The Ecorys Study

Ecorys recently concluded a study for DG Enterprise on food taxes, “Food taxes and their impact on competitiveness in the agri-food sector”. The study focused on three main questions: whether taxes on fatty/unhealthy foods work, whether they affect the competitiveness of the agri-food industry, and whether they affect employment and trade within Europe.

The study was difficult to research, as the food sector is highly complex, featuring a diverse range of actors, such as consumers, producers, retailers, producer-retailers (which often compete with own-brand products with producers), small businesses and large multinationals.

In the Ecorys study, demand for food products targeted by new taxes fell, but it was unclear to what extent this was caused by those tax changes. Consumers are not solely motivated by price, so these impacts will always be difficult to measure. Denmark, Finland, Hungary and France were studied as examples of countries which introduced new taxes. Ireland and Italy were studied as examples of countries which proposed but finally decided not to introduce new taxes.

Participants familiar with this study called for a broader, more scientific, data-driven study. Such a study would rely less on anecdotal evidence and interviews, and also contain data on the impact of food taxes on low-income groups. The Ecorys study looked into issues of competitiveness more than public health. Future studies should aim to look into both more deeply.

Hungary’s experience

Outside of this recent study, participants also pointed to other evidence – both anecdotal and statistical – which is of relevance to the debate. Hungary introduced a food tax in 2011. Their tax was primarily levied on sugary drinks and foods, some alcoholic drinks, and energy drinks. Rather than focusing on end products, the tax was levied on ingredients of specified end-products. It had two stated aims: a fiscal aim to raise funds for the health budget, and an aim to change consumption and production patterns. It was based on the principle that the producers of unhealthy foods should bear an extra burden for public health costs which their products have helped create.

The Hungarian government, working with the WHO, has been monitoring the impact of this tax, and performed a consumer and producer survey in 2013. According to the study, 30% of Hungarian consumers had changed their consumption patterns due to the tax’s introduction. Also according to the study, approximately 80% of this change was due to the price change, with the remainder due to increased awareness and other factors. The change was observed across all income and age groups, although not evenly. A greater change took place among younger people and lower-income groups.

According to the representative from Hungary, 40% of manufacturers changed their ingredients, of which 30% completely eliminated the taxed, unhealthy ingredients. 90% of manufacturers raised prices on affected goods, and sales of the taxed products decreased by 27%. On a fiscal level, about €60m is raised annually for the health insurance fund, which is roughly in line with the Hungarian government’s expectations. The government has declared that it is broadly satisfied with the tax.

Food industry’s meta-analysis

The food industry representative had performed a meta-analysis of over 40 studies on the topic of food taxes. This analysis confirmed that the demand for food and beverages is extremely complex.  Many studies point out that, as opposed to tobacco and alcohol, food and beverages are essential to life and consumers will hence continue buying them.  This also makes substitution among products highly unpredictable. In France, for example, a soft drinks tax lead to an increase in consumption of salted crisps, possibly as consumers sought other examples of unhealthy “treats”. Consumers may also continue to purchase taxed goods, and subsidise this consumption by purchasing less other products. The meta-analysis also showed that food taxes will often result in down-switching to cheaper brands and product switching to neighbouring untaxed alternatives. As a result, although food taxes may impact the consumption levels of products subject to taxation, this will lead to only a very small overall caloric intake reduction.  It can also be concluded based on the meta-analysis that the overall effects on public health are even more difficult to assess, as caloric intake is just one factor of many to consider.

In Finland, taxation is based on product categories. This means that ice cream is taxed, but other products that have a comparable impact from a health perspective (including frozen or chilled desserts) are not. Here, despite good intentions, consumption of ice cream simply shifted elsewhere, with little appreciable public health benefit. This tax design meant that only 14% of total sugar consumption in Finland was taxed.

In Hungary, also certain salty goods are taxed. In the end, as a result of the tax design and according to industry analysis, only 6% of total salt consumption in Hungary is taxed, so the public health impacts would presumably be limited. Consumers were also likely to switch to lower-cost products within the same category.

Mexico is one of the most recent countries to levy food taxes. There have, in industry’s view, been negative impacts there too. According to studies, demand has not significantly switched, but is to some extent being subsidised by lower demand for other product categories. Consumers are prioritising soft-drinks, snacks and cookies, despite the food taxes, to the detriment of other (taxed and non-taxed categories), including even personal healthcare products.  Additionally, the taxes have shown to be regressive.  Lower-income families normally spend a higher portion of income on food and, additionally, tend to proportionally eat more of taxed categories.  Prior to the tax, 12% of their income was taken up by products to be taxed. Afterwards, 15% of their income was spent on taxed goods.

It was on the basis of these results that food industry representative voiced strong reservations to existing design food taxes, above all because they see them as too narrow a tool to effectively address obesity and related Non-communicable diseases (NCD).  Plus substitution effects can never be precisely predicted and taxes risk generating several unintended consequences.

