Administrative rules in Canada’s dairy supply management system differ significantly across provinces. P5 provinces (Ontario, Quebec, New Brunswick, Nova Scotia, and Prince Edward Island) have made efforts to standardise their policies as they review and harmonize their quota policies every five years. Nonetheless, their quota policies differ significantly, in particular when it comes to quota transfer. Western Milk Pool (WMP) provinces (British Columbia, Alberta, Saskatchewan and Manitoba) also have different quota rules.
Quota transfer rules affect farm competitiveness, especially in the P5. I have discussed some of the economic impacts of quota administration in a previous post. In it, I consider scenarios in which the quota allocation is efficient or inefficient. Following previous literature, I should have described the quota allocations as competitive and non-competitive. The post shows that a non-competitive quota allocation supports a larger number of smaller farms and raises the cost of production.
This post is about the administration of dairy supply management in Canada, specifically quota transfer rules. Provincial boards oversee quota trade. The rules include provisions regarding the sale and purchase of quota between farms, mergers, and quota transfers to family members.1
Quota transfer rules affect the competitiveness of dairy production. Competitiveness is crucial to a resilient Canadian dairy sector. The existence of supply management depends on its acceptance by stakeholders, including farms, dairy processors and consumers. If the administration of supply management introduces too many inefficiencies, it may lead to a loss of support for the system. We will see other consequences of inefficient supply management, including those on milk prices and competitiveness with imports.
This post begins by showing how quota transfer rules differ across provinces. Then, I show how the rules affect farm size, technology adoption, and production costs across provinces. I discuss some of the consequences of strict quota transfer rules. I conclude with a final argument about technological progress and farm consolidation.
Comparison of quota rules across provinces
Most quota transactions occur in provincial quota exchanges. Some provinces allow for quota trades outside the exchange.
In Quebec, all quota transactions must go through a central exchange system (SCVQ). Producers who want to sell quota must offer it on the SCVQ. The offers to purchase quota on the exchange greatly exceed sale offers because of the binding quota price ceiling of $24,000 per kg of butterfat per day (I’m going to forgo writing per kg of butterfat per day going forward). Rules to allocate quota between buyers are necessary given the shortage. First, purchase offers cannot exceed the maximum of 7 kg, or 10% of the total quota owned and borrowed of a producer. Second, 50% of the quota offered for sale is allocated among buyers equally (method referred to as iteration) and the 50% remaining is allocated proportionally to the purchase offers. This should be change to 100% iteration soon.
Table 1 summarizes the main quota policies in P5 provinces.2 I admit that sometimes I had trouble understanding the differences in rules between the provinces, in particular those concerning the ownership of one or many farms. The language used to describe the rules differs across provinces and as I’m not a legal expert, I’m not 100% confident my interpretation of the rules is always correct. I did my best to summarize those rules in the table below, and hopefully I got them right. If I got something wrong, please let me know and I’ll make the correction.
| Quebec | Ontario | New Brunswick | Nova Scotia | Prince Edward Island | |
|---|---|---|---|---|---|
| Ceiling price | $24 000 | $24 000 | $24 000 | $24 000 | $24 000 |
| Limit to purchase offers | Max of 7 kg or 10% of quota | 10 % of quota | No limit | No limit | No limit |
| Quota purchase allocation | 50% by iteration and 50% proportional to purchase offers* | 50% by iteration and 50% proportional to purchase offers | 100% by iteration | 100% by iteration | 100% by iteration |
| Number of licenses | A producer can own one quota, and only one quota can be exploited per farm | A producer can be licensed at more than one facility | A producer can own multiple farm sites | A producer can own many farms under one registration number | A producer may hold interest in only one quota |
| Quota merging | No | Yes, under some restrictions | No | No | No |
| Going concern sale | Yes, production on-site must continue for 5 years, buyer cannot own other quota | Yes, production on-site must continue for 5 years | Yes | Yes but limited | Yes but limited |
| Note: | |||||
| *Will be 100% iteration soon. |
Comparing quota transfer rules between P5 provinces, the rules are more restrictive in Quebec because they do not allow for a producer to own multiple licences, and merging quota is not possible. Ontario has the most permissible rules in the P5. Rules in the Maritime provinces compare to those in Quebec. A key difference in New Brunswick and Nova Scotia is that these provinces allow a producer to own multiple licences.
