonable claim under Australian law. In this case, therefore, the Court considered that the jurisdiction of Victoria was appropriate for the action to proceed.
The importance of the case is noted by Thompson and van der Toorn (2003), who observed that Dow Jones, in attempting to argue against the jurisdiction of the Victorian Courts in this matter, were seeking:
The "single publication" concept is a rule found in some legal jurisdictions that makes the distribution of a single communication, even across different media forms, one act of publication. This is the case in over twenty states in the United States of America. In Australia, however, defamation is deemed to be "bilateral"; that is, that a statement is not published until one or more third parties consumes it (Transnational Law Associates, 2002). This opens the prospect for material to be considered "published" in a range of jurisdictions, presenting the prospect for multiple actions of defamation, or the strategic selection of legal jurisdiction based on the plaintiff's assessment of likelihood of a successful outcome.
The SCAG Working Group of State and Territory Officers explicitly considered this issue in its development of the model code, and determined that the existing approach is acceptable:
While much has been made of the impact of Dow Jones & Company Inc. v. Gutnick on online publishing, the Committee notes a number of aspects that limit the reach of the precedent in this case:
Legal action is unlikely in jurisdictions in which the defendant has neither residence nor assets, due to the limited likelihood that damages could be pursued
Mr Gutnick's choice of Victoria for the suit reflected the jurisdiction in which his social and professional acquaintances were concentrated; it was not chosen because it necessarily had the most favourable defamation laws for the plaintiff per se
The publication at the heart of the suit was one where the publisher had control over, and knowledge of, the location of readers (it required paid subscription to access it, which included location information of subscribers). (Beyer, 2004)
However, the Committee does note that the implication of Dow Jones & Company Inc. v. Gutnick means that material generated through online consultation and participation in Victoria could be subject to action in other jurisdictions - if the plaintiff can show that it was accessed in that jurisdiction. In the case of online consultation and participation processes, the State of Victoria would be considered the publisher in the context of defamation actions that arise as a result of the publication of a message, document or statement online via a citizen of Victoria.
While this and other cases have stimulated a discussion of the effectiveness of technological systems for filtering, particularly where specific jurisdictions have general or specific laws that would prohibit the publication of certain material (Macgregor and Vincent, 2003), the effectiveness of these technologies is limited to cases in which:
The filtering is based on a closed subscription service that maintains an accurate record of the location of participants (as was the case in Dow Jones & Company Inc. v. Gutnick)
Filtering is most effective between nations (focusing on the second-level domain address of recipients) and tends to be less effective within nations, requiring a significant investment in capacity at the ISP or network level to develop. The capacity for the State of Victoria, which is not the telecommunications regulator in Australia, to achieve this would be negligible.
In addition, the Committee is of the opinion that it is inappropriate for a State Government to systematically attempt to filter Internet content in order to prevent access to material stored in Victoria by a user in another State or Territory within Australia. The Committee feels this is somewhat counter-productive to the federalist project in Australia. Therefore, the Committee finds that the relevance of this filtering to protect the legal interest of the State and citizens of Victoria is very limited, particularly where the most likely risk of litigation would be either local (Victorian), or domestic. While the eventual design of some of these participation systems may entail the "closure" of content, by nature many will be open publishing environments, readable across the network.
Given the increasing number of citizens engaging in online publication, the Committee is of the opinion that the normalisation of laws across Australia is necessary, and the Committee commends the work undertaken by SCAG to resolve ambiguity in this area. While this will add clarity to the publishing decisions of people online, it does not, however mitigate the need for clear moderation to protect the State and its citizens against defamation charges - local or domestic.
In presenting the evidence above, the Committee also wishes to note that the role of moderation need not be simply considered one of risk mitigation. Moderation, if skilfully and reflectively undertaken in a transparent manner, can improve consultation outcomes for all participating parties. This finding has relevance across a number of levels.
At the most fundamental level, the submission of the National Forum - an organisation that provides a moderated public discussion service - reflects upon the role of moderation in increasing the public's awareness of issues of defamation and the implications of publication online. In his submission on behalf of the National Forum, Mr G. Young notes that:
In one way this represents a necessary educative process borne out of the application of poor metaphors. Many of the initial proponents of electronic discussion lists for civic dialogue explicitly used the imagery of the "agora", or romanticised town square of ancient Greek democracy. This likened discussions online to that of small conversations between friends or acquaintances. While this is partially true, in that these discussions can be causal, personal interactions between people who come to know their interlocutors well, the reality for many of these lists is that they are either archived on publicly accessible servers, or have much larger membership than the active participants assume (including non-participating readers or subscribers, often referred to as "lurkers"). The more appropriate metaphor is that of an informal public meeting, at which the conversation is transcribed and published for all to see.
At a deeper level, however, the role of moderation can be seen as positive and important for Government to get the most from the consultation process and for users to understand the rules of participation. In his submission on the Same Sex Relationships electronic discussion list, Mr J. Buckley observed that:
This view is reiterated and expanded in evidence from the United Kingdom, where the Hansard Society has taken a leadership role in championing online consultation and providing training for moderators. In her report into pilot Parliamentary online consultations undertaken by the Society, Ms N. Hall (2001) summarised the argument, not for moderation, but for the degree to which the moderator takes an active and interventionist role in the flow of dialogue and debate. She provided a spectrum of moderation to consider in the development of the project methodology:
Keeps the consultation on-topic
Keeps the flow of the discussion going (by throwing in new questions or adding new ideas)
Can explain issues to users
Can set negative and positive tone levels for discussion (what is not acceptable, what would be useful)
Gives a friendly face (first point of contact if any difficulties arise)
Provides assurance that someone is paying attention
Difficult position on key issues - how far should the moderator defend the MPs' role?
