"And What Do You Do Between Elections"

Address by Mr Ed Killesteyn, Electoral Commissioner

The 23rd Annual Government Business Conference

19 July 2011

When I was first appointed Australian Electoral Commissioner some 2.5 years ago, there were 2 questions that I found were repeatedly asked of me – one was, "when are we going to be able to vote electronically", and the second is the title of my presentation today: "and what do you do between elections"?

I'm happy to debate the first of these questions at another time, but the second question is perhaps one that many of you may have wondered about every 3 years or so when, after casting your ballot, you wander off from a polling place safe in the knowledge that you won't have to do that again for a few years.

While it didn't take me too long to get an appreciation of the answer to the question, I have to admit in the early days of my term I did wonder about it myself. The revelation, or at least a way of responding effectively to the question, came one day in talking to a wise, some might say wizened, Divisional Returning Officer who, when I was relating the experience of being repeatedly asked the same question, offered the following:

"The answer to that question is the same as the answer to the question about what firemen do between fires – you get ready for the next one".

Now this disarmingly simple answer belies a much more complex mix of activities that bears some explanation, because it points not only to the obvious exciting, some say glamorous, work of the AEC, of running elections, but to the less obvious work between elections and which, in some respects, is the foundation of a successful election. Equally, this less obvious work exhibits long term trends which point to some sobering challenges for Australia's democratic model and, in turn, opens up a debate about the role of the Australian Electoral Commission in this area of public policy.

Now it is likely that many will conclude that "getting ready for the next election" is about the logistical challenge of staging the largest peace time event in Australia every three years or so.

This is certainly an accurate assessment. Consider the following volumes of resources consumed at the last Federal election, at a cost of around $106M, not including public funding for political parties.

• Over 43 million ballot papers were printed by a handful of printers, absorbing about 300 tons of paper and printed in the space of a few days.
• These ballot papers were distributed across the length and breadth of Australia using hundreds of contractors, big and small, to around 7 500 polling places, many located in metropolitan areas, but many located hundreds of kilometres from any major town, and all Australian overseas diplomatic missions.
• In these polling places we provided 80 000 ballot boxes, 150 000 voting screens, 140 000 recycling bins (including those recycled from the 2007 election), 100 000 pencils and 140 kilometres of string, some of which is used to keep the pencils from straying too far from the voting screens.
• Some 65 000 polling officials were employed, most of whom had to be recruited, trained and paid for what was generally no more than 1 day's work
• And these polling officials, apart from managing the polling place on polling day, counted 11M House of Representative votes on Saturday night after the polls closed at 6pm. Importantly, counting the large majority of first preference votes on election night allows, in most cases, early clarity on who is likely to form Government. The 2010 election – that delivered an uncertain outcome on the night despite counting those 11 million votes – was a rare exception in this respect. And of course we know what resulted was a 'hung Parliament', the first so elected since 1940, where neither of the major parties gained a majority in their own right.
• All of that work was supported by a highly complex and integrated suite of computer systems – the AEC is perhaps the most highly computer dependent electoral management body in the world. For example, those computer systems allowed us to provide real time election data, with 43 million hits on our Virtual Tally Room over 6 hours, with updated results published around every 90 seconds on election night.

Now I could continue to regale you with hopefully what you believe are impressive statistics about volumes of activity administered by the AEC, but my point is that none of this happens overnight.

All of these "products" reflect years and months of preparation, planning, tendering, evaluation and procurement, contract administration, testing, election simulation, training, process improvement, legislative change and implementation, by over 850 full time and 1000 part-time staff who are dedicated to electoral administration, spread over 130 locations across metropolitan and rural Australia.

They are an impressive group of people, whose commitment to delivering a credible and trusted election result for the citizens of Australia, where Government can change from time to time in a peaceful and orderly fashion, is a primary motivating force.

It is that phrase: "for the citizens of Australia" that embodies the aspiration of our Australian democratic model, but also what I believe to be the defining challenge at this time for our unique model of democracy.

It is worth noting that both these sections have a longstanding pedigree: compulsory enrolment was enacted in 1912, and compulsory voting in 1924. And here in QLD, the pedigree is even longer with compulsory voting being enacted in 1915.

Together, the provisions of the Constitution and the CEA create the unique Australian model of democracy where all Australian citizens 18 years or older have both a right and an obligation to enrol and to exercise their franchise at the ballot box.

From the perspective of electoral administration, this contemporaneous duality of entitlement and obligation is reflected in a task we in the AEC call "roll management". Along with preparing for elections, roll management is the primary activity that keeps electoral administrators busy between elections.

Once again, the disarmingly simple phrase 'roll management' masks a complex set of interactions with electors and prospective electors which is more than simply recording new electors as they become eligible, or updating details of existing electors as they change residence.

Roll management also includes the very important task of maintaining the address register. This is perhaps one of the most important and enduring assets created by the AEC: a high integrity register of all addresses in Australia, which not only ensures that addresses are recorded consistently and faithfully no matter the identity of the resident who may occupy that address from time to time, but also acts to provide confidence that fraudulent addresses are not being used by electors. As at June 2010, there are nearly 9 million enrollable addresses on the register, growing by over 800 000 in the last 5 years.

