Insight, Agriculture & Adaptation

The Trough Before the Dividend

A European assessment of 51 farms finds climate-resilient practice cuts diesel by around half and production costs by around 40 per cent, and that farms are most economically vulnerable during the transition itself. The second finding is the one that decides adoption, and it is measured by nobody.

Line-art scene: a farm income curve drawn as a physical earthwork across a field, dipping into a visible trough in the middle before rising higher than it began, with a farmer and a tractor standing at the lowest point of the dip and a bank building visible on the far horizon.
The dividend is real. Whether a farm reaches it depends on who can stand in the dip.

The most important sentence in Europe's farm resilience evidence is the one nobody is measuring

The European Environment Agency's briefing Building climate-resilient agriculture in Europe: an economic perspective synthesises 51 farm-level case studies, running from the United Kingdom to Ukraine, across soil and water management, crop system diversification, landscape-level management and livestock system redesign.

The headline numbers are strong. Reduced tillage is reported to cut diesel use by around 50 per cent, production costs by around 40 per cent, and labour needs by roughly 25 to 30 per cent relative to conventional systems, case dependent. A conservation agriculture system in Spain reported saving around 2,000 litres of diesel a year and roughly 10,000 euros in fertiliser and herbicide costs.

Then there is a sentence that appears once, is not in the headline, and matters more than any of the figures above it. The analysis finds that farms are often most economically vulnerable during the transition to climate-resilient systems.

That is the whole adoption problem in a line, and it is the thing the evidence base is least equipped to describe.


A profitable investment and a survivable one are not the same thing

The economics of a practice change like reduced tillage are not a step from one steady state to a better one. They are a curve.

Year zero involves capital: a direct drill or seeder, possibly new sprayer capacity, and the disposal or idling of equipment that already exists and is often still being paid for. Years one and two frequently involve a yield penalty while soil structure, biology and weed pressure adjust, and while the operator learns a system they have not run before. Agronomic support in that period is not optional and is not free. Only after that does the input saving become the dominant term and the curve rise above where it started.

Chart, 'Farm margin through a transition to reduced tillage'. A single smooth coral curve plotted against a horizontal axis labelled year 0 to year 8 and a vertical axis labelled farm margin, with a horizontal cobalt dashed baseline for the pre-transition margin. The curve drops below the baseline after year 0, reaches its lowest point between years 1 and 2, crosses back above the baseline around year 3 and rises steadily to sit well above it by year 8. The area between the curve and the baseline where the curve is below is lightly shaded coral. Annotation inside: 'the trough: capital spent, yield adjusting, learning'. Annotation on the rising section: 'the resilience dividend, and where every published figure is measured'. Footnote: Schematic. Depth and duration vary by system and are not measured. Transitions Lab, 2026.
The reported economics live on the rising half. The failure mode lives in the dip.

So the practice can be unambiguously profitable across ten years and unaffordable in year two. Those are different questions and only the first is being answered.

A farm with retained earnings, low debt and a supportive lender can stand in the trough. A tenant farm, a heavily indebted farm, a farm whose operator is five years from retirement with no successor, or a farm whose entire margin is the subsidy payment cannot. Adaptive capacity is not primarily a knowledge variable. It is a balance sheet variable, and it is distributed in a way that is well known and rarely reported alongside the practice results.

Which produces the uncomfortable corollary. If the ability to cross the trough is concentrated among larger, better capitalised, owner-occupied farms, then a programme that improves average adoption accelerates consolidation while reporting success, which is the same structural point we made about European drought adaptation and it recurs here with an economic mechanism attached.


Fifty-one cases, and the ones that are not in the set

There is a methodological issue with the evidence base that is worth naming carefully, because it is not a criticism of the EEA's work, which is explicit about being a synthesis of case studies.

A case study is written about a farm that completed the transition. Farms that began the change and reverted after two bad years do not become case studies. Nobody writes up an abandoned conversion, no advisory service nominates it, and the farmer has no reason to volunteer.

So the reported economics are conditional on persistence. The figures describe what the practice delivers to a farm that got through, and they say nothing about the probability of getting through. Both numbers are needed to make a decision and only one exists.

Line-art diagram: a wide horizontal band representing a transition period. Twelve small identical farm icons enter from the left. Eight continue across the band and emerge on the right inside a bordered box in cobalt labelled 'written up as case studies'. Four turn downward partway across and exit the band via dashed coral lines to a faint unbordered area labelled 'reverted, and not written up', with a question mark above them. The four departing icons are drawn at reduced weight so they read as absent.
The case set is what got through. The informative group is what left the band, and nobody looks for them.