When does encouraging people to get involved become too forceful?
Can interfere with the natural flow (if it is too regular it becomes overpowering and detracts from the interaction between participants)
Difficult to get the tone right - it is easy to patronise people. (2001:23-4)
Clearly, effective moderation requires alignment with the audience and subject matter. Consultation and participation processes held for a small, professional community may require minium oversight as a means of providing due diligence against litigation, as the group has established professional norms of behaviour.
On the other hand, for highly emotive issues, a stronger role may be required for the moderator, in order to ensure that the discussion does not degenerate into simple polarisation or abuse. The means by which the appropriate approach to moderation should be determined should include:
Discussions with stakeholders prior to the initiation of the processes
Determining "ground rules" with participants
Having a policy framework for moderators
Having an appropriate process for handling disputes between participants and moderators.
The Committee notes that the last is a particularly useful means by which to address concerns about censorship. The Brisbane City Council, which runs the Your City, Your Say discussion list, makes it a regular policy to:
Correspond with authors who post messages that are not acceptable, to attempt to negotiate an appropriate form of words that are not defamatory
Has an appeals process to a senior manager for disputes that are not resolved through negotiation.
Similarly, the United Kingdom Government, through the Office of the e-Envoy, has established a generic policy framework and guidelines for the use of discussion lists by departments and agencies (Office of the e-Envoy, 2002) which provide an overview of different approaches, with suggested management approaches and elements for acceptable use guidelines.
This perspective has also been recognised in Victoria. In evidence presented to the Committee on behalf of the City of Darebin, Mr P. Shanahan noted that this municipality's experience with moderation and participation of public employees on the discussion list triggered a need to review and reiterate the expectations of the Council in the dealings with the public (Submission No. a12, p. 7).
The Committee considers that a clear policy framework for staff undertaking these processes is necessary for the State of Victoria as a means to balance the concerns of:
Providing a structure which prevents the misuse of moderation to unnecessarily censor participants
Protects participants (and third parties) from the risk of defamation and litigation
Encourages effective consultation and participation processes through the facilitation of dialogue
Guidance of public servants in their role as moderator within the conditions laid down within Section 95 of the Constitution Act 1975 which relate to the appropriate conduct of public servants in areas of political discussion associated with their department or agency.
The moderation of online consultation and participation processes needs a clear policy statement for staff that provides direction as to:
The purpose of moderation
The degree of moderation appropriate, depending on circumstances
Policies or procedures for resolving disputes
The necessity for the policy statement to be available to participants.
Given the discussion above, and following a careful review of the evidence provided to the Committee and supported by the evidence collected during the 54th Parliament, the Committee is of the opinion that the integration of ICTs into consultation and participation processes, particularly the more interactive aspects of participation such as deliberative and direct democracy, requires the application of skills that are not strictly transferable from experience with conventional processes.
The development of consultation and participation processes that make effective use of new technology tends to require expertise in four areas:
Specific expertise in the nature of communication, interaction and deliberation in online environments (an understanding of the social processes that occur when people interact through new media)
Specific expertise with the consultation and participation methodology being applied
Specific knowledge of, and capacity to communicate, information associated with the specific policy issue under consideration
The technical skills required to develop, test and maintain the ICTs employed, combined with an ability to communicate technology related issues to users.
While there are individuals who may possess all four of these specific skill sets, it is the view of the Committee that the Victorian Government cannot expect these skills necessarily to exist in all of the departments and agencies that seek to introduce technology into their consultative and participatory civic engagement activities.
In addition, the Committee notes that the area of online consultation is subject to a number of difficulties and considerations that require specific expertise to manage effectively. Given the limited experience of the VPS to date, there is no reasonable basis upon which we can expect the skills to have established themselves across the State Government. The evidence from the Local Government arena is that the development of the skills is a process that is currently being conducted with a degree of trial and error. Yet, the area is one of considerable complexity. Mr A. Davies of the VLGA commented that:
In the Committee's discussions with the Online Consultation Technologies Centre of Expertise in Canada, Ms Elizabeth Richard pointed out some of the complexities associated with converting conventional offline consultation processes into the online environment. Ms Richard highlighted two specific issues:
The conduct of online consultation can sometime be undermined through poor communication between service providers and policy officers. Too often, policy officers bring technical staff into the consultation development process too late, and toolsets are either not appropriately configured to the needs of the consultation process or are applied without appropriate user testing and refinement
Online consultation required moderation of a different type to that employed for face-to-face moderation. Because online consultations are often limited to a textual environment, the processes of social interaction online can be very different to that experienced in a face-to-face meeting. Sometimes this is the result of anonymity of the process, which may result in uncivil conduct that discourages participation (Davis, 1999:177), or can result from issues of scale or social factors online (such as the loss of paralinguistic cues, facial expressions, that serve as an important part of the communication process in many meetings).
