Geopolitics & international relations

Can Canada Really Break Free from Washington?

By Abderrazzak Jirari

· 10 min read

The Gordie Howe International Bridge under construction over the Detroit River, between Windsor, Ontario, and Detroit, Michigan
Photo: AHeneen

On January 20, 2026, in Davos, Mark Carney delivered the most widely discussed Canadian foreign policy speech in decades. The Prime Minister argued that the rules-based international order was going through "a rupture, not a transition," that great powers were now using economic integration as a weapon, and that middle powers had to act together or risk finding themselves on the menu rather than at the table (Carney, 2026). Borrowing the image made famous by Václav Havel, that of the greengrocer who displays in his shop window a slogan he no longer believes, he announced that Canada was taking the sign down.

Ten months later, the phrase has become a doctrine. It has acquired a vocabulary ("strategic autonomy," "values-based realism," "variable geometry"), a reference document in Ottawa's first national security strategy since 2004, and its first achievements, including the Alliance for the Future concluded with the European Union in Montreal on October 30. The question this article asks is easy to state and hard to settle: can a country that still ships nearly three quarters of its goods exports to the United States genuinely achieve strategic autonomy, or is the middle-power doctrine primarily a posture? International relations theory offers precise tools to answer it, and its answer is more nuanced than either the doctrine's champions or its skeptics suggest.

What "Middle Power" Actually Means

The idea of the middle power has been central to Canadian diplomatic thinking since 1945, to the point of becoming part of the national identity (Nossal, Roussel and Paquin, 2007). It nonetheless remains conceptually fragile. Adam Chapnick (1999) showed that the term covers at least three distinct models: a hierarchical model, which ranks states by their material capabilities; a functional model, which grants a state influence in the areas where it holds particular capabilities; and a behavioural model, which defines a middle power by the way it acts rather than by its size. Chapnick also established that the story of a Pearsonian golden age, in which Canada supposedly shaped the United Nations as a middle power, owed a great deal to national myth, since the policymakers of the time were far more cautious than the official history suggests (Chapnick, 2005).

The behavioural model has dominated the study of Canadian foreign policy. For Cooper, Higgott and Nossal (1993), middle powers can be recognized by their preference for multilateral solutions, their inclination toward compromise and their commitment to "good international citizenship." Writing as the Cold War ended, these authors already stressed that such behaviour is not automatic: it depends on a state's capacity to exercise intellectual leadership and to build coalitions around specific issues.

The most useful definition for reading the Carney doctrine, however, remains Robert Keohane's (1969). In a seminal article, he proposed defining states not by their resources but by how their leaders perceive their role. A middle power is a state whose leaders consider that it cannot act effectively alone, but that it can have an impact on the international system within a small group or through an international institution. The Davos speech is an almost perfect illustration of this definition. It does not claim that Canada can stand up to Washington on its own; on the contrary, it states that it cannot, and that it must therefore act with others.

What sets the 2026 version apart from the Pearsonian tradition is its relationship to institutions. The classic middle power relied on the United Nations, the GATT and NATO, that is, on permanent institutions underwritten by American hegemony. Carney instead proposes different coalitions for different issues and, borrowing the phrase of Finnish President Alexander Stubb, embraces "values-based realism." The change is one of kind: the middle power is no longer an auxiliary of a liberal order guaranteed by the United States; it is seeking protection against that very guarantor.

From Words to Deeds

At the level of declarations, the consistency is striking. Speaking at the United Nations General Assembly on September 26, Anita Anand restated the diagnosis of rupture and set the objective: greater strategic autonomy, pursued through issue-based coalitions of small states, middle powers and great powers (Anand, 2026). On October 29, The Plan for a Stronger Canada gave this orientation official form. The strategy deliberately begins with economic security, organizes government action around eight strategic capabilities (emerging technologies, critical minerals, defence, energy, transportation, pharmaceuticals and biotechnology, food security, and payments, settlement and investment) and explicitly commits to variable geometry coalitions (Government of Canada, 2026). Putting the economy at the heart of a national security document is no small matter. It is an acknowledgment that the main threat to Canada does not take the form of an invasion, but of commercial coercion.

Deeds have followed. In February 2026, Canada became the first non-European country to join SAFE, the European Union's defence financing instrument. The Montreal summit of October 29 and 30 launched an Alliance for the Future designed to create a common space of prosperity and economic security, with components on critical minerals, digital infrastructure and artificial intelligence, the Arctic and support for Ukraine (Canada and European Union, 2026). At the same time, Ottawa has re-engaged with Beijing: Carney's January visit produced a new strategic partnership, and as these lines are written, the Prime Minister is on a tour of Ireland, the Philippines, Cambodia and China.

Trade figures also appear to vindicate the strategy. According to Global Affairs Canada (2026), exports to markets other than the United States grew by 11.1% in 2025, raising their share to its highest level since 1981. In July 2026, merchandise exports to those markets reached a record $25.6 billion, or 33.7% of the total (Statistics Canada, 2026).

The Constraint Theory Predicted

These figures nevertheless call for caution, for two reasons. The first is composition: a significant share of the increase reflects exports of non-monetary gold, notably to the United Kingdom, whose value tracks the price of the metal rather than any lasting reorientation of the productive economy. The second is scale: in 2025, the United States still absorbed 72% of Canadian goods exports (Global Affairs Canada, 2026). Diversification is advancing, but from a very low base.

