Payment by Results - our first pancake
One of the key research questions of this test & trial was to develop a 'payment by results' approach which could operate on commons as well as home farms.
Originally intended as a smaller element of our work it was the Advisory Team of farmers and landowners guiding the test & trial who wanted to use the scorecard we developed as a land management plan to explore delivering the whole of ELM on a payment by results basis.
The Advisory Team also wanted to look at payments based on the inherent value of public goods delivered, rather than simply the costs of delivering them, as the foundation for payments by results but, alas, we were not sufficiently resourced to do this.
Instead what we did was go a bit Blue Peter and borrow some data that the Duchy had made earlier. The Duchy have been producing natural capital maps of their land and then, with some of their demonstration farms, working with Andersons to identify the costs to the farm of delivering that natural capital.
So when the second iteration of our scorecard was ready we used it to assess one of the Duchy's demonstration farms on Dartmoor. We then used the financial data for the costs of delivering natural capital on that farm to assign a value to their scores as an initial marker of what payment should be set for a particular score.
One farm is not particularly representative so we sense checked the costs for that farm against the Farm Business Survey data for Dartmoor, Bodmin and Exmoor. We included Bodmin and Exmoor because there are only 10 farms in the survey data for Dartmoor and they're all pretty different from each other so not a representative sample. Bodmin and Exmoor have different styles of commoning to Dartmoor but using a slightly bigger pool of data still helped produce more representative figures. It should be noted that management costs for commoning don't exist. The Defra Farm Business Survey doesn't collect any. This is currently being addressed by the Our Common Cause project.
Andersons then used their extensive farm management knowledge to build costs up and down the scoring spectrum from the original starting point.
We then used the SWEEP map, created for this T&T, to work out what proportion of the holding was covered by different habitat types and multiplied it all together.
So the score for the farm for a particular habitat type, like pasture, identifies a payment rate per ha, that payment is then multiplied based on how much of the holding is covered by that habitat type. Do that for all the different habitat types and you work out what that holding would be paid.
Our first attempt, based purely on what it cost to manage specific habitats, would have put most of Dartmoor's farms out of business the payment was so low. It also created some perverse incentives e.g. it costs more to produce leys than permanent pasture, so leys were coming out as being worth more, even though they deliver less public goods.
So we didn't bother showing that version to farmers because it was terrible.
The version shared with you today is our first vaguely acceptable but still not right/would need a lot more work version. It's not just based on management costs but includes distribution of fixed costs across the habitat values as well. Because unless farms are viable enough to stay in business they won't be there to deliver the management of these habitats. The fixed costs are also not distributed equally, an effort was made to use the distribution of fixed costs to address any perverse incentives.
As well as working out the scores from one farm we then used the scores from a couple of other farms to see what the impact would be for their farm business (I would flag that Farm C only covers 30% of the holding because the SWEEP map cuts out land not within the National Park). We also modelled what it might mean for commons.
I cannot stress enough the endless ways this isn't right as it is and would benefit from an extensive amount of adjustment. What it's useful for is highlighting the broad principle of how using a scorecard for a land management plan can then link to a payment by results approach.
We ran it past representatives from each of our three trial commons, who had also tested the scorecard on their home farms, and this is what they had to say:
A hybrid payment by results approach would be better to deal with upfront costs and experimentation
Prefer spreading out the costs, per each individual score, to brackets
It justifies the payment because the delivery is evidenced
Taking samples for no purpose is annoying, this has purpose
Have a levelling off of payments at the higher level because perfection may be unattainable
Take big capital costs for peat and heathland out of per/ha payment and have it in a separate pot
Archaeology, public access and other public goods should also be valued
A base payment might give confidence about payment by results
Bigger payment brackets at the bottom
Hybrid management model
Top payments are "unicorns", need exterminating
Risk is spread by having payments for multiple public goods
Hybrid development is recommendation
Re-grade brackets or mix brackets for top and bottom and have individual scores in the middle
More of the money needs to go slightly lower in the scoring, between 25-75% score
Some of the current payment jumps between brackets are too big
Overall the consensus in the room was that having tested the scorecard and now looked at this payment model the details might not be right but the approach in principle is sound.
If this has piqued your interest in payment by results and it's not something you've come across before then you may want to watch this film from the Burren, who have been running a payment by results approach for the last decade. Recommendations from the Burren farmers to the Dartmoor farming community about their experience probably explains why our T&T ended up going down this route.