Access to sufficient, nutritious food is a basic human right and is necessary to achieving the United Nations’ Sustainable Development Goals. We demonstrate that international food trade, in the current global system, is essential to nutrient access and enables some poorer countries to be able to nourish up to hundreds of millions of people. Protectionist trade policies could therefore have serious negative consequences for food security.
Because human health requires a balanced diet of macro- and micronutrients, there is growing consensus that optimizing food systems for nutrients could more effectively address undernutrition than increasing total food production1,2,3. Yet there is debate about the relative importance of different drivers of nutrient availability, such as increasing production4, 5 versus improving access and distribution5, 6. Here we quantify the relative contribution of international trade, waste and conversion to non-food uses to explain patterns in the nutritional potential of the global food system, for crop and animal commodities. We also estimate the potential of the global food system to meet current and future nutritional needs.
We show that current global food supply—the amount of food available for consumption after accounting for trade, losses and conversion to other uses—is sufficient to meet average nutrient demands for the aggregate global population, if equitably distributed (Fig. 1a; Supplementary Table 2). Supply of vitamin B12 is sufficient to meet the needs of 16.79 billion extra people and protein is sufficient to meet the needs of 11 billion extra people (Fig. 1a). By contrast, folate supply is barely sufficient to meet current needs (70 million extra people; Fig. 1a) and is the only nutrient for which losses from production are greater than food supply (Supplementary Table 2). Marginal gains or losses in folate supply, therefore, could strongly impact global adequacy of supply. Global food production—the amount of food produced before accounting for trade, waste and conversion to other uses—is sufficient to meet the needs of an even larger number of people, ranging from between five times (protein) and two times (calcium) the current global population (Supplementary Table 2). The gap between nutrient supply and production suggests that gains in crop yields are likely to be insufficient to meet nutritional needs without policies ensuring that gains in nutrient production translate to gains in nutrient supply.
Countries differ in ability to meet nutrient needs with current food supply (Fig. 2a, Supplementary Fig. 1; Supplementary Table 3). For protein, all countries have enough protein to meet current nutrient requirements. For vitamins A and B12, countries range from around 8-times and 12-times their population’s needs, respectively, down to less than 1. For minerals, the lowest 20 countries (Supplementary Figure 1) are all deficient, whereas the highest 20 countries (Supplementary Fig. 1) are sufficient. Thus, although certain macronutrients (for example, protein) are sufficient for all countries, many countries are not able to meet nutrient needs for micronutrients.
There are several potential mechanisms that determine aggregate nutrient supply and equitability among countries. Globally, waste plus conversion to non-food uses is associated with a loss of nutrients equivalent to the needs of between 10% (calcium) and 32% (protein) of the global population, depending on the nutrient (Supplementary Table 2). Among countries, trade accounts for a greater amount of total food supply, in most cases, than does waste, an oft-cited driver of inefficient food distribution, and conversion to seed (Supplementary Table 3). Under our no-trade scenario, nutrient distribution among countries was always less equal (Fig. 1b; Supplementary Table 4). The disparity of nutritional needs potentially met among countries was much higher under no-trade scenarios than with actual food supplies (Supplementary Figs. 2,3). Lower-income countries tended to gain nutrient potential with trade, for almost all nutrients except iron and folate (Supplementary Figs. 2,3; Supplementary Table 5). Higher-income countries were more variable than low-income countries in which nutrients increased with trade (Supplementary Figs. 2,3; Supplementary Table 5). Based on our trade versus no-trade comparison, trade contributes to between 146 and 934 million people being potentially nourished for certain countries, depending on the nutrient (Fig. 2).
Our results highlight the need for agricultural and food systems to explicitly consider nutrients, not just aggregate production. Given that all countries have an adequate aggregate supply of protein, but not of many micronutrients, emphasis should be placed on increasing availability of key micronutrients, either through intensification of target crops rich in those nutrients or trade for target food items, and ensuring adequate distribution of these nutrients at the sub-national scale. Our findings also demonstrate that current global food trade plays an important role in distributing nutrients among countries and improves the ability of many countries, especially poorer countries, to meet their nutrient needs. On the other hand, some countries had lower nutrient potential with trade—such as iron in low-income countries, which is associated with major malnutrition challenges. On average, protectionist trade policies that limit the flow of food items would likely have serious negative consequences for global nutritional security.