Similarly, discussions between Professor S. Coleman of the Hansard Society and the delegation of the 54th Parliament observed that:
Moderators are required for online forums.
E-democracy requires a connection with deliberative and civic skills.
Commercially available software is manifestly inadequate for e-democracy. (SARC, 2002b:8)
And appropriate planning and management of these processes was needed because of risks associated with:
Failing to manage expectations
Hijacking of the debate
Abuse by MPs: Some reportedly regard e-democracy as another tool to get re-elected; and/or as a way to corral debate on an issue. (SARC, 2002b:8)
It is clear that the characteristics of the different types of technologies that can be employed, combined with the range of consultation methodologies available, create different operating environments for consultation staff and facilitators. Different consultation methods examined by the Committee resolved these approaches in different ways. One of the earliest electronic consultation and discussion systems of note, Minnesota E-Democracy, resolved some of these problems through:
Focusing discussions at a local level
Including journalists as participants, to provide informed coverage of the event
Incorporating decision makers in the process
Limiting the number of messages posted to the discussion each day
Placing rules around appropriate behaviour (Dahlberg, 2001).
However, the Committee notes that the experience of governments around the world in the effective ways to develop and implement online consultations shows that a range of approaches can be effective, depending on the nature of the issues under consideration and the target audiences for the consultation process. Thus, the Committee observed a range of approaches to online consultation, including:
Un-moderated, wide-ranging discussions (the Canadian Foreign Policy Dialogue)
Projects designed to capturing personal stories (Northern Ireland Affairs Committee Inquiry into Hate Crime)
Deliberative decision-making (direct democracy) conferencing (the World Trade Centre redevelopment dialogue)
Structured, survey-like approaches (Victoria's Work Family Balance Consultation).
It is important that Victoria avoids the trap, identified in other jurisdictions that have invested heavily in online consultation development, of over-focusing on the rollout of new consultation websites with limited consideration to the significant lessons that still need to be learnt about the effective use of the online environment in consultation processes. As Ann Macintosh and Angus Whyte (2002) of the International Teledemocracy Centre (Scotland) noted:
The Committee warns that it is too early for a specific "best practice" manual, or single toolset, to be developed. Online consultations, when they are effective, are designed reflectively, in consultation between technologists, policy makers and stakeholders, and they use the tools appropriate for the target audience (electronic mail, online forms, instant messaging, SMS, etc.).
Online consultations should provide detailed reports on outcomes, including problems, to encourage learning across Government and the civil society.
Where possible and feasible, online consultation systems should be developed in open source, with the source code widely available to other Government agencies and civil society groups.
From the discussion and recommendations made above, it is clear that the range of electronic democracy activities and enablers relevant to fostering greater political participation, awareness and interest are broad. While the State of Queensland, for example, has adopted a specific, focused electronic democracy policy with key initiatives, the Committee considers that this approach is unsuited to the more amorphus and dynamic characteristics of technological change envisaged for Victoria, specifically given the mutability of the technologies in question (and applications yet to be developed) and the early state of research of the impacts of ICTs on political processes, institutions and public participation.
In addition, the Committee concedes that periodic Parliamentary inquiries into electronic democracy may have difficulty maintaining the pace of reform, and that a body tasked with oversight in this area would serve the State well by systematically collecting best-practice information, distributing it and fostering projects and initiatives. In addition, it may be necessary for Victoria to focus initial expertise in ICT-enabled engagement skills and moderation in a central agency, which can then act in an advisory or consulting capacity throughout the public sector.
As each political jurisdiction is different, due to different political institutions, laws, customs and norms, the Committee feels that local understanding and experimentation is required if electronic democracy is to evolve in an effective manner.
It is the Committee's view, therefore, that a centralised agency to oversee developments in electronic democracy practice in Victoria is necessary. During the delegation to the United Kingdom, Members of the 54th Parliament were impressed by the work of the Hansard Society in working with the national parliament to develop online consultation that included a strong component of evaluation. This evaluation was used to develop the proof of concept, refine practice and form the basis of online mediation training offered by the Society. This training has been employed by the Canadian federal public service, which has established a best-practice unit within the Department of Public Works and Government Services, to develop expertise and act in an advisory capacity to other agencies interested in applying online consultation methods in their work.
The Committee, therefore, recognises a need for a centralised clearinghouse for information on electronic democracy practices, but without the restrictive ambit of a top-down implementing agency or body. In the submission from the CCNR (Submission No. a5a, p. 8), it was recommended that an academic body, such as itself, would be best placed to provide the management for electronic democracy initiatives, particularly those focused on community based projects, with the advantage of freedom from "bureaucratic inertia" and onerous reporting requirements.
While the Committee recognises and accepts the advantages of a strong research basis, this approach was not deemed as optimal, since:
Any co-ordinating body would need to administer public funding to projects within and outside of Government; the placement of this agency outside of Government would not mitigate against due accountability procedures
Administrative expertise resides within the VPS
The short- to medium-term potential for electronic democracy initiatives is likely to be driven by the public sector (as per the experience in other Westminster democracies, as opposed to experience in the United States) and a Government department or agency has greater authority with respect to the performance of implementing agencies.