Today, there are slightly more than 14m registered electors on the electoral roll. The AEC completes around 2.5M enrolment transactions of all kinds per year, more in an election year, for persons coming on to the roll for the first time; for those whose details change because of changes in residence, marital status, and so on; and for those who are deceased and must be removed from the roll.

"Maintenance" of the roll is not a difficult processing task; the volumes of transactions are easily accommodated with efficiency and integrity by well trained staff, which we have, using intelligent computer based processing systems linked to confirmatory data bases of other agencies, as we also have. Where electors voluntarily comply with their obligations to enrol, to update their details, and to vote, the AEC completes the relevant administrative task quickly, efficiently and with a minimum of fuss and inconvenience to the citizen.

The challenge for the AEC, and perhaps for the wider community to consider, is in respect of those who do not voluntarily comply, either wilfully, through laziness, through misunderstanding or through incapacity brought about by some form of social or other disadvantage.

Consider the following. Based on Australian Bureau of Statistics data of eligible Australian citizens 18 years or over, the AEC estimates there are 15.4M Australian citizens eligible for enrolment and voting.

With 14M people actually on the roll, around 1.4 million people are not, and do not, indeed, cannot, vote. The participation rate, as it is known, that is, those enrolled as a percentage of those eligible to enrol, currently stands at around 90%.

At the 2010 election, slightly more than 13M people cast a ballot, nearly 1M less than the number on the electoral roll. The turnout rate, as it is known, was 93.22% for the 2010 election.

And, of those who did cast a ballot, around 730 000 had their votes excluded because of informality – that is, their ballot was defective for some reason, including around 48% who we can reasonably assume did so deliberately. The informality rate was 5.5% at the 2010 election, up from 3.95% in the 2007 election.

It is somewhat sobering in my view, for a country that has had compulsory enrolment for nearly 100 years, and compulsory voting for 87 years, to reflect on the fact that, if I add those not on the roll, to those on the roll but not turning out, and those whose vote was not counted because of informality, around 3M people, or 1 in 5, did not, for a range of reasons, effectively exercise their franchise at the 2010 election. Nor is this a recent phenomenon; the participation rate has exhibited signs of long term decline since about the mid-1990s.

Against the background of these figures, roll management becomes a principal element of electoral administration carried out between elections. Indeed, some practised electoral administrators would say that an election is only as good as the electoral roll on which it is based.

Now I hasten to add this is not a criticism of the legitimacy of the 2010 election, nor is it a call for more or harsher regimes for enforcing enrolment. Indeed, the overwhelming philosophy influencing electoral administration, especially in the context of roll management and the challenges of sustaining and improving overall effective participation, is to promote voluntary compliance.

The AEC does not have an organisational position on the merits or otherwise of compulsory enrolment. It is the law, and ultimately a matter for the people of Australia to determine if change is wanted. It is worth noting here the research on popular support for compulsory voting. The first Australian Election Study done after the 1996 election showed 74% of respondents supported compulsory voting at federal elections. The Australian Election Study after the 2004 election showed 74% in support. A Morgan poll in 2005 showed 71% support, and an Ipsos-Mackay study, also in 2005, showed 74% support. The 2010 Australian Election Study, just released, showed support for compulsory voting at 69%, down from 77% support after the 2007 election.

In my view these statistics give a strong lead to us as electoral administrators that roll management programs, heavily weighted in favour of encouraging voluntary compliance, are best suited to the Australian context, noting that occasionally some level of enforcement may be necessary in the most wilfully recalcitrant cases.

Over time, roll management programs have evolved to reflect the demographics of Australia and technological opportunities. In the last century, the primary means of roll management was through "foot soldiers". Literally, armies of door knockers weaved their way through the streets and suburbs of Australian electorates reminding people of their obligation, dropping off and collecting enrolment forms. This strategy was appropriate at a time when electorates were largely self-contained, serviced by a single Divisional Office in a relatively stable Australian demographic.

For the last decade, roll management has been practised through something we call the Continuous Roll Update program, or CRU. Using the opportunities provided by matching of 3rd party data with our own electoral roll, targeted letters are issued to people who we believe are eligible to be on the roll and who are not, or those who are on the roll but not at their latest known address. These letters invite the recipient to complete and return an enrolment form. In recent times over 3 million letters per annum have been despatched to existing and would-be electors as part of the CRU.

Driven by even greater demographic shifts, characterised by high rates of mobility, evidence of dwindling response rates to CRU letters, and with even greater technological opportunities associated with the use of the internet, our voluntary compliance strategies are evolving once more. Indeed it is my view that there is considerable urgency in making the transition to newer approaches to promoting voluntary compliance because of the electoral participation rates that I referred to earlier.

