Parker v Amalgamated Society of Carpenters and Joiners of Australia

[1992] FCA 538

Re: JOHN PARKER
And: AMALGAMATED SOCIETY OF CARPENTERS AND JOINERS OF AUSTRALIA
No. N I1 of 1992
FED No. 538
Industrial Law

COURT

IN THE FEDERAL COURT OF AUSTRALIA

NEW SOUTH WALES DISTRICT REGISTRY
INDUSTRIAL DIVISION
Wilcox J.(1)

CATCHWORDS

Industrial Law - Registered organisation - Proposed amalgamation with other registered organisations - Ballot of members - Application for inquiry into alleged irregularities - Irregularities established: some members wrongly deprived of the vote, associate members wrongly allowed to vote - Whether result of the ballot "may be affected" by the irregularities - Principles to be applied in determining that question - Relevance of fact that a major proportion of disenfranchised members came from one Branch - Ballot set aside.

Industrial Relations Act 1988, s.253M

HEARING

SYDNEY

#DATE 31:7:1992

Counsel for the Applicant: S. Rothman

Solicitors for the Applicant: Taylor and Scott

Counsel for the Respondent: M. Adams, QC and J. Phillips

Solicitors for the Respondent: McClellands

ORDER

THE COURT ORDERS THAT:

1. The declaration of the ballot made by Michael E Small on 19 December 1991 concerning the possible amalgamation of the Amalgamated Society of Carpenters and Joiners of Australia with the Federation of Industrial Manufacturing and Engineering Employees, the Australian Brushmakers' Union and the Australian Rope and Cordage Workers'Union be set aside.

2. A fresh ballot be conducted in relation to the possible amalgamation.

3. The respondent, the Amalgamated Society of Carpenters and Joiners of Australia, within 21 days of this day apply to the Australian Industrial Commission for directions concerning the conduct of the fresh ballot.
Note: Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.

JUDGE1

This is an application for an inquiry regarding a ballot of members of a registered organisation, the Amalgamated Society of Carpenters and Joiners of Australia ("the ASCJA"), upon a proposal to amalgamate that Society with the Federation of Industrial Manufacturing and Engineering Employees, the Australian Brushmakers' Union and the Australian Rope and Cordage Workers' Union. The ballot was authorised by Deputy President Williams in the Australian Industrial Relations Commission on 30 October 1991 and conducted by Mr M E Small, a Returning Officer employed by the Australian Electoral Commission.

1. The result of the ballot was declared by Mr Small on 19 December 1991. The certificate made by him on that day gave the total number of ballot papers issued as 6101 and the number of ballot papers returned for scrutiny as 2206; of which 20 were rejected at preliminary scrutiny leaving 2,186 for final scrutiny. Of those 2186 ballots, 1589 favoured the proposed amalgamation, 552 were opposed and 45 were informal. However, as a result of a check made during the course of this proceeding, the Australian Electoral Commission now states that the figure of 6101 issued ballot papers was erroneous. Apparently this figure included some ballot papers issued to members of two of the other organisations involved in the proposed amalgamation. The correct figure of issued ballots is 5581. Nothing turns on this error. The other certified figures are correct.

2. The Application initiating this proceeding was filed on 17 January 1992, the applicant being a member of the ASCJA, Mr John Parker. The Application makes no criticism of the work of Mr Small or the Australian Electoral Commission. The complaint arises out of the roll of members used in the conduct of the ballot, it being said that some members who were entitled to vote were denied the opportunity to do so; whilst some other persons, not entitled to vote, were admitted to the ballot.

