The Four Percent Lever: The IMF Estimate and Canada's Internal Trade Gap
The lead: four percent of GDP, three decades of inaction, and the lever Ottawa keeps not pulling
The IMF's central estimate is that fully removing interprovincial trade barriers would add roughly four percent to Canadian GDP. The 2019 working paper by Jorge Alvarez, Ivo Krznar, and Trevor Tombe puts the range at three to seven percent, with the central case at four. That four percent figure has been roughly stable across replication studies for years, which matters because the most common rhetorical response to the headline number is that it is exaggerated. The lower bound of the range is still larger than the projected GDP impact of CUSMA. The math is settled. The policy has not moved.
Canada has an internal trade problem that has been documented, debated, and ignored across multiple governments. The 1995 Agreement on Internal Trade was supposed to fix it. The 2017 Canadian Free Trade Agreement was supposed to finish what AIT started. Neither did. The Bank of Canada has cited internal trade as a chronic productivity drag in successive Monetary Policy Reports. Every premier has signed a communique pledging action. The wedge between the consensus economic estimate and the realized policy outcome has stayed roughly the same width for three decades.
The Trump tariff moment is the first external shock in a generation that might change the political calculation. When external trade is constrained, the value of internal trade rises. That logic is sound. It has also been the logic for sixty years. What is different is the magnitude of the external constraint and the political cover it provides for action. Whether Ottawa is willing to use that cover is the open question.
What the productivity gap actually looks like
GDP per capita varies widely across the federation. Alberta sits at roughly $96,500 in 2024 nominal terms, anchored by oil and gas. Saskatchewan is close behind at $90,400, with a similar resource profile. Newfoundland and Labrador, at $77,200, is the resource-driven outlier in the Atlantic group and breaks any clean east-west narrative. British Columbia sits at $75,700. Ontario, despite being the largest provincial economy in absolute terms, comes in at $74,100 per capita. Quebec is at $68,600. Manitoba is at $64,400. The other three Atlantic provinces cluster lower: Nova Scotia $60,300, Prince Edward Island $60,600, New Brunswick $56,300. (Statistics Canada Table 36-10-0222-01)
Some of that range is geography, demography, and resource endowment. None of those are policy choices. The portion that is a policy choice is internal trade.
Statistics Canada reports interprovincial trade in goods and services at roughly $532 billion for 2023, the most recent full year of data. International merchandise trade for the same year was $1.54 trillion, with exports of $768 billion and imports of $770 billion. Including services, total two-way international trade approaches $2 trillion. By any reasonable metric, Canada trades more with the rest of the world than it does with itself. The United States trades far more with itself than internationally. So does Australia. So does the European Union internally. Canadian provinces are unusual on this dimension. They should not be.
What internal trade barriers actually look like
The popular framing of the problem (you cannot ship Quebec wine to Alberta) obscures the more economically significant barriers. The headline barriers are the boring ones.
Mutual recognition of professional credentials is uneven. An Ontario nurse cannot automatically practice in Quebec without re-credentialing. A New Brunswick electrician cannot automatically work in Alberta. Each province maintains its own occupational licensing regime, ostensibly for public safety, in practice often for protection of provincial labour markets.
Procurement preferences tilt provincial government contracts toward in-province bidders. The 2017 CFTA added some procurement liberalization, but with significant exceptions, and most provinces have used those exceptions liberally.
Building codes, electrical codes, and technical standards differ across provinces. A modular home built to Alberta code cannot necessarily be installed in British Columbia without modification. The cost of these differences is largely invisible because it is absorbed before products cross borders. The cost is real.
Trucking operates under different provincial rules on weight, length, hours of service, and licensing. Cross-country logistics adds compliance cost that mostly pays for nothing.
Agricultural marketing operates under provincial supply management for dairy, eggs, and poultry, and through provincial marketing boards for several other commodities. These are legacies of mid-twentieth-century farm policy, and they persist for political reasons that have nothing to do with current efficiency.
Alcohol distribution is the example everyone remembers because it is so absurd. The 2018 Supreme Court decision in R v Comeau effectively narrowed Section 121 of the Constitution to permit provincial alcohol monopolies. The economic cost of provincial alcohol distribution is small in the grand scheme. The symbolism is large.
Why nothing changes
The standard explanation is institutional inertia. The deeper explanation is political economy. Each individual barrier protects a constituency: nurses unions, electricians unions, dairy farmers, provincial procurement officers, regional manufacturers. Each constituency votes. The benefits of removing a single barrier are diffuse, and they accrue mostly to consumers and to out-of-province producers who do not vote in the province imposing the barrier. The political asymmetry has been the binding constraint for decades, and it has not changed on its own.
What could change it is bundling. A premier who tries to remove provincial dairy supply management alone gets crushed by the dairy lobby. A premier who, with cover from a federal-provincial first ministers agreement, removes a basket of barriers in exchange for federal infrastructure spending or tax-point transfers can plausibly survive politically. The 1995 AIT and 2017 CFTA were attempts at this kind of bundle. They underdelivered because the bundle was thin.
The Trump tariff moment offers a thicker bundle. The federal government can frame internal trade reform as part of a national economic resilience package, alongside the $25 billion Canada Strong Fund and the CUSMA July 1 review. Whether the package is genuine or rhetorical depends on the specifics.