Encouraging voluntary compliance has two principal elements. The first is ensuring that the citizen, in this case, understands the obligation. The second is making it as easy as possible for the citizen to comply with that obligation. Our research tells us we need to work on both fronts.

Our pre-election research (Enrolment Triggers Report 2010) found that there was little interest in voting and less in enrolment. Only three quarters of respondents believed they needed to be enrolled before election-day, with over 10% believing they could enrol on the day.

Less than three quarters of the respondents reported that they were aware that both enrolling and voting are compulsory in Australia, with one in ten believing only enrolling is compulsory, and one in eight believing that only voting is compulsory. Although most were aware that there is a fine for not voting, only around half believed there is a fine for not enrolling.

Most people reported that enrolments needed to be updated each time they moved, but few actually did so unless an election was imminent. Many people report not thinking to update their enrolment details when they move. In addition, some do not realise they need to update their enrolment when they move within the same suburb or state.

Education of the community is an ongoing component of what we do between elections and subject to its own necessary reform program, but ease of compliance with enrolment obligations is also a key priority for reform.

To underline this point about ease of compliance, it was only last year, on 19 July to be precise, that legislation was finalised to allow electors to update their enrolment details on-line – a practice which has been commonplace with other Government agencies for many years. If 19 July sounds vaguely familiar, it was also the day on which the Governor-General issued the writ for the conduct of the 2010 federal election and needless to say, there was little time to implement that very welcome change to our enrolment facilitation tool kit [And those people who remained standing some little time ago in my audience participation exercise are invited to use that facility immediately after my presentation]. I'm pleased to say that we have already reached nearly 25% of enrolment transactions being completed on-line.

This seemingly simple change has opened up a wealth of collaborative opportunities with other government and private sector agencies to make it easier for electors to comply. One of the most exciting is the development of a partnership with the Australian Taxation Office. This year, as the expected 2.4M plus Australian taxpayers are completing their 2010 income tax returns on-line using e-tax, those who show a change of residential address will be automatically provided with a hyperlink to the AEC on-line change of address facility. While not entirely seamless, the conjunction of several related transactions with Government in a single online space is sound administrative practice offering considerable elector convenience. Like any expectant parents or grandparents, we are eagerly awaiting the outcome of this new collaborative approach to voluntary compliance, although the fact that it is early days is not stopping us from developing similar collaborative models with Australia Post, Centrelink and Medicare Australia.

Our reforms are however lagging behind those of state electoral jurisdictions. Driven by the same concerns about elector participation rates in both NSW and Victoria, legislation has been enacted which gives authority to the respective Electoral Commissions to directly update enrolment details, or directly enrol a person not on the roll, based on data provided by 3rd party agencies. The AEC is of the view that these reforms, although by no means offering a panacea for reversing the participation rates currently being experienced, would add to our tool kit of voluntary compliance initiatives.

The relevance of such reforms would seem to be well placed from the perspective of the community. Again, our research tells us that a high proportion of electors (54%), particularly in the under 36 category (69%), believe that they can change their details by calling the AEC, and a smaller proportion (24%, 34% under 36 year olds) believe that they do not have to do anything until the AEC contacts them.

Perhaps most concerning is that 9% of people of all ages incorrectly believe that the AEC already directly updates enrolments. This number will likely rise as State authorities move to direct enrolment of electors, as New South Wales and Victoria have already done.

The research suggests that people do not find the process of updating their enrolment particularly burdensome, but rather they do not think to do it. The research found that 85% are very or fairly confident about how to enrol and 86% of those who had enrolled described the experience as very or fairly easy (Enrolment Triggers Report 2010).

Convenience is a significant factor, and an increasing number of electors prefer to deal with the AEC online. During the 2010 election, for example, the AEC's enrolment check facility received 9.5 million page views (compared to 2.6 million in 2007). The AEC also received 170 000 emails during the election period, compared with 37 000 in 2007. As I have already indicated, recent changes mean that is now possible to update enrolment online, and three quarters (75%) of the population appear to be aware of this. People under 36 are more likely to be aware of this option than they are of filling in an enrolment form.

The efficacy of these reforms is not free from divergent political views however. I note that the majority recommendation of the Joint Standing Committee on Electoral Matters, in its recently released report on the conduct of the 2010 Federal Election, about the need for the Commonwealth to enact similar legislation to that now in place in the two largest states, was split along party lines. This illustrates one of the potential pitfalls of a "non-political agency operating at the very centre of a partisan contest", but the AEC's support for better tools to facilitate higher levels of enrolment is motivated solely by a concern over the electoral participation rates we are currently experiencing, the expectations for which are clearly spelt out in the Commonwealth Electoral Act – everyone, barring a few that the Act excuses for incapacity, must be on the roll. Nor is there anything new about this; the AEC has had a publicly stated target of 95% participation rate for at least 20 years, with a suite of strategies that have been implemented, fine-tuned, changed and evolved over time in pursuit of this objective. In pursuing these new strategies however, I am very conscious that we don't undermine the confidence in the integrity of the electoral roll and the conduct of elections.