3. When the Application first came before the Court on 31 January 1992, both the applicant and the ASCJA were represented by counsel. They agreed that their respective clients and instructing solicitors would cooperate in ascertaining the extent of any errors in the roll. They have done so, with the result that substantial agreement has now been reached on these matters, to the extent that they are capable of ascertainment. Although there is some uncertainty about the numbers, there is no matter of fact which I need to resolve. The contest between the parties relates solely to the application to the agreed facts of s.253M of the Industrial Relations Act 1988. I am grateful to the parties and their legal advisers for the time and effort they have expended in investigating the position regarding members wrongly excluded from the ballot and persons wrongly included. This work has saved considerable court time. I hope that I will not be thought ungracious if I say that I am concerned that this process took so long: about five months. All litigation should be conducted efficiently and completed as quickly as possible, consistently with the overriding imperative of fairness. But, as it seems to me, litigation concerning the amalgamation of registered industrial organisations has a special claim for expedition. Until that litigation is completed, uncertainty must exist as to the future of the proposed new organisation. This uncertainty affects not only the members of the organisation whose ballot is under question but also the members of other organisations involved in the proposal. Delay may cause serious problems.

4. In fairness to the present parties and their legal advisers, I should say that, in this case, there were real problems in obtaining the necessary information. During the five month period, the matter was before me on several occasions. On each occasion I was informed of the steps being taken to ascertain the facts. The parties agreed about the need for expedition and I am sure that they did what they could to finalise their investigations as quickly as possible. As I understand it, their problem was the manner in which the organisation's membership records are kept. Apparently, with the exception of those of the Victorian Branch, the membership records are kept manually, the system varying from Branch to Branch. Even in the Victorian Branch, where the membership records are on a computer base, they do not readily disclose the date upon which a particular person's membership was renewed. It was necessary to resort to receipts and other financial records. This problem appears to be not uncommon amongst registered organisations. No doubt one result of the increasing professionalism of union management, especially in the larger organisations, will be a greater resort to electronic data systems, standardised throughout the country and so programmed as readily to provide full information about each member's relationship with the organisation.

5. Under the ASCJA rules membership dues are payable quarterly or yearly in advance, the membership year commencing on 1 October. The roll of voters provided by the organisation to Mr Small for use in the amalgamation ballot was compiled on the basis that it included only members who owed nothing to the organisation by way of fees, dues or levies as at 30 October 1991 - the date fixed by Deputy President Williams as the commencing day of the ballot. The list also included associate members. It is agreed between the parties, I think correctly, that the roll erred in both these respects. First, rule 10(a) of the ASCJA provides that a "member who owes to the Society any fees or dues for a longer period than one calendar month after same has fallen due", or any levy or fine for a longer period than one calendar month after the due date of payment, "shall be deemed unfinancial and while he so remains unfinancial he shall be excluded from all privileges (including the right to vote), but not from the obligations of membership". Accordingly, a member who was financial until 30 September 1991, but who had not by 30 October 1991 paid membership fees for the year commencing 1 October 1991, was not yet ineligible to vote. Yet the roll supplied to Mr Small excluded all those members.

6. Despite the endeavours of the parties and their solicitors, the exact number of persons wrongly excluded from the ballot remains obscure. The first investigation made by Mr S J Boatswain, the applicant's solicitor, suggested a figure of 2101 (1543 in Victoria, 454 in South Australia, 84 in New South Wales and 20 in Tasmania). The ASCJA has no branch in Western Australia. It has Queensland members, at least at the present time. In an affidavit made by him, Mr Boatswain referred to a statement in the Industrial Relations Commission by Mr Ian Buckley, who identified himself to Deputy President Williams as "the agent in Queensland for the ASCJA", that the organisation had about 500 to 600 members in that State. Mr Boatswain said that no roll of Queensland members was provided to Mr Small. So far as I am aware, there is no separate Branch of the organisation in Queensland and, it seems, Queensland members were not included in the ballot. On the basis of this evidence, the figure of 2101 appears an underestimate. However, evidence was given by Mr Michael Borowick, an industrial officer employed by the organisation, that the Queensland membership applications were not handed to the Federal Secretary until February 1992. If this is so, they were not members at the time of the ballot and their exclusion was not erroneous.

7. After his first affidavit, Mr Boatswain found an additional 594 Victorian members whom he thought to have been incorrectly excluded from the ballot. This figure was challenged by Mr Borowick. Mr Borowick gives reasons, which seem to me to be good, for reducing the figure of 594 to 410. Even so, this leaves a minimum of 2511 members who were wrongly excluded from the ballot; that much is common ground.

