What Might Be the Impact of an Australian Soda Tax?
With the implementation of high profile sugar and soft drinks taxation in France in 2012, in Mexico in 2014 and Berkley, California in January of this year, the global debate concerning the purpose and efficacy of excise tax proposals on sugary beverages is inevitably moving in to other high per capita markets for carbonates.
Presently, the topic is on the agenda in Australia, a top 10 market for carbonates consumption in terms of per capita retail volume sold, and yet another country where obesity and other public health concerns are driving interest in added taxation as a potential policy solution. Fifteen years of volume and value sales data for carbonated drinks in the Australian market, published as part of Euromonitor’s non-alcoholic drinks research program, allow us to speculate on the potential impact of a soft drinks tax by considering the historic impact that price increases have had on Australian retail sales of carbonates, with a focus on cola.
Cola as a case study
As part of non-alcoholic drinks research published this January, Euromonitor International employed an inductive demand model to aid in five-year forecasting. The forecast model attempts to identify several measureable and statistically significant demand factors (including retail price) from historic data sets of the 80 markets researched. These factors are tested against historically available data for retail and on-trade beverage category sales, and then weighted to assist in building 2015-2019 country forecasts.
For Australia, the results demonstrate that a 1% increase in the retail selling price of regular, full flavour cola carbonates can be expected to yield just a 0.2% decrease in retail volume. Consequently, even a relatively substantial (and hypothetical) 7% increase in pricing in 2015 would yield only half a percentage point difference in expected declines: from a 6.5% forecasted reduction in off-trade regular, full-flavour cola volume for 2015, to a 7.0% reduction in 2015 under a soda tax scenario.
Discounting other factors, this finding suggests a weak relationship between price hikes and volume declines in Australian standard cola. However, this finding is simply based on observable data from the market and should not be oversimplified. In constant 2014 Australian dollars, retail unit prices for cola carbonates (including regular and low calorie cola alternatives) have fallen consistently over the review period – by 17% in total over 2000-14. There is greater uncertainty over the impact of a substantial soda tax in Australia, because there is simply no precedent for a substantial price shock in the Australian retail market. Furthermore, the introduction of such taxation would necessarily be accompanied by a high profile health and public policy debate in the media that may further impact consumer attitudes and behaviours towards the cola and wider carbonates category for reasons other than simple price.
Australia: Forecasted Off-trade Volume Forecast for Regular Cola Carbonates
Source: Euromonitor International
Note: Forecasted scenario based on Euromonitor Industry Demand Model (2015ed)
Would there be a public health benefit to a soda tax?
The policy argument for excise taxation on carbonates – or similar Pigovian taxation on other products, including alcohol and tobacco – is that taxes ultimately raise prices to the consumer, driving down overall consumption of unhealthy products. The low sensitivity of standard, regular cola retail volume consumption to changes in retail price in Australia and the relative importance of other demand drivers makes it difficult to draw hard conclusions about the immediate impact such a tax might have on consumption and health.
Additionally, as a developed soft drinks market, consumers in Australia have a wealth of diet, low-calorie, zero calorie, and other non-cola alternatives to replace regular cola carbonates in their diet. In fact, Australian low-calorie carbonates have gained considerably on regular cola over the review period. Crucially, for the first time in 2014, Euromonitor’s data suggests that low-calorie cola outsold regular cola carbonates in terms of retail volume in Australia.
We know that there have been substantial declines in standard, full flavour cola (down 22% in off-trade volume over 2000-2014) and wider carbonated beverages in Australia over the recent review period. Interestingly, these declines have taken place in an environment of flat or declining prices in real terms and have been accompanied by consumer migration to low calorie cola (and non-cola carbonate) alternatives. Recent volume declines, independent of observable category price increases, have had an impact on sugar consumption received from soft drinks, according to Euromonitor International’s Nutrition system. In 2011, Australians received an estimated 12.62g of sugar per capita, per diem from cola carbonate beverages. By 2014, this figure has fallen to 11.83g of sugar per capita, largely as the result of a 3% decline in cola carbonates retail volume over that same period. The amount of per capita, per diem sugar from cola carbonates is expected to fall to 10.28g by 2019, independent of excise tax legislation.
Is a soda tax necessary?
It may be worth considering whether consumers in Australia – and indeed in many developed markets – are addressing well publicised concerns about the category by exiting cola for other alternatives, independent of price considerations and motivated instead by health or taste considerations.
There is a weak observable relationship in historic volume data between cola consumption and price in Australia. This is primarily because there is little precedent for substantial price increases in retail cola, supported by a strong consumer expectation for discounting that has kept the price environment in the category flat or declining. It may be the case that a substantial price shock could have a disruptive and unexpected impact on consumption. However, even in a low price, discount oriented environment for full flavour cola, volume sales have declined substantially as consumers migrate to alternative beverage categories, including low-calorie colas (led by the brands Pepsi Max and Coca-Cola Zero).
Cola consumption (regular and low-calorie) is expected to decline by 9% over 2015-19, with regular, full-flavour cola expected to decline by a staggering 25% in just five years, independent of any tax increase. Consumption of total carbonates is expected to decline by 5% in retail. While soda taxes will gather political and media attention as a response to public health issues across food and beverages, it is worth considering whether consumers are already responding to health concerns in their soft drinks, largely independent of price considerations. Regular standard cola carbonates in Australia have declined, to the benefit of low-calorie alternatives, with a positive impact on per capita, per diem sugar consumption.
Australia Standard Regular Cola Off-Trade Historic Volume and Forecast, 2009-2019
Source: Euromonitor International
The chart above demonstrates this point clearly. In light blue, we have an industry demand model forecast estimate built only on core economic factors that influence consumer goods: population growth, average income growth, price and habit persistence (a lagged effect of growth in the previous year). If these factors alone were used to predict growth in Australian cola, a flat performance might be expected over the next five years. However, in dark blue, the actual published Passport forecast shows a CAGR of -5.4%, in stark contrast to the -0.6% CAGR expected by the industry demand model. The consensus forecast is revised down by 4.8 percentage points to account for unmeasured factors outside the demand model, most prominently rapidly changing attitudes to health, sugar and lifestyles. These consumer-led factors are expected to be the driver behind declines in the Australian cola category. It is therefore worth wondering whether sugar and soda taxation proposals are seeking to address a health question to which the Australian consumer has already found an answer.