John Gibson: Your blog on my Marsden Fund project

This is an unedited Right of Reply.  I will also not comment.  You may see that John has misconceptions about our commenters – which isn’t unusual.  If you’re going to complain about it, then you’ll end up proving his point.  I suggest you stick to the topic at hand and make me proud.  Thanks.

Dear Cameron

You raised some questions today in your blog about my project which the Marsden Fund will support from 2015-17. Let me try to answer them. Please note that I am directly e-mailing you because I don’t comment on blogs. Some commentators on blogs seem impervious to evidence, so my commenting would probably be a waste of time. When I produce evidence that might be of interest to a blogger I do sometimes send it to them directly (e.g. on rising public sector wage premiums to David Farrar and on school zone effects to Eric Crampton)  so I am not opposed to blogs per se but I do prefer to limit my engagement with them. Also note, that while I am an empirical economist and so will support conclusions based on what the evidence indicates, politically I would be broadly considered as right-of-center (not that this should be relevant) and am well aware of the limits to state action and the importance of individual responsibility.

a) what is already known in NZ and why is the study necessary?

A series of studies by Wellington School of Medicine researchers along with economists from Otago and NZIER used two types of data from Stats NZ: the Household Economic Survey (HES) and Food Price Index (FPI) data. They modeled the effect of prices on the shares of the household budget allocated to a number of food and beverage groups, with “energy drinks” as one group and “carbonated soft drinks” as another. Their model examined how households altered their budget shares for each item as prices changed. Based on their results, they concluded that “a 10% tax on carbonated soft drinks could lead to a 13% decrease in population purchases of these products”. The reduced purchases were assumed to all be in terms of quantities purchased, and this evidence, along with similar overseas studies has been influential with the NZMA etc in recommending a 20% tax on fizzy drinks as one way to combat rising obesity.   

The problem is that the HES does not, and cannot, tell us the quantity of soft drinks consumed, yet it is the quantity consumed, and not the amount spent, that affects calorie intake and potentially obesity etc. Consumers may respond to higher prices by either reducing quantity (as assumed by this existing study and almost all others around the world) or by reducing quality (switching from Coke to Pams etc), or some combination of both.

While this insight might seem very simple to some of your readers, who seem to know the answers a priori, it is not something that research in either NZ or elsewhere has considered. Why does it matter? Lets say, for the sake of argument, that half the response by consumers to price rises is quality downgrading and the other half of the response is cutting quantity. If that were the case, then any proposed tax to moderate quantity consumed would be only one half as effective as expected, or alternatively, to get to a particular reduction in quantity the rate of tax increase would need to be at least twice as big. And since taxes cause costly distortions knowing what the ratio of quantity responses to quality responses actually is should be helpful to Treasury, MoH and others who might debate these issues and design policy based on the published evidence.

b) why is the study not using NZ data?

Because NZ does not have the sort of data that are needed, and it would go well beyond the funding support that the Marsden Fund can give to gather such data. In order to work out consumer responses on the quality and quantity margin, we need data on three things: (i) price, (ii) spending or budget shares, and (iii) some index of the quality of what is bought within each of the food or drink groups. In NZ we do not have existing statistical evidence on (iii), we only have (i) or (ii). To get at quality of what is purchased in NZ, one could use scanner data from bar codes, but then if we used that method of research we don’t know anything about the characteristics of the consumer (age, gender, ethnicity, education, income etc). That is the reason why most of these studies here and overseas use household survey data, like the HES, because there is interest in looking at results by different population and income groups.

Also, there isn’t very much price variation over space in NZ – the price of 600ml Coke in my local New World supermarket is probably the exact same price as that item in your local New World supermarket. So it becomes very hard to identify these price responses in NZ.

In contrast, countries like Mexico and Indonesia have huge variation over space in the prices that consumers face, because of poor marketing systems, bad infrastructure, lack of national chain supermarkets and so forth. Moreover, in Indonesia for cigarettes and tobacco, and in Mexico for soft drinks, we do have access to the required data on (i), (ii) and (iii) – demand, price and quality. The statistical agencies in those countries have gone further than almost all other countries in gathering the sort of data that we need. Partly that reflects the importance of the items in these settings, with very high rates of smoking among Indonesian males and very high per capita consumption of soft drinks in Mexico. In order to inform the practice of researchers around the world, it is important to work in settings with the best data and where the issues are most salient.

I also should say that quite a large part of the proposal, which isn’t discussed in the newspapers at all, is experimenting with low cost ways of gathering price and quality data in countries where these are not available (e.g. using smart phones and crowd-sourcing techniques). The findings from that part of the project are as applicable in NZ as anywhere else.

