Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Joint Standing Committee on Electoral Matters
Conduct of the 2016 federal election and matters related thereto

BONHAM, Dr Kevin James, Private capacity


CHAIR: Welcome and thank you very much for your written submission to the committee. Although the committee does not require you to give evidence under oath, I should advise you that this hearing is a legal proceeding of the parliament and therefore has the same standing as proceedings of the respective houses. The giving of false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. The evidence given today will be recorded by Hansard and attracts parliamentary privilege. I now invite you to make a brief opening statement before we proceed to questions and discussion.

Dr Bonham : In my opening statement I want to cover a few matters that I did not have time to include in my submission or that arose after I completed my submission and also to amplify a few matters in my submission. The major thrust of my submission concerns the impact of the Senate voting reforms, their successes in solving the problems under the previous system and concerns that were held about problems that they might create. Most of those concerns are proving to be unfounded or exaggerated, but there were some issues caused by the crossover between the old system and the new system and also caused by the speed with which the new system was implemented that I think can be improved for the future. I have made some recommendations concerning those.

I also want to endorse some recommendations from Antony Green's submission. Antony gives a recommendation about leaving ballot papers with an elected candidate for the purposes of surplus votes if the ballot papers would otherwise exhaust so as to give more value to the votes that will continue. I endorse that recommendation completely.

He also recommends optional preferential voting for the House of Representatives and then says that, if you are not going to have optional preferential voting, you should at least accept a simple ordered sequence of preferences so that if someone puts 1, 2, 3, 4, 6, 7, 8, 9 and that is the only error, their vote should be counted as formal rather than treated as informal. I think that that particular reform is an absolute no-brainer and should definitely be done. Whereas you have the problem of people voting 1, 2, 2, 3. You do not have a fully ordered vote. So it is about counting a vote where the voter's intention is clear.

The AEC published a table in their submission—table 23—of informal votes by category. I wanted to note that a large proportion of the informal votes were just completely blank ballots. So some of the increase in informal voting, particularly in the Senate, may have occurred through intention, but we do not have an exactly comparable set of figures for 2013. The category that was of most concern to me was that some votes with repeated numbers became informal under the new system that would have been formal under the old one as a result of the nature of the new savings provision.

CHAIR: Could you just start that again, because some of us were looking for the table. So that table is on page 40 of the AEC submission and on page 51 of our papers. Please continue, Dr Bonham.

Dr Bonham : Under the BTL section, the table notes a significant number of papers that were treated as informal because they had repeated numbers. There were over 45,000 informal votes because of repeated numbers. Quite a lot of those votes probably would have been formal under the old system. I saw one vote treated as informal—and it was informal under the new rules—where someone had put 1, 2, 3, 4, 5, 7 and they had numbered every square.

Senator IAN MACDONALD: This is in the Senate?

Dr Bonham : Yes, in the Senate. Under the new system, that is informal. So I think there is a need for liberalising the savings provision slightly. I have made a recommendation in my report about that.

I want to speak a bit about my experience as a scrutineer. I scrutineered the Senate count for about 15 hours at the counting centre. I can take questions about that. It was very interesting to have watched the process. It is a very difficult process for parties and candidates to have the resources to scrutineer fully. Numbers appear on the screen and disappear very quickly. The data processers get faster and faster through the process. You have all these people sitting at their computer terminals typing what comes in on a screen, and through the count the speed of the operators at the centre more than doubled. At the end I estimated that they were clearing the average ballot paper in about six seconds. I was scrutineering and trying to sample preferences. I am very fast, and I was flat out to notice where something went for three or four votes before it went off the screen because it was done. Concerns have been raised in some submissions about it being desirable to audit the process and check the accuracy, and I think that is a good idea, because it is not practical under this system for scrutineers to see every vote in processing.

One more matter I will raise that came up after I completed my submission is the matter of the section 282 advisory recounts for filling Senate places. There are a lot of ambiguities about the rules for those. A few people who have been looking at that were surprised by some of the numbers in the distributions of preferences that eventually came out. And I think if this method is to be used in the future it will be useful to have the standards for deciding which votes make it into the recount and to what extent documented clearly. I think I will leave it there for now and take any questions.

CHAIR: On the issue of section 282, given that you have additional comments over and above your written submission, is it possible to ask you to put a further submission in on that issue in terms of where you see the issues are and the ambiguities so that we might give that some further consideration as well?

Dr Bonham : Yes. I am happy to do that.

