Madam Chair and members of the committee, thank you for the invitation to contribute to your work on an important issue for Canada: Canada's trade with North and West Africa.
My name is Thierry Warin. I'm a data science professor for international affairs at HEC Montréal and a fellow at CIRANO.
Canada is an open economy, but it is also an economy that operates in close proximity to the largest market in the world, the United States. Talking about trade with other regions today also means talking about trade with the United States.
If you draw on theoretical frameworks, you might be tempted to pit two classic explanatory models against each other. On the one hand, what is known as the Heckscher-Ohlin-Samuelson model—I’m sorry, I am a professor—which emphasizes factor endowments and specialization. On the other, the gravity model, which explains trade flows by the size of economies and, above all, by distance.
In Canada's case, it is quite clear that the gravity model is empirically dominant, but this dominance does not imply the optimal allocation of trade flows. It also reflects frictions, notably information asymmetries.
This is where we need to take a step back and consider data science. Countries do not trade; companies do. Decisions to export or import are not taken at the aggregate level, but at the company level, subject to very concrete constraints: imperfect information, prospecting costs, uncertainty regarding partners and difficulty in identifying the right inputs and the right markets.
In this context, the geographical focus on the United States is partly the result of rational trade-offs in the face of uncertainty. However, it may also mask untapped opportunities.
The real question is no longer simply who to trade with, but which products, which inputs and within which value chains. It is from this perspective that diversification towards Africa must be understood. Diversification does not simply consist of adding partners; it consists of precisely identifying the final products or inputs for which there is genuine complementarity. When we examine the figures now, four key elements emerge.
Firstly, in terms of volume, trade between Canada and Africa remains modest, and its dynamics vary across sub-regions. With North Africa, total trade rose from around $3.8 billion in 2015 to $4.7 billion in 2024. In West Africa, the dynamic is different. Total trade rose significantly from $2.5 billion to just over $5.8 billion in 2024. That growth is entirely driven by Canadian imports.
Secondly, geographical concentration is particularly pronounced. In North Africa, three countries—Algeria, Morocco and Egypt—account for the bulk of Canadian exports, with combined shares exceeding 90% depending on the year. In West Africa, the concentration is even greater on the import side. In 2024, Nigeria, Mauritania and Côte d’Ivoire accounted for over 90% of Canadian imports from the region. We are therefore dealing with trade structures dominated by a very limited number of partners, which potentially creates significant dependencies.
Thirdly, concentration by product is extremely high for Canadian imports from West Africa. In 2024, the top three product categories accounted for over 93% of these imports, dominated by mineral fuels, precious metals and cocoa. Conversely, Canadian exports are slightly more diversified, but remain concentrated. In North Africa, the top three products still account for around 63% of Canadian exports and the top five for nearly 78%. We are therefore dealing with situations where a few products currently account for the bulk of trade flows.
Fourthly, in terms of specialization, Canada mainly exports plant products, particularly grains, as well as inputs such as fertilizers and, in some cases, equipment. Conversely, Canadian imports are dominated by primary products: energy, minerals, specific agricultural products or inputs. This asymmetry is robust, even if its precise form varies by country. It reflects a vertical organization of trade, rather than intra-sectoral trade.
Finally, if these elements are combined, several trade patterns can be identified. An agri-food corridor with North Africa, which is relatively diversified and based on agricultural complementarities; an energy corridor with West Africa, heavily concentrated around Nigeria and characterized by a dependence on fuels; and industrial input corridors, notably with Morocco or certain West African partners.
What this essentially means is that diversification towards Africa cannot be considered solely in geographical terms. It must be considered in terms of flow structures, products and value chains. This is precisely where data science becomes crucial. It enables us to identify, at a granular level, the relevant complementarities—those that are not visible when viewed at an aggregate level.
Thank you for your attention. I will be pleased to answer your questions.