8. Secondly, it was erroneous to include associate members on the voters' roll. Rule 37 provides for the grant of associate membership to members who retire from the trade of carpentry and joinery because of incapacity or age. This grant is subject to certain conditions, including that the associate "shall not be entitled to vote on any matters before the Chair, nor be entitled to hold any official position or vote in any election for any official position or vote in any election for any official position in the Society". The parties agree that at least 34 associate members were wrongly admitted to the ballot.

9. The figures certified by Mr Small show that the majority in favour of amalgamation was 1037 (1589 less 552). If all those who were wrongly deprived of a vote had voted, and in the negative, the amalgamation proposal would have been soundly defeated. The issue of the ballot to the associate members could not itself have affected the outcome. However, to the extent that they did in fact vote, they may have affected the extent of the majority for "yes". If more voted "yes" than "no", the result of their inclusion in the ballot was to increase the "yes" majority; conversely, of course, if more voted "no" than "yes". There is no way of knowing how many associate members did vote, or in what manner.

10. Section 253M is included in Division 7 of Part IX of the Industrial Relations Act. Part IX deals with registered organisations; Division 7 is concerned with amalgamations. Subdivision E (ss.250-253P) contains a detailed procedure for approval of amalgamations. This procedure includes the conduct of a secret postal ballot of members of affected organisations (s.253J). Section 253K specifies the level of acceptance necessary for the proposal to proceed. Where there is a community of interest declaration in force in relation to the proposed amalgamation, pursuant to s.241 of the Act, the only requirement is that more than 50% of the formal votes favour the amalgamation. Where there is no s.241 declaration, an additional requirement applies: at least 25% of the members on the roll of voters must vote. It appears from the reasons for decision of Deputy President Williams that, in the present case, there was no s.241 declaration. The votes admitted to final scrutiny, 2186 out of 5581, comfortably achieved the 25% participation requirement. If the figure of 5581 is adjusted by deleting therefrom the 34 ballots sent to associate members and adding 2511 ballots which should have been sent to ordinary members, the 25% requirement would still be met, although only by a small margin. This statement remains true even if one assumes that all of the associate members voted, so that 34 votes should be deducted from the participation figure of 2186. If it is correct that, on top of the agreed 2511 excluded votes, there was an exclusion of some Queensland members, the 25% requirement might not have been fulfilled. However, the evidence does not satisfy me that there were any wrongly-excluded Queensland members.

11. Section 253L deals with a further ballot if an amalgamation is not approved. It is not presently relevant. Section 253M is, however, relevant; indeed critical to the case. That section reads:
    

"253M(1) Not later than 30 days after the result of a ballot under this Division is declared, application may be made to the Court, as prescribed, for an inquiry by the Court into alleged irregularities in relation to the ballot.

(2) If the Court finds that there has been an irregularity that may affect, or may have affected, the result of the ballot, the Court may:

(a) if the ballot has not been completed - order that a step in relation to the ballot be taken again; or

(b) in any other case - order that a fresh ballot be conducted in place of the ballot in which the irregularity happened; and may make such further orders as it considers necessary or desirable.

(3) The regulations may make provision with respect to the procedure for inquiries by the Court into alleged irregularities in relation to ballots under this Division, and for matters relating to, or arising out of, inquiries."

1. There were two irregularities in relation to the subject ballot: the exclusion of at least 2511 persons who were entitled to vote and the inclusion of 34 persons who were not entitled to participate. Given the affirmative vote's margin, the latter irregularity could not have affected the outcome. But it is obvious that, theoretically at least, the first irregularity may have done so. If a sufficient number of the excluded persons had chosen to vote in opposition to amalgamation, the proposal would have been defeated.