More than eighty years ago, Albert O. Hirschman (1945) offered the framework for understanding what this asymmetry means politically. In National Power and the Structure of Foreign Trade, he showed that foreign trade produces an "influence effect": the more a country gains from its trade with a dominant partner, and the more costly it would be to find alternative outlets, the more leverage that partner holds over it. Hirschman drew a consequence that is often forgotten: the dominant power has an interest in preventing its partner from diversifying its trade, since its power rests on the scarcity of alternatives. Keohane and Nye (1977) extended this insight by distinguishing a state's sensitivity to external shocks from its vulnerability, that is, the cost it would have to bear to adjust if the relationship were broken.

The events of summer 2026 read almost like a demonstration of Hirschman's argument. On July 1, at the first joint review provided for in the Canada–United States–Mexico Agreement (CUSMA), Washington declined to renew the agreement, placing it under a regime of annual reviews until 2036 (USTR, 2026). The bilateral negotiations that followed collapsed on August 21, and 50% tariffs hit roughly US$20 billion worth of Canadian goods (NPR, 2026a). According to the Prime Minister, one of the American demands that sank the deal sought to restrict Canada's ability to conclude trade agreements with other countries (Khaosod English, 2026). In other words, Washington sought to obtain, in exchange for tariff relief, precisely what underpins Canadian autonomy: the ability to build alternatives. This demand reveals, better than any speech, that diversification is not a symbolic gesture. The dominant power sees it as a threat to its leverage.

Farrell and Newman (2019) add a dimension that the Canadian doctrine takes seriously. Their theory of "weaponized interdependence" shows that global economic networks (finance, information, supply chains) are organized around a few central nodes, and that states with jurisdiction over those nodes can use them to monitor their partners or to cut off their access. For Canada, vulnerability therefore extends beyond the destination of its exports. It also involves payments, cloud computing, semiconductors and artificial intelligence models. Seen from this angle, it becomes clearer why the national security strategy lists payments, settlement and investment among its eight strategic capabilities, and why the Montreal statement refers to "trusted and sovereign" cloud and compute infrastructure.

Finally, Stephen Clarkson (2002) reminded us that North American integration is not merely a flow of goods. It is a legal architecture that has gradually narrowed the room for manoeuvre of the Canadian state. A strategy of autonomy therefore cannot be limited to finding other customers. It requires rebuilding public and industrial capabilities that three decades of continental integration had made seemingly unnecessary.

The Limits of Coalitions

If asymmetry with Washington is the structural constraint, coalition logic has fragilities of its own, which Keohane (1969) also identified. Collective action among states that can do nothing alone runs into the classic free-rider problem: each has an interest in the coalition existing, but none has an interest in bearing its cost alone in the face of a great power. The Montreal summit illustrated this. The idea of making Canada an "associate member" of the European Union, floated by the President of the Commission, was dropped from the final statement at the request of Germany and a few other member states, which argued that the concept does not exist in the treaties (Al Jazeera, 2026). Since Article 49 of the Treaty on European Union reserves membership for European states, the partnership advances through joint programs rather than integration. That is a great deal, but it is not an alliance in the sense that NATO is one.

The rapprochement with China raises a different kind of difficulty. Variable geometry allows cooperation with partners that do not share Canadian values, and values-based realism openly accepts this. The literature on interdependence nonetheless calls for consistency: replacing one asymmetric dependence with another produces no autonomy. The provisional anti-dumping duties Beijing imposed on Canadian canola in 2025, a year after Canada's 100% surtax on Chinese electric vehicles, are a reminder that trade weapons are not Washington's exclusive preserve. Strategic autonomy requires multiplying partners, not changing patrons.

Finally, American politics must be taken into account. The November 3 midterm elections gave Democrats control of both chambers of Congress (NPR, 2026b). A week later, Anita Anand said that this change in political mood had opened the way to détente and that Canada would seek to ease trade frictions with Washington (The Edge Malaysia, 2026). Some will see this as a return to normal. That would be a misreading of the doctrine. Within Hirschman's framework, détente and diversification are not opposites: a Canada with genuine alternatives negotiates from a stronger position with its neighbour, whoever occupies the White House.

Measuring Autonomy

Canada's middle-power doctrine is neither an illusion nor a revolution. It rests on a sound diagnosis, consistent with what international relations theory has taught since Hirschman: in an asymmetric relationship, power belongs to whoever holds the best alternatives. It nevertheless runs into an equally well-documented reality. Those alternatives take a decade to build, not a single mandate, and coalitions among middle powers remain fragile as soon as their costs have to be shared.

Strategic autonomy will therefore not be measured by the share of exports going to the United States, still less by a few months of gold shipments to London. It will be measured by the cost of refusal: what it costs Canada to say no to an American demand, as it did in August. Three indicators will make it possible to judge the doctrine in the months ahead: the share of non-U.S. exports once gold is excluded, the outcome of the next annual CUSMA review in July 2027, and the conversion of the Montreal and SAFE commitments into contracts, investments and real industrial capabilities. If all three improve, the Davos "rupture" will have marked the starting point of a genuine strategy. If not, it will remain a great speech.

(The author)

Portrait of Abderrazzak Jirari

Abderrazzak Jirari

Strategy, Digital and AI Consultant

Trained in political science (from an honours bachelor's to a master's at Université de Montréal), I'm now completing a graduate diploma (DESS) in private- and public-sector management at HEC Montréal.