We predict that a 10% tax on SSBs in Ireland would result in around 9,900 fewer adults with obesity and around 14,400 fewer adults who are obese or overweight. The greatest reductions would be seen in young to middle-aged adults (25–44 years). The tax appears to have a broadly similar impact, in terms of obesity, across different income groups. This study contributes to discussions of the proposed tax in Ireland using country specific data on population, consumption and body mass index.
Strengths and limitations
This study has a number of strengths. Firstly it uses a widely published model (PRIME), based on a validated set of equations for how energy intake affects BMI, to estimate how the predicted consumption changes impact on obesity and overweight. Secondly, it explores the differential effects of an SSB tax, both by income and age. Third, it is one of the first studies to explore the effects of an SSB tax outside the US, where patterns of SSB consumption and obesity prevalence may be different .
A limitation of this study is that it does not use empirical estimates of own and cross-price elasticities for Ireland. This is because there are no Ireland specific data that could be used. Furthermore, the modelled tax is based on two categories from the SLAN food frequency questionnaire: “fizzy soft drinks” (not low calorie), and “fruit squash”. Within the SLAN dataset, there are no specific categories for energy drinks or fruit juice with added sugar so changes in consumption of these drinks have not been explicitly modelled. Nonetheless the point-estimate of −0.9 for own-price elasticity is more conservative than estimates for Ireland,  and similar  or less than that used in the US modelling studies reductions [15, 22, 29]. It is also consistent with a recent meta-analysis estimate of −0.93 . A more recent review from US studies a price elasticity of demand for SSBs of −1.21 (range −0.71 to −3.87), averaged across 12 studies, although the definition here tended to be carbonated drinks with added sugar rather than any soft drink with added sugar . We also note elasticity estimates for sports drinks and fruit juices (both of which form part of the government Special Action Group on Obesity’s definition of SSBs but are not explicitly collected in the SLAN food frequency questionnaire) are relatively large (−2.44 and −1.41 respectively) and greater than the estimates we applied here .
While empirical cross-price data exists for other countries, it was not thought appropriate to use such data as the substitution effects (and cross-price effects measured as a percentage change) are likely to be heavily dependent on baseline consumption of the substituted product, which may vary markedly between countries. For this reason using data from the US, or another country, would be highly unreliable. The incorporation of cross-price effects would be likely to attenuate the observed effects. Consistent with this, some animal and human experimental data suggest that sugar consumption is addictive which might suggest that if people consume fewer SSBs,  they may seek to increase sugar consumption in other parts of their diet. However, part of the rationale for taxing SSBs is that their consumption leads to ‘passive over-consumption’ , and that it should be possible to reduce the consumption of SSBs without stimulating appetite and causing compensatory consumption of other calories [5, 10, 24]. The evidence of the relative impact of these two effects is mixed. Some US data suggest that an SSB tax would not lead to a shift towards sugary foods,  although other trial data suggest some compensation with food calories may occur . We also note that substitution towards fruit juice and milk in response to an SSB tax, both of which would attenuate the impact of the modelled tax on calorie reduction, have been described [28, 47, 49].
Although our study provides estimates of impact by age and income group, these are not based on age and income specific price-elasticity estimates. This is because age and income specific estimates are not available for the Irish population. Other studies have found that the price elasticity will vary by income group with some indicating that people in lower income groups will be more price-elastic,[50, 51] and others finding higher income groups to be more price-elastic [49, 52]. We have also not considered the impact by baseline consumption and there is some data to suggest that high consumers may be less price-elastic than low consumers [47, 53]. Given that in Ireland people from lower income groups consume more SSBs, if they were less price-elastic we may be overestimating the effect on population obesity, however, if they were more price elastic our results may be an underestimation.
Price responsiveness by age has not been explored. Differential price responsiveness by age is possible and if the price response was greater among the older population this might attenuate the age pattern we have observed. However given that consumption is so strongly patterned by age (68% of 18–24 year olds drink one serving a month compared to 16% of those aged over 65 years),  we feel that it is unlikely that large differences in price responsiveness will occur between different age groups such that the age-effects might disappear. Moreover differential price-responsiveness by age may be a less important factor as some SSBs are brought for consumption in the home (and are not purchased directly by the individual who consumes those drinks).
The pass on rate was assumed to be 90%, in line with reports from Ireland and France [30, 54]. However while this assumption may be reasonable there may be differential pass-on rates for different sectors which may impact differently on different groups [32, 55]. These effects have not been captured by our work. The tax we modelled was an ad valorem tax, as that was the proposed tax by the Irish minister, and also the typical form of SSB tax introduced [34, 56]. However a fixed price rise per unit volume tax has been proposed as a better health tax. This removes an incentive to shift to bulk buying and ensures comparable absolute price increases on cheaper products which may deter people from shifting to cheaper products to maintain consumption . The focus of this paper is the impact on obesity, other health outcomes (e.g. dental caries, type 2 diabetes) have not been considered. Nor have we considered the effect on non-health outcomes, such as disposable income. Like other indirect taxes, this tax is likely to be regressive [34, 49, 57].
This study suggests that a tax on SSBs in Ireland would have a small but meaningful effect on obesity. While such a tax would be perceived as affecting the whole population, from a health perspective the tax will predominantly affect younger adults who are the main purchasers and consumers of SSBs (assuming similar price elasticity values across age groups). Although improving the health of young people offers the potential for life-long benefits, immediate health savings from reductions in obesity may not be great, as the complications of obesity like cardiovascular disease and cancer tend to occur in middle to older age.
While the average change in calories may appear small, an average excess energy intake of 100 kcal to 200 kcal per person per day may be sufficient to explain the obesity epidemic [43, 59]. Furthermore, a recent meta-analysis has shown that increased sugar consumption is associated with weight gain in adults . An SSB tax is one means of contributing to reversing this excess. It is an example of the prevention paradox: population interventions that aim to improve health have relatively small benefits to the health of most people however the overall positive effect is greater than if only targeting those who drink large quantities of SSBs .
Given that SSB consumption in Ireland is relatively low, it may be prudent to consider taxing a broader range of unhealthy food items. However this would be dependent on identifying an appropriate range of unhealthy food items that could be readily taxed, are clearly associated with harm, and have limited potential for unwanted substitution effects. A tax on SSBs should not be seen as a solution in its own right, but should be part of a broader approach to tacking diet-related diseases.