Table 2 summarizes quota rules for WMP provinces. The rules are much less restrictive than for P5 provinces. Most notably, there is no ceiling price in WMP provinces, although price adjustments in British Columbia are rule based. Producers can also operate multiple farms and going concern sales are allowed. Overall, weaker rules in WMP provinces result in more liquid and competitive quota markets.
| British Columbia | Alberta | Saskatchewan | Manitoba | |
|---|---|---|---|---|
| Ceiling price | Monthly price adjustment rule based on sale rate in the previous auction | No ceiling | No ceiling | No ceiling |
| Limit to purchase offers | One bid less than 50 kg | Max of six offers totalling less than 30 kg | Max of 3 bids per monthly auction | No |
| Quota purchase allocation | Allocated equally up to 10 kg, then at the pro rata of buying offers | Supply and demand | Supply and demand | Supply and demand |
| Number of licenses | A producer can own only one farm | A producer can control multiple dairy operations | ||
| Quota merging | Yes, must be approved by the board | Yes | 90% of quota can be transferred, subject to approval by the board | Yes |
| Going concern sale | Yes | Yes | Yes | Yes |
Quota prices
Table 1 shows a quota ceiling price of $24,000 in P5 provinces while Table 2 shows there is no price ceiling in WMP provinces. This difference in quota transfer rules across the two pools is important. The price ceiling is one of the main obstacles to competitive quota markets in the P5.
The price ceiling in P5 provinces has been in place for more than a decade. Significant inflation has occurred during that period. Expressing the quota price in real dollars, by removing the effect of inflation, will allow us to see that the price ceiling has become more restrictive over time.
Figure 1 corrects for inflation by showing monthly quota prices by province in 2025 dollars. In the P5, the quota ceiling price was 25,000 $ in 2011 and then was dropped to 24,000 $ in February 2016. Figure 1 shows that quota prices in 2025 dollars in the P5 have steadily declined because of inflation. The 25,000 $ ceiling on quota prices in 2011 would equal about $35,000 measured in 2025 dollars. In WMP provinces, quota prices in 2025 dollars have been higher and more volatile than in P5 provinces, but have not trended down. Prices are highest in Alberta, recently surpassing $60,000. In British Columbia, the quota price in 2025 dollars increased between 2011 and 2016, and has since been trending down. In Manitoba, the quota price trended down between 2011 and 2018 and has since been trending upward. In Saskatchewan, quota prices in 2025 dollars have cycled up and down. The quota price gap between P5 and WMP provinces has increased since 2011.
Figure 1 suggests that the price ceiling in P5 provinces has been getting more and more restrictive, producing a quota allocation among P5 farms has been increasingly moving away from a competitive allocation. A binding price ceiling necessarily implies a shortage. That’s what we observe in P5 provinces: farms do not sell quota because the price is too low and farms who want to purchase quota at the ceiling price cannot obtain all the quota they want. That leads to an inefficient quota allocation. At a higher quota price, more farms would be willing to sell their quota, fewer farms would be willing to purchase quota, but more farms would fill their quota needs. This would lead to a more efficient quota allocation where the total quota is produced at a lower total cost. An efficient quota allocation requires a competitive market for quota, free of price control, that leads to the demand for quota to equal the supply of quota.
Economic impacts of restrictive quota transfer rules
In this post published in May 2025, I use simulations to explain that there are 1) more farms when rules restrict quota transactions; 2) that these farms are smaller; and 3) they produce milk at a higher average cost. Those three predictions can be tested empirically. To that end, I begin by comparing farm sizes across provinces. Next, I compare dairy farm technology adoption across provinces. In an article I co-authored with Bruno Larue and Alphonse Singbo, we show that the technology farms adopt depends on their size and capacity to expand production. Thus, if we find that farms vary in size by province, we should expect variations in technology adoption as well. Finally, I compare production costs using data publicly available.
Smaller farms in the P5
As discussed, strict quota transfer rules stunt farm growth. A more liquid quota market means that farms can more easily expand to exploit economies of scales. With strict quota transfer rules in place for several years in the P5, we should see larger farms in the WMP than in the P5.
One caveat before looking at the data. Farm size cannot be solely attributed to quota rules. There are other factors that can explain differences in farm size across provinces, including environmental regulation and the cost of inputs. But, I believe, and that shouldn’t be too controversial of an opinion, that quota rules are the main factor explaining differences in farm size across provinces.
Figure 2 shows the average dairy farm size by province. The figure emphasizes Quebec and WMP provinces to make it easier to understand. Clearly, farms are bigger in WMP provinces. The average farms in WMP provinces are more than twice as large as those in P5 provinces. Farms in P5 provinces are about to reach the size of farms in WMP provinces in 2011. Quebec had, by a good margin, the smallest average dairy farm in 2011. In 2024, only the average farm in Prince Edward Island was smaller than the average farm in Quebec, and it appears it could be a blip in the data.
