Statistics NZ have announced that for the first time they’re seeking public input into what questions should be asked in the next census and have launched a discussion forum so people can have their say.

For the first time, New Zealanders are being encouraged to go online to say what they think should be asked in the next census, with a new online discussion forum on the Statistics NZ website.

The six-week discussion forum has been set up as part of wider consultation on the content of the 2018 New Zealand Census of Population and Dwellings.

Census General Manager Denise McGregor says it is the first time Statistics NZ has engaged online about the census with the public.

“We know New Zealanders really care about what’s asked in the census, and the information they can get from it. This is part of a comprehensive review of the census, including how previous questions performed, a stocktake of what the data is used for, who uses it, and considering international trends.

“People can get in and respond to our initial recommendations and thinking, share their views, and discuss issues that matter to them with other Kiwis.”

Statistics NZ will be facilitating discussions, and will analyse emerging themes to help with making decisions on 2018 Census content. Factors such as statistical quality, and the length and complexity of the questionnaire will be taken into account before making any changes.

“The census needs to reflect the changing nature of society, and we have to balance that with being able to compare data over time and track trends.”

Statistics NZ is using the online forum to encourage a wide range of people to talk about the census ahead of the formal submission period.

“We will listen to suggestions raised in the online forums, but the best opportunity to influence census content is still to make a formal submission,” Denise McGregor said.

2018 Census has information about how to join the discussion or make a formal submission. The online discussion forum will be live until 10 June. People can make a formal submission from 18 May until 30 June.

Following consultation, Statistics NZ will analyse the submissions, and aims to confirm final content in early 2017 for the 2018 Census.

I think it’s good that they’re considering new ways to ensure the right questions are asked.

When it comes to transport we’ve long had issues with the data that is collected. At the last census the two questions related to transport were

  • Individual – What was the one main way you travelled to work – that is, the one you used for the greatest distance?
  • Dwelling – How many motor vehicles (not counting motorbikes) do the people who live here have available for their use?

We’ve previously highlighted some of what we consider to be issues with these questions and how they’re used, some of which are below

Main means of travel to work

While useful it ignores tens of thousands of trips for that are for education or other purposes. Particularly for education there’s a high likelihood that there will be a much higher share of non-car modes than there is for workers. Further as our transport system evolves I think there will become an increasing issue in only asking for the mode people travelled the longest on. This is because we’re increasingly going to see more multi-modal trips occurring thanks to developments like the New Network which will see an integrated network and many people transferring to complete their commutes. One other minor issue that is mostly unique to Auckland is that ferries are counted in the ‘other’ category and it would be useful to have that split out.

Number of motor vehicles

There’s not such an issue with what is collected but what it’s used for. Stats NZ feed the data into their Deprivation Index where they assume that households with fewer cars tend to be more deprived. That of course ignores that now in many places such as the City Centre, residents simply don’t need a car as they have many other options for getting around.

In their preliminary view Stats NZ acknowledge the issue of journey to education and are recommending it be considered to be added however in my view that still misses out trips for other purposes. They also suggest changing the question from the main means of travel to work on the day of the census to the usual means of travel to work. They say that it was intended to give a snapshot on transport volumes for a particular day but they note that many people now fill the form in early and in 2013, 46% completed the online form before Census day.

There’s more information on Stats NZ thoughts in the link above and you can also have your say in the online forum they’ve set up and I see there’s already some good discussion on these issues from some of our regular readers.

The actual census is three years away but it’s already fascinating to think how it will change. Between 2016 and 2013 we started to see some significant changes in mode-share and with the patronage growth we’re seeing and PT improvements planned those trends are likely to continue.


Share this


  1. Matt,
    Will TB be putting in a formal submission on these question to the Census 2018 panel as suggested or will you leave it to individuals to submit themselves?

    I figure this is not a popularity contest and organisations like AT and AC who use the data have a greater sway than any old citizen who wants something.
    And as TB is a well known organisation I think its submission would carry a lot of weight relative to anything I or anyone else would submit.

  2. I’m ‘deprived’……sniff.
    Actually it’s noticeable how much more money I have left at the end of the fortnight with no car. I’m off to the French Cafe to console myself.

  3. How many cars a family or dwelling has often bears no relationship to how many are, or can be, driven at any one time. Those sorts of stats are often used to portray views that are often not correct (the nonsensical headlines of “there are now X number of cars in Auckland, growing at x thousand per month). Due to my work and location of residence I have 5 vehicles, but at any one time can obviously only physically drive one. I do not believe my situation is abnormal, many households have personal, hobby and company vehicles yet all are not capable of being driven at the same time. There needs to be some sort of additional logic to the questions. I humbly suspect there may be on average one vehicle per person in Auckland, but those who are too young (infants, school kids) and those who are at the other end of the scale (elderly, incapacitated) are unlikely to be driving them. So the potential real numbers of vehicles on the road at any one time are actually substantially less. Vehicle ownership in itself does not correlate to actual usage. Food for thought.