The missing statistic is the reversion rate. Of farms that begin a transition to reduced tillage, what proportion are still practising it in year five, and what distinguishes them from the ones who stopped. That single number would do more for European adaptation policy than another synthesis of successful cases, and it is obtainable, because advisory services, machinery dealers and cooperatives know exactly who went back.

This is a specific instance of a general problem in the evidence base. Selection into the trial was the subject of the earlier piece. This is selection into the write-up, which happens later, is less discussed, and biases in the same direction.


A note on how a number travels

One small illustration of why this matters, drawn from the reporting of this briefing itself.

The EEA's press release states that labour needs were reduced by roughly 25 to 30 per cent. A summary from the EU CAP Network describes a Hungarian reduced tillage system that lowered labour requirements to around 25 to 30 per cent of conventional levels.

Those are very different claims. One is a reduction of a quarter to a third. The other is a reduction of two thirds to three quarters. The underlying case may well support the second while the generalised statement supports the first, and both may be accurate in their own context.

The point is not that anybody made an error. It is that a figure detaches from its case within one step of republication, and that a farmer or a policymaker reading a summary has no way of knowing which claim they are holding. Anyone building a business case on these numbers should go to the briefing and to the case behind the figure rather than to the summary, and any evaluation quoting them should say which farm, which system and which baseline.


What the trough implies for instrument design

If the binding constraint is a temporary cash flow problem rather than a knowledge or attitude problem, then the standard instruments are aimed at the wrong stage.

Advisory services, demonstration farms and information campaigns address knowledge. Capital grants address the equipment cost, which is real and is only the first part of the curve. Almost nothing addresses the two or three years in the middle, which is where farms actually fail.

Line-art timeline with three segments labelled 'decision', 'the trough, years 1 to 3', 'dividend'. Above the timeline, existing instruments drawn as small labelled objects with arrows onto it: an open book labelled 'advisory services' and a demonstration field sign labelled 'demonstration farms' landing on the decision segment, and a coin stack labelled 'capital grant' landing at the very start of the trough. The middle of the trough segment is conspicuously empty with a dashed outline of a missing object labelled 'nothing lands here'. Below the timeline, four proposed instruments in sky blue with dotted leader lines up into the trough: 'bridging finance', 'transition insurance', 'front-loaded payments', 'longer tenancy'.
The existing instruments land on the wrong end of the curve. The middle is the whole failure zone.

Four instruments would, and none is exotic.

Transition bridging finance, sized to the trough rather than the equipment. A facility that covers a modelled income shortfall over three years, repayable from the subsequent saving. This is the same logic as the working capital argument in funding an inventory loop with the right instrument: the shape of the need should determine the shape of the money, and a capital grant does not match a revenue trough.

Yield insurance for the transition period specifically. A defined, time-limited product covering the learning years removes the failure mode that actually causes reversion, and it lapses once the practice is established.

Payments phased against the curve rather than the calendar. Support that is largest in years one and two and tapers as the input saving arrives, rather than an even annual payment across a five-year agreement.

Tenancy terms that survive the trough. A tenant on a three-year rolling agreement cannot rationally invest in soil that will pay back in year six. Tenure length is an adaptation instrument, it is regulated, and it is almost never discussed as one.


What would settle it

The reversion rate, by farm type, tenure and size. The single most valuable missing number.

Trough depth and duration, measured rather than modelled. Actual farm accounts across the transition, not projections. Depth in euros, duration in seasons, and the variance across farms, which is what determines whether a bridging product can be priced.

What the ones who reverted say. They are the informative group, they are not in the case set, and they can be found through machinery dealers and advisers. This is the discovery work that no results framework requests.

Who financed the crossing. Own resources, bank debt, family, subsidy, or a buyer's advance. If the answer is mostly own resources, the practice is only available to farms that already have them, and the policy conclusion follows directly.

The EEA briefing has done the harder half of this by establishing that the economics work. The remaining half is establishing who can get there, and that requires talking to farms that are in the middle of it and to farms that stopped, over several seasons, which is field research rather than synthesis.

The Lab works on this in regenerative agriculture and climate and ecosystems, and on the question of who can afford a practice change through measuring change.

If you are designing support for an agricultural transition and need the trough measured before the instrument is fixed, tell us what you need to know.


Sources


This is an independent insight piece by Transitions Lab. For the Lab's applied work, see Regenerative Agriculture. See also Europe Has Enough Demonstrations on why a pilot cannot tell you what will scale, The Benefits Nobody Was Looking For on outcomes no results framework asked for, Equity Is the Wrong Money for a Warehouse on matching the instrument to the shape of the need, Whose Field Becomes a Wetland on the same instrument mismatch when the change to the land is permanent, and Adoption Is Not the End of the Research on why sampling the farms that reverted is the finding rather than a footnote. To discuss a study, see Contact.

Read more Articles & insights See all articles → See it in the field Case studies See all case studies →