The left panel in Figure 3 shows how the size of the average farms has grown in absolute value, that is, in terms of kg of quota. The right panel of Figure 3 shows how farms have grown in percentage relative to their size in 2011.
Farms have grown in absolute value since 2011 much faster in WMP provinces, with Manitoba and Saskatchewan leading the way. In absolute value, farms in P5 provinces have grown about half as much as WMP provinces. When examining relative growth, the story is different. Manitoba and Saskatchewan are still the fastest growing, but Quebec arrives third. This is not too surprising because the average farm in Quebec was the smallest in 2011, and had a lot more room to grow.
The data show that dairy farms in WMP provinces are much larger than dairy farms in P5 provinces. This is consistent with quota transfer rules being much more restrictive in the P5 than in the WMP.
The exit of dairy farms happened in parallel to the growth of the average farm. Figure 4 shows the number of dairy farms by province since 2011. Quebec has by far the largest number of farms, but also lost the most dairy farms; more than 2000 between 2011 and 2025. Ontario lost nearly 1000 farms over that period. The number of farms in other provinces is much smaller and, as expected, the number of farms lost is smaller. The loss of farms is generally monotonic, meaning that the number of farms decreases every year. Exceptions are Saskatchewan and, to a lesser extent, British Columbia and Prince Edward Island, where the number of farms increased momentarily.
Figure 5 shows the relative decline in the number of dairy farms by province since 2011. The vertical axis shows the number of farms relative to 2011. All provinces have lost at least 20% of their dairy farms between 2011 and 2025. That is, number of farms remaining in 2025 is less than 80% of the number of farms in 2011. The decline was strongest in Manitoba, with more than 35% of dairy farms disappearing. Quebec is second with a 34% loss. The P5 lost nearly 30% of its dairy farms between 2011 and 2025 while the WMP lost about 24% of its dairy farms over that period.
The data for the number of dairy farms are interesting but they should be taken with a grain of salt. They only cover the period between 2011 and 2025. There is nothing special for that period other than those are the data that were easily accessible. The number of farms may have declined relatively faster in the WMP before 2011, explaining a slower decline after 2011.
Technology adoption by larger and small farms
More restrictive quota rules affect the technology farms adopt because of scale economies. That is, farms need to be of a minimum size for them to profit from adopting, for example, robotic milking systems.
Figure 6 shows how different types of barn adopt robotic milking systems by province. Tie stall barns do not allow for the free movement of animals and are typically smaller. That type of barn will no longer be permitted in 2027 according to the new code of practices. Free stall barns allow cows to roam freely. These barns must be larger and thus require a greater capital investment. They provide better animal welfare and are more suitable for robotic milking systems.
The difference is striking between P5 and WMP provinces. In P5 provinces, the majority of farms do not have robotic systems, and tie stall barns are more common. In WMP provinces, a much larger share of farms have robotic systems, and free stall barns are the norm. These findings are consistent with automated systems being adopted by larger farms with free stall systems, which are more common in western provinces.
Production costs and farm size
There is a causal chain between quota transfer rules and costs of production: quota transfer rules affect the capacity of farms to grow; hence their size, which affects the profitability of adopting certain technologies, which sets the costs of production. Moreover, there is empirical evidence of economies of scale in dairy farming, including articles by Larue and Singbo and Larue, Singbo and Pouliot.
To my knowledge, there is no publicly available study that reports dairy cost of production by individual farms across Canadian provinces. I am aware of three cost of production studies with publicly available summaries. The first is the annual cost of production survey from the Canadian Dairy Commission (CDC). The 2024 survey covered 254 farms, with about 3/4 of them in the P5 and 1/4 in the WMP. The second is Alberta’s Dairy Cost Study. The Alberta study is smaller in scale and surveyed 16 farms in 2024.3 The third is from the Ontario Dairy Farm Accounting Project, a co-operative project between the Canadian Dairy Commission and the Dairy Farmers of Ontario. The 2025 survey comprised 81 farms sampled.
I will compare the outcome of the three studies, but I must make a few caveats before proceeding. The CDC survey covers farms in the P5 and the WMP. It is not possible to distinguish average production costs in P5 provinces and WMP provinces from that study. However, given that roughly 3/4 of the farms surveyed are from the P5, the outcome of the study largely reflects average costs in P5 provinces.