What to watch
Three concrete signals would tell you whether internal trade reform is real this time.
Mutual recognition of credentials. If the Council of the Federation announces a binding multi-province mutual recognition framework for major trades and professions, that is the highest-impact action available within current constitutional arrangements. The CFTA has a labour mobility chapter. Binding it with enforcement teeth would change practice on the ground.
Procurement. If federal infrastructure money becomes conditional on provinces opening procurement to all Canadian bidders without provincial preferences, that is real. If procurement language is "best efforts," it is not.
A specific dollar target. If the federal-provincial communique includes a quantified commitment ("aim to add two percent to national GDP by 2030 through internal trade reform"), that is something measurable and falsifiable. If the commitment is "deepen internal trade integration," that is vapour.
The next First Ministers' meeting is scheduled for early summer 2026. The IMF estimate is unchanged. The political constraint is the same one that has held since the 1990s. What is up for grabs is whether the Trump moment proves wide enough to drive a different kind of bundle through it, or whether the federation continues paying a four percent productivity tax on itself out of habit.
The brief
Telecom Q1 prints: ARPU growth versus the legacy drag (BCE IR / Telus IR): BCE and Telus Q1 results land in the first half of May. The watch item is the divergence between wireless ARPU growth and the drag from legacy media and wireline divisions. Both companies carry dividend-yielding capital structures that increasingly depend on asset rotation rather than free cash flow. Whether either signals dividend strain in Q1 commentary is the question. The yields tell you the market is already worried.
Canadian grocer Q1: the supply management headwind no analyst names (Loblaws IR / Metro IR): Loblaws and Metro fiscal Q1 results land in mid-May. The structural pressure no major chain will call out by name is dairy and protein margins under provincial supply management. The dairy lobby is too well-organized for anyone to take the swing. Internal trade reform on dairy alone would be the largest single move available to lower Canadian grocery prices, which is also why it is the least likely to happen before the 2027 federal election.
Federal critical minerals strategy update expected mid-2026 (Natural Resources Canada): A federal critical minerals strategy update is expected in mid-2026. The consequential question is whether export-credit and grant support is conditional on processing in Canada, or whether it is structured as a pure subsidy on extraction. The former changes where value capture sits. The latter funds the same extractive economics that have left Canada exporting raw resources for a generation. The wording determines which one this is.
Q1 labour productivity: the unread Canadian release (Statistics Canada): The Q1 2026 labour productivity release lands in early June. Canadian productivity has trailed the United States consistently since the early 2000s, and internal trade barriers are one of the most-cited causes. Whether the Q1 print extends the trend or breaks it is a useful read on whether tariff anxiety has changed corporate investment decisions. Productivity prints are the single most under-followed Canadian macro release.
TSX outperformance: structural or relative weakness? (Globe and Mail markets): The TSX has outperformed the S&P 500 over the first four months of 2026, anchored by gold, energy, and relatively low valuations. Whether that holds through the CUSMA July 1 review depends on whether tariff-exposed sectors price in further downside or whether the safe-haven trade in materials and staples can absorb the weight. The cross-asset signal that matters is CAD direction. A strengthening loonie would suggest structural confidence in Canada; a weakening one would suggest the TSX outperformance is a function of relative U.S. softness.
By the numbers
4 percent: the IMF's central estimate of GDP gain from removing interprovincial trade barriers. (IMF Working Paper 19/158, "Internal Trade in Canada: Case for Liberalization") For context, that is roughly equal to the projected GDP impact of CUSMA. The biggest deal Canada has not signed is with itself.
3 to 7 percent: the working paper's full estimated range of GDP gain from internal trade liberalization, with the four percent central estimate at the midpoint. (IMF WP 19/158) The range matters because the case against the headline number is usually that it is exaggerated. The lower bound is still larger than what most external trade agreements deliver.
$532 billion vs $1.54 trillion: Canadian interprovincial trade in goods and services in 2023, against international merchandise trade in the same year. (Statistics Canada Internal Trade Hub) Including international services, the gap widens further. For a country of 41 million people across ten provinces, the ratio is unusually low. Australia, Germany, and the United States all show internal trade exceeding international trade. Canadian provinces trade with the rest of the world more than they trade with each other.
Worth reading
- Internal Trade in Canada: Case for Liberalization (IMF Working Paper 19/158, 2019): The four percent estimate, the methodology, and the productivity-growth mechanism. Authors are Jorge Alvarez, Ivo Krznar, and Trevor Tombe. The paper that anchors most subsequent work on the question.
- Canadian Free Trade Agreement (CFTA Secretariat): The text of the 2017 agreement, the schedule of exceptions by province, and the scorecard of progress. The exceptions list is the most useful single document for understanding what has not been liberalized.
- Canadian Internal Trade Data Hub (Statistics Canada): The hub for current interprovincial trade data, including the 2023 and 2024 internal trade releases. The summary numbers cited above come from this hub.
Worth pairing with Issue #2 on the Canada Strong Fund and Issue #1 on the CUSMA review, which between them sketch the federal economic-resilience package this analysis fits inside.