7. Discussion & Conclusions
Summary of findings
97. The evidence provided by the Scottish Government and by the LFASS/ ANC Working Group allows some key findings to be offered under each of the specified review topics.
98. Land abandonment. Although the pattern is uneven and any estimates are subject to some uncertainty, agricultural land abandonment has occurred in recent years. This is despite on-going support under both Pillars of the CAP (i.e. SFP and LFASS) and reflects the influence of a range of factors, including the decoupled nature of most support, low market returns to farming and rising incomes in other sectors. Abandonment has been most prevalent on land with the poorest capacity for agricultural production.
99. Payment distribution. The previous (coupled) headage-based systems have continued to influence the distribution of (decoupled) support payments, with better quality land receiving greater funding. However, revisions to LFASS have increased payment rates on poorer quality land, as will the (phased) introduction of the BPS. Within a given payment category, area-based payments offer greater overall funding to physically larger farms (provided that land is actually claimed rather than abandoned). Although LFASS and the BPS operate separately, it is their joint distribution that is relevant.
100. Sustainable farming systems. On average, farming enterprises within the LFA are unprofitable and hence heavily dependent on support payments. This applies when looking at net margins, but also for many farms even when considering only gross margins or cash incomes. However, business and/or household viability is also influenced by the availability of other income sources, including diversified enterprises, off-farm employment and pensions/investments. On average, LFASS is less important than other income sources and attributing viability specifically to LFASS is difficult.
101. Calculating compensation payments. Although intuitively appealing, the concepts of income foregone and additional costs are hard to operationalise.  Specifically, it is difficult to identify appropriate comparators because observed production systems evolve to reflect prevailing local conditions: extensive production systems with less reliance on purchased inputs are more common in remoter areas with lower land quality and lower land costs.
102. The review findings summarised above have some implications for how a Scottish ANC scheme could be designed and implemented within the rules prescribed by EU regulations. The RDR rules are summarised as:
"Payments to farmers in mountain areas of in other areas facing natural or other specific constraints should, by encouraging continued use of agricultural land, contribute to maintaining the countryside as well as to maintaining and promoting sustainable farming systems. In order to ensure the efficiency of such support, payments should compensate farmers for income foregone and additional costs linked to the disadvantage of the area concerned."
103. The Scottish ANC Working Group principles are expressed in similar fashion:
"The ANC scheme should be focussed on ensuring that an appropriate level of support goes to areas facing natural or other specific constraints which contribute, by encouraging continued use of agricultural land, to maintaining the countryside as well as to maintaining and promoting sustainable farming systems. The scheme will be based on a calculation of additional costs incurred and income forgone as a result of the natural or other specific constraints".
104. However, at a practical level, difficulties in defining relevant comparators mean that estimates of additional costs and income foregone are subject to considerable uncertainties and likely errors - an "appropriate level of support" is difficult to gauge given heterogeneity of farming systems and the influence of factors other than biophysical constraints. Either further research (possibly based on case studies) will need to be undertaken or it will have to be accepted that simplistic estimates which neglect problems of self-selection and endogeneity will lead to systematic under or over compensation. 
105. Yet regardless of how additional costs and income foregone are calculated, the effect of LFASS support on the viability of individual farm businesses and households (rather than individual farm enterprises) is contingent on a number of other factors. In particular, most (but not all) farms receive Pillar I support. Indeed, given the apparent unprofitability of grazing activities within the LFA, viability is determined largely by support payments and other income sources. This means that it is the combined, joint distribution of support that matters rather than the specific value of LFA support viewed in isolation. If ANC rules preclude consideration of interactions with other determinants of business viability, support will not be well targeted and its effectiveness muted relative to other possible approaches.
106. The requirement to base payment rates on additional costs and income foregone also sits awkwardly with the decoupled nature of support. Payments rates based specifically on the additional costs or income forgone incurred through production are only relevant if production is actually occurring. Yet, by definition, decoupled payments are not supposed to directly influence production decisions, acting only indirectly as an income support or safety net mechanism. In which case, again, payments would be more logically calculated with reference to actual additional income support required to maintain business or household viability. However, a specific focus only on additional costs and only on biophysical constraints precludes this, excluding consideration of (e.g.) other income sources but also (e.g.) effects of scale and remoteness.
