Polity: Endangered! Polls without cell phones

Written By: - Date published: 5:07 pm, March 13th, 2014 - 28 comments
Categories: Politics, polls, us politics - Tags:

From Rob Salmond at Polity, this post on an interesting graph.

Here’s a pretty astounding graphic from the US about cellphone vs landline use, via Anzalone Lizst Grove Research:

The US is a few years ahead of New Zealand on mobile adoption and decoupling from landlines, but I think within 5 years we will see these kinds of proportions in New Zealand. This will make current pollsters’ policies of refusing to call cell phones hugely problematic – they will cut out almost half the population. No amount to weighting can reliably undo a sampling frame that unbalanced.

I know some pollsters are trying out online panels as a way to get the cell phone only population without the expense of talking to lots of cell phones. That is an innovative approach, and worth exploring. The sampling comparisons would need to be super robust, though, for it to be a long term solution.

lprent: Geographical variations probably already make this already the case in parts of Auckland. In the leadup to the 2011 election, there were electorates in South Auckland that had less than 50% of households with listed landlines. The Auckland isthmus had electorates with less than 60%. Whereas the North Shore electorates were largely over 75%. By contrast, some South Island rural electorates had close to 90%.

I’ve also observed a distinct age variant about what households have listed landlines. You’d have to argue pretty hard and with a lot of evidence to convince me that there isn’t a set of societal differential already strongly in play in NZ already distorting the polls without systematic cell coverage.

28 comments on “Polity: Endangered! Polls without cell phones ”

  1. Clemgeopin 1

    That is so revealing and astounding. After each poll, many people usually question the sampling phone methodology used by the polling companies here.

    As an aside, I am going to predict (guess) my estimate of the possible final election result on Sept 20, from now on up to the election, after each of the regular polls.

    Based on the published poll results so far, my first guess of the party vote is as follows:

    NATIONAL…..43.5 %
    LABOUR……..33.3 %
    GREENS………10.0 %
    NZF……………06.9 %
    CONS………….02.7 %
    MAORI………..01.3 %
    MANA…………01.0 %
    ACT……………00.2 %
    OTHERS………01.1 %

  2. Draco T Bastard 2

    I know some pollsters are trying out online panels as a way to get the cell phone only population without the expense of talking to lots of cell phones. That is an innovative approach, and worth exploring. The sampling comparisons would need to be super robust, though, for it to be a long term solution.

    I know Horizon Poll ask for your full information to join their polling panel. Name, address, income, etc. When they do a poll they match the randomly selected people against StatsNZ demographical information.

  3. lonelyavenger 3

    Roy Morgan call cell phones at least as part of their general election polls.

  4. weka 4

    The cell-only people in NZ, what are they doing for internet access?

    • Tazirev 4.1

      I only have a Skinny cellphone and tstick, combined that gives me 260 Mb of data and 60 minutes of calls a week for $8 a week, more than ample for my needs

    • lprent 4.2

      I have a landline. It hasn’t been listed since 1991 and I seldom give that number out any more because we’re often not at home. When I’m at home alone working, I seldom answer it unless I know the number. That is because phone calls and coding aren’t that compatible. But anyway the only people who use it much are Lyn’s parents on the weekend. Seven times out of ten they get me and I really only pick up because I know their number.

      And it is VOIP on naked broadband anyway and I see that last month there were 15 incoming calls on it. I was just thinking about dropping the VOIP from the fibre when it gets installed.

    • Draco T Bastard 4.3

      Naked DSL?

      I keep wondering why land-lines are still provided.

      • Ergo Robertina 4.3.1

        ‘I keep wondering why land-lines are still provided.’

        You wouldn’t be wondering for long if you experienced a natural disaster which outed power to charge electrical devices and your landline became essential.

          • Ergo Robertina 4.3.1.1.1

            It’s not just the individual device; cellular systems are more vulnerable in a disaster than landlines. Even though cell networks are getting stronger, it’s logical to have two independent systems so if one goes down we have the other.
            In America, the Federal Communications Commission has said the old landline system is unsustainable because of cost. But New Zealand must not follow America, where basic services like rubbish collection, street lighting, fire fighting, law enforcement, and road maintenance are being cut as cities go broke, but the rich get richer.

