In news sure to annoy the government, a recent survey shows that almost 2/3rds of Aucklanders back the City Rail Link.

According to Research New Zealand, 63 per cent of Aucklanders surveyed are in favour of the link, while 29 per cent were against it and eight percent didn’t care.

63% is a pretty high level of support for a project with a cost like the CRL however it would be interesting to see how the questions were asked. Here’s what Len Brown has had to say about it.

“The vast majority of Aucklanders understand just how critical this piece of infrastructure is to the future of our city because it will reduce journey times across the entire rail network,” Brown said.

“Common misconceptions that the benefits won’t be felt beyond the central city are finally being dispelled and we can now move to an educated debate on how we pay for this and a number of other major infrastructure projects necessary to get our city moving.”

I’m not so convinced that the wider public are really that aware of just how much the CRL will benefit the rest of the region and that was made pretty clear in feedback to the RLTP where even some local boards, who would have seen more information than most of the general public, didn’t understand it’s benefits. That is part of the reason why we have created a permanent page for the project on the blog  and are working to come up up with new ways of explaining the project that AT could potentially use.

In other news, today the Auckland Council will debate and should adopt the final version of the city centre master plan. This plan is pretty important and sets out how the CBD will develop and with the exception of a few wording changes is pretty much the same as what was initially put out and one of the key projects is of course the CRL. We have had quite a bit of coverage on topics like what service patterns the trains should run, how many stations we should have etc. but one thing we haven’t really had is just how much development the CRL allows for. Now this information was actually in the draft city centre master plan in a very graphical format but it is worth just stripping out the numbers to highlight them. This is the indicative growth potential for each of the stations:

To put that into perspective, the growth potential is equivalent to about 50% of all of the land in the CBD, which is traditionally considered to be the land currently boarded by the motorways and the harbour. That percentage would increase if you took out all of the land taken up for streets and parks. The commercial space for an extra 30,000 workers represents an increase of employment of over 33% while an additional 22,000 people living in the area would be almost double what is there now.

And here are some images the council put together to represent this.


K Rd


Share this


  1. Great to see that public support for the project has survived. Hopefully the Herald will now shut up about a few grumpy submitters to the LTP meaning that support had supposedly dropped.

      1. That’d be breaching copyright.
        The summary is that the NBR thinks the survey results are a con because the sample was only 119 people with 75 positive responses. There’s no talk of the margin-of-error or the sampling methodology, so no evidence whatsoever to suggest that the positive support is wrong. Even an error of 8.8% leaves over 50% support.

    1. Even with such a small sample, this result is still statistically significant. The margin of error is commonly taken as the 2*the standard error, that is, 2*sqrt(0.63*.37/119) = 8.8%.

        1. SE=(p*(1-p))/n
          ie 63%*(100%-63%)
          The commonly reported Margin of Error is just calculated at 50%, because that is the probability at which it’s biggest. So really you can just do a quick hack with the n. Off the top of my head n=110 gives ~10%, n=500 gives ~5%, n=800 gives 3.5%. And after that, increasing n gives very little improvement. So unless you have some subgroups you want to analyse (e.g., splitting out estimates by age), there is really no need to sample more than 1000 people.

  2. Wow, I’ve never looked at the NBR blogs before — what a load of partisan shit to be coming from professional journalists.

    Here’s a quote from the blog Capt Stubbing links to :

    Auckland mayor Len Brown has conned the public.

    In the latest piece of survey propaganda from his office today mayor Brown trumpeted: “Two out of three Aucklanders back the CRL.

    This is not true.

    Not only is it not true, it is grossly misleading.

          1. Rob – it’s not so much an article as uninformed drool. They take the sample size, multiply it by the support rate, and calculate that 75 people supported it. They then claim that 2 out of 3 people support it can’t be true, and berate some people for good measure. Whoever wrote it is statistically illiterate.

  3. It would be good if this blog did an online survey of it’s own soonest, as that may well give a more accurate picture of current support for the CRL – for the positive or for the negative.

    Such a survey would certainly have more than 119 people participating and would more transparent I believe, as questions would be seen online before and after the survey is conducted.

    If the result is in the positive, then it confirms changes in the mindset of the Auckland public at large in regard to rail-based PT. If the result is in the negative, then that means more information on the CRL is needed in the public domain, for people to make an informed decision on the project.

    1. An online survey here or elsewhere would give a wildly inaccurate picture of current support for the CRL. The quality of a survey is almost entirely governed by the quality of its sampling and has nothing to do with its sample size. If a survey was posted on this blog, and not reported elsewhere, the response would be in favour. If it was picked up by some different types of blogs, it could become much more polarised. As long as the sample is self-selecting, and is finding it via forums such as this that don’t reach a representative audience, it is essentially meaningless, even if it were to receive tens of thousands of responses. The key to getting a good picture is good random sampling. Research New Zealand (the firm that ran the survey) sample 500 New Zealanders every month across a broad variety of questions. My guess is that there were 119 people from Auckland out of the 500 that answered this question. It would obviously be more transparent if we knew what the question was, but as it stands, this seems a good estimate.

      119 is getting reasonably small. I think I’d be uncomfortable with less than 100, but the increased precision of going to 500-800 (the usual size of a poll) only halves the margin of error, and is probably going to produce a similar estimate of support.

  4. I’ve been out of New Zealand too long it seems. I’m used to Japan where online surveys are commonplace, well implemented / managed and produce accurate reflections of public opinion.

    1. Hi Rob – that’s entirely fascinating. When I read that, I started wondering if there was some sort of cultural difference, and it turns out there is. Traditional best practice sampling turns out not to work in Japan, which is why online surveys are popular. However, online surveys here tend to produce horribly biased results, which is why I’d emphatically come out in favour of this method, even despite its small sample size. I’m therefore sorry about jumping so far down your throat (and will probably even use Japan as an example in my research methods course).

  5. I noticed this morning on q + a, Michael Barnett commenting on the need for the government to lead on investment in infrastructure to support economic growth in Auckland. I was waiting with baited breath, he mentioned the waterview connection and the convention centre, but not the crl.

    Apparently no economic benefit there from a rail link that would put people pretty much at the door of the convention centre (which I think should be at the waterfront but anyway). Incredible the amount of community disruption involved in putting these roads in. For waterview it’s meant peoples houses being demolished, a local school.

Leave a Reply