Photo by Sabrina Rizzo on Unsplash
Twenty years of criticism, nearly twenty years of comparative data, and one simple question: does Quebec get enough for what it pays?
Those who know me know that my criticism of the Quebec model This isn't a recent phenomenon.
I've been blogging about these issues since 2006, initially with The Quebec Minarchist, then with Counterweightbefore launching Free North/True NorthIn other words, I didn't wait for Éric Duhaime or the 2026 election to wonder if Quebec was really getting its money's worth with its particularly interventionist state model.
In September 2012, I was already writing about our healthcare system that in the absence of a monetary price, scarcity does not disappear. It simply finds a substitute: time.
Thirteen years later, this idea has taken on a somewhat too concrete form for me.
In July 2023, my family doctor detected a heart murmur. My mitral valve needed to be repaired. The surgery finally took place on February 25, 2025.
Nearly twenty months after the initial discovery.
My family had already experienced a similar situation. We had registered our daughter for a spot in a subsidized daycare center. Fortunately, we had in the meantime found a private, subsidized daycare that used different admission criteria. The letter informing us that a spot in the subsidized daycare center was finally available arrived while she was already attending elementary school.
In one case, rarity was measured in years on a list. In the other, in months before surgery.
An anecdote obviously proves nothing. But it can illustrate a mechanism that the data can then be used to test.
And that is precisely the exercise I want to do here.
Twenty years of criticism do not constitute proof.
If the Quebec model works better than I thought, the data should be able to contradict me. If some of my intuitions were correct, they too should stand up to scrutiny.
At the same time, we can apply exactly the same rule to the diagnosis put forward today by Éric Duhaime and the Conservative Party of Quebec.
Quebec has truly chosen a more cumbersome state
Let's start with the least controversial point.
In 2024, consolidated spending by provincial and local governments represented 33.68% of Quebec's GDP, compared to 25.35% in Ontario, 26.56% in British Columbia, and 20.40% in Alberta. This data comes from Consolidated Canadian Public Finance Statistics from Statistics Canada.
Public sector employment represented 24.14% of Quebec employment in 2025, compared to 19.70% in Ontario, 18.55% in Alberta and 20.76% in British Columbia, according to the Statistics Canada table 14-10-0027-01However, several smaller provinces show even higher proportions.
The same observation applies to the tax side.
According to Statistics CanadaThe tax burden on provincial and local governments, which includes taxes, duties and social security contributions, reached 21.2% of Quebec's GDP in 2024compared to 16.8% in Ontario, 15.5% in British Columbia and 9.9% in Alberta. Quebec thus had the highest tax burden among the provinces according to this measure. See the Statistics Canada table.
We could debate at length whether this level is desirable.
The observation, however, is quite difficult to avoid:
Quebec channels a much larger portion of its economy through the state than its major neighbours.
The interesting question begins after this:
What do we get for the difference?
Education: good results, ailing buildings
If the objective was simply to demonstrate that all public spending ends up in an administrative black hole, the data would be rather contradictory.
Education is probably the best counter-evidence.
Quebec was spending ,087 per student from preschool to secondary school in 2022-2023, according Statistics Canadaand obtained in parallel 514 points in mathematics at PISA 2022, a very high result in the Canadian comparison. The Council of Ministers of Education of Canada formulates methodological reservations concerning in particular the non-response, but the result remains important enough to have to be taken seriously.
But looking only at educational outcomes would give an incomplete picture.
The physical capital of the school network tells a much less rosy story.
In its 2024-2025 annual report, the Ministry of Education indicated that only 47% of school buildings were considered to be in good conditionThe ministry specifies that a new calculation method, more detailed than the old one, has modified this measure and notes an improvement in the ratio of the asset maintenance deficit to the replacement value.
The Auditor General of Quebec nevertheless assessed the school building asset maintenance deficit projected at .5 billion in 2024-2025, compared to 4.5 billion in January 2019. He specifies that a new standardized inspection methodology and a revision of unit costs explain part of this increase. So it's not eight billion worth of physical deterioration that someone might have missed while passing by the schools. See the VGQ follow-up.
The VGQ nevertheless considers the progress unsatisfactory on several recommendations concerning the maintenance of the park and the space requirements.
This case is particularly instructive.
