Value chain collaboration in practice: how farmers and processors share risk

On the dairy farm where I grew up, the milk truck came every second day and nobody at the kitchen table thought much about who drank the milk. The price was set elsewhere, the quality tests came back in the mail, and the relationship with the processor was, for practical purposes, a relationship with a pipe. Many farm products still move that way. But a growing share of Canadian food is produced under arrangements where the farm and the buyer plan together, share information and agree in advance how risks will be split. People call this value chain collaboration, and the phrase covers everything from a simple contract to a partnership that runs for decades.

This piece sets out what the term means, the main types of arrangement, how they share risk, and why so many of them fall apart. The examples are illustrative composites, not descriptions of real firms.

Supply chain or value chain

The two terms are often used as if they meant the same thing. They do not.

A supply chain is the physical sequence of steps that moves a product from farm to consumer: growing, hauling, processing, packing, distributing, selling. Every food product has one. Each link can deal with the next purely through price, buying from whoever is cheapest this week.

A value chain is a supply chain whose members coordinate on purpose because they believe they can create more value together than separately. They agree on what the end customer wants, they share the information needed to deliver it, and they agree how to divide the gains and the losses. The physical steps may be identical. What differs is how the links talk to each other and how long they expect to stay together.

The difference matters most when products are specialised. A farmer selling standard feed barley can sell to any elevator. A farmer growing a specific variety of pea for a particular processor, or raising hogs to a retailer's welfare specification, has made investments that only pay off if that buyer keeps buying. That dependence is what makes coordination valuable, and also what makes it risky.

The main arrangements, compared

Most collaboration starts with a contract. The table below compares the common forms.

Marketing contracts

A marketing contract settles the sale before the product exists. A vegetable processor might agree in winter to take a set number of tonnes of sweet corn from a grower in August, at a stated price per tonne, with deductions for poor quality. The grower still owns the crop and still bears the weather risk. What the grower gains is certainty of a buyer and a known price. In some provinces, prices and terms for processing vegetables are negotiated collectively by a growers' marketing board rather than farm by farm, which changes the balance of bargaining power.

Production contracts

A production contract goes further. The buyer supplies the animals or the main inputs, sets the practices to follow, and pays the farmer a fee for the work and the use of buildings. The farmer's income no longer depends much on the market price of the product. Instead it depends on performance measures such as feed conversion or mortality. The farmer has traded upside for stability.

Cost-plus and formula pricing

These are two ways to set the price inside a contract. Cost-plus starts from an estimate of the farmer's cost of production and adds a margin, so the farmer is protected when feed or fertiliser prices rise. Formula pricing ties the payment to a reference market and adds premiums and discounts for quality, which keeps the price close to the market but rewards better product. Our earlier piece on how cattle and hogs get priced shows how grid formulas work for livestock.

Many arrangements combine the two. A formula might follow the market most of the time but include a floor and a ceiling, or a cost-adjustment clause that kicks in when feed prices move beyond an agreed band.

Sharing information

Contracts set the terms. Information is what makes them work.

The most useful data flows in both directions. The processor knows its sales forecast, its plant schedule and what customers are complaining about. The farmer knows planting dates, herd health, crop progress and what the yield is likely to be. When each side keeps that to itself, both plan blind. When they share it, the processor can schedule the plant better and the farmer can plant or breed to match.

Quality data is the clearest example. A packer measures carcass weight, lean yield and defects for every animal it buys. A farmer who sees that data animal by animal, linked back to genetics and feed, can improve. A farmer who only sees the cheque cannot. The same is true of a processor that reports to growers which fields produced the best pack-out and why.

Information sharing also cuts waste. Overproduction that cannot be processed in time, and specification failures that send product to a lower use, are among the larger sources of loss before food reaches a store. The piece on food loss versus food waste explains how those losses are counted.