CHAIR: Thank you.

Mr GILES: Thank you, Dr Bonham, for a fascinating submission and a very instructive opening statement. There are just a couple of matters I would not mind exploring with you before yielding to other members. Firstly, picking up on your comment about a voter's-intention approach to informality, I am wondering whether you have a sense of how many votes in the House of Representatives may be informal because of a skipped number or whether that data is available.

Dr Bonham : I suspect that data will be available, because the AEC does regular informality surveys of the Reps. I know it is obvious that when you have a large number of candidates for an electorate the informality rate goes up a lot.

Mr GILES: If that is a matter that you are able to address in a further submission, and indeed anything else on the workability of an expanded savings provision in the Senate but also looking at the Reps on that, then that is something that I would find of great interest.

The other matter was really just teasing out your thoughts on your first two recommendations. I know that you touch on the fact that in northern Victoria in the last legislative council elections at a state level a candidate was elected through the operation of the inclusive Gregory system who would not have been elected. I am just wondering whether that has ever happened in the federal sphere or in other jurisdictions.

Dr Bonham : I believe that 2014 case in Victoria was the first clear-cut case of it happening in a state or federal election. This refers to the way in which votes are transferred as ballot papers rather than with their previous values. As a result, some people's votes effectively increase or decrease in their total value as the count goes on. I mentioned in my submission that close to one per cent of the votes in Tasmania went into a certain point worth something, helped someone get elected and then came out worth more, which is silly. I agree with Antony Green's recommendation that there should be a report commissioned into how to fix this problem, because some day it is going to elect the wrong person.

Mr GILES: For those of us for whom maths is not their special subject, like me—although I have probably spent too much of my life dealing with schedule D proportional representation elections under the Victorian branch of the Labor Party rules—what objections if any have been advanced to the change that you have recommended in recommendation 2, which seems to me at face value to be straightforward? That is the progressively-reducing quota for Senate elections.

Dr Bonham : I have not seen any objections advanced to it since people started suggesting it in the context of this particular system. I have seen this work in other systems: as the count goes on, you progressively reduce the quota because votes are going out of the system. There is often a reluctance to change the system, because of issues with it being supposedly too difficult to do manually if you make too many changes. I would say that it is totally impractical to do it manually anyway, so I do not think it is a valid objection in the computer age. There are people who can try to replicate what the AEC has done. You have someone, Graeme Bowland, who completely replicated all the election counts independently.

Mr MORTON: In a previous role I was the state director of the WA Liberal Party during the last Senate by-election. Your daily blogs in relation to what was going on there were most helpful, so it is a pleasure to meet you. In relation to your experience as a scrutineer, I think the scrutiny of the Senate count, given the new changes, is something that this committee should look into little further. As a scrutineer, if you wanted to, could you see every ballot paper? Not that you would have eyes to look over every paper—

Dr Bonham : That is the main issue.

Mr MORTON: But if you had unlimited scrutineers, is every ballot paper available for inspection by a scrutineer?

Dr Bonham : Every ballot paper comes through the scanning system so that the scrutineers can see it, is my understanding. My understanding is that if a candidate were able to have 30 scrutineers in the room at once, which I am not sure would be allowed, they could actually see every ballot being processed. In some cases they would see every ballot being processed multiple times, because ballots that are tricky have a series of escalations: you see the ballot paper appear on one screen and then you see it on the other side of the room. It is like, 'Oh yes I have seen that one before,' because it has been escalated to a higher level of—

Mr MORTON: In relation to that escalation, I know that in some states there is a 'calling' of the physical ballot paper to come before the officers there—the actual piece of paper, not the scanned version. Did you have any experience in actually calling upon—seeing the actual paper version of the ballot paper? And how was that dealt with?

Dr Bonham : I did not see anyone demand to see a paper version of the ballot. I saw paper ballots being scanned in a side room, where you could go to stand and watch the paper ballots as they were being scanned in. But you would not be able to get any sort of sense whether most of them matched or not from doing that.

Mr MORTON: Okay. In relation to your submission, you talked about the reports of the 105 Senate ballot papers that were deemed informal because in WA's case I think some Victorian ballot papers issued, not Western Australian ones. This is an issue that I am very concerned about. You mentioned that in two cases in the Western Australian case, the final candidates for major parties were excluded by less than that margin, and that under the old system if these exclusions had affected the result that could have been voided again.

