Three observations about the Labour leadership election

For weeks Jeremy Corbyn had been expected to win the British Labour leadership race but the sheer scale of yesterday’s victory has shocked many.

While some thought it would go to a second round, Corbyn won in the first round with 59.5%. His closest competitor, Andy Burnham, only received 19%. It was a larger victory than Tony Blair’s in 1994 who won with 57% (although the selectorate was far larger in 1994).

It has left no doubt about Corbyn’s mandate amongst the Labour membership and the sheer number of registered supporters have quelled concerns about entryism.

Personally I am sceptical that Corbyn is the long term answer for Labour but neither were the other candidates. His victory does highlight, as Tim Lyons has noted, that the centre-left needs to deal with its inability to seek with moral clarity and a clear purpose. I can only hope that the disruption that his victory brings may force many within Labour to rethink and move on which it sorely needs to do.

There will be a plethora of thinkpieces about what Corbyn’s victory means over the next few weeks and months. It’s too early to tell what the long-term implications of his victory are but there are three quick observations I have from the leadership election based on the results.

1. The landslide victories were built on recruitment campaigns

Corbyn won 49.6% of the membership in the first round but he absolutely dominated the registered supporters section with 83.8%. The size of his victory amongst registered supporters is staggering but not a surprise.

Unlike the other leadership candidates, his campaign team actively recruited. Similarly Sadiq Khan actively recruited registered supporters and it helped to secure a landslide win over Tessa Jowell who was seen as the favourite to become the London Mayoral Candidate. She was well-regarded, had a good public profile and was associated with the successful London Olympics.

The moves by winning candidates to expand the selectorate reminds me of Canadian leadership elections. In the three major parties, members directly vote fo the leader and party memberships are actively sold to supporters during leadership contests by each campaign team. Those who can often recruit and organise the most supporters win.

The lesson for future Labour leadership elections will be the need for genuing organising and fieldwork and to bring potential supporters into the tent.

2. Union turnout was low

I have been sceptical of opt-in affiliation for unions and whether it would actually engage members of affiliated unions. While initial registration figures made me question my scepticism, the low turnout suggests that union member engagement, even amongst those who opted-in, was not high.

While turnout has not been officially revealed, based on those who registered to vote, only 48.2% of union members who opted in did. In contrast, 93.6% of registered supporters and 83.8% of members voted.

In the end, affiliated voters constituted only 16.9% of total votes (down from 27%) with 71,546. Rank and file members were 58.1% (up from 53%) and registered supporters made up 25% (up from 20%).

If there is no greater Labour Party engagement amongst affiliated union membership, it raises questions about what opt-in affiliation has really achieved beyond reducing funding from unions.

3. There will be a rethinking of support for primaries

The great irony of this result is that that the New Labour Right were the biggest advocates of primaries. They have admitted that it was an attempt to dilute left-wing member and union influence (through opt-in affiliation and primaries). Meanwhile the Left opposed the Collins Review that suggested recommended.

The problem is that their understanding of how primaries would work relied on America and the idea that a significant portion of the general population would participate. A better example to study would have been the party primaries run by the Italian Democratic Party (PD). The model was the same with a small fee and a pledge that voters needed to sign.

Academic studies have been done on participants in PD primaries, profiling them and their attributes. These studies found that participants have been more political than the average voter and more left-wing with no guarantee of party loyalty.

It is quite likely that the Labour Left will become converts to primaries whereas elements of the Labour Right will rethink their support. Already Dan Hodges has admitted that the Labour Right got it wrong.

There will be ripple effects here. It is likely to strengthen opposition to further democratisation of the Labor Party and the use of One Member One Vote being used to select leaders. Nick Dyrenfurth has already highlighted concerns about such a direct election model and is unlikely to be the only one to do so. The global trend, however, seems to be towards greater democratisation and primaries which poses a challenge for opponents.

Whatever does happen next, British Labour and social democratic politics will never be the same again after Corbyn’s victory.

Only a small number of union members are opting into (British) Labour Party affiliation

I have previously written about the push to change how unions affiliate to the Labor Party. Currently unions affilate on behalf of their membership but there have been calls to change this so individuals must opt-in to be counted for affiliation. British Labour has adopted it and there are advocates within the ALP pushing for it to be embraced here.