Joint problem solving

The most developed value chains go beyond contracts and data. Farmers and processors meet regularly to work on specific problems: a recurring quality defect, a new product, a cost that is growing faster than anyone likes. Engineers or field staff from the processor visit farms; farmers visit the plant and see what happens to their product. The model has roots in manufacturing, notably the Japanese supplier networks discussed in a piece on the keiretsu supplier model.

Joint problem solving needs a rule about who keeps the savings. If a farmer finds a way to cut the processor's trim losses and the processor keeps all of the gain, the farmer will not bother again. Successful chains usually split documented savings by an agreed formula.

Two illustrative chains

An illustrative pork chain

Picture a regional pork program built around a processor and about 30 independent hog farms. The program sells fresh pork to a grocery chain under a specification covering genetics, feed and animal care. The farms raise hogs to that standard, and the processor pays a premium of, say, $8 per hog over the reference grid price, plus quality bonuses.

The numbers here are round and invented. A farm marketing 10,000 hogs a year earns an extra $80,000 in premiums. In return, it has committed its barns to a specification that costs, say, $5 more per hog to meet, for an extra $50,000 in costs. The net gain of $30,000 is real but modest, and it depends on the program's continuing. Each farm also receives carcass data on every hog, which helps it earn the quality bonuses.

The risk sharing comes from two clauses. A feed-cost adjustment raises the premium when feed costs rise beyond a set band, so the processor absorbs part of a feed shock. And a five-year term with a notice period means neither side can walk away in a bad month. Farms gain stability. The processor gains a reliable supply of hogs that meet its customer's specification.

An illustrative vegetable grower and processor

Now picture a grower with 200 acres of processing tomatoes and a cannery 40 kilometres away. Each spring they sign a contract for an expected tonnage, with planting dates staggered so that harvest matches the plant's capacity day by day. The cannery's field staff advise on varieties and spray timing. At delivery, loads are graded and paid by quality.

The collaboration shows in the hard years. When a wet spell delays harvest, the grower and the plant agree on which fields to pick first. When a hot summer brings the whole crop in early, the plant runs extra shifts rather than refusing loads. Neither side writes those decisions into the contract. They come from years of working together and from the knowledge that next spring's contract depends on how this autumn was handled.

Why collaboration fails

Most value chain partnerships that collapse do so for a small number of reasons.

Power imbalance

When one side has many alternatives and the other has few, the stronger side can change terms at renewal, delay payments or tighten specifications without sharing the cost. A farmer who has built barns for one buyer has little bargaining power once the barns are built. Without some protection, such as a long term, a notice period or collective bargaining, the arrangement turns into dependence.

Data hoarding

A processor that collects detailed quality and yield data but reports only summary figures to farmers keeps the value of that information to itself. Farmers notice, and they stop sharing their own information in return. The reverse happens too, when farmers withhold expected volumes to keep their options open.

One-sided contracts

Contracts that allow one party to cancel at short notice, change prices unilaterally or reject deliveries on vague quality grounds put all the risk on the other side. They may look efficient on paper, but they invite the weaker party to leave at the first better offer.

Unclear sharing of gains

If nobody wrote down how savings or premiums would be split, each side will remember the agreement differently. Disputes over money are usually disputes over expectations that were never set.

Turnover of people

Value chains rely on trust between particular individuals. When a plant manager or a field buyer leaves and the replacement treats the farms as interchangeable suppliers, years of goodwill can vanish in a season.

What tends to work

Reading the arrangements that last, a few features come up again and again. Terms long enough to justify investment, with clear exit rules. Pricing formulas that both sides can check, with the data behind them shared. A regular forum to deal with problems before they become disputes. An agreed way to divide gains. And a product whose value to the final customer is high enough that there is something worth dividing.

None of this removes market risk. A collapse in retail demand, a disease outbreak or a closed export market will still hurt everyone in the chain. What collaboration changes is who takes the first hit, and how fast the chain can adjust. A farm and a processor that already share forecasts and quality data can change a plan in a week. Two parties who talk only through price find out about the problem when the cheque arrives, and by then it is usually too late to fix it. More on these arrangements is gathered on the value chain management hub.