A miller and a baker in the same German state apply for a grant towards the same machine. The miller can receive up to 25 per cent. The baker, if the business counts as small, up to 20. If it counts as medium-sized, up to 10.
The difference is not size, not region, not need. It is a list.
On 2 September the German federal agriculture ministry presented the market-structure measure of the Gemeinschaftsaufgabe — the joint federal-state funding framework — at an event run by the national rural networking agency. One slide answers the question of what the grant rate depends on. The first item: the end product. More precisely, whether the end product is an agricultural product within the meaning of Annex I to the Treaty on the Functioning of the European Union.
Flour is in Annex I. Bread is not. Sugar is in it, sugar confectionery is not. Cocoa beans are in it, chocolate is not.
The line follows no logic of processing depth. Sausage is in Annex I — chapter 16 is listed in full; a bread roll is not, because chapter 19 does not appear in the list at all.
Processing inside the list earns an SME up to 25 per cent, with room to stack: up to 15 percentage points if all products are quality products — organic, say, or protected geographical indications — and up to 10 more for a regional value chain. The bonuses combine. Processing outside the list earns a small business up to 20 per cent, a medium-sized one up to 10, and not a single bonus. Every figure is a ceiling, and the states implement the measure through their own guidelines.
So the sharpest comparison is not 25 against 20. It is the medium-sized business: up to 25 per cent on one side of the list, up to 10 on the other, with nothing to stack.
Two caveats belong in this paragraph rather than in a footnote. The flour-and-bread distinction is not on the slide; it follows from Annex I itself, and the slide expressly names mills, bakeries and slaughterhouses alike as eligible applicants. The boundary does not run between the businesses but between their products. And the study this note returns to below is not yet published. What is quoted here is a conference presentation.
Produce your own and you fall out
A slide later comes a sentence with an exclamation mark. Processing and marketing businesses may not simultaneously be engaged in agricultural production.
That is the second boundary, and it cuts deeper than the first, because it attaches to the type of business rather than to the product. A farm milling its own grain is processing inside Annex I and is still excluded, because it also produces.
What it looks like when a state fills that gap was on show at the same event. The Bavarian State Research Centre for Agriculture presented its offering for direct marketers: free advice at 32 regional offices, three regional platforms, some 90 qualification courses, an umbrella brand, a competition. It is a serious programme, and it contains no investment grant.
Produce and process, and you get advice. Process only, and you get money — provided the product is on the list.
The independent study puts it more plainly
At the same event, a research team from the University of Freiburg presented the first independent nationwide study of structural change in German food processing: twelve months of work, more than 400 sources reviewed, some 30 interviews with people from businesses, associations and research.
One slide is headed “Regulatory obstacles” and consists of four lines about small and medium-sized enterprises. Subsidies: often excluded. Administrative burden: disproportionately high. Tax burden: no relief. Permit procedures: slow and resource-intensive. Beneath them the conclusion, in four words: regulation oriented towards large enterprises.
The talk closes on four recommendations. One of them is to orient funding and subsidy policy towards SMEs and craft businesses.
Two presentations, one event. One explains the instrument, the other recommends rebuilding it.
The favoured stage is the one disappearing faster
This is where it gets uncomfortable — for the obvious reading of this note.
Anyone who has read this far now expects the sentence that says this funding logic is killing the baker’s and butcher’s trade. The figures in the Freiburg presentation do not support it.
Between 1998 and 2023 the number of mills fell from 1,271 to 506, down 60 per cent. Bakeries went from 21,406 to 9,252, down 57. Butchers from 25,492 to 13,522, down 47. Meat cutting was hit hardest: from 1,008 to 396, down 61 per cent. And confectioners: from 3,922 to 3,433, down 13.
Now sort those same figures by the list. Mills, butchers and meat cutting all process inside Annex I. Bakeries and confectioners sit outside it.
So the steepest decline is on the favoured side. The shallowest by a wide margin is on the other. Confectioners, whose products appear in no chapter of the list and who can therefore claim not a single bonus, lost 13 per cent of their businesses in a quarter of a century. The subsidised mills lost 60.
So the funding asymmetry does not explain structural change. It does something else, and that is the actual finding: it sorts on a criterion that has nothing to do with need. Whether a product appears in Annex I is a question of tariff nomenclature. Whether a business survives the next ten years is a different question.
An instrument distributing on the first criterion may hit the second or miss it. It does not notice either way, because it never asks.
Apparatus: grant rates, bonuses and the exclusion of simultaneous production come from the presentation by the German Federal Ministry of Food, Agriculture and Home Affairs, unit 712, on the market-structure measure of the Gemeinschaftsaufgabe “Verbesserung der Agrarstruktur und des Küstenschutzes” (funding area 3.A, measure 2.0), given at the event “Strukturwandel im Lebensmittelhandwerk — Stand und Perspektiven” of the Deutsche Vernetzungsstelle Ländliche Räume on 2 September 2026. All rates quoted are ceilings; implementation rests with the individual states under their own guidelines. The classification of individual products follows Annex I TFEU itself and is not on the slide: chapter 11 (milling products) and chapter 16 (preparations of meat) are listed in full; from chapter 17 only headings 17.01 to 17.03 and 17.05 — sugar confectionery (17.04) is absent — from chapter 18 only cocoa beans and cocoa shells, and chapter 19 does not appear in the list at all. Business counts for 1998 to 2023 from the Zentralverband des Deutschen Handwerks (2025), cited in the presentation by Sophie Buckwitz, Arnim Wiek, Bianca Blum and David Sipple, Humboldt Professorship for Sustainable Food Economy, University of Freiburg, at the same event; the underlying study was still in preparation and unpublished as of September 2026 — a presented working state, not a publication. The direct-marketing offering from the presentation by the Bayerische Landesanstalt für Landwirtschaft, same event. The Gemeinschaftsaufgabe is a German instrument; it does not apply in Austria.