The big question has always been: how many affiliated union members will opt-in? In the 2010 Labour leadership election, 238,618 union members cast a vote but over two million ballots were mailed out. Any opt-in process would mean the number of voters would fall significantly. Estimates had ranged from 25,000 to 80,000 but a recent news article suggests it might be lower.

The New Statesman has reported that in London, only 1,197 members of affiliated unions have opted in. To give some context, Labour’s largest affiliated union, Unite, has 200,000 members in London and the most recent publicly available data indicates that 21% of Labour Party members are in London.  With Labour’s membership now in excess of 220,000, affiliated union members seem likely to be less than 10% of the vote in the upcoming British Labour leadership contest.

There are still two months left for union members to opt-in for the leadership ballot with 12 August being the last day to register. It, however, seems that even 25,000 may not be reached considering Unite has only just started getting members to opt-in since the election.

If the number of union members that affiliate to British Labour for the contest is tiny, it seems likely that the enthusiasm for opt-in affiliation amongst many in the ALP will disappear. It is possible that it will evolve into a much clearer debate about whether the current labourist model should be ditched and the party should move to a One Member One Vote model. Each model has its problems and while “breaking the link” is often floated by those who decry trade union influence, the Nordic experience of cutting links shows that the result is not always predictable and may not be the best outcome.

Whoever wins in the British Labour leadership won’t really matter to the ALP but the number of union members who opt-in and participate in the election will. We should pay attention to the contest because it will shape the ongoing debate about the ALP-union link and whether the “labour party” model is truly dead.

UPDATE 16/6/15: GMB have said 10,000 members have registered as affiliated members and Unite is said to have similar numbers, however, Labour says only 2,500 have completed the process & paid a fee.

UPDATE 24/06/15: LabourList has been provided with the numbers of members and supporters. So far only 9,115 registered supporters and 3,788 affiliated supporters will be able to vote. Total Labour Party membership is 246,469.

UPDATE 12/07/15: Unite is claiming 50,000 members have opted in as affiliated supporters.

UPDATE 15/07/15: Unite has stated it is aiming to get 70,000 members to become affiliated supporters by August 12.

UPDATE 17/07/15: The Evening Standard is claiming it has seen figures suggesting 65,000 union members have registered.

UPDATE 12/08/15: The latest breakdown is 70,000 registered supporters (16%), 92,000 affiliated supporters (21%), 282,000 members (64%)

UPDATE 13/08/15: The Labour Party has revealed total figures of those who applied for a vote: Affiliated: 189,703, Registered: 121,295, Members: 299,755, Total: 610,753.

Portuguese Socialist Party holds open primary for Prime Ministerial candidate

Over the weekend, the Portuguese Socialist Party became the latest European social democratic party to experiment with open primaries. It held a primary on September 28 to select the party’s candidate for Prime Minister in the upcoming 2015 general election. The primary was adopted in June to address factional dispute within the party between the candidates after disappointing result in European elections.

The two candidates were Antonio Jose Seguro, the current General Secretary of the Socialist Party and Antonio Costa, the mayor of Lisbon. With an eligible electorate of 9,753,568, turnout was 174,516 or 1.8%, a similar figure to turnout in some of NSW Labor’s recent community preselections. Costa won the primary in a landslide with 67.88% (118,454) while Seguro received 31.65% (55,239). Following this result, Seguro resigned as General Secretary and an election will be held for the General Secretary’s position by mid-December.

In order to participate, voters had to pre-register to vote by mid-September. Voters who were not party members had to be on the electoral roll and had to sign a declaration supporting the values of the Socialist Party. Unlike other primaries in Europe, there was no fee to participate. The rules for the contest are available here.

Portugal now joins France and Italy as European social democratic parties that have used open primaries to select candidates. The next big open primary will be held by Spain’s Socialist Workers Party in November for its Prime Ministerial candidate. Given the lack of electoral success that many European social democratic parties have experienced in recent years, the increased use of open primaries is likely to continue into the future.

Mapping the left-wing of global social democracy

Whether called factions, currents, tendencies or platforms, there are often organised and sometimes formalised left-leaning groups within or seeking to pressure social democratic parties. Within the Australian Labor Party, it’s the Labor Left but what about elsewhere in the world?