Quebec can achieve good educational results while simultaneously accumulating a serious problem with the maintenance of its infrastructure.
There is no contradiction.
"Education spending" is a huge aggregate that mixes teachers, professionals, administration, maintenance, renovation, construction and several other items.
An average can therefore appear excellent while one component of the system is deteriorating.
This is precisely why the amount spent is such a poor measure of performance.
The right question is not simply: How much do we spend on education?
But : What are we getting in each of the areas that this money is supposed to fund?
Redistributing and producing are not the same thing.
Poverty provides another serious argument in defense of the model.
In 2024, 7.0% of Quebecers lived below the official poverty linecompared to 12.5% in Ontario and 13.0% in British Columbia. Statistics Canada.
This social performance, however, coexists with a lower overall income level.
The median after-tax income of families and individuals not living with families was ,800 in Quebec in 2024, compared to ,500 in Ontario, ,300 in Alberta and ,800 in British Columbia, in constant 2024 dollars. Statistics Canada specifies that these differences do not take into account regional factors such as the cost of living or age structure.
This is a fundamental tension within the Quebec model:
Reducing poverty and producing more wealth are not exactly the same goal.
A model can be relatively effective in redistribution without necessarily being as effective in income creation.
The real question: the efficiency of the state
Health also shows why simply counting dollars is not enough.
According to theCanadian Institute for Health InformationProvincial health spending represents a particularly large share of Quebec's economic capacity.
However, available capacity does not always keep pace.
Between 2021 and 2025, the number of long-term care beds in Quebec increased by 11%, but the population aged 75 and over 13%The number of beds per 1,000 people in this group thus increased from 54.9 to 53.7.
This does not prove that the additional spending was wasted. Needs may have become more complex, costs may have increased, and home care may have replaced some hospital beds.
But one question stubbornly remains:
What did we achieve with the additional resources?
An expense is not a result.
A result is a patient treated, a student trained, a permit issued, a road built, a building maintained, or a problem effectively solved.
Public dollars do not become productive simply because they have passed through the doors of a ministry.
Daycare: affordable does not mean accessible
Quebec's network of subsidized childcare services also provides a strong defense of the model. Its low fees reduce the costs borne directly by families, and economic literature has documented effects on mothers' participation in the labour market.
But the network also illustrates a persistent capacity problem.
At May 31, 2025the government still counted 30,688 children in immediate need of a place and not currently occupying one in a recognized childcare serviceThis is the last data of this type available before the change in collection method related to the new portal. See the statistics from the Government of Quebec.
After nearly three decades of network expansion, the problem of rationing had therefore not disappeared.
We must avoid taking an overly simplistic approach. The low subsidized rate stimulates demand, but capacity also depends on the number of educators, real estate costs, the construction of places, their location, and the rules of the network.
One rather stubborn economic reality remains:
When the quantity demanded exceeds the available capacity, a mechanism must decide who gets the scarce resource.
If this mechanism is not primarily price, it will be lists, priorities, criteria and time.
That's exactly what my family went through.
A position can be very cheap for the person who gets it and completely inaccessible for the person who doesn't.
Affordable price and accessibility are not synonymous.
In health, the price can be paid in time.
The same principle appears in health with one fundamental difference: clinical prioritization is necessary.
A patient whose life is immediately threatened must obviously take precedence over one whose condition is less urgent.
The problem is therefore not that a queue exists.
The problem begins when there is insufficient capacity to provide care within reasonable timeframes.
This is what I was already describing in 2012 when I wrote that waiting could become a substitute for monetary price.
My cardiac experience does not allow me to determine whether each of the months that passed before my surgery was medically avoidable.
However, it highlights something that public accounting measures poorly:
Time also has a price.
Worry has a cost. Pain has a cost. An inability to work can have a cost. So can the time spent by loved ones.
The absence of a bill upon discharge from the hospital does not eliminate these costs. It simply removes them from the RAMQ statement.
Tax in order to then subsidize
Economic intervention provides another test of the model.
In 2024, consolidated subsidies to private companies represented 1.385% of Quebec's GDP, compared to 0.967% in Ontario, 0.665% in Alberta, and 0.698% in British Columbia. Quebec was not the province that subsidized the most as a proportion of its GDP, but it used this tool significantly more than its major comparators. The data comes from consolidated public finances of Statistics Canada.
