The underlying premise is that "smart" voters know their candidates. Yet in a party-dominated system like Australia, the prevailing understanding is that voters know their parties primarily, and only indirectly their candidates.

Smartvote Australia, one of a family of technologies known as voter advice applications, operated at the 2019 federal election and is an adapted version of a platform first developed in 2003 by Politools, an interdisciplinary scientific network based in Bern, Switzerland. Politools continues to support the smartvote project, which is funded by the Australian National University and led by Professor Patrick Dumont of the School of Politics and International Relations. The Canberra Times is a media partner.

The guiding assumption is that democracy is strengthened and trust in the political system is increased by greater information about the policy positions of those standing for election being made available to voters. As a candidate-focused platform, it is especially well-suited to proportional representation electoral systems such as the Hare-Clark system used in the ACT. Voters are provided electorate-specific results for each of the ACT's five electorates.

The project depends initially on the co-operation of candidates and parties, and by Tuesday this week 101 of the 137 candidates (74 per cent) had participated. Their participation shows an acceptance of the guiding assumption of the project, and is a brave move in a world of tightly controlled political campaigning, where a word out of place can be used by political opponents.

There are 32 questions, ranging across broad policy areas from economic to social policy. The questions range from "Do you think increasing the ACT's population is needed for our economic growth?" to "Should pill testing at music festivals be prohibited in the ACT?"

Participation by voters follows, and more than 16,000 voters have already taken part. Only they know how useful their participation has been in terms of voting choice as distinct from satisfying their curiosity or increasing their general competence. Both candidates and voters have provided not just yes and no answers, but answers ranked on a scale of 1-4, or sometimes 1-5. Candidates have also provided further information explaining their answers, making this a sophisticated exercise.

It cannot be denied that greater information means a more informed vote, which is an important step forward.

One limitation is the relationship between parties and their candidates. It does not apply, by definition, to independent candidates, who are not constrained by party membership. Is a smart vote one for political parties which form governments, or only for particular candidates of that party?

One strength of the Hare-Clark system is that voters have a choice between the individual candidates put forward by each of the political parties. Such candidates attempt to distinguish themselves from their colleagues through their campaigning, but most of that effort must be within the bounds of party policy. In the ACT this means Labor, the Liberals or the Greens in particular.

Candidates tell their life stories, including their education, profession, community service and connection with the electorate. Candidates can also distinguish, explicitly or implicitly, their personal characteristics, such as their gender, age, sexuality or ethnic background. These personal characteristics appeal to voters who want an Assembly which is more representative of the community, regardless of party affiliation or policy. Already in this campaign attention has focused on these characteristics as one way to distinguish between the major parties.

READ MORE:

In the main, the three major ACT parties dictate the policies of their candidates. Broadly speaking, Labor offers their candidates the least room to move, followed by the Liberals, with the Greens being the most flexible. Only on a limited range of issues is a free vote, sometimes known as a conscience vote, allowed to MPs. In a survey such as that conducted by smartvote, this constraint applies to all candidates, although some flexibility is allowed for variations in intensity of support and personalised comments.

The complex relationship between party policy and candidates was illustrated in the case of Dr Brendan Long, Labor's candidate for Murrumbidgee. It emerged that Labor candidates were able to provide their personal views as distinct from party policy only on women's reproductive rights and end-of-life choices. Dr Long is opposed to euthanasia, and this informed his smartvote answer, which was mistakenly altered by the party secretariat to conform to the pro-euthanasia party policy. Later the party secretary corrected the record.

The smartvote exercise accepts that other factors not covered by a policy-oriented questionnaire may be at play in any informed choice by a voter. They include leadership, trust, management capacity, government record, government longevity and strategic voting. Some of these have already featured in this ACT campaign.

The relevance of these complex factors to voting choice can be sketched in some simple questions. Should Andrew Barr or Alistair Coe be our next chief minister? Is majority government the best option, or would it be better to elect independents or minor parties to constrain the winning major party? Regardless of their views, whom do I trust the most? Regardless of their views, who has the better capacity to govern? Regardless of my traditional party loyalty, is it time for a change?

Smartvote potentially makes voters more informed, but many other factors remain in play.