While the Queensland model positioned this role in the Department of the Premier and Cabinet for the initial introduction of that state's electronic democracy strategy, the Committee considers the Department for Victorian Communities is best placed to house this body in the Victorian context. While some may argue that a technology focused agency would be more appropriate (such as the OCIO or MMV), the Committee regarded the DVC as best suited, because:
Issues of community building and participation should be the primary focus of this agency, followed by technology concerns as facilitators of social objectives
The DVC has specific experience with, and is responsible for, facilitating action across Government and through community partnerships
The DVC reports to a large group of Ministers and can therefore expand its educational and service-delivery assistance widely.
An electronic democracy co-ordinating body should be established in the Department for Victorian Communities.
In identifying the DVC as the preferred location for the co-ordinating body, the Committee recognises the need for wide-scale participation and collaboration in this area (L. Stillman, Submission No. a5, p. 6); not simply to incorporate an external research component, but to develop and locate projects at the local level and in conjunction with local community groups and municipalities.
It has been demonstrated in the collaboration between the DHS and Swinburne University of Technology to establish the Wired High Rise project, that forming linkages between Government departments, academic institutions and private providers can be an effective means by which to develop community based information projects. This model appears worthy for emulation by this agency. In addition, good project co-ordination can provide for increased access to Federal Government research funding via the Australian Research Council (ARC) Linkage Grants scheme, with ARC funding already employed in Victoria to provide IT training and access to senior members of the community (through the Department of Veteran's Affairs' partnership with academics of the RMIT University of Technology).
The Committee therefore considers that this body should be specifically responsible for the development and maintenance of good linkages across the three levels of government (Federal, State and Local), between the community and academic communities, and incorporate, where possible, industry representation. This would develop and strengthen ad hoc networks of practitioners that have already begun to form around this area of practice (such as the practitioner networks maintained by the VLGA and the Australian Government Information Management Office (AGIMO); the City of Darebin City, Submission No. a12), and provide continuity and longevity to projects and the lessons gathered from them.
The Department for Victorian Communities, through the electronic democracy co-ordinating unit should:
Forge and maintain ongoing networks of practitioners and scholars, from across all levels of Government.
Develop a set of evolving benchmarks and performance measures for online consultation, to assist in establishing and advancing best practice.
Include seconded membership from across the public sector.
Be advised by industry and academic experts.
Have active participation from key Local Government organisations, either directly or through relevant peak bodies.
It is anticipated that the outcomes of the two recommendations above include the creation of general and specific resources for public managers and community groups regarding the facilitation of ICT-enabled consultation and participation processes and technologies. Through the work of the co-ordinating unit, the Committee recognises the need for the systemisation of evidence collected and resources available to new and experienced practitioners.
Across Australia, and internationally, governments have been attempting to underscore their commitment to consultation and participation innovation through the development of centralised policies and guidance documents for their public-sector organisations. Examples of this approach are:
The creation of the Western Australian Consulting Citizens: A Resource Guide by the Department of Premier and Cabinet (2002)
The Australian Capital Territory's 2001 Consultation Manual (Chief Minister's Department, 2001)
The United Kingdom has developed a Code of Practice on Consultation that provides advice, as well as a set of criteria that must be presented to the public as minimum standards for government consultations (Cabinet Office, 2004).
Victoria has not been negligent in this area, and the Committee has identified a range of materials generated at the whole-of-government level (A Guide to Electronic Discussion Groups, Department of Premier and Cabinet, 2002), sectoral (Best Value Victoria Community Consultation Guide, 2001; the Local Government Consultation and Engagement website), and various departmental, agency or audience-specific reference material (for example, A Guide to Community Consultation in Rural and Regional Communities; DHS, 2003).
The Committee therefore gave consideration to the need for a State-wide approach to community consultation and participation strategies, in line with practice in other jurisdictions. The advantages of a centralised manual and policy framework would be:
Announcement of a clear intention by the Government to pursue community consultation as an aspect of good governance practice
Clear direction to departments and agencies in order to minimum expectations of the Government as to the conduct of consultations
Ability to produce up-to-date and broad-ranging documentation that includes new or rarely used approaches (participatory forms of shared decision making, use of ICTs)
Elimination of redundancy in the development of departmental, agency or program-level guidance documentation.
However, the Committee also recognises problems with this approach. These include:
The need to develop documentation with application so wide as to limit its directive and instructional value
Lack of relevance to specific audiences or policy areas
Co-ordination and promotional costs of development and distribution of whole-of-government documents.
Thus, the Committee prefers the creation of a best practice reference guide for the VPS in electronic form, rather than a singular, centralised manual. This view is based on consideration of the range of consultation and participation methodologies available, the diverse nature of the public sector and its stakeholder base and the important recognition that considerable expertise exists in departments and agencies which have, and will continue to, conduct, review and refine consultation and participation processes (on- and offline) into the future.
Overall, therefore, it is the view of the Committee that a document that provides:
An overarching rationale for consultation and participation
Typological classification of methodologies
A review of technologies associated with consultation
A clear, reviewable process for implementation of effective consultations
A detailed description of consultation and participation methodologies, with case examples
A discussion forum for consultation managers and staff
Up-to-date links to existing resources of a specific nature for agencies to access and use
would provide a better resource for the VPS, Local Government and community-based organisations in Victoria. In effect, this would serve as a clearinghouse, with clear expectations that content would be co-created by all Government departments and agencies, and their partners in Local Government and the community sector.
