Canada and the United States may be at the verge of a new trade war, and this one appears to be much nastier than the spat about steel and aluminum in 2018 that eventually led to a new trade agreement.

Free trade has characterized our trade relationship for most of our history. In 1855, when Canada was still a British territory, free trade was introduced through the Reciprocity Treaty, but it was the newly-founded Canada that turned to protectionist policies under John A. Macdonald's National Policy. Macdonald's tariffs ranged between 17.5% and 20% and were aimed at heavy industry and mining. Despite calling for more free trade for decades, it was the Liberal Party under Wilfrid Laurier that raised tariffs to over 30% for some goods. When Laurier returned to the electorate in 1911 with a new free trade agreement in hand, voters rejected him and voted in Robert Borden, a Conservative politician who had campaigned famously on the slogan "No truck or trade with the Yankees" (Norrie & Owram, 1991). Free trade returned in earnest only in the 1950s, leading eventually to the Canada-United States Free Trade Agreement (CUSFTA) in 1988, which expanded into the North American Free Trade Agreement (NAFTA) in 1994. Its successor, the Canada United States Mexico Agreement (CUSMA) came into effect on July 1, 2020. CUSMA is primarily a modernized version of NAFTA with expanded scope to cover intellectual property and digital trade,and stricter rules of origin for automobiles made within CUSMA.

With a looming trade war, Canada finds itself in the unfavourable position of having to design retaliatory actions. The objective is to inflict economic pain on the party that started the trade war without inflicting too much economic pain on our own country. Retaliatory tariffs, while important and effective, inevitably hurt Canadian consumers as well as firms that rely on imported intermediate goods from the United States.

There is one other important tool in our tool box: export surcharges. They essentially work like tariffs in reverse. Exported goods are saddled with a surcharge, and this higher price is passed on to foreign consumers. This, of course, reduces domestic output and may imperil domestic jobs. However, there is a situation when this policy instrument can be quite effective and actually economically beneficial: when the export industry is somewhat (weakly) oligopolistic, but where the export country in its entirety holds stronger market power. Then it is possible to extract monopolistic rents by controlling exports nationally, rather than through individual firms' actions. In essence, oligopolistic competition is turned into an export monopoly. This notion is well understood in the international trade literature, but it is usually not given much coverage as it is rarely used, and of course it is anti-competitive in nature. But in extremis, it is a powerful tool.

Below I explain how it works, with a bit of mathematical economics. If you are not interested in the math, skip forward to the end for some policy conclusions.

⚛ ⚛ ⚛

Consider two countries, home \(H\) and foreign \(F\). Each demands a commodity, but at different levels \(q^H\) and \(q^F\), according to linear demand functions \[\begin{array}{rcl} p & = & a^H-b^H q^H\\ p\cdot(1+\sigma)&=& a^F-b^F q^F \end{array}\] Here, \(p\) is the domestic price, while the foreign price is subject to the export surcharge \(\sigma\). The parameters of the demand function are \(a^H\) and \(b^H\) for the home country, and \(a^F\) and \(b^F\) for the foreign country. Think of the \(b^H\) and \(b^F\) as the market size of the home and foreign country. We need to combine the demand functions, and it will turn out that two new aggregate demand variables will be super helpful. They are \[A\equiv\frac{a^F b^H+a^H b^F}{(1+\sigma)b^H+b^F} \quad\mathrm{and}\quad B\equiv\frac{b^H b^F}{(1+\sigma)b^H+b^F}\] The two demand variables \(A\) and \(B\) are weighted averages of the original parameters. The \(n\) firms in the domestic industry produce \(q=q^H+q^F\) in total. The export surcharge \(\sigma p q^F\) is collected by the government.

Individual producers in the home country operate oligpolistically, and compete in Cournot fashion. They maximize net profits individually \[\pi_i = (1-\tau)\cdot (p - c_i)\cdot q_i\] where each firm faces its unique production cost \(c_i\). Profits are subject to a corporate income tax of rate \(\tau\). It will be useful to define the average cost and standard deviation of the cost of the \(n\) firms in this industry as follows: \[\bar{c}=\frac{1}{n}\sum_{i=1}^n c_i \quad\mathrm{and}\quad \tilde{c}=\frac{1}{n}\sum_{i=1}^n (c_i-\bar{c})^2\] Furthermore, the industry size \(n\) plays an important role in the Cournot model, and therefore the additional two variables \[\nu\equiv\frac{n}{n+1} \quad\mathrm{and}\quad \mu\equiv\frac{1}{n+1}\] come in handy to simplify notation. We are now able to determine the equilibrium through the first-order condition for the profit maximum, and summing the \(n\) first-order conditions to solve for the equilibrium price and quantity. The equilibrium price is \[p=\nu \bar{c}+\mu A\] while the equilibrium quantity is \[q=\nu\cdot\frac{A-\bar{c}}{B}\] The individual firm produces quantity \[q_i=\frac{p-c_i}{B}\] This all looks very much like a conventional Cournot model, with three twists: the presence of two countries, the application of the export surcharge, and the presence of a corporate income tax.