There are also challenges and limitations with current trade practice. First, the foods that are most easily traded—such as staple grains—are often lowest in composition of micronutrients in which many countries are deficient. Relatedly, the nutrient composition and quality of food items could differ when items are consumed fresh versus processed or harvested early for an international supply chain7. Second, food items that arrive in a country by trade are preferentially consumed by individuals with revenue to purchase imported items. Thus, there is no guarantee that people with greatest need for imported nutrients will have access; instead, the distribution of food items—and associated nutrients—would depend on ability to pay. Relatedly, trade of food products could impact domestic producers differently from wage-earning consumers. Consumers who purchase goods benefit from lower-cost imports, whereas domestic producers may not. Third, increased specialization associated with trade can increase nutrient supply, but could also lead to environmental damage if specialization is focused on environmentally harmful crops, such as large-scale oil palm plantations3. Both nutrition and environmental outcomes are central to the Sustainable Development Goals. Finally, trade can increase nutrient supply but can also make countries vulnerable to sudden changes in global trade patterns8, 9.
Our results highlight the importance of trade in the current global food system to countries’ abilities to meet their nutritional needs. Our analysis, however, is limited to the current global food system and does not take into account changes in food security that occurred with historical changes in trade regimes. Rather than implying that free trade is always better, our findings suggest that trade in the current global food system is associated with greater equality of nutrient access. Put briefly, the ability of many countries—especially low-income countries—to meet their aggregate nutritional needs in today’s world would be less without trade.
To quantify nutrient gaps in the global food system, we merged Food and Agriculture Organization food balance data10, food item nutrient composition and country-specific nutrient requirement estimates. Food balance data include production, conversion to seed, conversion to feed, waste, trade balance, and other uses. Food supply is the amount of food available for consumption, defined as production plus imports and changes in stocks minus exports, conversion to seed, conversion to feed, waste, and other use. ‘Other use’ includes non-food products, such as oil for soap production. We analyse five-year average values (2007–2011).
To determine nutrients in each food balance sheet component, we combined food balance data with USDA Food Composition Database nutrient data on each food item, excluding beverages, spices and some others (Supplementary Table 1). We multiplied the amount of each food item in each component of the food balance sheet by a refuse fraction that captures the non-edible amount of each food item11 and multiplied the resulting edible quantity by the concentration of nutrients in each item. We focused on eight nutrients: protein, energy, zinc, calcium, iron, vitamin B12, folate and vitamin A.
To estimate the potential to nourish each country’s population, we multiplied nutrient amounts as described above by country-specific dietary targets. Targets are a weighted average of individual dietary requirements based on the distribution of a country’s population into different age classes—each of which is associated with different dietary requirements11. FAO nutrient requirements were used because they have been derived specifically for global application12. We merged food balance sheet, nutrient concentration and dietary target data to calculate the fraction of a country’s population’s nutrient needs that could be met. These estimates assume equitable distribution within countries, which is not achieved in reality, but allows us to identify if nutrient supply is insufficient under any circumstance. We measured the equality of nutrient potential among countries with the Gini coefficient, which is a common measure of income inequality and ranges from 0 (perfectly equal) to 1 (perfectly unequal).
We created a hypothetical no-trade scenario to estimate the impact of the current international trade system on nutrient supply. We subtracted trade balance (imports – exports) from food supply. Thus, we started with the number of people whose nutrient requirements could be met for a particular nutrient. We subtracted the quantity of nutrients imported (in terms of number of people potentially nourished) minus amount exported to other countries. This is not a realistic scenario as a world without trade would look quite different from the current world. As such, this scenario cannot evaluate the complex nature of the effect of trade on food security.
Food balance sheet data have limitations. Data quality differs by country, often with lower-income countries having low-quality data. In some cases, national agricultural statistics may not fully capture the production of smallholder farmers13. Production statistics are generally more reliable than harder-to-measure outcomes like food waste and post-harvest losses. FAO data also are poor in capturing processed foods and are most relevant for commodities. Still, FAO data are the best available data on national level food production, exchange and loss.
Statistical code is available through GitHub (https://github.com/swood-ecology/nutrient_trade).
Data are available from the corresponding author upon request also through the Knowledge for Biocomplexity repository (https://knb.ecoinformatics.org/#view/doi:10.5063/F1542KR1)14.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors would like to acknowledge M. Bradford, D. Kane, S. Kuebbing, E. Oldfield and C. Palm for helpful comments. S.A.W. was supported by a NatureNet Science Fellowship.
Supplementary Figs. 1–3, Supplementary Tables 1–5