While factionalism does not exist to the same extend within British Parliamentary Labour Party, there are left-wing pressure groups such as Compass, Next Generation Labour and those much further to the left such as the Labour Representation Committee. There are others that organise slates for internal elections such as Centre-Left Grassroots Alliance and the Campaign for Labour Party Democracy. In Scotland, there is the Campaign for Socialism.

In other sister labour parties, left-wing groups have emerged and disappeared. For example, inside Irish Labour, left-wing members established the Campaign for Labour Policies trying to outline an alternative to the austerity agenda being pursued by the current Fine Gael-Labour Coalition Government. In Canada, there was the New Politics Intiative that believed the New Democratic Party was moving too far to the Right and sought to integrate social movements into the party that disbanded in the mid-2000s.

Outside of the English-speaking world, left-wing groupings are far less known due to language barriers.

The German Social Democratic Party, for example, has a Parliamentary Left grouping that has existed since the 1970s and an aligned think-tank, Forum Democratic Left (DL21).

Other more factionalised social democratic parties have a more diverse range of left-wing groups that appear and disappear. Within the French Socialist Party,  there appear to be a range of other institutionalised left-wing groupings at a party and youth level. The main left faction appears to be Maintenant la gauche who have opposed Hollande’s austerity agenda. Another recent notable grouping to emerge is the Socialistes Affligés led by former MEP Liem Hoang Ngoc and university professor and Guardian columnist Philippe Marliere which includes those outside of the French PS who are seeking a possible “red-rose-green” alliance to force a left turn and avoid emulating PASOK.

Some other parties have recognised factions exist and incorporated their recognition into party rules. For example, the Spanish Socialist Workers Party (PSOE) has recognised the Izquierda Socialista or Socialist Left faction and even provides space on the party website. Under the Italian Democratic Party’s rules, groups can be established and can operate autonomously. They have been actively promoted as it has sought to become a broad-based party that includes all those from the centre to the left. The main left-wing grouping that currently exists is Remake Italy.

Given the known links between the moderate or “right-wing” of social democratic and labour parties (e.g. Labor Right and British Labour’s Progress) and the transfer of ideas via these relationships, the social democratic left should also seek to identity the “left-wing” of sister parties, build and strengthen these relationships.

UPDATED 16/08/17: The list would now obviously include Momentum but also soft left groups like Open Labour within British Labour.

Community preselection success in Balmain

I am on the record as a sceptic of community preselections. I’ve previously outlined many concerns and have not been convinced they are necessarily a solution to Labor’s problems.

While I am a sceptic, the turnout for the Balmain community preselection this weekend was nothing short of impressive with over 5600 voters or a turnout of over 11%. It met the target I previously suggested it needed to meet and considering Labor only managed a primary vote of 30% at the 2011 NSW election, it can be deemed a genuine success if turnout is the main measure of success.

Darcy Byrne beat Verity Firth in the party vote 183-174 (51%-49%), however, Verity romped home in the community vote with 3104 votes (60.7%) compared to 2006 votes (39%) for Darcy. With the affirmative action loading on the party and community vote, Verity ended up winning both sections with 53% of the party vote and 65% of the community vote.

While I’m still not completely sold on community preselections, particularly as a significant cost is born by the party and its impact will still need to be measured at the ballot box in March, there are definitely lessons to be learnt from the Balmain community preselection. How to get community participation in party processes and the importance of high-profile, quality candidates are the two that spring to mind.

Whether this success is a one-off because of particular local factors or the ideal model of how conduct community preselections has been found is still to be determined. The only thing we can be sure about is that NSW Labor will continue with the community preselection experiment as it would feel vindicated after these results.

What turnout should be expected for Labor’s primaries?

I have previously stated that the participation of thousands of voters should not be a measure of success for NSW Labor’s community preselections. Low barriers to participation will inevitably mean a higher level of participation. This has been demonstrated by the open primaries held by the British Conservative Party five years ago.

In 2009, the British Conservative Party trialed two open primaries. The primaries were held for the seats of Totnes and Gosport with every voter on the electoral roll sent a postal ballot. Both open primaries had a high level of voter participation, Totnes had a turnout of 24.6% of registered voters while Gosport had 17.8% of registered voters participate. As a comparison, turnout in state wide primaries in the United States was 17.3% in 2012.

While further open primaries using postal ballots have not occurred since, they do give an indication of what turnout should be expected by a trial of primaries in Australia where voter is mailed.