Even more striking, these subsidies amounted to approximately 70% of provincial revenue comes from corporate income tax, compared to approximately 45% in Ontario, 39% in Alberta and 35% in British Columbia.
This ratio obviously does not mean that the same companies pay a dollar of tax in the morning and get 70 cents back in the afternoon.
It shows something more interesting:
The state takes a large sum based on profits made, then reallocates a considerable amount of capital according to other criteria.
Some interventions may be justified. A subsidy can correct an externality, fund research, or support an activity whose social benefits exceed those captured by the company.
But the public allocation of capital also carries risks.
A OECD econometric study published in 2025 concludes that, in its sample of large manufacturing firms, subsidies increased the market share of recipients without producing corresponding gains in investment or productivity.
Of the research published in 2026 on the global steel industry arrive at a related conclusion: more subsidized firms can gain market share even when their productivity, cost efficiency or financial performance is lower, while reducing the incentive to invest in some less subsidized competitors.
These studies do not specifically focus on Quebec.
They demonstrate why “Economic development subsidy” should never be automatically considered synonymous with “economic development”.
Yield must be measured.
Meanwhile, investment remains relatively low
In 2024, gross fixed capital formation by businesses represented 17.25% of Quebec's GDP, compared to 18.32% in Ontario, 19.22% in Canada excluding Quebec, 21.10% in British Columbia and 21.16% in Alberta, based on Statistics Canada's Provincial and Territorial Economic Accounts.
The investment is important for a fairly simple reason: better equipment, software, technologies and infrastructure generally allow you to produce more with the same amount of work.
There Bank of Canada He also points out that Canadian companies have been underinvesting for a long time and establishes an important link between this weakness in investment and poor Canadian productivity performance.
However, our own figures have forced us to qualify the Quebec picture somewhat.
According to Statistics Canada, the hourly productivity of the business sector reached in 2024 .40 in Quebec, compared to .00 in Ontario, .30 in Alberta, and .90 in British Columbia. Quebec recorded an increase that year of 2.0%, while Ontario, Alberta and British Columbia were falling back.
So no, Quebec is not condemned to remain motionless while all its neighbors disappear over the horizon.
The Catch-Up Trap
Our long series also shows that Quebec is not engaged in a continuous economic decline.
Between 2000 and 2024, its real GDP per capita grew faster than Ontario's.
But the word catch-up It deserves to be handled with care.
In recent years, several major comparators have themselves slowed down. Part of Quebec's convergence therefore stems from better relative performance, not necessarily from a dramatic acceleration.
Catching up with someone doesn't always mean running much faster. Sometimes, the other person slows down even more.
This destroys two comfortable narratives at once.
Catching up does not prove that the Quebec model has become a machine for prosperity.
But it also contradicts the idea that it mechanically condemns Quebec to perpetual decline.
Numbers have this annoying habit of not asking permission from ideologies before being released.
Large state equals low growth? The data doesn't give us that shortcut.
This is probably the most important conclusion of this research.
Our dossier does not allow for establishing a simple causal relationship between the size of the state and the set of observed economic performances.
Our exploratory analyses of the North American panel did not produce the wonderful downward-sloping line that would have allowed us to write:
One more state point = that much less growth.
Too bad about the simplicity.
So much the better for rigor.
The relationship seems to be mediated more through institutions.
Taxation can alter incentives. Regulations can change entry and investment costs. Subsidies can change capital allocation. The way a service is financed can influence its ability to meet demand.
And, as the case of schools has just shown, A system can succeed in one dimension while allowing another to degrade..
The real question then becomes less: How much does the state spend?
that : What does each dollar produce, and what does it prevent us from producing elsewhere?
This is the question of opportunity cost.
And what about Duhaime in all of this?
This is where you need to carefully separate them diagnosis And prescription.
Several problems currently described by Éric Duhaime correspond to phenomena observable in our data: high weight of the public sector, significant tax pressure, substantial use of subsidies, relatively low private investment and persistent capacity problems in certain services.
This does not automatically turn the Conservative Party's program into a search result.