One last point: the Marsden Fund is paid for by NZ taxpayers, and I am very grateful for their support. However, its goals are to contribute to leading edge research which is not necessarily restricted to research that occurs in NZ or solely for NZers. Just as we in NZ benefit from fundamental research which has been done overseas, and was funded by the taxpayers of other countries, so too is our Marsden Fund research designed to be part of this global public good that can have benefits more widely than just these shores. There are a number of other funds where the NZ taxpayer supports research that is more solely aimed to benefit just NZ (e.g. MBIE, HRC and so forth).

Thanks for your time.


John Gibson



THANK YOU for being a subscriber. Because of you Whaleoil is going from strength to strength. It is a little known fact that Whaleoil subscribers are better in bed, good looking and highly intelligent. Sometimes all at once! Please Click Here Now to subscribe to an ad-free Whaleoil.

  • mike

    Why do these studies at all would be my response.

    I’m a layman but I know that too much sugar or fat is potentially bad for me (too much of anything really), everything in moderation should be the key tenet of a healthy life… and sometimes that includes moderation, it’s ok to over indulge now and again.

    I don’t need a $800k study to tell me that jacking up the prices may lead people to buy cheaper drinks. Nor do I need a $1m study to tell me that too much sugar is bad.

    What I’d prefer is for the scientists to focus on things that really matter and stop wasting my tax dollars.

    • WeaselKiss

      As I read your words Mike the thought struck me that many many of the people who are in the group of our society who are worst afflicted by (among other things) glugging too many sugary drinks, would not be able to write a post such as yours, perhaps some could not read it either.
      My point: uneducated people are less likely to make the right drinks choices in this regard.

      • mike

        That may be so, but will they listen to the results of an $800k study?

        Or do we all have to suffer a cost increase in a treat because they are so stupid?

  • dumbshit

    5 bob chucked at parental enlightenment would go a long way

  • mommadog

    Thank you for the explanation. I can now see the gap in knowledge that this research is planning on addressing and I am now clear that the Marsden Fund is for research overall – not just research in and for New Zealand. Still I have three concerns that remain.
    One is that regardless of results if this research is not done in NZ the results will not translate well here because we are a different country with different cultures and different habits. I suspect the results will mean that someone else will apply for more research funds to repeat the study here eventually anyway as one study leads to another.
    My second concern is the topic – even with a strong personal interest in nutrition and the above explanation I just don’t see that studying soft drinks and energy drinks is vital research that is going to make any difference to our health as a society and how we live. We as humans are more than two items we potentially consume and there is already a large amount of knowledge of what is healthy, what is not and why. The need is for more education in some quarters rather than more research. Although I do see how the results of this study will make some residents of NZ very excited and give them another political tool to use. Particularly those who don’t believe we have the capability of making up our own minds about what we consume and should not be left to decide.
    My final concern that remains is not related to the research but to the Marsden fund itself. Because it is a research fund that comes solely from tax-payer dollars according to the above letter I am even more convinced that it should only be used for NZ research and NZ based researchers. Until the rules/regulations of this fund change then good on researchers applying for it as in this case. There needs to be change here.

  • BlitzkriegNZ

    He explains his case very well. I think he misses the point that a huge number of us are angry that this is our tax payer funds going towards a study on how to more effectively control what we do and take more of our money off us for doing something that we consciously choose to do, which is consume crazy amounts of sugar. We all know it’s bad so bad luck if it helps put some of us in an earlier grave, we’re adults and throwing our own money at finding the best way to control what we do is wrong.

  • Dave_1924

    Some points in response

    Well done WOBH for continuing to broaden out and publishing dissenting or clarifying posts. Encouraging a wider exchange of views is useful

    The thrust of the post seems to be there isn’t the data available in NZ or if available its not of the right quality. So the study will be based on overseas datasets – fair enough I suppose.

    However this will always leave the study open to the concern that the dataset used for the study has been selected to arrive at an outcome that supports a specific goal. And obviously makes it highly subject to attack by lobbyists for the industries invovled.

    How will the study control for the different genetic pattern of the NZ population compared to the Mexican and Indonesian populations? How do we know whether different population groups process the calories/sugar in Fizzy differently andf therefore how can we compare across population groups to the NZ context as this study proposes to do?

    What other methods, other than taxation, should be pursued to tackle the defined problem namely rising obesity?

    Why would a compulsory 1 hour physical activity session at every school, every day not tackle the problem by raising calorie burn over?

    Why would teaching children how to prepare satisfying meals that are nutrional balance not help alleviate the problem?

