Before the Last Shift Ends

For nearly twenty-five years, I have lived in Northern Ontario. During much of that time, I watched one of the province’s foundational industries contract around us.

The decline of forestry, and particularly of pulp and paper, was not an abstract economic trend. Mills that had shaped communities for generations curtailed production, shed shifts, stopped investing, closed machines and eventually disappeared. Contractors lost work. Suppliers lost customers. Skilled workers left or accepted jobs that paid considerably less.

Some communities adapted remarkably well. Others never fully recovered.

Today, watching the escalating trade conflict between Canada and the United States and the extraordinary pressure being placed on Ontario’s automobile industry, I find myself wondering whether there is anything useful in that experience for Southern Ontario.

It is an uncomfortable comparison, and it is also one that should not be made casually. A pulp mill is not an automobile assembly plant. Newsprint is not a pickup truck. The forces that devastated Ontario forestry twenty years ago are not identical to those confronting auto manufacturing today.

Most importantly, parts of the paper industry were facing a genuine destruction of their underlying market as digital media displaced printed newspapers. North Americans, by contrast, are not about to stop buying automobiles.

So this is not an argument that Ontario auto manufacturing is destined to become Northern Ontario forestry. Nor is it an argument that a trade war destroyed forestry and will therefore destroy auto. That would be historically wrong. Forestry was hit by a combination of forces: a rapidly appreciating Canadian dollar, high electricity and fibre costs, international competition, aging facilities, weak capital investment, the softwood lumber dispute, declining newsprint demand and, eventually, the collapse of the United States housing market and the global financial crisis.

The similarity lies somewhere deeper. It lies in the way a large industrial ecosystem can weaken gradually, and then suddenly. It lies in the difference between saving a factory and preserving the economic system that makes the factory worth operating. And it lies in the uncomfortable fact that by the time a permanent closure is announced, much of the important economic damage may already have happened.

That is the lesson from Northern Ontario that deserves Southern Ontario’s attention now.

Ontario did not fail to recognize the forestry crisis. One of the easier stories to tell about the collapse of forestry would be that governments simply failed to understand what was happening until it was too late.

The historical record does not support that conclusion. Ontario created the Minister’s Council on Forest Sector Competitiveness in 2004. By the time its work was released the following year, its diagnosis was remarkably stark. The council concluded that the industry was in crisis. It identified delivered wood costs, electricity, transportation, capital investment, international competition and other structural disadvantages. It identified a group of Northern Ontario mills as being at risk and estimated that their loss could eliminate thousands of direct jobs, many more indirect and induced jobs in the North, and substantial employment in Southern Ontario as well.

In other words, this was not a retrospective analysis written after the mills were gone. Ontario had been warned while there was still an industry to protect. Just as importantly, policymakers understood that forestry was not a collection of independent factories. Pulp mills, sawmills, harvesting operations, transportation companies and energy systems were economically connected. A sawmill did not simply produce lumber. It also produced chips and residual fibre that could be sold to a pulp mill.

A pulp mill closure could therefore damage the economics of a sawmill that remained physically capable of producing lumber. Nothing needed to go wrong inside the sawmill itself for its economics to deteriorate.

That relationship sounds obvious now, but it is central to understanding what happened. The loss of one important facility could weaken another one, which could in turn reduce the economics of a third. The damage could move through the system. Ontario understood that mechanism, and it also did not simply stand aside. By the middle of the decade, the province had assembled more than a billion dollars in forest-sector support. The measures included loan guarantees intended to stimulate capital investment, a Forest Sector Prosperity Fund, significant support for forest access roads, electricity assistance for northern pulp and paper mills, stumpage relief and other programs intended to improve competitiveness. Ottawa was acting as well, with programs aimed at community adjustment, worker retraining, technology, new products, export diversification and industrial transformation.

It would therefore be unfair to say that governments neither understood nor responded to the crisis. The more interesting question is why such a significant response could coexist with such a profound industrial contraction.

That is where the lessons become useful.

There is a tendency in public policy to discuss industrial support as though it were a single thing. A billion dollars of government assistance becomes evidence either that government is heroically defending an industry or recklessly subsidizing it, depending on one’s ideological preference. The forestry experience suggests both framings are too crude.

Different interventions were solving fundamentally different economic problems, and their effectiveness depended heavily on whether the underlying facility still had a viable future.

