Change that counts

Supplementary Submission to the Victorian State Parliament
Electoral Matters Committee

Functions and Administration of Voting Centres

Review of the report submitted by the Victorian Electoral Commission Submission 4a - dated May 27, 2010

 

By Anthony van der Craats

 

May 31, 2010



Supplementary submission

 

This submission paper seeks to address a number of issues and concern in relation to the paper published by the Victorian Electoral Commission dated May 27 2010 – Functions and Administration of Voting Centres.

The Victorian Electoral Commission gave an undertaking to fully disclose information related to the conduct of the election

On Page 8 of the report the Commission indicated that “A Full account and Analysis of the election (2006) were documented in the VEC‘s Report to Parliament”. This statement is false and misleading.  Information was missing and data supporting a full account of the elections results was not provided.  No detailed explanation has been given to explain the changes in the total number of ballot papers recorded at the primary count in comparison to the Secondary Final count for of the Western Metropolitan Region and the fact that the total number of votes does not tally with the information tabled in relation to the voting centre returns and the corresponding lower house districts. Whilst the errors in data entry in relation to the Northern Region had been explained as a result of a lack of due diligence and failure of the electoral commission to reconcile ballot papers recorded with the expected number of ballot papers listed in the voting centre returns.

Copies of the preference data file pertaining to the primary count had not been made available and the data had been destroyed/overwritten with no data backup process in place.  The Commission failed to respond in a timely fashion in providing copies of the preference data-files as requested prior to the commencement of the count.  As a result of omissions and errors by the Commission full disclosure and independent analysis was unable to be undertaken.  The report by the Commission does not address this issue or the issue of data protection and security of electronic data.

Most important the commission has not outline the process and availability of access to copies of electronic preference data other then limited access to printed reports.  Access to copies of the preference data file is essential to enable independent analysis and proper scrutiny of the transcribed data entry.

The Electoral Commission has not outlined in a comparison chart the costs and man hours required to undertake a data-entry electronic counting procedure as compared to a manual counting procedure. The number of scrutineers allowed per candidate is restricted to the number of data entry personal and does not take into consideration the number of support staff undertaking other administrative duties in preparation for the data-entry process.   Whilst many tasks can be parallel process they come at the cost of a loss of openness and transparency as the information recorded cannot be readily verified by scrutineers.  Provision of the preference data files allows Scrutineers to undertake independent analysis and review of the data files and the ability to properly scrutinise the results of the election.

 

Comments in relation to the Victorian Electoral Commission’s report - Appendix 3 – Upper House Count

Introduction

Election manager

Ambiguity and uncertainty exists as to who holds the position of Returning officer as outlined in the Election Act. No explanation has been provided as to why there is not a duly appointed Returning Officer who is responsible for each lower-house district.

There appears to be an excessive middle layer of election officials between the chief Commissioner and the eight regional returning officer/managers.

It is noted that the commission is proposing a double data-entry process (something that was not undertaken in previous elections) this proposal is welcomed as it should improved on the quality and accuracy of the data-entry process. The Commission has not outlined in detail how this process will be undertaken.   Access to copies of the electronic preference data-files for each data-entry process must be made available to scrutineers with certified copies published at the conclusion of each stage of counting and prior to the commencement of the subsequent stage.  Without access to this information it is impossible for scrutineers to verify to the full extent the validity and accuracy of the recorded data. (See below)

Count Overview

Proposed commencement of counting postal votes

The Commission has proposed to commence the counting of postal votes at 6PM on Election Day. 

The early commencement of counting postal votes will limit the number of resources available to provide proper scrutiny as candidate’s resources are generally tied up with overseeing the conduct of voting centres. 

Given that the number of postal votes may not have been received by the Commission It is recommended that the counting postal voting commence no earlier then the Sunday following the Election Day.

Voting Centres Election Night

Publication of Voting Centre return statistics

The voting centre return statistics including the tally of Upper house above the line and below the line first preferences should be published on the Commission’s Internet site as a matter of course on election night.  This would allow for scrutineers and election official to reconcile the number of ballot papers issued and returned prior to being transferred to a regional centre.  This information would also allow candidates to plan and allocate resources required for further scrutineering of the elections results at later stages of the count.  It will also provide a means of ensuring that all ballot papers have been accounted.  The Victorian Electoral Commission failed to publish this information during the 2006 Election and as a result it prevented scrutineers and election officials from reconciling the number of votes with the number of ballot papers recorded in the computerised counting process.

Information published on the Commission’s Internet site should also include statistics on the number of Formal, and informal votes the total number of ballot papers issued for each voting centre.  

In addition each voting centre should also include statistical information for absentee ballots issued for each district. This information is essential in order to ascertain in advance the number of absentee ballot papers that are outstanding and waiting to be received by the district office.  This information should also be available as part of the published voting centre statistics.