    Why not a campaign targeting parents demonstrating the negative impacts on their children of excess sugar and other foodstuffs not help alleviate the problem?

    Why always Tax?

    Checking to see if tax modifies behaviour is of interest across a number of areas.

    But the approach raises the question: If I consume fizzy drinks, to follow the phrase above, but do so responsibly and in moderation i.e. to splice the odd Bourbon I may have every now and then, why should I be punished via an increased cost for this activity because others can not control themselves?

    Fundamental it is a maternal/paternal-istic approach that doesn’t require people to consider their choices.

    Freedom of choice, in knowledge of the facts, is in my mind at least, much more preferable to constant state intervention

  • Timboh

    The research topic he is addressing falls into the “so what” category. Then using. offshore data that has already been collected and no doubt analysed to death already speaks to the waste of time and money.

  • LesleyNZ

    This is a generalisation but everything has the potential to be bad for you. Some things are “more” bad that others. Smoking is bad – very bad. Too much sugar is also very bad. Taxing sugary products will not stop people buying the sugary products. The only thing that will stop people smoking and eating too much sugar is knowing the damage the smoking and excess in sugar does to your body. Often then it is too late. Recently I saw a display showing how much sugar goes into certain products including fizzy drinks. I intend to avoid fizzy drink as much a possible. Many years ago when I was about 12 our class went to the Epsom showgrounds. There were displays to do with diseases and health issues. I remember seeing hydatids in a jar. What really sticks in mind was the “Smoking Joe” mannequin guy. A lit cigarette was put in his mouth. The smoke swirled down to his glass jar lung. After he had smoked the cigarette the person in charge took out the filter and showed us the sticky yellowy brown tar. I remember thinking – why on earth would anyone want to smoke and put this gooey into their lungs. John – you don’t need to do a study on taxing sugar. It is said “seeing is believing”. Educate the kids and and educate the parents so that good healthy choices will be made. Educate the manufacturers to STOP putting so much sugar (and also salt) in processed food/drink. There is far too much sugar in the popular breakfast cereals including top of the range Muesli (except Weetbix which has a low sugar content). A few years ago the President of Fiji visited a very remote village. He was thirsty when he got there and sat down while someone got him a drink. A villager came with a can of soft fizzy drink. They thought they were giving their honoured guest, the President of Fiji, the best drink of all as it cost them a lot of money – so in their mind it had to be good for you. The President then pointed to a lemon tree and told them the best drink is made from the squeezed lemons from the tree. I remember reading this in one of the first speeches the President of Fiji made. He told the villagers to save their money and not waste it buying fizzy drink which is not good for them and to make their own lemon drink from the beautiful lemons – far better for them. It is all about education. Far better to put taxpayer money into educating – rather than researching what we already know.

  • ozbob68

    You are trying to use economics to influence human nature. If history has taught just anything, nature always finds a way.

  • symgardiner

    A detailed explanation that could have been done much more succinctly.
    Essentially he is saying that we need this research because its quite possible that raising the price of fizz won’t make any difference to demand. It would be better to figure this out before hand rather than go through the hassle of doing an Aunty-Helen-ban-the-light-bulb approach.
    That said, I do worry whether his approach is a good one. If the NZ market has very little price variability between suppliers and actually even product lines, I do wonder of the usefulness of using data sets from markets that do have this variability. Surely that would be an important dynamic.

  • cows4me

    What a load of hoo har . I think we all know what the problem is, obesity. The country doesn’t need another doorstopper of a report, most with more than a few brain cells have already figurer out.

  • 1951

    Very gallant JG to come into the Whale’s den. Like BlitzkriegNZ below, it is not you per se, it is the idea of supporting those who wish to dictate to us on how we should live our lives. Those experts/elitists who depend on our enforced donations/taxes for their very existence and usually much more comfortable lifestyle. I would have no criticism if you were to take your expertise in economics and apply the same energies in holding a blow torch into the like of those mentioned in Cam’s article. Those who have spent a life time being the ambulance-at-the-bottom-of-the-cliff in stead of the fence-at-the-top. The cost to us of keeping 75 Health Scientists, who I bet are very comfortable, who appear to be unable even think as individuals, go on attack if the narrative does not fit their agenda. If the Marsden Fund was used for something more fruitful rather than lead to attempts to sway market forces, it might be more acceptable.

  • arlo

    Problem = obesity
    Proposed solution = bigger government, higher taxes
    Predicted result = obesity continues unabated
    The real problem = inactivity

    How will you tax people into moving more which is the real cause of the problem?

