Who Owns the Second Harvest?

AA. A patent is normally granted over an object that cannot make more of itself. Seed is the awkward exception: the thing bought in spring is also the machine that produces next year's stock, and the law has never settled comfortably on what that means for ownership. Two instruments now sit side by side. The older is the breeder's certificate, formalised by the Lindau Convention of 1964, which rewards a breeder for a variety that is distinct, uniform and stable, but preserves what the drafters called the farmer's privilege: the right to keep part of a harvest for sowing on the same holding. The younger is the utility patent, extended to plants by several jurisdictions after 1987, which carries no such privilege and reaches not only the finished variety but the individual trait inside it. A grower may therefore hold two sacks of seed that look identical under the same lamp and owe entirely different duties on each.

BB. Quantifying the resulting conflict is harder than the headlines suggest. The Marlow Institute assembled a register of 412 disputes filed between 1998 and 2019 in which a seed right was asserted against a grower. Only 24 reached a full trial; 71 per cent were settled privately, and a further tenth were withdrawn once the parties had exchanged evidence. The register also corrects a common impression about what the arguments concern. Deliberate reproduction for sale accounted for fewer than one claim in six. The great majority involved seed saved and sown on the grower's own land, conduct that would have been lawful under a breeder's certificate and was not lawful under the patent that in fact covered the material. Median damages, in the small number of cases where a figure was recorded at all, came to 9,800 euros, well below what it costs to defend such a claim to judgment.

CC. The single decision most often cited is Delacour, handed down in 2004, and it is cited by both camps for opposite reasons. A regional court found that a grower whose field carried a patented herbicide-tolerance trait was not liable merely because the trait was present; pollen and spilt grain travel, and presence alone proved nothing about intent. Liability, the judgment held, was confined to growers who had identified the trait in their own harvest and had then deliberately selected and resown it. Rights-holders read the ruling as confirmation that a patent survives an accidental crossing. Grower organisations read the same paragraphs as a shield. What the court did not do, and what commentators on either side regularly forget, was to say anything about which party should meet the cost of testing a harvest to find out what it contains in the first place.

DD. The economic argument has hardened into two positions that rarely meet. Petra Vollmer, an agricultural economist, points to private breeding expenditure in the cereals sector, which she reports as rising from 190 million euros in 1985 to 1.1 billion in 2015; without a right that survives the first sale, she argues, no firm would fund a programme whose product reproduces itself for free. Aidan Kessel, an agronomist, does not dispute the expenditure figures. His objection is that the same period saw the share of certificates held by the four largest firms climb from 38 to 63 per cent, and that money spent is a measure of inputs rather than of varieties actually delivered to growers. Vollmer's reply is that concentration by itself proves nothing about the pace of improvement. Neither has produced a study tracking yield gains against the strength of protection across comparable countries, which is the comparison the quarrel actually requires.

EE. A separate literature asks what the regime does where it is barely enforced at all. Across much of eastern Africa, informal exchange between neighbours still supplies an estimated 78 per cent of what is planted, and action against such growers is close to non-existent: Kessel's field survey of 1,140 households in two neighbouring districts of the same highland province found that only three had ever been contacted by a rights-holder. The survey nevertheless recorded a change in behaviour. Thirty-four per cent of respondents said they had grown more cautious about passing seed over the fence, and a quarter of those cited advice from an extension officer rather than any legal document. Whether that caution has yet altered what actually goes into the ground, the survey did not attempt to establish.

FF. The next round of disputes will probably not be about sacks of seed at all. Sequencing has made it possible to assert a claim over a stretch of genetic code and then hunt for that code in a harvest, which shifts the argument from what a grower did to what a laboratory can detect. Three responses are on the table: a compulsory licence at a fixed royalty, a research exemption allowing protected material to be used freely for further breeding, and a scheme obliging a rights-holder to publish the marker relied on before enforcing anything. The Marlow panel that reviewed the register recommended the third as the least disruptive, noting that it meets the evidential difficulty the Delacour court left open while settling none of the questions about ownership underneath it.