NSW council amalgamations – proposals released

Screen Shot 2015-12-21 at 5.09.31 pmThe NSW government has released its plans for council amalgamations following a lengthy of period of reviews and submissions by local councils.

The government is proposing cutting the number of councils in the Sydney region from 43 to 25, as well as merging other councils in rural NSW.

In this post I’m going to focus on the changes to the region stretching from Port Stephens to Shoalhaven, covering the vast bulk of New South Wales, including about 6 million residents.

This region includes 53 councils, and the NSW government proposes reducing this to 32 councils, with only 14 councils unaffected.

I’ve done some analysis of the political make-up of each new local government area, examining allegations of gerrymandering, and posting some maps showing the stats for each proposed new council.

You can download the Google Earth map of the proposed boundaries here. (link broken)

You can also download the dataset I used here.

My map does not cover rural areas – it only stretches from Port Stephens to Shoalhaven.

Firstly, I’m not going to deal with the question of whether councils should be amalgamated, whether they should be forcibly amalgamated, or specifically whether the NSW government’s approach is correct. I’ve previously analysed the flawed IPART report, and will instead focus on describing the councils as proposed.

For this analysis, I have used federal election results to analyse the underlying political make-up of any council. This is the easiest data to access – it would have also been interesting to see the council election results for the new councils, but council results are often quite dependent on who runs, and in a bunch of councils there are no political parties running, which would limit the value.

I calculated the Liberal two-party-preferred figure (vs Labor) for the 2013 federal election. I then subtracted 3.5% from the Liberal vote to adjust the vote to a situation where the NSW vote was 50/50.

This map shows the political make-up of each council before the mergers.

There’s some interesting trends in areas affected by the redistribution. The old Parramatta council is marginally pro-Labor (54.1%), while Burwood, Strathfield and Canada Bay are also quite marginal – ranging from 54.8% Labor in Burwood to 51.5% Liberal in Canada Bay.

In the eastern suburbs, Randwick is marginally pro-Labor while Waverley and Woollahra are solidly pro-Liberal.

Now let’s look at the same data for the proposed councils.

Marginal Liberal Ryde has become solidly Liberal with the addition of Hunter’s Hill and Lane Cove.

Parramatta remains marginally pro-Labor, although less so than the old Parramatta.

In the east, the council I have nicknamed ‘East Sydney’ is 55.7% Liberal, wiping out any pro-Labor leanings in Randwick.

I think there are some interesting cases where you can argue that the Liberal government may have engaged in some gerrymandering:

  • Burwood and Canada Bay agreed to merge with Auburn, which would have required Strathfield to be included as well. Instead the government has created this new council without Auburn. With Auburn the council would have been 54.3% Labor, without 50.2%.
  • Parramatta has been expanded slightly north to take in pro-Liberal suburbs south of the M2 (which seems to make sense), but instead of expanding south to take in pro-Labor Holroyd and Auburn, the government will remove South Granville and Granville (some of the most Labor-friendly parts of Parramatta) and roll all those Labor areas into a strange-looking council to the south of Parramatta. If you included these areas into a larger City of Parramatta, it would increase the Labor margin from 50.4% to 55.4%.
  • Randwick is marginally Labor, and combining it with pro-Liberal Waverley and Woollahra has produced a much larger pro-Liberal council. Inclusion of Sydney or Botany Bay (as proposed by the independent panel) would have weakened this considerably).
  • Kogarah is marginally pro-Liberal, and will be merged with Hurstville to create a marginal pro-Labor council, but it would have made a lot of sense to include Rockdale in a larger St George council, and they didn’t. Excluding Rockdale reduced the Labor margin from 52.6% to 51%, while also significantly weakening Labor’s hold on Botany Bay council (which considering its democratic deficit, is probably a good thing).
  • Ryde council, which has always been marginal, will shift more towards the Liberal Party with the addition of more parts of the North Shore.
  • The old Newcastle had a pro-Labor margin of 63.9% and with the addition of pro-Liberal Port Stephens this has dropped to 58.4%. The alternative option of merging with Lake Macquarie would have not produced any weaker margin.