There are some key differences that limit the comparison between the Tories open primary experiment and NSW Labor’s community preselections. The Tories mailed ballots while NSW Labor will only be mailing out instructions about how to participate. Both seats were not in a major city and were safe Conservative seats which meant greater media focus on the primaries. Unlike Australia, the United Kingdom has voluntary voting (only 82.3% of the population was registered to vote in 2011) and there is still a rank-and-file component in Labor’s community preselection.

Still, based on the experience of the Tories’ trials in the United Kingdom and considering the differences, it would be reasonable to expect that the NSW Labor’s community preselections should get at least 10% or around 5,000 participants in the primary component. Anything less would not be a success considering the cost and resources invested in this process.

UPDATE 31/3/14: According to The Australian, there were 1,061 non-member voters in Campbelltown community preselection and 1,345 non-member voters in Newtown or approximately 2% of each electorate.

UPDATE 21/9/14: The Ballina community preselection had 2,317 votes cast (no member/non-member breakdown is available yet) or approximately 4% of the electorate.

Are primaries really the answer to Labor’s woes?

Last year, New South Wales Labor announced that preselections for five “winnable” seats would be conducted by a community pre-selection. Fifty per cent of the vote would come from rank-and-file members of the ALP, with another 50 per cent to come from an open primary where members of the public on the electoral roll can vote.
These “primaries” are currently being held for the seats of Newtown and Campbelltown and will soon be held for the seat of Balmain. The primary for the seat of Strathfield is still to be announced while the proposed Londonderry primary did not proceed as only one candidate, Prue Car, nominated.

While there has been enthusiasm by some Labor Party members about this process, are these “primaries” really what Labor needs?

Considering how resource intensive primaries are and the limited resources that exist for party organising, it is essential that the success of community preselections can be measured. The lack of discussion as to what those measurable outcomes are, beyond greater numbers of people participating, is concerning.

Isn’t greater participation a good thing? Turnout is not a good measure of success for community primaries. It is inevitable that primaries will have far more people involved when barriers to participation are low — for example, unlike similar primaries in Europe, the lack of a fee to participate.

The involvement of 5,000 residents out of 100,000 enrolled voters in the City of Sydney community preselection seems like a lot, until you learn that all electors were mailed about the primary and realise that student elections, at the University of Sydney for example, routinely pull a few thousand voters out of approximately 33,000 undergraduate students.

What outcomes should primaries be measured against? In my view, the success of community primaries for Labor should be measured using three tangible outcomes:

  • Increased party membership
  • Increased primary votes
  • Increased number of volunteers involved in a campaign.

It is unclear if any of these outcomes were met thanks to the City of Sydney community preselection trial. To my knowledge, there has been no assessment of whether the trial met any of these outcomes.

If none of these outcomes were met, the question must be asked: why is Labor adopting primaries? It only leaves three other plausible reasons:

  • Voter self-identification
  • Greater transparency and openness
  • A wider variety of candidates preselected

Voter identification was flagged as a reason for adopting a primary model for electing the NSW State Parliamentary Leader by NSW Young Labor, in their submission to the NSW Labor commission on the direct election of leader.

Greater transparency and openness was raised by City of Sydney councillor Linda Scott during a session at the SEARCH Foundation’s Left Renewal Conference. This is understandable given NSW Labor’s public standing and the scandals being investigated by Independent Commission Against Corruption.

The shrinking pool from which candidates are preselected has been widely discussed for the past decade and continues to be an issue across all parties. Tim Soutphommasane has suggested that primaries may be one way of ensuring our parliamentary representatives are more diverse and representative.

But it is unclear whether community preselections will achieve these other outcomes. There are also questions about whether a primary is a good use of scarce resources.

At 47 cents per letter (currently) for pre-sorted mail with approximately 100,000 enrolled residents in the City of Sydney, it would have been a considerable expense to mail every person enrolled in the City of Sydney to inform them about the primary.

An online portal to allow electronic voting and polling booths to vote in person were also established, adding to the cost. For the same amount spent on a primary, it is likely Labor could have hired a call centre to do voter ID work and identify an equivalent number of people.

There is a need for greater transparency and openness to regain public confidence but it is difficult for that to be measured other than by an improved electoral outcome.

Ensuring a diverse and representative range of preselected candidates is important but it is unclear whether community preselections would do that. A successful candidate would still need to have strong links in the local branches and be able to cobble together a campaign machine pretty quickly.