2. Counsel for the ASCJA read an affidavit of Mr Ian Nivison-Smith, a statistician formerly employed by the Australian Bureau of Statistics but now in private practice. Mr Nivison-Smith made a number of calculations based upon differing assumptions. One assumption was that 2511 persons were wrongly excluded from the ballot. He said that, from the established response rate of 38.3% amongst those to whom ballots were sent, "it would be expected that 963 of the persons allegedly excluded would have returned a completed ballot paper. With the established proportion of 'yes' and 'no' votes, it would be expected that 715 would have voted 'yes' and 248 would have voted 'no'". He said that the "probability of achieving 1,038 or more 'no' responses in this case is very small, much less than 0.02%". On the basis of this evidence, counsel for the ASCJA argue that the Court should not find that the exclusion of the 2511 votes "may have affected" the result of the ballot. This submission raises the question of the meaning and application of the phrase "may have affected" - a subject upon which there is a deal of authority.

3. Section 253M was added to the Industrial Relations Act only in 1991. It had no counterpart under the Conciliation and Arbitration Act 1904. But the phrase "may have been affected" is used also in s.223(4) of the Industrial Relations Act, in connection with inquiries into elections for offices in registered organisations. It was earlier used in s.165 of the Conciliation and Arbitration Act. In those contexts courts have had to consider what is involved in saying that a result "may have been affected" by an irregularity. In Australian Timber and Allied Industries Union; ex parte Black (1991) 39 IR 106 Ryan J applied the election inquiry cases to a disputed amalgamation ballot. Having regard to the similarity of the language used in s.223(4) and s.253M, I agree that this course is appropriate.

4. The election inquiry decisions should be read against the background of what Griffith C.J. described, in Chanter v Blackwood (No.2) (1904) 1 CLR 121 at 129, as "the Common Law of elections". That case concerned a disputed Parliamentary election. The petitioner challenged a declaration of the respondent's election. He was successful in obtaining a recount, which gave him a majority of 67 votes. However, it transpired that the recount included the votes of 91 persons who had no right to vote and that 2 persons entitled to vote had been deprived of the opportunity to do so. Additionally, there were a number of absentee votes left uncounted because of a mistake of the electoral officers. If counted, their net effect would have been to reduce the petitioner's majority to 55. Griffith C.J. applied a test stated by the Court of Common Pleas in Woodward v Sarsons (1875) LR 10 CP 733 at 744: whether "a majority of the electors may have been prevented from electing the candidate they preferred". (Original emphasis). His Honour made no inquiry into probabilities. He said at 131:
   

"... the numbers being as they are, it is impossible for me to say that the majority of the electors may not have been prevented from exercising their free choice".

Emphasising the point that a theoretical possibility of a different result was enough to require the setting aside of the election, Griffith C.J. commented that, if the number of excluded votes had been only 56, "the result would have been the same".

1. In Kean v Kerby (1920) 27 CLR 449, another Court of Disputed Returns case before a High Court judge, Isaacs J disagreed with the test applied by Griffith C.J. in Chanter v Blackwood, saying that it was not enough that the result may have been affected. He referred to s.194 of the Commonwealth Electoral Act 1918 which provided that an election shall not be avoided on account of an error of an officer "which shall not be proved to have affected the result of the election".

2. It is not relevant for present purposes to consider whether the test applied by Griffith C.J. correctly reflected the Commonwealth Electoral Act. But it is interesting to note the difference between Griffith C.J.'s application of that test and the decisions in this Court applying the test propounded by the two industrial statutes. According to the latter decisions, it is not enough to find a mere theoretical possibility that, absent the irregularity, the election (or ballot) may have resulted differently.

3. The early cases in this Court rejecting the simple mathematical approach were referred to by Keely J in Re Vehicle Builders Employees' Federation of Australia (SA Branch) (1987) 13 FCR 350 at 354-355. They are Troja v Australasian Meat Industry Employees' Union, Victoria Branch (1978) 46 FLR 340; Re Australian Postal and Telecommunications Union (New South Wales Branch); ex parte Wilson (1979) 28 ALR 330; Kelly v Amalgamated Metal Workers' and Shipwrights Union (1981) 56 FLR 124; Re Penhallurick (1983) 5 IR 470; Re Bragg; ex parte Australasian Society of Engineers, South Australian Branch (1985) 6 FCR 304; and Re Ferguson; re Inquiry into Election in Australasian Meat Industry Employees Union, Western Australian Branch (1986) 17 IR 208. In this last case ballot papers were sent to 110 ineligible persons. There was no evidence as to how many of them voted. The overall participation rate was 54%. The margins between the successful and unsuccessful candidates, in relation to particular offices, varied from 68 to 264. Referring to the language of s.165 of the Conciliation and Arbitration Act, Toohey J said at 210:
   