107. The decoupled nature of LFA/ ANC payments also affects the ability of support to achieve any of the stated economic, environmental and social objectives. Specifically, although total land abandonment might be avoided through sufficiently high payments, this would not by itself necessarily translate into significant farming activity. Rather, because conditionality criteria (e.g. "active farmer", cross-compliance) can be met with minimal production activity, the influence on commodity output is not guaranteed and therefore the associated effects on (e.g.) retaining labour and skills or maintaining appropriate land management for environmental benefits are not guaranteed either. Hence decoupled LFA/ ANC support is at best a weak instrument, further blunted by difficulties in targeting to address heterogeneity in farming systems and circumstances.
108. Notwithstanding the problems articulated above, some form of Scottish ANC policy is expected to be required, at least in the short-term. Although detailed consideration of specific designation criteria and payment rate calculations is beyond the remit of this report, some generic observations about design options and choices can be made.
109. First, different constraint criteria will lead to designated areas of different geographical shapes and reach. Inevitably, there will be an imperfect overlap with the current LFA designation. This will lead to the eligibility of some farms changing, both by moving (in either direction) across the external boundary and by moving between categories. Although a degree of change is unavoidable (not least with loss of the Fragility Tiers), attempts are typically made to minimise it and/or deploy transitional arrangements for affected farms.
110. Second, accounting for heterogeneity within the total designated area is difficult. Although imperfect, the current grazing categories (A to D) were a pragmatic solution (as were the Fragility Tiers). Other descriptors, such as the Land Capability for Agriculture ( LCA) classes or the individual variables underlying the classes might be more accurate  but will give slightly different designation patterns. Retention of grazing categories within the ANC as an already familiar system, and one also essentially adopted for regionalisation of the BPS, would offer greater continuity and avoid additional complexity.
111. Third, however, information to permit calculation of additional costs and income foregone arising solely from biophysical constraints for different degrees of disadvantage/constraint is not generally available. Moreover, such data as are available are generally not collected or reported on the basis of categories likely to be used for sub-dividing the total ANC area.  As such, unless bespoke data collection exercises are undertaken, any calculations are likely to be rough approximations. This may, nevertheless, be adequate for the purposes of proposing payment rates for domestic consultations and for presenting them to the European Commission ( EC). That is, provided that data and estimation methods are presented transparently and caveats are noted, crude approximations may be sufficient if supported by stakeholders and verified by independent analysts.  Subject to observing the minimum payment of €25/ha (c.£18 to c.£20/ha), setting payment rates below calculated costs, as permitted under the regulations, might further ease acceptance by the EC. 
112. Fourth, some complexity can be avoided by limiting sub-divisions within the total ANC. For example, at is simplest, the minimum permitted ANC payment of €25/ha could be applied uniformly within the total designated area. Ignoring the current enterprise mix uplift, this would represent an increase for Grazing Category A land everywhere and Category B in Standard and Fragile areas, but a decrease for categories C and D everywhere. If the total area claimed reached the 2007 level, total expenditure would be just within the current budget.
113. Fifth, the numerical example above reveals that sub-categorisation of the total ANC area into different degrees of constraint can be accompanied by differential payment rates only if either the total ANC is smaller than the (claimed) LFA and/or the overall budget rises - paying more than the minimum anywhere will not be possible otherwise. This highlights that changes to the existing distribution of support is inevitable and any design will lead to a degree of change.
114. Sixth, if more than the minimum payment is offered, a degree of degressivity is compulsory - paying a higher rate per ha on the first few ha of a farm but a lower rate on successive ha. It may be possible to use degressivity to impose an upper limit, a cap, on the maximum area that can be claimed by an individual business (i.e. a payment of £0/ha beyond some further threshold), in which case differential payments might be achievable within the existing budget and the likely overall extent of the ANC (i.e. if the area cap excludes sufficient land to allow some budgetary headroom for varying payment rates).  However, it is not yet clear whether the regulations permit use of such a cap if it leads to an average payment of less than €25/ha for an individual business - much depends on which area is used as the denominator, all of their ANC land or only their ANC land up to the cap threshold.
