The Australian Paradox

Professor Jennie Brand-Miller and Dr Alan Barclay

Questions and Answers

An Australian economist has posted a large volume of "questions" about the Australian Paradox paper. In reality most are vitriolic statements. We have tried to answer what we think were the key issues raised:

1) Figure 5A from The Australian Paradox shows a 30% rise in sugary softs drinks not a decline. Figures 5B and 6 provide the complete picture. The composition of nutritively sweetened beverages changes constantly as manufacturers introduce new varieties and new flavour variants of existing varieties (for example, Pepsi Next contains 30% less sugar than regular Pepsi, but it would still be counted as a sugar sweetened beverage in figure 5A). Consequently, the sugar content of nutritively sweetened beverages changes over time. This fact is not picked up in the volume of sales data presented in Figure 5A. However, Figure 6 shows clearly that the total amount of added sugar (tonnes) in soft drinks decreased from the mid 1990's to the mid 2000's. This fact has been confirmed with the recent publication of new research into Australians sugar sweetened beverage consumption which found that over the 15 year period of 1997-2011 "Sugar contribution from water-based beverages and soft drinks fell from 9.2 to 7.6 kg per person and from 8.4 to 6.2 kg per person, respective".

2) Figure 5A proves that rather than having "decreased by 10%" nutritively sweetened beverage consumption has "increased by 30%". This is disingenuous. Figure 5 contains two parts for a reason. The 10% decrease relates to Figure 5B. Figure 6 provides the complete picture as discussed above.

3) The ABS "apparent consumption of sugar" series was discontinued by the ABS after 1998-99. This is correct. All apparent consumption data ceased at that point in time, with the exception of alcoholic beverages, due to resourcing issues at the ABS. FAOStat extended the series to 2003, and FAOStat data were used to enable a direct comparison between Australia, the United Kingdom (UK) and the United States (US) in The Australian Paradox. Other notable authors were still quoting ABS apparent consumption data at the time The Australian Paradox was written. Most notably the 2003 Dietary Guidelines for Australian Adults (pages 171 - 191).

4) Did you not notice the fact that the database underlying your preferred FAO "apparent consumption of sugar" series had been discontinued in 1998-99? The research underpinning The Australian Paradox was conducted in 2009 and the most recent data available at that point in time were utilised. As one of the primary aims was to compare sugar consumption patterns in Australia to that of the US and the UK, FAOStat was utilised and at that point in time it had data to 2003. ABS Apparent Consumption data and FAOStat data for Australia were essentially the same up until 1998-99, so it was assumed that they were using very similar methodology.

5) You cannot reasonably publish the story of a 30-year decline - 1980 to 2010 - as fact when your preferred series is invalid for over one-third of the relevant time-frame? The Australian Paradox looked at 3 primary sources of information 1) National nutrition survey data (1983-2007); 2) Beverage consumption data (1997-2006); and 3) FAOStat data (1980-2003). Collectively these data spanned 27 years. Furthermore, the abstract of The Australian Paradox states very clearly that "In Australia, the UK and USA, per capita consumption of refined sucrose decreased by 23%, 10% and 20% respectively from 1980 to 2003".

6) In your initial rebuttal of my critique in March, your main specific argument against my analysis was that the uptrend in "sugar availability" in Figure 1 (ABARE data) was a poor indication of human consumption because cars not humans were consuming up to 14kg per person of sugar via the rapid growth in ethanol production. A very early draft of our rebuttal contained one sentence and a footnote that included calculations for biofuel conversion because our initial investigations provided reason to believe that some Australian sugar was being diverted to ethanol production in the early 2000's. It amounted to 3 sentences (148 words) out of a 2,082 word document and as such was by no means the "main specific argument" against the economists fallacious arguments as claimed. Further investigation prior to publication of our detailed rebuttal indicated that sucrose was not used for biofuel production in Australia, so the sentence and footnote were removed. It is now apparent that much of the additional growth in sugar availability seen in ABARE data over this time period was due to exports of sugar in the form of value-added ingredients to Japan. ABARE data are simply not a valid measure of Australian's sugar consumption.

7) Why did you not agree to formally correct or retract your entire error-ridden Australian Paradox paper? None of the sources utilized in The Australian Paradox papers have been retracted including the ABS apparent consumption data, FAOStat data, National Nutrition Survey data or beverage consumption data. Furthermore, there is new independent evidence that the apparent consumption of sugar continues to decline in Australia, and that sugar from sugar sweetened beverages is declining, adding weight to the conclusions of The Australian Paradox papers. Finally, the Australian Governments Australian Health Survey has confirmed that total sugar consumption is decreasing in Australia.