Consider forest access roads. Ontario progressively assumed more of the cost of constructing and maintaining them. This was not particularly glamorous industrial policy. There was no revolutionary new product involved. It simply removed a structural cost that Ontario producers faced. The reason industry regarded the program favourably is easy to understand. A viable forestry company did not need to invent a new business plan to benefit. Its operating cost simply declined.

Electricity assistance operated on a similar principle. Northern pulp and paper mills are extraordinarily energy intensive. Reducing electricity costs and encouraging efficiency could improve the economics of an otherwise viable operation. But electricity relief cannot create a customer.

A cheaper road cannot reverse the decline of newsprint. A lower power bill cannot give a mill a competitive product if its market has disappeared. That boundary between a competitiveness problem and a business-model problem is central to understanding what happened.

Ontario’s capital programs illustrate the problem particularly well. The province created large loan-guarantee and capital-support programs to encourage modernization, energy efficiency, cogeneration, value-added manufacturing and other investments.

On paper, these were major interventions.

In practice, substantial portions of the funding were never deployed. Projects were delayed or abandoned because the industry’s economics continued to deteriorate. Companies that were already questioning the future of a mill were understandably reluctant to commit major amounts of new private capital simply because government financing was available.

There is an important paradox here. Government had created programs intended to help companies obtain capital at precisely the moment when deteriorating markets were making companies less willing or able to invest capital. A financing program can reduce the cost of a good investment. It cannot make a bad investment good.

That distinction is worth remembering as Ontario considers how aggressively to support automobile manufacturing today.

The short-lived resurrection of the former Cascades fine-paper mill in Thunder Bay offers an unusually clear illustration. Governments supported the effort to reopen the facility as Thunder Bay Fine Papers. The objective was understandable. A significant industrial asset already existed. Skilled workers were available. Restarting it offered the possibility of restoring hundreds of high-paying jobs rather than accepting the finality of a closure. The mill did restart, but it struggled to generate sufficient sales in its coated-paper and catalogue markets. Production stopped again, workers were laid off, additional financing was required and the company ultimately entered receivership.

The global financial crisis undoubtedly made a difficult situation worse, and it would be unreasonable to claim that government support caused the failure. The useful lesson is narrower. Government financing had succeeded in getting the machinery moving again. It had not repaired the economics that justified moving the machinery.

A plant is not economically valuable simply because its lights are on. Its value comes from a web of relationships extending far beyond the walls: customers who want its products, suppliers capable of serving it, workers with specialized skills, transportation systems, engineering capability, capital investment and, most importantly, a credible reason for its owner to invest there again.

Once enough of those relationships disappear, reopening becomes considerably harder than keeping a viable plant operating in the first place.

Fort Frances provides one of the clearest examples of how industrial decline actually unfolds. Its pulp and paper operation did not move overnight from full employment to closure. Employment declined progressively. Machines were idled. Product markets weakened. A major pulp customer disappeared. Production was curtailed further. The operation was eventually reduced to a skeleton workforce. Only later did permanent closure become official. By the time the announcement everyone remembers occurred, a long series of smaller economic decisions had already transformed the facility.

Kenora, Red Rock and Smooth Rock Falls each followed their own path, but similar patterns appear repeatedly. Machine closures often preceded mill closures. Employment reductions preceded final shutdowns. Capital investment weakened. Market relationships changed. The economic rationale for operating the facility eroded in stages. Then came the secondary effects. When the anchor facility disappeared, contractors lost volume, suppliers lost customers, transportation demand declined and skilled workers moved away or changed industries.

Some communities adapted better than others. Kenora is an important counterexample to simplistic predictions of community collapse. It lost its large paper mill and much of its traditional forestry employment, yet the community adapted through tourism, services and other economic activity. Thunder Bay also absorbed the loss of thousands of forestry jobs without suffering the kind of demographic collapse experienced in smaller single-industry communities. This distinction matters enormously. A community can remain economically viable while losing an industry.

Brampton would survive the disappearance of automobile assembly. Oshawa survived GM’s 2019 closure.

Ontario itself would plainly survive the loss of several plants. But municipal survival and industrial survival are not the same thing. The relevant question is not whether Brampton would continue to be a prosperous city. It is whether Ontario would still possess the integrated manufacturing capability necessary to win the next automobile platform.

That is where the analogy becomes uncomfortable.

Ontario’s automobile industry is not presently in a generalized collapse, and it is important to say so plainly because exaggeration would weaken the argument. Windsor has recently added production and employment. Ford is preparing Oakville for Super Duty truck production. GM is investing in Oshawa as it prepares the plant for the next generation of full-size pickups. These are not the characteristics of an industry that has already ceased to be competitive.