All of the above information should be readily available as it should be included in the Commission’s EMS database and form a part of the voting centre return/declaration.  If it is not collected or included in the Commission’s EMS system then changes to the system should be required to ensure that the above information is collected and available on election night.  Copies of the voting centre summary report should be provided to scrutineers.

The Commission’s report failed to include or mention of the right of scrutineers to place seals on the parcels of ballot papers being transferred to central counting centre. 

 

 

District Election Offices

Primary Counts

Primary count s of absent, early, postal and unenrolled sealed ballots other then the tabulation of the number of sealed ballots received should not commence on Election Day but be deferred until the following day so as to allow for the presence of scrutineers if required.  No sealed ballot papers should be opened in the absence of appointed scrutineers.

Region Recheck Centres

Copies of the reconciliation report, including any changes or alterations to the voting centre declaration should be recorded and copies of the report made available to scrutineers and published on the Commission’s Internet site.

Computer Count Process

Prior to the Count

Copies of the reconciliation report, including any changes or alterations to the voting centre declaration should be recorded and copies of the report made available to scrutineers and published on the Commission’s Internet site it is important that reconciliation issues are resolved prior to the commencement of data-entry.   The total number of votes recorded should not change between counts.

The appointment of a Reconciliation Coordinator is noteworthy and hopefully will go a long way to avoiding the lack of due diligence undertaken during the 2006 Election count.  Scrutineers or Candidate’s representatives should have access to all the reports and reconciliation documentation as signed off by the Reconciliation Coordinator prior to data-entry commencement.

During the Count

The Commission has proposed to undertake a double entry counting process in the first instance.

This is a welcomed change, although highly resource intensive, and should help prevent the mistakes that occurred in the 2006 Counting process.  However the report does not provide details as to how the double entry comparison will function. 

It is desired that second round of the double counting process should not commence until the completion of the first round of processing.  It is noted that the Commission has proposed this in its outline.  (Paragraph 3 page 5)

As the number of primary voters should already be determined Scrutineers should be able to formulate a strategy plan to monitor more closely the data-entry of candidates’ votes that be of interest.

Scrutineers right to access and obtain a copy of the electronic preference data files.

There is ongoing concern that the report submitted by the Commission does not make mention of access and provision of copies of the data-file in electronic format.  This needs to be clearly spelt out as the Commission has been relucent or unable to provide copies of this information in past elections, both State and Municipal.

Copies of the preference data files allow Scrutineers to monitor and identify any changes in the recorded preferences and if necessary single out any ballot paper requiring further examination during the second data-entry process.

Scrutineers must be provided with copies of the preference data files at regular intervals during the count and at the conclusion of the first stage of the data-entry process.  Scrutineers should be able to process the data-files and generate independent reports highlighting any ballot papers that may require further examination.  The provision of a copy of the data files should allow for independent analysis and review. This must not be dependent or restricted to reports provided by the Commission. 

Certified copies of all data files should be published on the Commission’s web site as and when they become available.

Format of preference data-file

The Commission needs to provide information on the file format which ideally should be in an XML file designed so as to provide a reference number to assist in indentifying the correct batch number and ballot paper.

Backup copies of data files.

The Commission must ensure that certified backup copies of all data files are made and kept prior to the commencement of the second data-entry process. In the 2006 election the Victorian Electoral Commission failed to ensure that copies of the primary data-entry count were maintained. The deletion of this information and the failure of the Commission to provide Scrutineers copies of the preference data files seriously undermined confidence in the election process.  

Second round of data-entry.

The second round of data entry verification should not be allowed to proceed until digitally certified copies of the electronic preference data are published and or made available to scrutineers and backup copies made

It is important that the Reconciliation Coordinator is able to produce a running tally sheet as to identify the batches that have been completed and those that are in the process of or waiting to be processed throughout the count.  This report should be available to Scrutineers in request.


Above the line batching

There is some confusion as to the system proposed by the Commission in relation to the processing of ‘Above the line’ Ballots. It is assumed that the Commission is proposing that these ballot papers be collated and entered into the Commissions EMS system in multiple batches.  It is difficult to ascertain on what basis this needs to be segmented and batched as outlined by the Commission.

It should be possible for the Commission to collate and tabulate the ‘Above the line’ votes manually and then record the total number of above the line votes allocated to each group as opposed to segmenting each group into parcels of 1000 or less ballots.   The system proposed by the Commission certainly does not appear to be efficient.  This process would need to be observed more closely.

Each ‘Above the line’ ballot paper should not require a separate data entry process.