            • Colonial Viper 4.3.1.1.1.1

              Oddly enough, law enforcement and rubbish collection don’t usually get cut in wealthy white counties.

            • Draco T Bastard 4.3.1.1.1.2

              Even though cell networks are getting stronger, it’s logical to have two independent systems so if one goes down we have the other.

              They’re not independent – they’re fed on the same trunk lines.

              What you keep bringing up is power. A power cut won’t immediately crash the exchanges because the exchanges have batteries and generators to keep going. Wouldn’t be surprised if the cell phone towers have similar for emergency purposes. The real big problem with the cell network is that it just doesn’t have the capacity for everyone to use them all at the same time which is why the telcos always start telling everyone in an affected area to stop using their cell-phones unless in an emergency. Thing is, neither does the land line network.

              Also, modern cell phones have an FM radio built into them so general information would be available whether the telephones were working or not.

              In America, the Federal Communications Commission has said the old landline system is unsustainable because of cost.

              The old land line system will be going the way of the dodo. Cost will have an effect upon that but mostly it will be that no-ones using it and so all that precious copper will get dug up and recycled.

              But New Zealand must not follow America, where basic services like rubbish collection, street lighting, fire fighting, law enforcement, and road maintenance are being cut as cities go broke, but the rich get richer.

              Of course we shouldn’t but we keep doing so anyway. For some strange reason the politicians, especially those in National, seem to think that the US works really well.

      • weka 4.3.2

        How available is naked DSL, location and price wise?

        • weka 4.3.2.1

          Vodafone’s naked BB starts at $85/mth. That’s not cheap, so I’m guessing that low income people that have given up landlines are not using naked BB.

          • Lanthanide 4.3.2.1.1

            That’s a good question.

            Minimum DSL packages in NZ really are quite expensive. IMO you should be able to get a really basic plan, in terms of speed and maybe 5Gb of traffic, for $30-40/month, but you simply can’t.

            So I guess people who don’t have naked DSL / landline, but have a cell phone, probably use the internet at their work, or the library, or maybe just a data plan on their phone. Or they don’t use it at all.

            • weka 4.3.2.1.1.1

              Mobile data (phone or t-stick) is about $20/Gb, so not cheap either, unless one is using small amounts.

              I suppose the next question is how many people have internet at home/in person. If the idea is that many low income people have cell phones and no landlines (hence one of the skews in the polls), does that also mean they don’t have good internet access?

              • Draco T Bastard

                Entirely possible that they don’t which brings about the question of just what Labour plan on doing to get it to them.

          • Tamati 4.3.2.1.2

            Do you know someone who has a cellphone contract with Vodafone? Get them to change their address to your place, then get them to call up Vodafone and tell them you want link up the two accounts. You get 30 bucks of your naked broadband but can have each account in a different name.

            Did it with my flatmate who moved out two years ago. Vodafone none the wiser.

        • nadis 4.3.2.2

          we use compass broadband mammoth plan. Naked broadband, no data cap at $85 per month. Our average monthly usage (2 adults, 4 teenagers/young adults) is around 150 to 200 GB per month. All of the kids swear it’s not them downloading tv shows and movies.

    • Tracey 4.4

      I think there are larger pockets than some imagine who do not have in-house internet, unless through their phones. Just guessing mind.

    • nadis 4.5

      we have no landline and an uncapped broadband plan (via ADSL) for $85 per month.

      6 cellphones in the immediate family, my wife and are on unlimited call to mobiles and landline plans, kids are on $19 a month 2 degrees plans. Use skype but mainly for work related video calls.

      Havent missed a landline at all – the only calls that really came with it were marketing calls.

      I’m generally a Nat voter so not all cellphone only users are lefties.

      As long cellphone/landline ownership is not a variable that pollsrters need to measure (i.e., not a variable that is highly correlated with political outlook – it may or may not be, I dont know) then statistically it is very possible to correct a survey for the block of people who are not contactable by the survey method. You know the population consists of (say) 10% males between 18 and 30 living in Auckland, your poll size is 1000, you keep calling landlines until your reach 100 that fit the demographic.