"Clearly the court is not required to make a positive finding but, equally, the court is looking at real not merely theoretical possibilities. I am not to be taken as accepting as a fact that no more than 50 of the 110 ineligible persons voted. I say no more than that I am not of opinion that the result of the election may have been affected by the existence of 110 ineligible persons to whom ballot papers were sent."

1. Of course, in that case, for the election of even the candidates with the lowest margin to have been affected by the irregularity, not only would the ineligible voters have had to participate at a rate markedly above the overall average; they would have needed to have voted overwhelmingly for the unsuccessful candidates having the greatest number of votes.

2. In the Vehicle Builders' case Keely J applied the test stated by Toohey J, "real not merely theoretical possibilities", but with a different result. In that case the winning margin was only 27 votes. It was proved that 59 voters were wrongly given ballot papers. Eight people were wrongly denied ballot papers. In the context of those numbers, Keely J was not prepared to place weight upon evidence that the overall participation rate was only 30%, so as to conclude that only 30% of the 59 actually voted and that only 30% of the eight would have voted.

3. The "real and not merely theoretical possibilities" approach was adopted by Gray J in Re Patterson; re Association of Railway Professional Officers of Australia (1987) 19 IR 373, although his Honour held that there was a real possibility of affectation and avoided the election. I applied it in Re Transport Workers Union of Australia, New South Wales Branch; ex parte Edwards (1990) 33 IR 436, in relation to the denial of voting papers to 1201 members included on a list of members with whom the organisation had at some stage lost contact. These people were referred to in evidence as the "O" coded members. In relation to some positions, where the winning margin was high, I held that the irregularity would not have affected the result; in others, that it may have done so. As that exercise demonstrated, the application of Toohey J's test to the facts of a particular case involves, not merely a mathematical calculation,but the making of a judgment in the light of a number of factors. The most important factor will always be the relationship between the winning margin and the number of votes infected by irregularity. But the participation rate and the pattern of voting, so far as this may appear from the evidence, will always be material. As it seems to me, the smaller the number of irregular votes, the greater the difficulty in a court feeling confident that the relevant voters acted, or would have acted, in line with the total electorate. The smaller the margin, the greater the difficulty in assuming that it would not have been bridged by atypical voting.

4. There is a further factor, one which I noted briefly in the Transport Workers case. If a conclusion that an irregularity did not affect the outcome of an election is to be drawn from evidence as to the voting behaviour of members who were provided with ballot papers, it must be shown that the excluded, or improperly included, voters were typical of the whole. In the Transport Workers case it was possible to draw that inference. The excluded voters were persons, on a separate list of members, with whom the organisation had at some stage lost contact. But it seemed that, in every other respect, they were a random sample of the whole membership. As I noted at 457, the "O" coded members did not "share any particular electoral allegiance or come from any particular geographical area".

5. As I have already mentioned, Ryan J applied the election inquiry decisions to s.253M in Australian Timber and Allied Industries Union. His Honour also applied the "real not merely theoretical possibilities approach" in rejecting the submission that the outcome, favouring amalgamation, may have been affected by the proved irregularities. In coming to that conclusion, Ryan J commented that, for the result of the amalgamation ballot to have been affected by the errors, "one has to assume that almost two-thirds of the 478 members who were disenfranchised would have cast valid votes all of them against the proposed amalgamation, and similarly, that almost two-thirds of the 956 persons who wrongly received ballot papers had cast effective votes, all of them in favour of amalgamation". The evidence revealed a participation rate amongst members of the organisation's four Branches ranging from 31.3% in New South Wales up to 40.1% in Tasmania. Supportive of his Honour's ultimate conclusion were the facts that most of the persons wrongly admitted to the ballot were members of the Victorian Branch (661 out of 956), the Victoria members having voted by a margin of 59.6% to 36.6% against amalgamation, whereas most of the excluded members were from New South Wales, where amalgamation was favoured by a margin of 62.2% to 35.5%.