There should be no need to determine formality of ‘Above the line’ votes other than to ensure that no more than one group has been marked on the ballot paper. The allocation of preferences is determined by the registered ticket and as such the number of ballot papers should be able to be batched entered with the total number of ballot papers allocated to each corresponding group.

Calculation of the result

The Commission must ensure that the total number of ballot papers recorded in the data entry process including the number of informal votes reconciles with the expected total number of a ballot papers issued and returned. This is something that was not undertaken during the 2006 State election. The number of ballot papers recorded in the 2006 Western Metropolitan Region did not reconcile with other published information, including voting centre declarations, the number of ballot papers allocated to the corresponding lower house district and/or the total number of votes recorded on the first count as compared to the second count.  There was a discrepancy in the information provided by the Electoral Commission of up to 450 votes that could not be reconciled.

Certified copies of the data files must be published and copies distributed to Scrutineers prior to the calculation of the result of the election.

Scrutineers and Transparency

The Commission had failed to mention the right and ability of scrutineers to obtain copies of the preference data files in electronic format periodically on request.  

Without access to this information Scrutineers are prevented from independently reviewing the quality of the data-entry process and or the conduct of the election count.  The Parliament must ensure that this information is readily available and that copies of the preference data-files are published and available for public review.

 

Number of scrutineers entitlements

The statements in the Commission’s report are misleading

There is ongoing concern as to the method used to determine the number of scrutineers allowed per candidate.  The Commission’s report indicates that there is an entitlement for one Scrutineer per Candidate for every election official.  This is not what is currently listed in the regulations. The number of scrutineers is limited to the number of data-entry operators.  With further limits applied to the monitoring of other administrative process.  In many cases there are more election officials then Scrutineers are permitted.

The Commission should be required to outline in full a list of election officials that will be engaged in the election process in order that the entitled number of scrutineers can be determined.   The Commission also need s to ensure that the facilities provided for the counting of the vote can accommodate the entitled number of scrutineers in order to meet Occupational health and safety requirements.

Feedback from Scrutineers involved in the 2008 City of Melbourne Municipal election indicated that there was insufficient room and ventilation provided in the counting room to accommodate the entitled number of Scrutineers.

 

Recounts

There are questions as to the published threshold of 80 votes as recommended and outlined in the by the Electoral Commission’s report (Paragraph 2 page 31) that would trigger a recount.  This figure is arbitrary and too low and should be reviewed,

Any threshold for consideration of a recount should be expressed as a percentage of the vote and not a number.

Consideration should be given to allowing a recount at the request of any candidate or party provided that the variance between any one candidate falls within 0.5% of the total vote.

In 2006 the Western Metropolitan Region’s primary count indicated a difference of 150 votes between the winner of the fifth place candidate and the sixth place contender. Following a recount of the ballot there were 450 less ballot papers record in the second count then were accounted for in the primary count. The results of the election had changed following the second recount of votes. As copies of the primary count preference data file were not provided it is impossible to fully ascertain where the discrepancy or changes in the outcome of the vote occurred and the full extent of errors in the primary data-entry process.  Errors that may have occurred in the second data-entry process were not tested.

If as a result of a recount the overall election results change as was the case in the 2006 Western metropolitan election then consideration should be given on request for a third recount to verify the accuracy of the second recount.

The mistakes identified in the 2006 Northern Metropolitan election arising from a lack of due diligence undertaken by the Victorian Electoral Commission staff should not occur as a result of the proposed requirements to reconcile the total number of ballot papers with the information recorded in the Voting Centre declarations.

A candidate or party requesting a recount should be able to demonstrate under what circumstances a change in the allocation of preferences would produce a change in the outcome of the election.

The availability and publication of the transcribed detailed preference data-files would allow Scrutineers to identify potential thresholds in the outcome of the election and as such provide justification and merit for a subsequent recount. It also provides a means of avoiding unnecessary recounts if the overall outcome of the election is unlikely to change.  

 

Other issues

Statistical information

Prior to Election Day

General Postal Voting (GPV)

The Commission should be required to publish the progressive number of postal vote application per district/region that have been issued and returned on a daily basis.

Early Voting in Centres

The Commissions should be required to publish the progressive number of Early Vote applications per district/region/Voting Centre that have been issued and returned on a daily basis.

Mobile Voting Centres

The Commissions should be required to publish the progressive number of Mobile Voting applications per district/region/Voting Centre that have been issued and returned on a daily basis. There is concern that the period of Mobile voting has been extended to two weeks.  Consideration should be given to transferring mobile voters to the GPV system.

Mapping Software

The commission should be required to publish on its web site the detailed report and statistics outlined in its report.  This information should be publicly available and should not require an application made pursuant to the Freedom of Information.

Electronic Roll Marking

Information should be provided in relation to the proposed Electronic roll with data made available to scrutineers periodically during the conduct of the election.