But other plants are flashing very different signals.

Brampton may be the most troubling example. The facility stopped vehicle production while being prepared for a new Jeep Compass mandate. The retooling was subsequently paused. Stellantis then shifted the future Compass program to Illinois. The plant remained idle, and Stellantis has since considered the possibility of selling the facility. Seen as a series of individual announcements, each development can be explained. Seen through the forestry lens, the sequence looks more consequential: an existing product ends, the plant is idled, a future product is promised, capital work is interrupted, the future product moves elsewhere, the facility remains without a mandate, and eventually the future of the plant itself comes into question.

A factory without a product mandate is still a building, but it is no longer fully an industrial operation in the economic sense. It is an asset waiting for someone to provide a reason to invest in it.

CAMI in Ingersoll raises a different version of the same problem. The plant was heavily modernized to build BrightDrop electric commercial vans, but the market developed much more slowly than expected. Production was suspended and the product was ultimately discontinued. GM did not simply move BrightDrop elsewhere; it ended the program and began assessing what might come next for the facility. That distinction matters because it demonstrates why the current automotive challenge cannot be reduced to tariffs. CAMI has encountered a product-market problem, an electric-vehicle transition problem, regulatory uncertainty and a product-allocation problem at the same time.

That should sound familiar to anyone who studied forestry. Industries rarely collapse because of one thing. They become vulnerable when several pressures begin reinforcing one another.

Oshawa provides a useful contrast. GM is reducing production from three shifts to two, which is plainly a negative signal for the workers affected. But GM is simultaneously investing heavily to prepare Oshawa for the next generation of full-size pickups. This difference may be among the most important lessons in the entire comparison. A shift reduction is painful. A temporary shutdown is concerning. A permanent loss of the next product mandate may be existential.

Industrial policy tends to focus on employment because employment is visible.

Governments can count jobs. Politicians can announce that jobs have been saved or created. Reporters can interview workers entering or leaving a factory gate. But jobs are often a lagging indicator of a much earlier capital-allocation decision. An automobile company may decide years before the final vehicle rolls off a line whether a plant will receive its next product. Tooling orders follow. Supplier investments follow. Engineering assignments follow. Hiring follows. By the time layoffs become visible, the decision that created them may already be several years old.

This was true in forestry. It is likely true in automotive manufacturing. The most important question Ontario can ask about any assembly plant may therefore not be how many people work there today. It may be what the plant will build next, and then what it will build after that.

Oakville is particularly interesting in this context because Ford confronted a changing market and changed the product rather than abandoning the facility. The plant had originally been scheduled for a major electric-vehicle transformation. When market conditions changed, Ford pivoted to Super Duty production instead, committing a high-demand and highly profitable product to the facility and bringing with it substantial employment and investment. That is what successful industrial adaptation looks like: a changing market, a new product, new investment, continued supplier demand and a continued reason to retain a skilled workforce.

It is essentially the reverse of the sequence Northern Ontario watched unfold in numerous mills.

Perhaps the most transferable lesson from forestry concerns the companies surrounding the anchor facility. A pulp mill closure could harm a sawmill because the sawmill lost a buyer for its chips. Nothing necessarily changed in the sawmill. Its equipment remained intact. Its workers remained productive. Demand for lumber might still exist. But one important source of revenue disappeared, changing the economics of the whole operation.

Automobile manufacturing has an even more complicated version of this relationship. An assembly plant is surrounded by Tier 1, Tier 2 and Tier 3 suppliers, tool-and-die firms, machine shops, plastics companies, electronics manufacturers, logistics providers, engineering firms and specialized service businesses. Many serve several customers. Their survival does not necessarily depend entirely on one assembly plant, but their future investment decisions may.

Imagine an Ontario supplier with three major programs. It loses one because a vehicle mandate moves to Michigan. Revenue falls by 25 per cent. The supplier does not close. Nobody holds a press conference. It simply decides that a new production line no longer generates an adequate return. Five years later it competes for another contract, but its technology is older than that of a U.S. rival. It loses the contract. Revenue falls again.

Another planned investment is cancelled. A skilled engineer leaves for Detroit. Another customer evaluating Ontario now sees a slightly thinner supplier base. The next investment becomes slightly more attractive somewhere else.