115. Seventh, if budget constraints and minimum payment requirements limit the scope for differentiating payments to reflect heterogeneity across the ANC, there seems little merit in seeking overly complex, variegated categories or deploying "other specific constraints" or "mountain area" options.
Conclusions and Recommendations
116. LFA support has a long history within Scotland and LFA payments represent a modest but important component of overall farm income. However, effectiveness in terms of stated objectives has been questioned.
117. Whilst the efficiency of previous headage payments and stocking densities may be debated, the switch to decoupled area payments has undoubtedly reduced the influence that payments can have on land management. Consequently, whilst the advent of ANCs offers an opportunity to redesign payment structures, the decoupled nature of payments will necessarily weaken their effectiveness. In addition, a narrow focus on biophysical constraints and additional costs will limit the degree to which payments can be targeted most effectively to achieve stated policy objectives.
118. The effects of 'flattening' and the move to a new BPS scheme will have significant redistributive effects, both from more intensive farmer to less intensive farmer operating similar types of production and from more intensive farming areas to less intensive farming areas. The forecast changes in the BPS distribution show variable but often substantial net gains for areas where there are greater constraints on farming systems. Whether these changes in the distribution of the BPS will be sufficient to sustain current stocking rates and farming practices is unclear given the low financial returns to livestock grazing enterprises, variation in farm business/household circumstances and the limited traction of decoupled payments on management decisions.
119. In this context, if an ANC scheme is to be adopted, it is recommended that further research be undertaken to provide a better understanding of how support needs vary. To comply with the ANC rules, particular attention will need to be paid to disentangling the effects of biophysical constraints from structural (scale) effects and remoteness (transport and local market) effects and, moreover, to account for endogeneity in observed farming systems and land costs. Any scope for flexibility within the ANC rules should be explored and the limitations acknowledged explicitly. Presentation to the European Commission of designation criteria and payment rate calculations should be clear about the data and methods used, including appropriate caveats.
120. Separately, adoption of an ANC scheme is not obligatory. Indeed, other parts of the UK have already moved away from blanket LFA payments in favour of more targeted support mechanisms (e.g. Glastir in Wales). Hence LFASS funding could be switched to non- ANC schemes. For example, to maintain an element of income support, a proportion (or all) of funding could be re-routed though Pillar I  whilst other Pillar II measures could be used to address objectives relating to skills or environmental management. Such changes would not be simple to design or non-contentious. Nevertheless, given that a degree of change is inevitable anyway, it is recommended that serious consideration be given to the merits of diverting at least a proportion of current LFASS expenditure to other support measures which can be more clearly targeted - to focus on desired positive outcomes rather than negative constraints.
121. This recommendation is based on the weak effectiveness of decoupled payments as a land management tool, but also on fundamental flaws in the logic of ANCs. In particular, focusing solely on biophysical constraints neglects a number of other important determinants of farm business/household viability. For example, market interactions which affect factor costs (e.g. land prices), access to other regional markets (e.g. for off-farm employment or diversification) and structural constraints (e.g. small enterprise sizes).
122. Moreover, the concept of additional costs is itself problematic. For example, it could arguably be interpreted as the total costs of land management where land has been abandoned. Equally, it could be applied to any activity, not just livestock grazing.  That is, whereas the focus is habitually on sheep and cattle enterprises because of their historical and continuing presence, a range of activities are possible technically but not economically. That other enterprises (e.g. most dairying, cereals, soft fruit) are not considered for support reveals that implicit judgements have been made about the appropriateness of particular activities. Yet, unless public goods are being targeted effectively, the logic actually flows the other way and undermines the case for supporting any loss-making activities where inertia payments are likely to hinder the structural adjustments and resource reallocations necessary to secure long-term business viability.
123. Although the EU has spent considerable time and effort in devising the ANC framework, overarching EU strategies and priorities emphasise the need to support sectors and communities in coping with long-term pressures for change. This requires a balance between insulating against shocks and encouraging gradual adaptation to different ways of working and living. Hence, the strategic emphasis on (e.g.) innovation, skills development and collaborative organisation, and the increasing emphasis on evaluation tied to outcomes and impacts achieved. Denying and attempting to avoid pressures for structural adjustments cannot continue indefinitely and sustainable rural development is unlikely to be achieved through decoupled payments alone.
Email: Eilidh Totten