    1. It’s a bit less than 1 per person, car ownership in Auckland has been pretty stable at ~0.67 cars per person in Auckland for many years now, so about 2 cars for every 3 people.

    2. As SNZ points out, councils knowing how cars are on a property [or nearby] is also very valuable information, yes in general you can only drive one at a time.

      But over the course of Census day you may drive several of them (for work and other purposes), Assuming as you do that you only drive 1, all day on Census day is not necessarily correct or valid.

  4. Better questions about people’s housing would be useful. Can’t remember what they ask, but remember that their housing questions were frustratingly uninformative. Also that – for all the effort of putting a census together, they asked very little. Would be happy to give far more info if they were interested in collecting it.

  5. I should note I have a slight dissenting view on this; I’ve argued against changing the “travel to X” question for the sake of consistency and respondent burden (if this means adding another question – and if it doesn’t, what takes priority if someone went to both education and work on census day?). I also note that only people aged 15+ fill out these questions, so you’re picking up some secondary school students but not others, or primary ones.
    SNZ seem to be on top of these issues from their discussion, and are also considering making the journey to education question something that is filled out by people of any age. So fair enough, and I agree that it’s valuable data. And maybe in the current Auckland situation, it’s a pressing issue so might be the one to go for if SNZ are adding new questions and given that they have to limit the numbers of questions they ask. But they also have to think about not just what’s useful in 2018, but what’s going to be useful 30 years from now, so they keep the questions consistent and useful over time. Maybe it’s worth it, and I’m coming around to the idea, but I’m not completely sold and would expect that SNZ are in a good position to weigh up these choices.
    Matt, note that the deprivation index is put together by the University of Otago, not SNZ. I’d agree with them that low number of vehicles still correlates with deprivation – e.g. note Peter’s (I think?) recent post that most carless households in Auckland are in low-income suburbs.

    1. John there are two competing changes,

      One is “Travel to X” question changes where they propose questions for both (paid) work and (currently) “Education” as two question sets to capture paid work and education related travel.

      Secondly there is a proposed change from “Travel On Census Day” – which what is asked now, to how do you “usually travel”.

      I have problems with both, and the “Usually travel” change is the one with more downsides as it breaks continuity with the existing data sets of Census Day travel as they all ask “On Census day how did you travel [to paid work]”.

      There is a problem however as they point out as 43% of people fill in the [online] Census questionaire’s before Census day so any question filled in like this is really a “prediction” of how they think they will travel, not an after the fact answer.

      To fix that I’d propose a new Yes/No question to accompany the “Usually travel” question changes, to the effect that “I did/will travel this way to X on Census day”.

      Its a valuable supplemental question – if implemented correctly gives rise to two very valuable sets of travel data (at least).

      And you can then use this additional question information to align the usually travelled answers back to prior Census using that additional question.
      Yes it has a cost but adding questions about travel adds cost and adds “burden” to the respondents.
      And this area of travel is one that actually needs a total revamp as the amount of money and spending that hangs off these answers for councils and governments throughout NZ means its critical to get this section as useful and valuable as possible.
      While still maintaining continuity backwards.

      For the “Travel to Education” questions for the “Travel to X” questions, Education as a travel question is way too narrow and needs to be broadened to capture all the reasons other than “paid work” as to why people travel. I suggest a “Write in” option of what the travel was for.

      I think leaving this as a “How do you *usually* travel to X” (where X is anything but paid work) is ok as its a new dataset as asking how you travelled to X on Census day is probably not important.
      Although doing so would allow capture of the same datasets as work related travel for the same time. So in that case adding the “I did/will travel this way on Census day” will fix that issue.

      While a lot of Census questions assume that it is filled out on Census day e.g. normally resident population questions as a snapshot which requires a specific time definition e.g. On Census day, many other questions are not.

      e.g. “Do you Smoke?”, or “Do you have a telephone, electricity, gas” etc as these are not time [Census day] critical answers as the answer would for the most part be the same whether answered before, on or after Census Day.

      While its great that SNZ are asking about these changes, there are a lot of fishhooks in what is proposed currently that need to be thought through.

      The one about how many cars is also going to have to change to reflect societal change, is “number of cars you have access to” – apply on the basis you can drive it?
      What about if I can hail an Uber or other ridesharing service? Does that count as “access to a motorvehicle” even though I maybe can’t drive/have no license?
      What about a car sharing scheme – does that count too? And/or taxis? do they count?

      I am sure inner city dwellers with strictly “no access to a car” as defined by the Census 2013 question answers they gave, mostly do not feel as deprived as those who live in the outer suburbs – who are technically in the same boat.
      And yet may well fall at different ends of the Deprivation index despite being lumped together by this answer and subsequent analysis.

Leave a Reply

Your email address will not be published. Required fields are marked *