What happens when the export surcharge is introduced? It is obvious that total output will shrink, as monopolistic power can only be exercised through withholding output. What happens to the equilibrium price? The derivative \[\frac{\partial p}{\partial\sigma}= -\frac{b^H(a^F b^H +a^H b^F)}{(n+1)[(1+\sigma)b^H+b^F]^2} < 0\] is clearly negative, which means that increasing the surcharge \(\sigma\) lowers the domestic price. This benefits domestic consumers, as they can now buy output at a lower rate. Domestic consumer surplus increases, at the expense of foreign consumer surplus.

Domestic consumer surplus is the triangle \[\Omega^H=\frac{a^H-p}{2}q^H =\frac{(a^H-\nu\bar{c}-\mu A)^2}{2b^H}\]

It is possible to determine an optimal \(\sigma\) that maximizes domestic welfare, that is domestic consumer surplus \(\Omega^H\), domestic producer surplus \(\Pi\), and government revenue \(\Sigma\). But any export surchage \(\sigma\) below the optimal level will generate the desired retaliatory effect, and the magnitude of this effect can be adjusted as long as it does not exceed the domestic social optimum. But there remains a problem. Raising \(\sigma\) lowers the profits of firms: \[\frac{\partial \pi_i}{\partial\sigma}=-(1-\tau) \frac{(\mu\zeta)^2-(\nu\bar{c}-c_i)^2}{b^F} < 0\] Without compensation, the producers will not be happy. We can also determine the total producer surplus as \[\Pi\equiv\sum_{i=1}^n\pi_i=n\cdot \frac{\mu^2(A-\bar{c}))^2+\tilde{c}^2}{B}\] It can be shown analytically, although it is messy, that raising \(\sigma\) lowers \(\Pi\). Note the presence of the standard deviation of the production cost \(\tilde{c}\) here. If the oligopoly is composed of firms of different cost (and thus size), then overall profits will be higher because there is more concentration and oligopolistic market power. When all firms have the same cost, the standard deviation term vanishes.

The government receives revenue equal to \[\Sigma\equiv \sigma\nu\frac{(A-\bar{c})(\mu A+\nu \bar{c})}{B}\] This expression is not quite proportional to \(\sigma\) because \(A\) and \(B\) are also functions of \(\sigma\). If the government wanted to maximize its revenue, it could raise \(\sigma\) until \(\partial\Sigma/\partial\sigma=0\). The government could keep the revenue \(\Sigma\), but that would make the producers very unhappy as they lose profits. That is the reason why I had introduced the corporate income tax \(\tau\) in the model above. The government needs to use some of the \(\Sigma\) revenue to (fully) compensate the producers. Importantly, if the trade distortion is expected to remain temporary, they can (and should) not lay off workers but instead put them on "short-time work" (wage subsidies) to keep them employed and ready to return to full work hours when the trade dispute ends.

Let me illustrate the model with a few made-up numbers. Assume \(n=10\), \(a^F=a^H=100\), \(b^H=1\), \(b^F=1/9\), \(\bar{c}=25\), and \(\tilde{c}=2\). With these numbers, the foreign market is nine times larger than the domestic market. Then initially, when \(\sigma=0\), the market price is $31.82, output is 681.8, producer surplus is 5048.8 (before tax), and domestic consumer surplus is 2324.4. Total domestic welfare is 7373.2. Now introduce a 25 percent export surcharge. Then the (domestic) market price will drop slightly to $30.15, while output will shrink to 630.7 units. Producer surplus drops to 3737.0, consumer surplus increase to 2439.6, while the government collects 4753.6 in revenue. Total welfare is This means that total welfare is now 10930.3, so a net gain of 3557. Producer surplus has shrunk by about 26%, or 1311. But there is now ample government revenue to compensate the producers. Corporate profits before the export surcharge were \((1-\tau_0)\Pi_0\), and we are interested in holding the producers whole so that \( (1-\tau_0)\Pi_0 =(1-\tau_1)\Pi_1 \). Therefore, the new tax rate needs to be \[\tau_1=1-(1-\tau_0)\Pi_0/\Pi_1\] Let's say the initial corporate income tax was 8%, and we know that \(\Pi_0/\Pi_1=1.35\). Then we need to institute a 24% profit subsidy. If the corporate tax rate was 15%, we would only need a 15% subsidy, And if the corporate tax rate is 28% (Canada's nominal rate after the federal tax abatement), then a tax reduction to 3% would be needed (but not a subsidy).