  • Sunshine

    There is quite a bit of research data in this area already. I get really annoyed that the researchers, when they
    carry out these research studies, never go so far as to monitor biomarkers of the subjects. What is the point of this research if, they do not medically look at whether taxes on unhealthy foods actually improves ones health. They already know taxes have a limited effect on the actual spending habits as in directing people to forgo “bad fizzy” for healthy fruit and veg. Studies so far have shown
    a) taxes can generate a huge sum of money for the government (that’s how it is promoted to policy makers)
    b) Long term spending habits in favour of choosing cheaper healthier foods over higher taxed unhealthy foods – a small statistically insignificant, percentage will make small changes for about six months. After one year, virtually no one has made permanent changes.
    c) taxing “unhealthy” food items does not make people buy healthy food items. Many choose the next unhealthy alternative instead.
    I once asked a professor, an active researcher in this area, “Did you collect medical data from the study subjects before, during and after” the study on how taxes affect spend habits. The reply was “no, that is too expensive and would be hard to do”
    If the taxpayers are to fund more research, it needs to actually include data on the biomarkers of the subjects, so that a comprehensive look into actual health benefits can be ascertained.

  • Day Day

    Scientific papers that propose political solutions should be classified under A for agitprop.

  • kiwiinamerica

    Short version – I have a degree and teach at Uni and know better than the great unwashed what’s best for them.

    I met a health nazi bureaucrat from Scotland at a touch ref’s tournament. He was doing similar research on how to boss the Scots around by using taxes on fat and sugar. The notion of individual choice and liberty were an anathema him. He was on a mission to save his people from the twin evils of fatty stews and fizzy.

  • RightofSingapore

    Wasn’t it Denmark that brought in the sugar and fat taxes yet saw obesity continue to increase?
    I shouldn;t have to pay more for foods I like or have my choices restricted just because some fatty boomsticks lacks discipline and self-control.

    • kiwiinamerica

      The fat tax was scrapped in Denmark. It made no difference to obesity, was difficult to figure out, was very unpopular and raised little revenue because, shock horror, the Danes bought their fatty food tax free in Germany or Sweden.

  • Teletubby

    Genuine question, maybe John Gibson could bend his personal rules a bit and answer.
    My understanding is that this study is going to take 3 yrs and I believe it involves 3 people for that time but the study will be done using data from overseas. Given that there is no data collection or observation stage in this study how is it going to take years?

  • Citizen

    Shouldn’t we be actually looking at why obesity is on the rise…?. Maybe ban video games, have we had a study comparing hours spent playing games versus playing outside in say…1970? Or television?. Or the evil sky TV…so many channels..and a remote control…could the remote be the cause? Just don’t come after bacon!!

    • Dumrse

      Bacon must get on the list eventually. Its the way the system works, when the job is done, find another. Just as Cam said, first it’s smokes, then it’s fat n fizz. Sure as god made little apples….light bulbs and shower heads will get another look in.

      • Citizen

        Maybe they ought to look at thats known as the French Paradox. Wine, Cheese $ Fat…but very low levels of obesity or heart disease.

  • I took some data sets from research and then modelled them.

    I didn’t actually talk to the people, after all its a computer model, what could go wrong.

    The model predicts less money on fizzy drinks, but does not determine quantity of fizzy drinks.

    They’re doing lots of research for smoking in countries with high prevalence of smoking, that’s like fizzy drinks!

    However you can’t do this for fizzy drinks because we only use data we don’t talk to people, and there is no way of knowing from the data what the demographics of the consumer is.

    Then I got confused “So it becomes very hard to identify these price responses in NZ.”

    Which says to me , pricing in NZ means that it’s difficult to affect consumption as pricing is not the variable that determines consumption

    This isn’t research is trying to use data in a way that fits your desired outcome

  • Dave

    Once again the academia has the ambulance half way down the cliff. My belief is simple! If your going to have a Tax system, have one?? If your going to have a welfare system, have one, likewise with a health system. BUT don’t intertwine them, run them separately. Likewise, if there is an obesity problem, look to fix it, not to intertwine Tax with soft drink and obesity, the lower wage earners will substitute until a few months as they adjust other spending and then revert to pre tax spending, look at cigarette and petrol price rises.

    Hence the study is not relevant, it’s fiddling with economics 101, and intertwining tax to influence consumer behaviour. If your serious about sugar and obesity then either educate or regulate, ban it at all schools and govt/council sport facilities. But please don’t forget the real issue is the amount of calories in to the amount of calories consumed, which begs the question, should you be looking instead of how many kids walk 2 km a day or more to combat obesity, and then tax school buses and ban car parks within 200 metres of any schools, Xboxes, TV etc etc.