I am very sensitive to results of elections being voided. I wonder if you could add to what you were talking about there in your submission in relation to the informality and the effect of Electoral Commission mistakes in relation to the issuing of ballot papers—and the effect that would have on the account as well?

Dr Bonham : The word 'major' in my submission there is an error. My apologies for that. It should be 'final candidates for parties'.

Mr MORTON: Rightio.

Dr Bonham : I think I meant to say the final top candidates for parties.

Mr MORTON: My concern remains the same.

Dr Bonham : Yes. The issue there was that under the old Senate system, if somebody missed out by a very small number of votes at a certain point, that could change the whole outcome, sort of like butterfly wings causing a cyclone in Taiwan—that kind of stuff. It is like the case we had in the 2013 Western Australian election: there was this very fine tipping point where what happened between two candidates determined two seats, and in the end the margin between those two candidates was so small that, even if the AEC had not lost ballot papers, you might have had trouble making that count stick. In this case, if there is some party that is coming 16th and some party that is coming 17th, and one of them goes out before the other one, it makes no difference, because they are both going to go out anyway. There is just no way for a party that has polled a very low vote to get elected, because they do not attract enough preferences.

Mr MORTON: Finally, what is the layman's explanation of the differences between the inclusive Gregory system and the weighted inclusive Gregory system?

Dr Bonham : The weighted inclusive Gregory system means that the total value of your vote ideally stays at one. Under the other system, because it treats votes in a simplistic way and ignores their past history, the total value of a vote through the count can go up or down from one vote. So some people's votes may end up being worth more than one vote and some people's votes may end up being worth less than one vote, in terms of what they add for particular candidates and what help they provide to candidates to get elected.

CHAIR: Thank you, Dr Bonham. On that particular issue, would you mind providing a little bit more explanation for us, because I think that is something that we are going to have to get our heads around further, and it is somewhat complex but important. Would you mind, on notice, providing us a little bit more detailed information on that and how it works.

Dr Bonham : Yes, I am happy to provide more detail. There is also an explanation of it in Antony Green's submission.

CHAIR: We have already taken a note that we will pursue this further with him.

Mr DICK: Dr Bonham, you said that you were a scrutineer. I am not asking you to say who you were scrutineering for, but you were obviously scrutineering on behalf of a candidate or party.

Dr Bonham : Yes, I was nominated by a particular candidate.

Mr DICK: In your submission, you talk about authorisation in terms of online materials, which we are particularly interested in, and I think you make a recommendation or suggestion around tweets—obviously with the character limits imposed—having a link or some sort of statement. Would that be for individuals, parties, candidates, organisations or all of the above?

Dr Bonham : I think ideally whoever is posting electoral material.

Mr DICK: During the election period itself?

Dr Bonham : Yes. It might get impractical to police if you are going to say, 'Well, everyone who comments on the election on Twitter should have an authorisation thing at the bottom of their website,' or something like that, but just in theory I do not see that there is a problem with it.

Mr DICK: But you certainly support some standardisation across the board.

Dr Bonham : Yes, I think there should be standardisation across different platforms in a way that does not make any of them impractical. There has been some concern with the Tasmanian legislation as to how you deal with authorising a tweet under the Tasmanian legislation. You cannot put an authorisation in a tweet.

Mr DICK: Would that go the same for bloggers?

Dr Bonham : Yes. A blog is more straightforward because it is just like a written thing that you are putting on the internet, and you can just whack an authorisation stamp and disclaimer at the bottom of the page or at the bottom of it. Because it is an unlimited length format, it is not a problem. I have something at the bottom of mine, 'Authorised by', and it has got my name, in a very abbreviated form.

Senator CAROL BROWN: Do you have a view on whether Senate candidates should be residents of the states that they seek to represent?

Dr Bonham : Ideally, yes. I have seen concerns raised about whether it is constitutional to limit them. I mentioned in the last session of the committee that I would like to see some form of state based nominator requirement for candidates so that you get rid of any candidate who has got no support in the state that they are running in and is just nominated centrally. I do not think we should have those candidates basically filling up the ballot papers and making the ballot papers too large. I believe that the answer to the question that you asked previously is that there were, by my count, four on the Tasmanian ballot paper who were not resident in Tasmania.

Senator CAROL BROWN: Who were they?

Dr Bonham : I would have to check.

CHAIR: We have got that question on notice to the AEC here. They will know.

Dr Bonham : Yes. They will be able to answer it.