The PCQ, for example, claims to be able to achieve billion in cumulative savings over five yearsThis includes reducing subsidies, cutting staff primarily through attrition, and further integrating artificial intelligence into government operations. These are the savings calculated by the party, not savings demonstrated by our analysis. Source: PCQOur comparison of financial frameworks is also available in Five financial frameworks, five paths to a zero deficit.
The same caution applies to his proposal to gradually increase Quebec's corporate income tax rate. 11.5% to 4.7%The PCQ estimates that this measure would create between 25,000 and 45,000 jobs over five years. This is, again, a projection by the party. Source: PCQ.
The proposed mechanism nevertheless deserves to be studied: simultaneously reducing taxes and subsidies amounts to shifting part of the intervention from "Tax then choose" towards a more general reduction in the levy.
The outcome would depend on the reaction of businesses, investors, workers, and public finances.
Duhaime's proposals must therefore pass exactly the same calculator test as the model he criticizes.
Health probably offers the best test
The PCQ proposes establishing time limits for major surgeries, specialist consultations, and diagnostic investigations. When the public system cannot meet these deadlines, the patient could be treated elsewhere in the public system or by an accredited private provider, at the expense of the RAMQ (Quebec Health Insurance Plan). The party also proposes 400 to 500 additional medical school admissions per year and the training of 1,500 new nurse practitioners during its mandate. Source: PCQ.
The principle of a guaranteed delivery time combined with the choice of another supplier is not a Quebec invention.
L'OECD It documents similar mechanisms in Denmark and Portugal, where patients can choose another provider, including a private one, when the maximum waiting time is approached or exceeded. It particularly emphasizes that strategies that have sustainably reduced waiting times have generally combined explicit limits with interventions that genuinely affect supply and demand.
That's the relevant test.
If a private clinic simply recruits the same professionals who would have worked elsewhere in the network, the total capacity may change little.
If a different organization allows for more procedures, better use of facilities, more hours of care, or the arrival of new resources, the result may be different.
Changing the logo above the door still doesn't create surgeons by mitosis.
The proposal will therefore need to be evaluated based on the capacity actually added or better utilized.
Same test in daycare centers
The PCQ proposes to maintain funding for the subsidized network while increasing the tax credit for unsubsidized daycares.
It also offers a taxable allowance that can reach 0 per week per childthat is ,200 per year, to certain families who do not have a subsidized childcare space and do not claim the childcare tax credit. The party also states that approximately 10,000 places are currently vacant. in unsubsidized daycare centers. Source: PCQ.
The proposed mechanism therefore targets an allocation problem: attempting to make existing places financially accessible.
But the figure of 10,000 places remains that put forward by the PCQ, and even if this capacity is indeed available, it is not necessarily located in the places, for the age groups and at the times where the demand is.
Once again:
The test focuses on the result, not the announcement.
The calculator test
After analyzing the Quebec model with figures, I don't believe the best question is simply:
"More government or less government?"
That's too crude.
Education is now revealing something even more interesting than at the beginning of our research: The same network can achieve good academic results while accumulating a significant debt for the maintenance of its buildings.
Daycare centers show that a program can reduce the price paid by families and make parents' work easier while still rationing some access.
Health shows that a universal system can absorb enormous resources while remaining limited by its capacity.
Subsidies show that a policy designed to help the economy can also change the mechanisms that direct capital.
So let's apply the same test to everything.
- What problem are we trying to solve?
- How much does it actually cost?
- What measurable result is obtained?
- What behavioral reactions are provoked?
- What are the second-order effects?
- What is the opportunity cost?
- Is there a less expensive way to achieve the same result?
I have been publicly criticizing the Quebec model since 2006.
Twenty years later, the data does not confirm all my intuitions.
Quebec is not in continuous economic decline. Some institutions within its model are achieving significant results. And the relationship between the size of the state and prosperity is far less straightforward than a simple comparison of spending rates might suggest.
But the figures also highlight persistent problems of efficiency, rationing, allocation, investment, maintenance of public capital and supply adaptation.
On several of these specific problems, the diagnosis put forward by Duhaime corresponds to phenomena that the data actually allow us to observe.
His solutions now face exactly the same challenge.
Because ultimately, the most useful question is neither "public or private?" nor "left or right?".
It is much more uncomfortable:
What do we actually get for what we pay?
About the Author
3 thoughts on “The Quebec model: what if Duhaime was right about the diagnosis?“