      Polling is pretty sophisticated and the chances of surprise outside the maargin of error from a properly constructed poll is pretty low i.e., the margin of error. Most commentators on blogs and in the media completely misrepresent what the margin of error is. National at 45% with a margin of error of 4% does not mean their result could actually be as low as 41% or as high as 49%. What it really means is that if you repeated the survey 100 times using the same techniques, 96% of the time the poll would indicate support for national in the range [41%, 49%], 4% of the time it would be outside that range, and depending on the mechanics of the poll (usually) normally distributed around an average of 45%.

    • Sanctuary 4.6

      I’ve got a 500GB connection without a landline. I use Skype and a bluetooth headset to make the odd PSTN call. Otherwise, I have a mobile with a 500MB data plan for all my whatsapp and Skype needs.

      I never answer mobile calls from numbers I don’t recognise, they can leave a VM and I’ll call them back if I want.

  5. Grumpollie has posted on this…

    I agree with Rob Salmond that within five years polling methodologies will likely change.

    …but adds some of his own thoughts, then concludes:

    Calling cells is not, and will never be, the magic bullet for opinion polling.

    http://grumpollie.wordpress.com/2014/03/13/rob-salmonds-post-on-cell-phone-polling/

  6. Steve Withers 6

    My experience has been that Kiwis have been ahead of Americans – generally – on cell phone adoption. We only pay for outgoing calls and txts. Canadians and Americans pay for incoming and outgoing calls and txts. We pay one rate for the whole country. They pay a myriad of rates local calls, local toll calls and long distance toll calls – inward and out ward. I had a cellphone in NZ for most of a decade for ANY of my family or friends in Canada had one. For all practical purposes, the Canadian cellular market progress has been essentially the same as the US cellular market due to the high degree of integration. Certainly adoption will higher in some regions than others….but overall, the high cost of using a cell phone in North America compared to a $10 / year pre-pay in NZ kept a lot of people there from going mobile for a long time. A pre-pay in North America might cost $20 / month – use or lose it in 4 weeks with calls costing 50 cents / minute outward local and $1.39 / minute for tolls….outward. Inward calls are cheaper – but you still pay by the minute. Pricing does vary widely and a monthly account of about $150 can see you with unlimited almost everything.

  7. Rosie 7

    Can’t argue with the above graph and this:

    “In the leadup to the 2011 election, there were electorates in South Auckland that had less than 50% of households with listed landlines. The Auckland isthmus had electorates with less than 60%. Whereas the North Shore electorates were largely over 75%. By contrast, some South Island rural electorates had close to 90%.”

    However, I want to know what I am missing. Our landline and internet connection with 80GB is $80 per month with Slingshot. I use my cell (old stylez, not smartphone) and use a 2 Degree’s $20 pre pay top up voucher maybe once every two months at the most. The other person in the house, Mr R, tops up with a $20 voucher once every 6 months.

    It is far more economical for us to use our landline than our cell phones.

    In households where there are say 4 people with no landline and 4 individual cell contracts/pre pays, isn’t that going to add up to more than the cost of the one landline, depending on use? Or is it that I just don’t see the need to be contactable at all times on a private individually owned phone?

    A friend texted the other day to say she was happy she was no longer with Vodafone as she was paying $40 – $60 per week!!! How could that be? Then on the other hand I have friends that don’t even have cell phones.

    I see there was a discussion above about the importance of non cordless landlines during an emergency. I can vouch for that. We acquired an old style phone for our emergency kit. It really came in handy during last years’ storm when our power was out for 19 hours!

    And finally, my rarely used cell phone actually rang the other day. It was Colmar Brunton, wanting to do a survey on ACC’s pain management services………

    • Rosie 7.1

      One more thought. I can, however, see the advantages in living in land line free accommodation if you’re living alone or living in a boarding house type situation. If you’re needing to keep costs down and want to use the internet and check emails etc you can use the free wifi at the library. You’d be using a cheapie pre pay cell and really relying on people contacting you rather than the other way around.

      While this is cheaper than a land line it’s not necessarily convenient and to a degree may inhibit your social interaction, as well as making you less available to polling companies under the current most common polling method. Therefore, these companies may be missing out on obtaining valuable information on this sector of society, which comes back to the essence of the article.

      Still though, why are multiple cell phones necessary in mulitple person dwellings? Isn’t this more expensive?

  8. Tracey 8

    free wifi at the library…. mt albert library has a very long wait to get onto a system, especially when unitec is back.

The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.