Senator CAROL BROWN: I want to ask about your activities on the day of the election. Did you scrutineer throughout the day or was it just when the count was being conducted?

Dr Bonham : No, I did not scrutineer on the day. I was appointed as a scrutineer afterwards, for the post-counting only. I did not seek to scrutineer on the day because I had far too much else to do on the day. So I did not scrutineer in the booth.

Senator CAROL BROWN: In your submission, you talk about the fact that there was an issue around the instructions people were given for Senate voting. How did you come to that view? What information came to you to suggest that was an issue?

Dr Bonham : I was gathering reports on Twitter. People started saying this, and I started saying, 'I've heard that this is an issue. Have you been hearing this?' A large number of people—this is nationwide—were saying that they had been just told, 'Vote 1 to 6,' or 'Vote 1 to 12.' There were a small number of cases—and probably I saw a handful of these through the day—where people were actually saying that they had been told only to vote 1 to 6 or 1 to 12. There were also some cases where people would report having discussions with the polling booth official, and the polling booth official did not quite know or gave incorrect advice—that sort of thing. I did sense that there was an issue there in terms of making sure that all polling staff were giving the correct instruction: 'At least this number'.

Senator CAROL BROWN: What about your experience? You voted on the day?

Dr Bonham : I was ready for this and hoping to find out the answer, but when I walked into the polling booth the person asked me: 'Are you familiar with how the new system works?' I cannot tell a lie. I said yes, and I found out no more. They gave me the ballot papers after I said yes. I do not know if they recognised me.

Senator CAROL BROWN: Maybe, Dr Bonham. You have a high profile in Tasmania.

Senator IAN MACDONALD: This is not directly to what you are talking about, but, in your scrutiny of Senate ballot papers—and I am really interested about this in Reps—are you as a scrutineer able to see declaration votes coming in in the envelope and to make a judgement, as the AEC apparently does, on whether from the outside of the envelope the ballot is properly lodged?

Dr Bonham : I did not have any involvement in that part of the process. I was commissioned as a scrutineer by a candidate who thought I would find it interesting to observe the process so that I could do more analysis of who was going to win. That was my main purpose for being there.

Senator IAN MACDONALD: I am not going to ask you but I think you might have been doing it for one of two senators in Tasmania. But I am not going to ask you; it is none of our business. I just thought you might have done the whole scrutiny bit, which involves checking validity.

CHAIR: Dr Bonham, just in relation to your recommendations and comments on authorisation: I note you were also talking about how opinion polling should have some form of authorisation as well during the campaign. Can you expand on that a bit further?

Dr Bonham : Yes. During campaigns there are a lot of polling calls—and at other times as well—made. People complete the call and then they are not sure who the call was on behalf of. Some polls contain some pretty subjective wording. It is quite common to see commissioned polls engage in what I called skew polling, which is loading the preamble to the poll to get a certain result that you want, and then you can report that so many per cent of the people agree with us. There is a sort of line where some of these polls start to run into political advertising themselves. I would just like people who answer polls to know who commissioned the poll.

CHAIR: That is actually a very interesting point—if you do not even know whether they are a valid pollster, so to speak, who has been commissioned on behalf of a third party. Where does that fall into push polling and actually just communicating, 'Do you think candidate X should stop beating their wife?' or, 'Do you think the government should get rid of its plans to cut Medicare?' or, 'Do you think the Labor Party should do something or other?' There is a bit of a blurred line there, isn't there?

Dr Bonham : I think we have very little outright push polling in Australia. The term is often misused for things that are not push polling. But I would just like that, if someone hangs on the line, at the end of the poll they should be able to hear some kind of authorisation statement—something that gives the poll some kind of identity and source.

CHAIR: 'On behalf of the Liberal Party, thank you,' or, 'On behalf of the Labor Party,' or the trade unions or whatever else. Just on that, in terms of whether you do it at the beginning or the end, the issue that was put to us last week was: where things go at the end—even like on a recorded message—people, quite often, will hang up before the very end, so they do not actually hear it. Whereas if you have it at the front, they know upfront. But, then again, that could skew the poll itself.

Dr Bonham : Yes. If you have it at the front that could skew the poll itself. If you have it at the end, even if only a small minority of people actually hear the declaration then knowledge that a poll with these questions was commissioned and authorised by so and so, whoever so and so is, will get out.