The Committee notes that the Federal Government, through AGIMO, has already developed a number of "better practice checklists" that focus on issues around online consultation and information management, and that the Canadian Public Works and Government Services Department has established a very broad range of materials on this and related subjects. Thus, an inter-governmental approach to resource sharing and rationalisation would appear to be of value to the VPS and its wider stakeholder groups, to gain access to and distribute, the best material available internationally[34] - particularly those based on experiential learning.
The Committee considers the advantages of this approach would be:
The addition of a co-ordinating body to the systematic creation, indexing and discovery of practical, experience-based information about consultation and participation processes in Victoria
Distributing the creation and maintenance of this information across the VPS, Local Government and participating community groups
Stimulating a community of best practice in and around Government with regard to consultation and participation strategies and techniques generally.
The Department for Victorian Communities should establish a best-practice reference guide for public consultations in Victoria, with supporting documentation, references and check lists, to provide an ongoing resource for the successful and effective conduct of consultation processes in the State.
This reference should be regularly reviewed and updated, and should form the basis of ongoing improvement in best practice throughout the State.
This reference should be supported by a professional network of public servants engaged in the conduct of consultations, as reference points for other agencies and officers.
Electronic democracy through online consultation is not, and should not be, restricted to policy-related consultation processes managed and initiated by Government. Just as the Committee recognised a valuable role for the community, non-profit organisations and the private sector in the development of a rich information commons for Victoria (see Part II), the Committee also notes that democracy is a value shared across a society. Democracy, therefore, can be seen as a political value, a civic activity and an industrial process.
On one level, democratic decision-making structures are a legal requirement of companies and community organisations, exercised through regular and formalised decision-making processes, annual meetings, elections and debate. Corporate and community organisations are often key contributors to Government consultations, formal and informal, with an important role played by industry peak bodies, charitable and non-profit organisations and citizens groups in representing their members' interests to Governments.
However, the role of industry bodies, peak organisations or non-profit groups is not without controversy. While there is often a general concern about the disproportionate influence of large interests (such as corporate and business) on Governments through political donations (Gettler, 2004), to which the greater inclusion of the public via online consultation can seek to address, another area of concern is the "representativeness" of representative organisations which often act as regular advisors to Government or participants in consultative processes.
As Dr R. Melville states in her 2003 review of non-profit peak bodies in Australia:
The Committee recognises that this can be problematic, producing lopsided policy outcomes, or necessitating the submissions of peak bodies to be substantiated from other sources. While Dr Melville notes the response of the Federal Government has tended to be forced or encouraged aggregation of peak bodies, or the specific "licensing" of one group to speak on behalf of an industry segment, there remains a need for genuinely representative organisations to demonstrate that they are acting on behalf of their membership base, regardless of who they are.
For many industry bodies, and the larger charities, this is achieved through regular meetings, member surveys and by their internal processes of election. In making representations to Government or the Parliament, these groups are often willing to show the basis upon which they are making claims or providing evidence. For smaller groups, or those organisations with more limited resource base, this can be difficult, and the Committee recognises that some of the work undertaken by Government to facilitate government-to-citizen online consultation would also be beneficial for member surveying and internal consultation within representative organisations.
Thus, it is the view of the Committee that the use of ICT-enabled consultation and participation methods developed by Government and the Parliament can be catalytic in improving the internal communications and democratic processes of member-based organisations across our community. Any skills and technology transfer between Government and representative organisations in this area is beneficial in:
Improving the democratic character of non-Government organisations in Victoria (a general benefit to all, resulting in a more democratic society)
Improving the flow of information from these organisations to Government (a specific benefit in improving the outcome of civic engagement processes).
The use of online consultation by Government departments and agencies should consider the value of these experiences for technology and skills transfer to stakeholder groups in the community, in order to develop and expand the use of online consultation within their governance practices.
Where practical and feasible, consultation processes should encourage stakeholder groups to poll or solicit member responses using the technology appropriate to solicit the widest possible participation.
Given the interest of the Committee in means by which the increased use of online consultation in, and by, Government can be transferred to organisations outside of Government, the Committee gave consideration to two questions:
Is there a source of online consultation toolsets available to corporate and community groups?
To what extent do existing tools available meet the Government's objective in assisting organisations outside of Government to adopt online consultation?
With regard to the first question, the Committee clearly has an interest in determining whether, like the provision of inexpensive or free online community hosting and CMS, a ready supply of online consultation tools exists. The Committee identified a range of tools available free or at cost. These include:
A wide array of email list management systems through existing "virtual communities" tools, including a number specifically targeted to policy discussion processes (e.g. Australian Public Policy Research Network)
An expanding number of open-source publishing tools, from CMS (e.g. PostNuke, Plone and many others that often include discussion list capabilities) to collaborative publishing and editing ("wiki") systems (such as TWiki or MediaWiki)
A small, but developing market for online consultation services (such as that provided by CommunityPeople, BigPulse's Opinion Markets or JustSys's deliberative discussion technologies)
A range of open-source e-democracy tools (e.g. GroupServer)
Corporate collaboration and deliberation tools (such as phpGroupWare, Blackboard, MeetingWorks or ThinkVirtual).