    Finally, let’s not forget wine and beer contain sugar, are they in the study as well, or should any proposed tax just be shifted to All Sugar.

  • Second time around

    It is good that John Gibson should choose to defend his research in the media. Research, necessarily, has to be of limited and well defined scope, otherwise it never gets completed. The Marsden Committee has chosen to fund his research for what must be sound reasons. All the same I have some misgivings. Commodity size packs of sweetened drinks, and the high sugar and high caffeine drinks are all a recent phenomenon. Coca Cola is advertised as if it was a dietary staple, and if the generic versions are cheaper than the branded product, it seems unlikely that pricing rules and extra taxes could have any meaningful impact on consumption. Taxes were not a solution in Denmark and although there is already a 15% tax in the form of GST in NZ (water in the same volume has effectively no GST) people still choose to purchase these drinks. My feeling is that the solution lies in education, not in taxation, and that society would enjoy greater benefit if the research study had had a different focus.

    • 1951

      Marsden Committee …research…sound reasons…? “You scratch my back, I scratch yours” comes to mind about these so called Health Scientists.

  • Never in the dark…..

    Have I read things correctly? That this isn’t a ‘study’ per se, but rather a set analyses on data collated by other people/organizations elsewhere, not in NZ.

    Sounds like easy moeny to me.

    As others have suggested, there are ridiculous duties, levies and taxes on tobacco, yet people still smoke. The biggest impact on tobacco consumption, from my anecdotal observation, is the smoke bans imposed in bars, restaurants etc. Partly why I quit. It was becoming too much of a hassle.

  • Hedgehog

    The problem I see is the overseas data and the following statistical analysis. We all know that If you start with a bias, the data analysis will be skewed accordingly. Any analysis can be managed to support the outcome that is desired. Therefore after 3 years and a whole bunch of money later all we end up with is an overlay of Mexican consumption of fizzy drink in New Zealand. In other words nothing meaningful.

  • dennis

    The plan is to suggest a tax on sugary drinks ( first stage ) then have some of that tax allocated to more fat research . Set up a fat foundation to educate fat people from another portion of the sugar tax and hey presto a job for life and the pleasure of telling people what to do.

  • Alright

    “Also note, that….I am an empirical economist……….”

    Excellent. So are (generally) people who fly hot air balloons.

  • andrewo

    I’m more than happy that new taxes are applied to soft drinks…because I don’t drink them (and I think you’re dumb if you do) but I would require those taxes to be fiscally neutral – reduce taxes somewhere else so the government doesn’t gain revenue.

  • MrBarrington

    Seems to me that you are conceptually right, and fundamentally wrong. There is NZ data on quality and income levels, however, these are inferred rather than from direct surveys. However, that doesn’t mean that an inferred variable is less valuable… you might pick up income and quality data from surveys, but you can measure this more accurately with direct measurement.

    Time to get out the tape measure and read up on how supermarkets work. Part of what a supermarket does is rent shelf space. So for the companies that rent this space, the returns from that ‘real estate’ have to be sufficient to make a profit. So for a supermarket in a wealthy area the amount of space given over to ‘fizzy drinks’ may be different from a low income area; think Devonport vs Mangere in Auckland…

    This should tell you something about consumer preferences. Also you should be careful to look at the % of total shelf space given over to this product group vs ‘healthy’ products such as vegetables/fruit.

    A further point is non-sugared drinks… I see in my local supermarket that there seems to be about as much space given to Coke Zero/Diet Coke vs regular Coke…

    Finally, I worry about the error value in your equation… I would want to see a 99% confidence level easily passed and a nice high T test.. if the error value is non trivial then your work will be suspect.

  • Alan Wilkinson

    Seems completely pointless since if it did indeed have any impact the market will simply respond with new product/price combinations to satisfy the demand. There is plenty of profit margin in sugared water to do so, let alone supplying the components separately.

  • I think this an excellent and valid response, especially if one takes a Bayesian approach. John is quite correct in that we do lack sufficient data in NZ to deduce and validate conclusions however the effect inter alia of price/tax vs reduced intake is only an assumption. And assumptions without facts remain assumptions. While we may puzzle at the research element it is simply another small piece of knowledge that may hopefully contribute more to decision making than what is surmised.
    When I completed my PhD my professor said to me that for seven minutes I knew more about my topic than anyone else on the planet. In seven minutes someone’s efforts would surpass mine. Yet it didn’t make my infinitesimal contribution any less valid. Hopefully John’s research will add something valid …