Food processing and global trade flows: What are the trends?

By Eriya Chea, Siramane Coulibaly, Mateo Ferrero, Chen Lu, Thomas Verbeet and Erik Wijkström

Economies around the world are increasingly adopting public health policies aimed at improving nutrition and reducing the onset of diseases such as heart disease, stroke and diabetes. In parallel, public health agencies and researchers are focusing more and more on whether the processing of foods (through the use of additives, preservatives and flavourings) negatively affects health. So, what are the trends for trade in various categories of food and what tariff levels do economies apply?

This blog examines trade flows and tariff structures to explore possible linkages between trade policies and types of food, by processing category, traded across borders. The aim is to highlight patterns in the data only. It does not imply that trade policy is an appropriate tool for improving health outcomes, nor that food processing itself necessarily poses health risks.

Tariffs and food processing categories

Using the NOVA classification, which categorizes foods by the degree of processing they have undergone, we have mapped tariffs to the level of food processing to examine possible trade and nutrition linkages.

The NOVA food classification system categorizes foods into four groups according to their degree of processing, ranging from unprocessed to ultra-processed foods. It has been widely applied in scientific literature, although critics argue that it is too broad and inconsistently applied.

The four NOVA categories may be grouped into three broad clusters:

• Unprocessed food (NOVA Group 1): This category includes unprocessed or minimally processed food products, accounting for about 60 per cent of all food products under consideration (e.g., fresh fruits and vegetables, unprocessed wheat, oats and other cereals, and lentils, chickpeas and other legumes).
•  Processed foods (NOVA Groups 2 and 3): This category simplifies the NOVA classification to include both culinary ingredients (e.g., oils, fat, sugar and salt) and moderately processed items (e.g., canned vegetables, cheese and bread). They represent about 32 per cent of the food products under consideration.
• Ultra-processed foods (NOVA Group 4): About 8 per cent of the food products under consideration (e.g., ice cream, sugary drinks, packaged snacks and instant noodles).

This classification was applied to a total of 797 tariff lines at the six-digit level under the Harmonized System 2022 used to classify traded goods, thereby covering most food products that are traded. The full listing is available here.

Trade flows and tariffs by degree of food processing 

From 1996 to 2024, global imports of unprocessed foods grew more rapidly than those of processed or ultra-processed foods, both in terms of total value and on a per capita basis. 

The increase in total value of unprocessed foods was most pronounced in Asia, reflecting the region's scale and population growth, while per capita imports rose most strongly in North America, Europe, and Oceania. In these three regions, per capita imports of processed and ultra-processed foods also rose, but not as sharply. 

For Africa, Asia, and South and Central America and the Caribbean, the trend in per capita imports of unprocessed foods was more stable and, in the last decade, even showed signs of declining. 

Overall, the data suggest that global trade patterns on a per capita basis continue to be shaped by demand for primary unprocessed or minimally processed food products, rather than for processed or ultra-processed food products.

Mapping international data on tariffs to the degree of food processing shows that average most-favoured-nation (MFN) tariffs tend to be lower for unprocessed foods than for processed or ultra-processed foods. However, this is not universally the case. Some economies apply lower tariffs to ultra-processed foods, which may potentially incentivize trade in these foods. Meanwhile, other economies have retained higher tariffs on such products. 

Most economies in the top 20 importers of food impose relatively low tariffs across the three food product categories although differences remain. For example, some economies impose significantly higher tariffs on ultra-processed food products vis-à-vis unprocessed food products, whereas other economies impose significantly higher tariffs on unprocessed food products than on ultra-processed ones.

Key takeaways 

Mapping tariffs onto the NOVA framework provides a useful way of examining how trade policies and nutrition may intersect. It shows that, although generally lower MFN tariffs tend to be applied to unprocessed foods, there is not a single global pattern. Tariff structures are diverse and complex. 

At the same time, the trade data show strong growth in trade in food products overall, with imports of unprocessed foods expanding much faster than those of processed or ultra-processed products, both in total value and in per capita terms. The findings may be  relevant for policymakers and researchers exploring how trade measures intersect with public health objectives and policies. 

Decisions on tariff levels could have indirect implications for diet quality and public health outcomes. As discussions on trade, food systems and public health continue domestically and internationally, a better understanding of how trade measures relate to different types of foods traded can lead to coherent policymaking across these domains.

This blog focuses on tariffs but non-tariff measures, such as mandatory nutrition labels, advertising restrictions or standards that require food manufacturers to alter the composition of their products to improve nutritional quality, also play an important role in shaping food consumption patterns. An increasing number of food-related regulations are being notified to the WTO under the Agreement on Technical Barriers to Trade (TBT), with some being raised as "specific trade concerns" in the TBT Committee.