At some point, industrial decline stops being caused solely by the original external shock. It begins producing some of the conditions for its own continuation.

This is the phenomenon Ontario should fear most: not the dramatic overnight disappearance of automobile manufacturing, but the gradual thinning of the ecosystem until the next investment goes elsewhere because the previous investment went elsewhere.

The present Canada-U.S. trade conflict makes this problem considerably more urgent. The immediate cost of a tariff is easy to calculate. The longer-term cost of uncertainty is much harder. Automobile plants operate on long product cycles. Companies allocate billions of dollars years before production begins. A plant competes internally against other plants owned by the same corporation. Ontario does not merely compete against foreign automakers.

An Ontario GM plant competes for capital against GM facilities elsewhere. A Stellantis plant competes against other Stellantis plants. Ford makes similar choices among its own manufacturing assets.

If management believes that producing a vehicle in Ontario introduces a persistent tariff disadvantage, political uncertainty or border risk that does not exist at an otherwise comparable American plant, the relevant loss may never appear in a customs report. It appears as an investment that is never made.

That is another lesson forestry teaches painfully well. A plant need not become physically incapable of operating before its owner decides that the next dollar will earn a better return somewhere else. Industrial decline often begins in the capital budget.

None of this leads me to the conclusion that government should refuse to support automobile manufacturing. Quite the opposite. There are circumstances in which failing to intervene against a temporary, externally imposed trade shock would amount to allowing economically viable productive capacity to be destroyed for reasons unrelated to its underlying competitiveness.

But forestry suggests that government needs to be much more precise about what it is purchasing with public money.

Suppose an automaker asks Ontario and Ottawa for substantial support to keep an aging vehicle program operating for another three years at declining volume. That may preserve jobs temporarily, but it does not necessarily preserve an industry. Now consider a different proposition. In return for public assistance, an automaker commits a next-generation platform to Ontario for eight years, invests several billion dollars in tooling, commits minimum production volumes, sources significant components from Canadian suppliers and locates engineering work here.

Both arrangements could be described politically as support for the auto sector.

Economically, they are entirely different. The first buys time. The second buys capability.

Forestry suggests that when public money is used for industrial policy, capability is the more durable asset. Major industrial assistance should therefore secure measurable forward commitments: future product mandates, capital expenditures, production volumes, Canadian supplier spending, engineering and research activity, and sufficient duration to ensure that public support purchases more than a temporary postponement. Clawbacks should apply when promised investment or production disappears, and policymakers should pay at least as much attention to the supplier network as they do to the assembly plant itself.

The objective is not to preserve a building. It is to preserve Ontario’s ability to manufacture complicated things competitively at scale.

Forestry also offers a warning against confusing two very different economic problems. A fundamentally competitive mill can experience a temporary shock. Currency can move. Energy prices can spike. A trade dispute can interrupt normal markets. Credit can disappear during a financial crisis. In those circumstances, temporary assistance can make enormous economic sense. It bridges the period between two viable states and prevents productive capacity, skilled workers and supplier relationships from being destroyed unnecessarily.

But a facility whose underlying product is disappearing presents a different problem. No amount of bridge financing can bridge permanently to nowhere. That was the difficulty facing parts of the pulp and paper industry as consumption of newsprint entered secular decline. The question for government was no longer simply how to keep making the same product more cheaply. It was how to create different products, different markets and different industrial uses for the capabilities that remained.

Ontario auto manufacturing is in a better position in one crucial respect. The underlying product is not disappearing. North Americans will continue to buy automobiles. The existential question is where those automobiles will be built.

That makes the present problem more preventable than the collapse of newsprint, but it also makes complacency less excusable. Ontario is not being asked to manufacture demand for an obsolete product. It is competing for its share of an enormous existing market, and that competition is increasingly about who wins the next capital cycle.

The forestry collapse also taught Northern Ontario that employment statistics can conceal significant economic loss. A displaced mill worker who eventually finds another job may be recorded as successfully re-employed. That does not mean the worker or the community has been made whole.

High-wage industrial employment often carries benefits, pensions, overtime, accumulated seniority and skills that are difficult to reproduce in other sectors. When workers leave those jobs permanently, their earnings can remain depressed for years.

The same issue applies to automobile manufacturing. A 52-year-old toolmaker who moves from a highly paid manufacturing position into a lower-paid service job is no longer unemployed, but productive capacity has still been lost. Skills embedded in experienced industrial workers should therefore be treated as capital. Programs that allow firms to reduce hours temporarily rather than disperse entire workforces make sense when the shock is genuinely temporary. Retraining makes the most sense when it occurs before a worker has become detached from industry and when it leads toward another credible industrial employer rather than simply toward a classroom.