The three cost studies differ methodologically. Notably, in the CDC study, the value of homegrown feed is implicitly priced at the cost of production. In the Alberta and Ontario cost studies, all feeds are valued at market price. This can make a significant difference in years when we see big swings in commodity prices. The Ontario study include revenues from the sale of crops and costs for the crop enterprises, I take the difference between the costs and revenues for crops and consider the outcome as feed costs.
Let’s now compare cost estimates from the three studies, knowing well that this is an imperfect comparison. We will use the CDC cost estimates and the Ontario cost estimates as proxies for the average production costs in P5 provinces, and the Alberta cost estimates as a proxy for the average production costs in WMP provinces.
Figure 7 compares cost of production estimates for the Alberta, Ontario and the CDC cost studies. The CDC adjusts its cost estimates to a standard milk composition, and accordingly I reversed that standardization to show cost estimates for the actual milk composition. The figure shows that the CDC and Ontario cost estimates are higher than the Alberta cost estimates except in 2023. That year, a drought affected Alberta, causing higher feed costs. Higher commodity prices since 2021 likely explain the narrower gap between CDC and Alberta cost estimates.
Assuming that the data in Figure 7 offer a reasonable comparison of average production costs in the two milk pools, the figure shows that production costs are higher in the P5 than in the WMP. This is expected given the difference in average farm size across provinces and scale economies in dairy farming.
Consequences
Strict quota transfer rules keep the number of farms higher, but farms are on average smaller. This is a desirable outcome for some. I can understand the attachment to small farms, and I hate to see farms cease operation. But, there are consequences to rules that keep farms small, while market forces push in the other direction everywhere else. We will see some of these consequences here.
Milk prices and profitability
The milk pricing system in Canada is complex. At its root is the annual adjustment based on the National Pricing Formula. The formula specifies that the annual farmgate milk price adjustment is set as the sum of 50% of the annual percentage change in the consumer price index and 50% of the percentage change in the cost of production as estimated by the CDC. That adjustment is at the national level.
The P5 has a smaller average farm with a higher average production cost than the larger average WMP farm. But also, the average P5 farm tends to adopt technology at a slower pace, and its production costs will tend to increase faster than the average WMP farm. Thus, in the annual cost of production study, the average P5 farm will tend to push the estimated costs up. On the opposite, the average WMP farm will tend to push the estimated cost down. With dairy prices set at the national level, while costs are lower in the WMP than in the P5, it means that the profit margin in the P5 will be smaller than in the WMP. Furthermore, if WMP adopt cost lowering technologies faster than farms in the P5, it also means that the gap between margins in the WMP and the P5 widens.
Table 3 shows a hypothetical example to help understand the point I’m raising here. I assume that farms in the P5 and the WMP have the same costs in year 1, but that costs tend to increase slightly faster in the P5. The column Cost change is a weighted average of the percentage cost change in the P5 and the WMP, with a 25% weight to the WMP. The price change is calculated from the inflation rate and the percentage cost change following the National Pricing Formula. I assume that the same farm price applies to the P5 and the WMP, although this is not exactly accurate in practice. The last two columns show the margins, calculated as the difference between the price and the costs, for the P5 and the WMP. The example shows the margin tends to decrease in the P5 while it trends upward in the WMP.
| Year | Cost P5 ($/hl) | Cost WMP ($/hl) | Inflation (%) | Cost change (%) | Price change (%) | Farmgate price ($/hl) | Margin P5 ($/hl) | Margin WMP ($/hl) |
|---|---|---|---|---|---|---|---|---|
| 1 | 82.40 | 82.40 | 84.40 | 2.00 | 2.00 | |||
| 2 | 84.13 | 84.05 | 2.05 | 2.07 | 2.06 | 86.14 | 2.01 | 2.09 |
| 3 | 85.81 | 85.64 | 1.75 | 1.98 | 1.86 | 87.75 | 1.93 | 2.10 |
| 4 | 88.39 | 88.30 | 2.10 | 3.03 | 2.56 | 89.99 | 1.61 | 1.69 |
| 5 | 90.60 | 89.62 | 1.85 | 2.25 | 2.05 | 91.84 | 1.24 | 2.21 |
The example in Table 3 captures what I believe has been happening. The National Pricing Formula, combined with strict quota transfer rules in P5 provinces, has created a situation where WMP farms benefit indirectly from the lower efficiency of P5 farms.
I’m quite confident that the situation described is happening. But I would need better data on production costs by province to confirm this. I would be very interested in having data necessary to verify if the margins in the WMP have been growing compared to the P5.