Mr MORTON: In relation to the simple sequence of the House of Reps and Senate votes and where someone leaves out a particular number: I have been made aware by reading other submissions that the number four carries quite a meaning for people of Chinese heritage in that some ballot papers are marked one, two, three, five, six because of the avoidance of the meaning of the number four for people of that heritage. That is why I think this is a very good issue to look at. In your experience, are there any other particular cultures, areas where there is a justification for the avoidance of a particular number or where sequencing would actually allow more votes to be counted as formal?

Dr Bonham : No, I was not actually aware of that one before I read it in one of the submissions. I think it was Jeremy Buxton's submission 12. I had noticed previously that there were issues associated with the No. 4. In the 2013 Senate ballots, I noticed when I was doing some analysis of the numbers that No. 4 is where you start to get problems quite a lot of the time. A high number of people put 44 instead of four, or whatever. I do not remember the exact details of that.

Mr MORTON: But certainly from the figures there, sequencing is an issue that does invalidate a number of votes. Thank you for that, Dr Bonham.

CHAIR: I have a question about the real-time virtual tally room in terms of your recommendation:

… one display of votes for each state that only includes those booths that have been fully processed …

Can you explain the thinking behind that for us.

Dr Bonham : The Tasmanian Senate race was a race not just between parties but also between candidates. Because of the very high below-the-line votes for Lisa Singh and Richard Colbeck, to try to project the outcome when it involved not only who would get elected within each party but also how many seats each party would get, you needed to know where the candidates stood. But because of the way in which the votes were being unfilled through the process—they start out as unapportioned and then they are distributed as above-the-line or ticket votes as they are called or below-the-line votes for individual candidates, what was happening in a lot of cases was that the votes on that particular booth were going through to the total. So, the party votes from a booth were going through to the total but the below-the-line votes for candidates from that booth were not going through to those candidates in the total that was appearing. As a result, it was looking like these candidates were on much lower votes than they actually were. We were projecting them to finish on four per cent, six per cent or whatever they finished, yet there were lots of people saying, 'So-and-so is only on one per cent.' It was because of the confusing way that different polling places were being added to the total.

Mr GILES: Dr Bonham, I might ask you to draw out your fifth recommendation, which is:

That the costs and practicalities of rotating the order of parties on Senate ballot papers be investigated.

As someone who is not from Tassie or the ACT, I am just wondering if you could first say whether such an investigation has been done previously to the best of your knowledge.

Dr Bonham : I am not aware of one. That is not to say it has not happened, but I am not aware of it.

Mr GILES: Do you have any initial thoughts on any of the practical difficulties in implementing such a regime?

Dr Bonham : If you bring in rotation, even a simple form of rotation, you have to print different designs of ballot paper. The computer scanning system is going to have to be able to recognise different kinds of ballots. It is slower for manual counting in booths on the night, because the candidates are not always in the same place every time on the ballot paper—these sorts of things. The AEC would have to provide advice on costs. I would have no idea of whether it is practical or not to bring in some form of rotation. The main reason I raise the issue is that, when you have a long list of candidates, preferences will tend to flow to nearby candidates to the preference source. So someone might gain a big advantage from being next to a really good preference source, and that is probably a bigger deal than donkey voting.

Mr GILES: I saw that elsewhere in your submission. Obviously, you touched upon evidence to the effect that micro-party preferences seem to be very poorly followed in terms of the how-to-vote recommendation. This appears to be a way through other distortions in the preferencing process, but it also appears to me that the practical difficulties at a federal level would be very considerable.

Dr Bonham : Yes, I think it is worth investigating to find out whether it is practicable or not. I do not know if it is.

Senator WATT: Is there a higher rate of informality where that does occur?

Dr Bonham : Where—?

Senator WATT: Where there is a rotation of ballot papers either here or internationally?

Dr Bonham : I am not aware of it being an issue. I do not think the informal rate in Tasmania is all that high. A lot of the informal rate that we do have results from savings provisions issues—the same problem that you have to get 1 to 5, once and once only, and you cannot make a mistake in that distance.

CHAIR: Unfortunately, our time has expired. We have gone a little bit over time. Dr Bonham, thank you very much for your attendance here today and for your very considered and very helpful insights for this committee on a range of issues. We have given you a little bit of homework for additional information. We would be grateful if you could provide it to the secretary by 25 November. You will be sent a copy of the transcript of your evidence and will have an opportunity to request corrections to transcription errors. Thank you again.

Proceedings suspended from 11:11 to 11:23