Thus, while the availability of online consultation tools to the wider community - particularly those groups with limited resources and skills - remains limited and tends to be focused towards general-purpose tools, the Committee does not regard it appropriate at this time for the State Government of Victoria to invest in the development of ICTs for consultation and participation purposes beyond its internal needs. This is because:
Demand is uncertain at this time
Some toolsets do exist, particularly at the high end (groupware) and low end (email lists) of the technological spectrum
The specific types of tool needed by representative organisations or community groups remains unclear, making investment beyond
However, in making this finding, the Committee wishes to note two mitigating factors.
First, it is clear that Recommendation 64, which allows third-party organisations access to the tools developed by Government via OSS licensing, will "grow" the available range of localised or locally developed consultation and participation tools over time. In addition, the use of open source will encourage organisations and individuals from around the world to develop and improve these tools, providing a net benefit in the rapidity of innovation and array of applications by which the tools can be applied.
Second, where resources are maintained by the VPS, it appears appropriate to share these systems with community organisations, where appropriate and in the public interest. This appropriateness may be where there is a demonstrable need to assist the group in canvassing members to feed into a formal consultation process, or where the organisation is under-resourced. In evidence presented by the Director of Government and Community Information of the DVC with regard to "Central Station", Ms J. Duffy (Minutes of Evidence, 17/02/05, pp. 5-6) observed that this is already underway to some extent.
Thus, the Committee notes that the Victorian Government should consider creating a policy environment which allows relevant community groups access to facilities hosted or maintained by Government where the sponsoring agency is not a direct participant.
Clearly, a public-interest test would justify this use of public resources, but a policy framework may be required to ensure that these consultations remain confidential and that an acceptable-use agreement can be provided to protect the State Government.
Awareness of the existence of a consultation mechanism is a key requirement to utilising these opportunities for accessing Government. Historically, Governments have requested participation through the publication of advertisements in specific and general-readership periodicals, through direct mail to existing lists of stakeholders, television and radio advertising, and through innovative marketing exercises like the production of free postcards, stickers and other merchandising.
In recent years, this approach has been extended to the Internet, with a wide range of departments, agencies and the Parliament advertising and promoting consultations online. Often these are published via departmental or agency press release websites, though organisations which have regular need for public participation will also maintain specific consultation pages on their websites, with current and previous consultations listed (for example, the Essential Services Commission maintains a "Consultation Centre" on its website with a link from the main page).
In addition, governments such as those of Western Australia, Queensland, the Australian Capital Territory and the United Kingdom have established consultation portal websites or directories whereby citizens can access a centralised list of government consultations and reports. This approach does not specifically mandate the use of online consultation methods, but increases the ability of citizens to locate, and participate in, government consultations. Individual departments and agencies are responsible for providing notification information into the system, broadening the reach of their consultation processes and distributing costs across government.
While these websites represent intelligent application of web-publishing technology, often built on database systems and producing dynamically updated content, this approach has been taken further by some jurisdictions. In the State of Virginia in the United States, the Virginia Regulatory Town Hall provides a subscription service for citizens interested in government regulatory decision making. According to the State of Virginia, this allows for a targeted and customised service for citizens to increase their participation in policy making:
The advantages of this approach have already been noted and detailed in the Victorian context. The SARC of the 54th Parliament made explicit reference to the value of this approach (a centralised system that aggregates information from agencies and provides an internal management process and means for public participation) in its September 2002 report Inquiry into the Subordinate Legislation Act 1994 (SARC, 2002a). In addition, evidence from other centralised access systems has shown that streamlining access provides efficiency benefits to government and can increase levels of public access (Sternstein, 2005)
The SARC report noted that the benefits of the model included:
Quick and easy access to regulatory proposals and related information
Capacity for people to log in to personal accounts and receive information about regulatory proposals with the appropriate time to participate
Allows personalisation of the site
Allows tracking of regulatory proposals through the decision-making process
Provides the public with greater capacity for input
Contains information about public meetings. (2002a:286-7)
The Committee considers that the views of the previous Committee are accurate, and recommends that this approach be adopted in Victoria. In particular, from evidence collected in the United States in 2004, the Committee would recommend this approach to:
Emphasise the development of a customised website (site personalisation) and email notification with a high degree of plasticity to user requirements (type of information required, subject, format, frequency)
Provide the capacity for registered users of the site to act as content re-distributors to "social networking" message construction: the easy capacity for users to "refer" others to content of relevance to them, either via "pass along" emails or "email this page" functionality. This approach, empowering subscribers to identify information of interest to their professional, familial or social networks and pass on content, was seen as a particularly valuable approach to electronic democracy design by Associate Professor B. Noveck, the Director of the Democracy Design Workshop at New York University Law School, whom the Committee met with during the delegation to the United States.
A centralised portal and subscription service for Government and Parliamentary publications, consultations and legislative work should be established, allowing free public access to Government policy processes in a timely manner, in order to encourage public participation.
Local Governments should be allowed to contribute relevant material into the subscription service.