Now it’s possible to argue some of these cases on grounds other than partisan gerrymandering – Ryde and the East Sydney proposal make a decent amount of sense, but the Parramatta proposal really only makes sense as a way to hive off strong Labor areas from a strategically-important city (in the same way past Liberal governments hived off Labor-voting suburbs in the City of Sydney to form various South Sydney councils).

Probably the only Liberal-leaning councils which could feel aggrieved by this proposal are Canada Bay and Kogarah, but these councils only lean slightly towards the Liberals (and are in areas where they did particularly well in 2013). In both cases the Liberal Party’s amalgamation logic wouldn’t have allowed those councils to stand along, and in both cases they kept out more pro-Labor councils which would have shifted the new councils further towards Labor.

It’s also worth noting that these margins suggest underlying trends and councils can vary from that. They also don’t factor in personal votes for federal MPs which could see certain councils diverge from their underlying ideology. Waverley in the past has had progressive majorities (and was half-progressive until 2012) despite being a strong Liberal area at a federal level. Newcastle likewise had a conservative majority until recently despite an overwhelming Labor vote federally, and a series of Labor-leaning Western Sydney councils are controlled by Liberals after the 2012 council elections.

But council amalgamations are not just for the next election cycle – they should be in place for a long time, and underlying political trends will influence their make-up. Putting a thumb on the scale to push a council a few points one way or the other might make a difference in which political grouping wins a majority.

One other thing I want to look at is population. One of the goals of this process was to have similarly-sized councils across Sydney.

Firstly, here is the map of existing council populations:

There’s a lot of variation. Only Blacktown is over 300,000, while there are eight councils in central Sydney under 50,000 population and a further ten in the centre of Sydney with a population under 100,000.

This looks much more equal under their proposals:

Most of these new councils aren’t some new creature – they’re roughly in line with the large councils of Western Sydney, with most having populations between 100,000 and 200,000 people.

There are two proposals that stand out as strange.

Firstly, the merger of Canterbury and Bankstown doesn’t appear to make sense when considering the rest of these proposals. While the councils have a lot in common, geographically it creates a strange shape, with about 20km between Earlwood and Chester Hill.

It’s also not necessary to achieve an even population balance. Bankstown is already over 200,000 residents: about the same as Fairfield, Liverpool and Penrith, and more than Campbelltown, yet none of those councils are being merged.

Canterbury’s current population of 150,000 is similar in scale to the proposed Hurstville-Kogarah, Botany Bay-Rockdale and Burwood-Canada Bay-Strathfield mergers happening around it.

Instead, by merging two councils which are already at a similar scale to neighbouring post-merger councils, Bankstown-Canterbury becomes the most populous local government area in New South Wales.

Secondly, the state government decided to split Warringah in half and give each half to Pittwater and Manly, instead of merging the three councils to form a Northern Beaches council. While the state government has attempted to clear up messy and non-intuitive boundaries around Parramatta, they have done exactly the opposite by drawing a new local government boundary through the backstreets of Warringah.

Finally, a note on council names. The state government hasn’t proposed names for any of these councils – and even if they did there’s a history of dry state-designed names being quickly replaced by a new council, such as when Eastern Capital City Region became Palerang.

I’ve made up names for most of the new councils for ease of use and wanted to explain them. For a number of the new councils, I have kept the name of the constituent council which is clearly dominant: Parramatta, Manly, Pittwater, Wollongong, Newcastle, North Sydney, Shoalhaven and Ryde. I’ve also retained the names Strathfield and Botany Bay, despite those councils making up a minority of the new council, since the names seem generally applicable.

I’ve made up the following names:

  • Central Coast – Gosford and Wyong
  • East Sydney – Randwick, Waverley and Woollahra
  • Hawkesbury Hills – Hawkesbury and the Hills
  • Petersham – Ashfield, Leichhardt and Marrickville
  • St George – Kogarah and Hurstville

I don’t have a good name for Auburn-Holroyd, Bankstown-Canterbury or Hornsby-Ku-ring-gai.

About Ben Raue

Ben Raue is the founder and author of the Tally Room. If you like this post, please consider donating to support the Tally Room.