The resources and time required would limit a realistic chance to be preselected to those who have been active party members for a considerable period of time and/or political professionals. Outsiders will have little chance.

The frontrunner for Newtown is Penny Sharpe, the current Shadow Transport Minister, who is expected to beat former South Sydney Councillor Sean Macken and refugee activist Natalie Gould. Former NSW Education Minister Verity Firth is making a comeback in Balmain and is up against Leichardt Mayor Darcy Byrne in what will be a much closer contest. Campbelltown will be the most unpredictable contest with no obvious favourite out of Ian Fulton, Greg Warren and Brendan Whitehouse.

There should be an open and robust discussion about the use of community preselections but questions will remain until these concerns about scarce resources and measurable outcomes are addressed. Avoiding these issues will do the Australian Labor Party no favours in the long-run.

Published in New Matilda on 14 March 2014

Spanish Socialists (PSOE) experiment with primaries

Last month, the Spanish Socialist Workers Party (PSOE) became the latest social democratic party to adopt open primaries. The primaries will be for their Prime Ministerial candidate at the next Spanish election in 2015.

Under the PSOE primary rules, the primary would be open to all Spanish citizens aged over 16. There would be a month to register to take part with the register closing six days before the vote. Those who register required to pay 2 Euros to vote. Candidates for Prime Minister will require nominations from 5% of members (but no more than 10%). A period of two to three weeks will be given to collect nominations and there will be no run-off round. Primaries may also be held for Presidencies of autonomous regions with the support of regional Executives. Valencia will be the first region to conduct an open PSOE primary for the Presidential candidate. A copy of the primary rules is available here (Spanish).

The PSOE’s Catalonian sister party, the Socialist Party of Catalonia (PSC) have also embraced primaries. It will also use an open primary for its Barcelona mayoral candidate at the end of March. The primary will run for six weeks and candidates will need to get a minimum of 1000 nominations from people on the primary electoral roll. A candidate will need to obtain 40% and win by more than 10% to avoid a run-off. More details are available here.

It will be interesting to see what effect an open primary will have. While the polls are close between the PSOE and its conservative rival, the People’s Party (PP), they are both polling well-below their voting share at the last Spanish election. Other groupings such as the left-wing United Left (IU) and social liberal Union, Progress and Democracy (UPyD) have grown their vote and combined are polling over 20% in the opinion polls. While the electoral system disadvantages these smaller parties, it is likely there will be a hung parliament. Other parties like the UPyD have also adopted the use of open primaries and the threat of abstentions is high.

Whether the primaries will have an impact on the level of support for the PSOE or throw up a surprise candidate is unclear but one thing is for sure, the momentum towards adopting primaries in social democratic parties shows little sign of abatement.

UPDATE 15/06/14: Jaume Collboni won the PSC primary to become their Barcelona mayoral candidate. There were over 5,000 participants (only a fifth were party members) in the second round of voting, down from over 7,000 in the first round.

How will we know if primaries work?

Every day the list of ALP figures advocating “community preselections” aka primaries seems to be growing. Today, former Victorian Secretary Nick Reece became the latest to publicly urge community primaries as part of party reform.

I have previously written about my concerns about community preselections and those who have spoken to me know I do not believe that Labor should institute primaries for the sake of it.

Considering how resource intensive primaries are (and the limited resources that exist for party organising), it is essential that the success of community primaries can be measured. The lack of discussion as to what those measurable outcomes are, beyond greater numbers of people participating, concerns me.

Isn’t greater participation a good thing?

Turnout is not a good measure of success for community primaries. It is inevitable that primaries will have more far people involved when barriers to participation are low. The involvement of 5,000 residents out of 100,000 enrolled voters in the City of Sydney community preselection seems like a lot until you learn that all electors were mailed about the primary and realise that student elections, at the University of Sydney for example, get a few thousand voters out of approximately 33,000 undergraduate students.

What outcomes should primaries be measured against?

In my view, the success of community primaries for Labor should be measured using three tangible outcomes:

  1. increased party membership;
  2. increased primary votes; and
  3. increased number of volunteers involved in a campaign.

It is unclear if any of these outcomes were met because of the City of Sydney community preselection trial. To my knowledge, there has been no assessment of whether the trial met any of these outcomes.

What are other outcomes of primaries?