The Committee regards this approach as overdue for Victoria, and one that may have a range of additional possibilities for content syndication and alternative access to the content database, such as access via telephone, issue tracking and means by which interested citizens could organise pre-consultation "caucuses" prior to providing formal responses to requests by Government for participation.
During the Inquiry, the Committee took evidence about the relationship between work being undertaken in Victoria and around Australia to assist local community groups and organisations in developing programs and strategies to address issues of local concern. This approach, community strengthening, was first introduced in the late 1990s in New South Wales, in recognition that:
Communities have an inherent interest in local problems and a willingness to fix them
Local communities and organisations are often best placed to respond to local concerns because of local knowledge and existing infrastructure
Partnerships between Federal, State, and local organisations are necessary to ensure that resources and skills are available at the local level, and that co-ordination between projects and initiatives is maintained.
The success of this approach, as well as the general acceptance of the need for greater Government involvement in community capacity building saw the establishment of the Victorian Government's Community Building Initiative in October 2001. The subsequent creation of the DVC the following year as the lead Victorian agency in this area, gave it responsibility for networking and co-ordinating across and between Governments, and with local organisations. This overall approach reflects a renewed interest by Government in developing, maintaining and sustaining partnerships between Government and the public that can empower local communities.
In New South Wales, the communitybuilders.nsw website was established in 1999 as an online clearinghouse for information and resources for local community groups to develop their capacity to respond to local community needs, access information about government assistance programs (training, funding, support) and to network with other groups to share information and form partnerships to address shared concerns. This site demonstrated that:
There existed strong community interest in an online resource to support community organisations' work
The positive benefit of ICTs in social networking between groups, to share information and form new partnerships.
In recognition of the value of the New South Wales model, a similar website was established in Victoria in 2003. This site provides information for community organizations, to assist in their internal management, applying for and managing government grants and other funds and developing and implementing programs.
At present, the DVC, in conjunction with Infoxchange Australia, has redeveloped the Community Building website, incorporating a new "newsboard" function, revised and updated library of online resources and new back-end functionality to increase information discoverability. The Committee commends the work of the DVC in continuing to develop this resource.
However, from evidence presented to the AGIMO Enabling Government, Engaging Communities: An Online Perspective conference held in Brisbane in late 2004, the Committee observed that there may be areas of overlap and duplication between states. In a presentation on the communitybuilders.nsw website by the senior project manager, Ms C. Leigh, approximately eight percent of users of the New South Wales website were located in Victoria, indicating that the Victorian resource may not yet provide the full range of material of interest to local groups, or that general awareness of it could be improved (Leigh, 2004).
This finding is not surprising. The New South Wales Government was an early leader in this specific area of community strengthening, recognising how an online information repository and citizen-to-citizen discussion forum could be beneficial in empowering local groups to address issues of local concern and form networks between public-, private-, and community-based organisations. Because of this, the New South Wales website has received considerable publicity, and it is reasonable to expect that it may take some time for the Victorian site to be fully promoted and embraced by local groups.
While the Committee notes that the success of the Victorian site's new design will take some time to determine, it considers that the current approach could be developed further in two ways:
Increasing the level of citizen-to-citizen interaction via the Community Building website
Better intergovernmental information sharing in the community building policy area.
With regard to the first area, the Committee observes that one of the most interesting features of the New South Wales model is the capacity for community groups to directly interact online. In the New South Wales model, an online discussion area is provided via the website, which provides practitioners with the capacity to share information and post requests directly to other groups. This moderated form allows direct information sharing between participants, as well as the capacity for government staff to respond with additional information and links to other resources.
The Committee asked the DVC about the utility of the model for Victoria. Noting that the DVC was aware and had been interested in the outcomes of the approach, Dr D. Adams, Executive Director, Strategic Policy and Research, DVC observed that the ICT enablement of community based projects was already underway in Victoria and that:
Dr Adams went on to say that:
In keeping with the general perspective that Government ICT use in community participation processes should, where possible, facilitate citizen-to-citizen interaction, the Committee commends this approach by the DVC.
Following the results of these pilot projects, the Committee considers that it would be appropriate for the DVC to consider means by which the project-specific online discussion spaces could be incorporated within Victoria's www.communitybuilding.vic.gov.au, but in such a way as to maintain the capacity for local projects to use private workspaces as appropriate, and have access to the wider community building sector for more general discussion, networking and information exchange. This may be achieved through open and closed discussion list threads, or by adding general discussion spaces "above" local project lists.
www.communitybuilding.vic.gov.au should adopt a moderated citizen-to-citizen discussion forum to develop and deepen linkages between community groups within Victoria.
In addition, the Committee notes that the value of the community strengthening approach in Victoria is being recognised by other jurisdictions. Queensland has recently established a Department of Communities, and the Federal Government has begun further work on developing community capacity and resilience. There is an increasingly large amount of online information being published by governments around Australia, aimed at assisting local groups to develop their capacity to engage in partnerships with government. It would appear that, while some information is specific to local jurisdictions, there is a clear tendency for overlap and duplication, and the prospect of "losing" Victorians to other resources that may be of limited relevance to local conditions.