6. In the light of the decisions in this Court, counsel for the ASCJA submits that it is not enough, in the present case, that the established irregularities may theoretically have affected the result of the ballot. Relying primarily on Mr Nivison-Smith's evidence, they say that there is no real likelihood of affectation; so the Court should not be satisfied that the result of the ballot "may have been affected".

7. Whilst I am grateful for his mathematics, I do not think that the type of analysis undertaken by Mr Nivison-Smith advances the resolution of the question whether an irregularity may have affected the result. As he conceded in his oral evidence, this analysis assumes that the excluded voters would have behaved in precisely the same way, both as to participation and manner of voting, as those who did vote. As Toohey J pointed out, that type of assumption cannot be made as a conclusion of fact. No-one can know what would have happened if the relevant irregularities had not occurred. Of course, once this assumption is made, it inexorably follows that the outcome could not have been affected by the irregularities. It is not necessary to resort to mathematics.

8. I agree that there are circumstances in which the results obtained in a poll of a proportion of a larger electorate will enable confident predictions of the attitude or likely conduct of the whole. This principle underlies public opinion polls. But the results of a smaller poll should only be extrapolated to a larger population when it is clear that the polled group is fairly representative of the whole. In the present case, there is no evidence that this is so. Indeed, it is unlikely to be so. It appears that 1953 of the 2511 excluded voters were members of the Victorian Branch. I do not know the ratio of Victorian members to the total membership of the organisation but it would be surprising if it was 78%. It may be that there is no difference in attitude to the proposed amalgamation amongst Victorian members than amongst the members as a whole. On the other hand, there may be a difference; as the Timber case reminds us, it is not unusual for people resident in one part of the country to take a different view about an issue than their equivalents in other parts.

9. Without pretending to put the chance in mathematical terms, it seems to me unlikely that the ballot would have resulted differently if all those who were entitled to vote were sent ballot papers. Even if the assumption is made that the associate members participated heavily in the ballot and were overwhelmingly in favour - and, of course, there is no reason to make either of these assumptions - the disenfranchised members would have needed to have voted strongly in opposition to have overcome the otherwise majority, of about 1,000, in favour of amalgamation. If the participation rate which applied to those sent ballots had applied to the 2511 members who were denied them, 963 additional votes would have been cast. Even if they were all negative votes, there would have remained a small majority in favour of amalgamation. If twice the proportion of disenfranchised voters had voted, over three-quarters of their 1926 votes would have to be negative for the proposal to be defeated. If the relevant question was whether the admission of the disenfranchised voters to the ballot would probably have affected the result, the answer must be "no".

10. However, this is not the test postulated by s.253M. That section empowers the Court to order a fresh ballot if it finds an irregularity that may have affected the result. As I have indicated, that test does not require the Court to give effect to a merely theoretical possibility. But it does suggest that the Court should set aside the result of the ballot where there exists a real possibility that it was affected by the proved irregularity or irregularities. Despite my view about the probabilities of the situation, I cannot say that there was no real possibility of a different result, absent the irregularities. As I have said, the excluded voters were predominantly members of the Victorian Branch. If there was antagonism to the amalgamation proposal within that Branch, it may have been sufficient to cause a greater proportion of the excluded members to vote than those who did vote, and to vote overwhelmingly against the proposal. I agree that there is no evidence of such antagonism, and that my supposition is an unlikely one. But there is no evidence to justify its rejection as a real possibility. I think that the only conclusion open to me is that the irregularities may have affected the result of the ballot. Accordingly, I order that the declaration of the ballot made by Mr Small on 19 December 1991 be set aside and that a fresh ballot be conducted in place of that ballot. I think that it is preferable for directions for the fresh ballot to be given by the Australian Industrial Relations Commission rather than by this Court. Accordingly, I direct the respondent, within 21 days, apply to that Commission for appropriate directions concerning the conduct of the fresh ballot.