If none of these outcomes were met, the question must be asked: why are we adopting primaries? It only leaves three other plausible reasons:

  1. Voter self-identification;
  2. Greater transparency and openness; and
  3. A wider variety of candidates preselected.

Voter identification was flagged as a reason for adopting a primary model for electing the NSW State Parliamentary Leader in NSW Young Labor’s submission to the party commission on direct election of leader.

Greater transparency and openness was raised by City of Sydney councillor Linda Scott during a session at the SEARCH Foundation’s Left Renewal Conference. This is understandable given NSW Labor’s current public standing and the scandals being investigated by ICAC.

The shrinking pool from which candidates are preselected has been widely discussed for the past decade and continues to be an issue across all parties. Recently, Tim Soutphommasane suggested that primaries may be one way of ensuring our parliamentary representatives are more diverse and representative.

Will primaries achieve these other outcomes?

It is unclear if a community preselection would achieve these other outcomes. There are also questions about whether a primary is a good use of scarce resources.

At 43 cents per letter for pre-sorted mail with approximately 100,000 enrolled residents in the City of Sydney, it would have been a considerable expense to mail every person enrolled in the City of Sydney to inform them about the primary. An online portal to allow electronic voting and polling booths to vote in person were also established, adding to the cost. For that same amount spent on a primary, it is likely you could hire a call centre to do voter ID work and identify an equivalent number of people.

There is a need for greater transparency and openness to regain public confidence but it is difficult for that to be measured other than by an improved electoral outcome.

Ensuring a diverse and representative range of preselected candidates is important but it is unclear whether community preselections would do that. A successful candidate would still need to have strong links in the local branches and be able to cobble together a campaign machine pretty quickly. The resources and time required would limit a realistic chance to be preselected to those who have been active party members for a considerable period of time and/or political professionals.

I am open to a discussion about community preselections but will remain sceptical until these concerns about scarce resources and measurable outcomes are addressed. Avoiding these issues will do the Australian Labor Party no favours in the long-run.

Where is NSW Labor likely to trial primaries?

Earlier this week, the NSW Electoral Commission released its draft redistribution with boundaries to be finalised later this year. Antony Green has done an initial examination of the redistribution and what the notional margins would be. If the boundaries do not change significantly, it gives a sense of which seats Labor are likely to win back at the 2015 state election.

State Primaries in 2015

In 2011, Labor amended its rules to allow community preselection in up to five “winnable” seats at the next state election. While Labor has not made substantial gains, recent polls seems to indicate an increase of about 3-4% in its vote since the 2011 election. Seats within a margin of 5% are likely to be candidates for a community preselection in 2015.

The push for community preselection is gaining momentum in NSW. There has been a growing chorus advocating its use over the past few months. Most notably, Mark Latham advocated the use of primaries in his recent Quarterly Essay earlier this year as has Tim Soutphommasane in a recent column. The NSW Young Labor submission to the Labor’s internal review on party leader selection even suggested a primary to elect the state leader. This all indicates that the pro-reform elements of the NSW Right have embraced primaries. It is not a far stretch to suggest that the push for primaries is likely to strengthen even more if Labor loses heartland Western Sydney seats in September.

Which seats?

Unless Carmel Tebbutt is the candidate, the most likely community preselection will be for the new seat of Newtown. There would be two purposes, firstly to build up candidate profiles in a politically aware electorate and secondly conduct voter identification in a notionally Greens seat.

The other seats that are likely to fall into the “winnable” category with a less than 5% margin are:

  • East Hills
  • Granville
  • Newcastle
  • Monaro
  • Prospect
  • Rockdale
  • Oatley
  • Swansea
  • Wyong
  • Balmain
  • Londonderry
  • Wyong

It is difficult to know how many seats and which seats will be chosen but I would be surprised if the trials take place in other seats.

One thing is clear, the NSW Right is moving towards embracing primaries and it will be happening. The question is not when but where.

UPDATE 30/06/13: It looks like I was correct about Newtown and Balmain being seats where primaries will be trialled, however, it seems there is a push to expand the number of primaries beyond five.

UPDATE 22/09/13: The seats where primaries will be trialled were revealed to be Strathfield, Londonderry, Campbelltown, Balmain and Newtown.

UPDATE 21/09/14: NSW Labor also held a community preselection in the seat of Ballina with 2,300 votes cast.