The Committee does not consider it necessary, nor appropriate, for Victoria to jointly develop and maintain a shared resource with other jurisdictions. However, it would appear prudent at this time if the developers of state-based online resources were to more closely collaborate to develop a method of syndication and redirection between the comparable websites around Australia. The advantages of this approach would be:
Low costs associated with the creation of material that is generic
Greater amounts of content available across Australia
Greater information sharing and collaboration between community building and strengthening agencies around Australia
Low rates of "loss" of users to websites in other jurisdictions
Possibilities for joint development of website functionality.
www.communitybuilding.vic.gov.au should further investigate methods by which the relationship between this initiative and similar initiatives around Australia (such as communitybuilders.nsw) can be strengthened to:
Syndicate relevant content between these services
Redirect Victorian users to relevant Victorian resources (where they exist).
Notes: |
|
|
Voting in Local Government elections in Victoria is only compulsory for residents under the age of 70. This figure is based on the complete electorate, which included those aged over 70 years and non-resident enrolled electors (e.g. property owners). |
|
|
Similarly, the Australian Bureau of Statistics (ABS) estimates that 3.8 percent of the Australian population would classify themselves as having "poor" reading skills (ABS, 1997). |
|
|
The Committee notes that the Victorian Government has sponsored a number of education programs and schemes which attempt to increase multilingualism in the VPS. |
|
|
In 2000, the Government of South Australia proposed the establishment of a "virtual electorate" for expatriate South Australians who would have voting rights for a seat in the upper house of the Parliament (Olsen, 2000). While the language used in this discussion focused on technology issues, and the concept was included in the "information economy" strategic initiative of the State in 2000 (Government of South Australia, 2000), the emphasis here is on the notion of the role of expatriates in the policies and social life of the jurisdiction. The Committee notes that the idea, while attracting limited enthusiasm in South Australia, is not unheard of internationally, particularly in the case of countries which have large expatriate populations. Portugal, for example, permits its expatriates to vote for MPs in a number of "emigration constituencies" via postal ballot (The Southern Cross Group, 2004). Other indirect representative mechanisms for expatriates also exist in Europe (e.g. France). Overall, the question of "virtual" or expatriate electorates, while possibly facilitated via ICTs, is a wider policy question for the people of Victoria and beyond the scope of the Inquiry. |
|
|
The Federal Voter Assistance Program has been established to assist overseas voters (particularly members of the military) in casting votes. Traditionally, this has been through the development of detailed voting guides for personnel, the facilitation of postal ballots and training for military officers charged with assisting their soldiers to cast a vote. The Program has also been active in working with counties in the United States to provide a policy environment favourable to the voting requirements of military personnel. |
|
|
The Committee met with, and received extensive briefings from, both the SERVE project team and Dr Wagner, one of the members of the SPRG minority group, while undertaking its delegation to the United States. |
|
|
Though the use of this approach may be problematic given WCAG considerations. |
|
|
This malware would need to combine a range of factors in order to be successful: it would need to be based on a new or redeveloped version of a piece of malware to evade common anti-virus software, the malware would need to function correctly (even though it has been newly developed) in its distribution vector, it would need to be distributed secretly in advance of the election to prevent counter measures being developed by election officials, it would need to interface with an encrypted voting client (the design of which the malware author may have limited knowledge of in advance), and act without the voter or electoral official becoming aware of its presence. |
|
|
The Committee of the 54th Parliament was informed of an allocation of A$10 million to the projects. |
|
|
The advantages of this approach would be considerable. The network is highly secure, matching its role as a diplomatic and security channel, and provides access to embassies and consulates around the world. |
|
It should be noted that, while there were serious flaws with the management of the provision of electronic-voting systems in the Californian example, both at the State, County and equipment-provider levels, the Committee did not find any evidence of attempts by any party to deliberately manipulate electoral outcomes. |
|
Based on a survey sample of 4,600 students from 154 Australian secondary schools (national). |
|
Application to become a general postal voter, Application to become an itinerant elector, Application to become an overseas elector, Application to become a silent elector. |
|
The committee also notes the first generation of consumer electronic devices to include biometric scanning systems have now begun to appear in the marketplace. |
|
This is an important consideration, and the Committee considered the example of voting via SMS - as one model used in the United Kingdom - incompatible with the mandatory preferential system of voting employed in Victoria. This system works well for first-past-the post voting only, and even for jurisdictions that have optional preferential systems, the Committee would express caution that it might encourage lower numbers of preference provision by the voter, altering the context of voting significantly. |
|
The Irish assessment is based on the Nedap/Powervote system, a commercial product. |
|
This is a fundamentally different technology to the mark-sensing systems used in first-past-the-post elections, which are based on the absence or presence of a single mark. These systems are much simpler, and have high levels of accuracy at low cost (Brady, Buchler, Jarvis and McNulty, 2001), but are not feasible for preferential voting environments. |
|
In this case using a five-level scale or typology. |
|
Note that a further discussion of these issues, pertaining to Parliamentary Privilege, is provided in Part IV. |
|
This would also permit the transfer
of technologies from other jurisdictions. For example, during the delegation
of the 54th parliament to Europe, the Committee received a briefing on
the development of the EDEN project, a complex electronic democracy toolset
development project sponsored by the European Union that focused on the
development of "natural language processing" technology to
assist in the communication between citizens and government through technologies
such as address guessing, translation and interpretation, and the development
of customised information provision through artificial intelligence. |
|