Market access
The current US administration has been very aggressive in its trade policies, and Canada’s supply management has been a target. The United States has demanded modifications to Canada’s dairy TRQs administration. I have my doubts that such modifications would lead to a significant increase in dairy imports from the US. I also have my doubts that the CUSMA review, set to begin soon, will bring about meaningful changes. Nonetheless, pressures on Canada’s supply management will not disappear and the risk is to see increased imports of US dairy products will remain. Preparing for a future where dairy supply management is weaker, or even no longer exists, is relevant, especially that it is difficult to predict what the current US administration might do.
The best way to prepare for increased market access is by making Canadian farms more competitive. This means producing milk at the lowest cost possible. Farms do their best to minimize costs while operating in the regulatory environment given to them. The quota transfer rules, in the P5 and in particular in Quebec, are not conducive to expansion and keep farms from producing at the lowest cost technology permits.
Farms are larger in the West and the evidence above suggests that they produce milk at a lower cost. Nonetheless, Western farms are still small compared to dairy farms in the United States. The average US dairy farm produces about 480 kg of butterfat per day, compared to less than 250 kg per day in the West (see Figure 2). That means that there is still plenty of room to grow, even for farms in Western Canada.
Farmers life quality
Being a farmer is choosing a way of life that is not always easy. Dairy farmers must tend to their cows from early in the morning until the evening, 365 days per year. They must also be available 24 hours per day during calving, and must also work in the fields, sometimes late at night. Dairy farming is not a single occupation; it requires doing a dozen different jobs all year-round.
Operators of larger farms may enjoy a better work-life balance because they can hire more labour to take work off their hands. That is more difficult for smaller farms, even though many employ labourers. Larger farms can more easily spread responsibilities across many people, such that having one or two people missing work does not affect operations too much. The mental and physical burden on a single person may be smaller on a larger farm because responsibilities can be spread across several people. Large farm operators can more easily afford to take days off and vacations than small farm operators. That is a meaningful advantage to the quality of life of large farm operators.
Many stressors affect farmers and many are struggling with mental health issues. Thankfully, an increasing number of resources are made available to them. My understanding is that financial problems are one of the main sources of stress. I believe that loosening quota transfer rules, in particular the quota price ceiling, would offer a way out to many farmers which could ease the mental health crisis.
Removing the quota price ceiling would cause the quota price to climb. If only based on the data in Figure 1, I would not be surprised if it climbed to $40,000 in the P5. What can that do to farmers’ life quality? It would give those struggling financially a decent way out. For example, suppose a small farm with a 50 kg quota. At the current ceiling price, that farm would obtain $1.2 million from selling its entire quota. Supposing a $40,000 quota price, that farm would obtain $2 million, $0.8 million more than with the price ceiling. That is a substantial difference. It could give that farm enough money to pay all its debts and perhaps enough to invest in another production.
Removing the ceiling price and other restrictions on quota transfer is not a silver bullet. There would still be farms struggling, but I’m convinced that it would give a way out for many farms in difficulties. The downside of letting the quota price climb is that entry in dairy farming would become more difficult. There are ways, however, that the industry can use to facilitate entry other than putting a price ceiling on quota price.
Progress won’t stop
This post argues for improved quota transfer rules. I think I’ve already brought good arguments, in particular related the quota price ceiling. I’m going to conclude by making a last argument.
Strict quota transfer rules have slowed down farm growth and farm exit in the P5, but have not stopped these trends. Technological progress is a major factor in farm consolidation. It will keep happening and, in response, farms will keep getting bigger until there are no more economies of scale to exploit. Strict quota rules only delay the inevitable, and that delay could have a very significant cost to the Canadian dairy sector if the threat of a significant increase in imports materializes.
I know the transition from the current rules in the P5 to a freer quota market might not be all smooth and will displease many. Nonetheless, I believe that this is one of the courses of action to take to ensure the best future for the Canadian dairy farms, and should be considered by provincial boards.
Footnotes
I’ve written before in collaboration with Bruno Larue and Alphonse Singbo about quota transfer and technology in Quebec dairy farms. In that article, we argue that Quebec farms are unable to quickly increase their output because of restrictions in quota tradability and this has been a barrier to the adoption of the more competitive milking technologies.↩︎
There are, of course, many policies not mentioned in the table. For example, some provinces have credit trading systems. Besides, Quebec producers recently voted in favour of a credit trading system with a price fixed at $3/kg. I also do not include clauses related to entrants.↩︎
The 2024 study is not available online. I reached out to the contact person for $AgriProfit who promptly shared the most recent studies with me. Thank you very much for the quick response.↩︎