⚛ ⚛ ⚛

Economic theory reveals that an export surcharge is a viable economic policy that limits domestic harm while inflicting significant economic harm on the trade partner. Normally, that would be a bad thing to do because we would expect the other side to retaliate, if they can. But if they can't pivot to alternative supplies, this can be a hard hit if there is significant import dependence. Under normal circumstances, using such a crude anti-competitive practice would be abhorrent. But in a trade war where the rules of engagement no longer stick to accepted international norms, the aim is to inflict short-term pain that is felt quickly, with the objective to force the foreign country to return to economic sanity—and free trade among the free trade partners. Effective retaliation involves inflicting maximum pain on the trade partner without causing (excessive) harm to one's own economy. Export surcharges in an oligopolistic market are among the least harmful measure that we can find.

The premier of Alberta does not see it that way. She purposefully absented herself from a joint statement of the Canadian premiers and the Prime Minister. Rather than standing shoulder to shoulder and providing united front, there was an empty chair at the joint press conference. Alberta, obviously, holds Canada's mightiest policy instrument: oil exports to the United States Canada exports about 4 million barrels per day (MMB/d), worth $115B in 2024, most of it shipped to PADD 2 (the Midwest). Even though the U.S. also exports oil and on balance is relatively energy independent (US EIA Facts), the refineries in the U.S. are tuned to processing Canadian crude. Pivoting is not so easy in the short term, and that gives Canada some market power.

In normal times, export restrictions often amount to resource nationalism, especially export bans. However, the World Trade Organization (WTO) has limited regulations in this space. There are no restrictions on imposing duties, taxes, or other charges on exports. However, some new WTO members were obliged to remove export charges during their accession negotiations.

A recent Globe and Mail editorial cautioned against using oil as a tool in the looming trade war, suggesting that it could trigger a national unity crisis. The editorial suggested that it would cause too much collateral damage. The editorial is wrong about that, as I have shown above with the help of some basic economic theory. It comes down to fine-tuning the response: combining export surcharges with appropriate compensation for producers. I agree with the editorial writer that an outright export ban would be harmful and nonsensical. However, a well-dosed export surcharge is not—especially if the new US administration exempted energy imports from tariffs. After all, oil buyers deal with volatile international market prices all the time. A well-dosed export surcharge does not trigger an economic doomsday scenario. The editorial's conclusion that national unity requires keeping export surcharges off the table is basically calling for defeatism. Any other province whose industries may be hurting could ask for the same exemption. But there are also other voices heard in The Globe and Mail that point to our resource industry as a powerful instrument for retaliation.

Make no mistake: a trade war will hurt all of us. When it comes to standing united, Alberta cannot claim a special status. Oil, as Canada's largest export good to the United States, is pretty much the only effective tool in our tool box. U.S. demand for Canadian oil is somewhat price-inelastic at least in the short term, giving Canadians market power. Rising gasoline prices would be felt quickly across the United States, putting pressure on U.S. politicians to intervene and end this harmful trade war. Canadians would be foolish to exclude our most forceful instrument from consideration, as a threat at first, and in practice if all else fails. Today is not the time for Alberta to stand aside and play regional politics. As explained, economic harm from an export surcharge can be mitigated effectively. It is time for all Canadians to stand together, across all provinces and across all political parties. Any show of disunity will only give the master of divide-and-conquer politics, who will be inaugurated in the United States today, the upper hand in the looming trade war.

If only economic sanity would prevail, and we could wake up tomorrow and find out there was no trade war, and it was all just a bad dream. After all, in a trade war, there are no winners. But if we must fight a trade war, we must fight it without keeping one arm tied behind our backs.