Once a specialized workforce disperses, reopening a plant becomes harder even if the building remains.

Again, the lesson is to intervene before the ecosystem disappears.

There is an understandable temptation, looking at the scale of Northern Ontario’s forestry contraction, to conclude that the billions spent by governments were wasted. The evidence does not justify that conclusion. Some programs clearly improved competitiveness. Some mills survived that might otherwise have closed. Some private investment was leveraged. Forest roads, energy programs and modernization support benefited operations that continued to have viable markets. Nor can we know the counterfactual. Without government action, the contraction might have been considerably worse.

The more defensible conclusion is that financial assistance has limits. Government can reduce a structural cost disadvantage. It can share the risk of productive capital investment. It can help an otherwise competitive company survive a temporary shock. It can support workers while demand recovers. What government cannot reliably do is subsidize an industrial ecosystem back into existence after the relationships that sustained it have disappeared.

That is why timing matters so much.

The cheapest auto plant to save may be the one that still has a product. The cheapest supplier to preserve may be the one that has not yet cancelled its next machine order. The cheapest workforce to retain may be the one that has not yet dispersed. The most important industrial-policy meeting may happen years before anyone stands in front of microphones announcing a closure.

For those of us who have lived in Northern Ontario through the forestry contraction, there is something unsettling about portions of the present automobile story. It is not that Brampton is Fort Frances or that Ingersoll is Red Rock. They are not. Southern Ontario’s automobile communities are generally larger, wealthier and much more economically diversified than the northern mill towns that absorbed some of forestry’s hardest blows. Automobile demand remains substantial. Ontario possesses globally competitive engineering, machining, tooling and manufacturing capability. Several major assembly plants are actively receiving new investment.

There is every reason to believe the industry can remain an important part of Ontario’s economy.

But there is also enough evidence to recognize the sequence. A plant loses a shift. An investment is delayed. A retooling is paused. A future product moves somewhere else. A supplier loses volume. A capital project is cancelled. A skilled worker leaves. Another future product becomes slightly harder to win. None of these events individually constitutes the collapse of an industry. That is precisely the point. Neither did the first closed paper machine.

Northern Ontario’s experience should not be invoked to predict that Southern Ontario automobile manufacturing is doomed. It should be used to understand how an industrial economy can move from a series of manageable problems to a self-reinforcing decline, and it should remind policymakers that the closure announcement is often the least useful moment to begin worrying.

If we were serious about learning from forestry, governments would measure the automobile industry differently. We would certainly continue counting employment, exports and production, but we would also watch the leading indicators. Does each assembly plant have a credible next-generation product mandate? Are committed five-year capital expenditures rising or falling? Are plants adding shifts or progressively removing them?

What are tooling companies seeing in their order books eighteen, thirty-six and sixty months ahead? Are Ontario parts suppliers investing in new equipment or merely maintaining what they already own? Is Canadian supplier content rising or declining? Are engineering and research jobs growing here? When companies make global platform decisions, how frequently is Ontario winning against competing U.S. and Mexican facilities?

Those questions tell us more about the future than a photograph of politicians standing beside workers at a funding announcement.

There is a tendency during an economic crisis to think the boldest policy is the one involving the largest number. A billion-dollar rescue sounds more serious than a modest intervention made three years earlier. Forestry suggests the opposite may often be true. By the time an anchor facility has closed, suppliers have lost revenue, workers have dispersed, capital budgets have moved elsewhere and the site has begun moving toward redevelopment, extraordinary sums can be required simply to recreate capabilities that once existed naturally.

The more sophisticated industrial policy is therefore preventative. Act earlier. Distinguish a temporary shock from structural decline. Reduce real cost disadvantages where they exist. Use public money to secure future products and future capital rather than merely extending yesterday’s production. Protect suppliers, tooling, engineering and skilled labour, not simply the assembly building. Treat the loss of the next product mandate as seriously as the loss of the current shift.

Northern Ontario learned these lessons at considerable cost. Southern Ontario now has the opportunity to use them.

Ontario’s forestry experience does not tell us that automobile manufacturing will collapse. It tells us something more useful: what industrial decline looks like before everybody agrees to call it industrial decline.

We should recognize it early enough to write a different ending.