So although open to any solution that would help address obesity, the food industry representative was convinced that a more holistic, long-term, multi-stakeholder approach will generate greater and more sustainable results.  Reference was here also made to the recently released McKinsey study (quote from study: “only a holistic, broad, and multipronged approach can be successful in reversing the obesity crises”).

Other studies

Research from the non-profit sector by WWF has been undertaken, examining which factors drive consumer choice of food. In this study, price was the most important factor in consumer choice. This report also argued that food taxes could “fiscally rebalance” the health costs of unhealthy foods to the food producers, in line with the “polluter pays” principle. Some participants in the meeting agreed with this sentiment, and expressed frustration that industry receives large amounts of money from consumers, yet consumers must pay for solving the resulting public health effects. The example of the tobacco industry was given; in particular the comparison between the industry’s large profits and the high costs borne by cancer victims and citizens who pay for public health systems.

The WHO has found that taxes on fatty foods in Denmark did lead to a decrease in consumption, but there were problems with substitution for other fats. In addition to this, some consumers purchased fatty foods across the border in order to avoid the tax.

More broadly, it was argued that – while the markets are different – the impact of taxes on alcohol and cigarettes was reasonably clear. Such taxes have been effective, and could be when aimed at unhealthy foods.

Oversimplification

A common issue which came up in interviews with industry stakeholders in the Ecorys study was that certain foods were labelled “unhealthy”, even though they may be healthy as part of a broader healthy diet or lifestyle. This concern was echoed by some participants in the discussion, who cited the example of the UK’s traffic light system as “simplistic” and something which could mislead consumers.

To highlight this, a speaker gave the example of olive oil, which was labelled as unhealthy in the UK, despite being part of a “healthy Mediterranean diet”.  Some participants disagreed with this analysis, however, saying that the Mediterranean diet would still be healthy with less oil, and its presence in a generally healthy diet is not an endorsement of oils.

While all speakers accepted that food taxation was more complex than other taxed products, such as alcohol or tobacco, some took issue with the idea that there is no clear definition as an unhealthy product, citing the example of a soft drink, which would typically contain more sugar than would be healthy to consume almost regardless of the surrounding diet.

Importance of other measures

While participants disagreed about the necessity or efficacy of food taxes, there was broad agreement that any measures to combat unhealthy diets should not be taken in isolation, but should instead be part of a larger package, which could include education, reformulation, portion-guidance, marketing guidelines, regulation and so on. Broadly speaking, the focus should be on making healthy choices easier for citizens. There was also some concern that the state should not “tell citizens what to do”, although the burden for maintaining healthy diets should not fall entirely on the end consumer.

Real-world examples

One speaker gave the example of Italy, where an unpopular soft drink tax was defeated because it was seen as a new tax, levied in isolation as a fundraising mechanism. To overcome this, governments should try to disentangle fiscal and public health objectives, and not treat food taxes merely as a revenue-raising measure.  Governments could, for example, lower taxes for some goods, to have a net zero balance tax take from a new tax.

Various measures aside from taxation were suggested in the discussion. One such measure suggested was that manufacturers change the ingredients in their products. This, however, might clash with consumer expectations of taste or quality. For example, some producers had attempted to replace palm oil ingredients. This change had some negative effects on taste, and was rejected by consumers. There was some disagreement on this point, however. Producers have, in the past, always argued that they cannot change the makeup of their products, a concern which can be overcome. Consumption patterns have shifted as a result of marketing, and that can be done again. A speaker gave the example of four-wheel drive cars, for which little market existed previously. Demand for these goods is now quite high, as a result of successful marketing.

Changing marketing by the food industry was cited as a clear example of a non-taxation measure which could have an impact on food consumption. One speaker said that, despite voluntary agreements against such practices, marketing of unhealthy foods to children is still rampant. The scale and sophistication of such marketing, it was argued, will simply always win when pitted against public health education projects. Various speakers pointed to the need for regulation in this area.

As another example of measures, Cyprus banned the sale of unhealthy foods in schools. The ban initially proved ineffective, as students simply brought unhealthy foods to school from home. Increased food education and public awareness reduced this problem. This example was taken as evidence of the necessity of a combination of measures.

This point was echoed by other initiatives which had similar problems with banning certain goods from schools. They found that small changes about the presentation and marketing of foods in the school, while still leaving unhealthy foods as possible options, were more effective.

Other measures worth examining include labelling and portion sizes, which have been a major issue in the United States.

Conclusions

The discussion highlighted the complexity of assessing the impacts of food taxes. Even though participants disagreed about the necessity and efficacy of food taxes, there was a common stance to avoid oversimplification and to take a broader approach when trying to bring about health gains. Education and public awareness, reformulation, portion guidance, marketing guidelines and adapted regulation to make it easier for consumers to choose healthier products are options which could be further explored in order to achieve this aim. The chair concluded that more research on certain elements such as consumer behaviour, product substitution as well as on the impact on low income groups and public health was needed.

Speakers: