Evaluation of less favoured area support scheme

3. Land abandonment

12. Land abandonment is an emotive term, but also one open to different interpretations. In particular, although land used previously but no longer active in agricultural production has been abandoned from a farming perspective, it may still be delivering social benefits in the form of other ecosystem services such as recreational enjoyment or climate regulation. Moreover, the possibility of returning land to agricultural (or other commodity provisioning service) usage in the future means that it retains an option value: unless land values drop to zero, it has not been truly abandoned.

13. Nonetheless, prolonged agricultural abandonment in the sense of no active farming management has been observed historically in Britain during periods of agricultural depression and more recently as a transitional effect of accession to the EU in some New Member States. Moreover, once abandoned agriculturally, some land may be very difficult to bring back into production due to (e.g.) scrub encroachment, drainage degradation and loss of local skills. Policy interest in this form of abandonment arises for a number of reasons, including the relationships to total agricultural production and agri-food supply-chains, to the viability of individual farms and rural communities, and to environmental conditions.

14. The impacts of any agricultural abandonment will depend on the nature of these relationships. For example, a diminished area of utilised farmland may reduce the volume of domestic agricultural output available for processing and consumption whilst any associated reduction in other inputs (notably labour) and income may impact negatively on local economies. Equally, various semi-natural habitats may be threatened by cessation of land management. Conversely, abandoned land may impact only marginally on production levels, local economies may be more dependent on other activities and some environmental conditions may be enhanced by reduced management intensity.

Measurement

15. Irrespective of the nature of the relationships, the first step is to identify and measure the extent of any land abandonment. Unfortunately, the main source of agricultural land use statistics - the June Agricultural Census - is only available with a degree of aggregation at a regional, parish or farm level. This means that intensity of use can only be calculated across the entire reported area, not for individual parcels of land, ^[1] for example, by assessing average labour or livestock per ha across the whole farm, parish or region.

16. Yet intensity of management does vary and, although wholesale abandonment is possible, it is generally individual parcels of land (or indeed parts within a parcel) that are abandoned first. For example, individual farmers will typically choose to retire their least productive land, such as parcels most compromised for production (e.g. prone to flooding) or most distant from the farmhouse, and to continue farming their remaining land. Abandonment also lies at one extreme of the spectrum of management intensity, and may be preceded by a gradual reduction in intensity. Consequently, the actual extent of any abandonment (both current and/or historic) is obscured by estimates of average intensity of usage which can, at best, only give an indication of likely abandonment.

17. Although active land uses are recorded in the agricultural census, there is no explicit category for land abandonment and hence abandonment can only be inferred from a reduction in the total area reported. This may be reasonable at the regional or national level, but is unlikely to be accurate at the parish or farm level since reductions in reported areas may simply reflect transfers between farms in different parishes, not actual abandonment. ^[2] Moreover, the annual census lacks full coverage due to a combination of less than complete compliance by farmers and deliberate less frequent inclusion of smaller farms. Consequently, some census figures are based on estimates rather than reported data and may or may not accurately reflect land use at a given location. Nevertheless, census data can provide broad indications of change.

18. As an alternative, claims for area payments under either Pillar I (i.e. the BPS or previous SFP) or Pillar II (i.e. LFASS) are more likely to be accurate and timely since there is a direct financial incentive to claim and penalties for mis-claiming. As such, reductions in areas claimed may be reasonably interpreted as indicating abandonment. ^[3]

19. Some farmers may adjust their average land use intensity by reducing the area claimed without altering other inputs (e.g. livestock), meaning that land may appear to be abandoned on paper but nothing need necessarily have changed on the ground. That is, for example, the same number of livestock may actually be run over the same area of land, but the area reported will change. ^[4] Nonetheless, payment claims are probably a reasonable indicator of abandonment.

20. Other means of measuring abandonment could include bespoke surveys of farmers or analysis of remote sensed land use data. The former would be time-consuming and expensive whilst also adding to the information provision burden on farmers. By contrast, remote sensed data are generally already available. Unfortunately, a recent attempt revealed a number of practical difficulties and was unable to generate reliable estimates (Aitkenhead, 2015). Hence identification of abandonment here is subject to the caveats noted above and based on the aggregate census and payment data provided by the Scottish Government.

Findings

21. At the national level, over the period 1992 to 2014, census data show that there has been a reduction in the area of LFA farmland associated with grazing livestock and in the number of livestock (see Table A). The reduction in livestock has outpaced reductions in area, so average stocking densities have also declined. The reductions have been steeper in those parts of the LFA receiving the lowest level of LFA payments (less than £20/ha, compared to £20-£40/ha for mid-range payments and greater than £40/ha for high payments).

Table A: Livestock and Farm Area Changes, 1992 to 2014

Source: derived from Table 2 in SG's Project 1 Summary Report on Land Abandonment.

22. This pattern is confirmed by LFASS claim data, albeit for the shorter period of 2007 to 2013. Specifically, there was a reduction in the total area claimed of over 460k ha (14%) with over three-quarters of this decline occurring on the poorest quality land, Grazing Category A (see Table B).

Table B: LFASS claimed land area 2007 and 2013, by LFASS grazing category

Source: Table 3 in SG's Project 1 Summary report on land abandonment.

23. Although the SG analysis concludes only that abandonment may have occurred, this seems unduly cautious - the figures are consistent with other commentary about general declines in agricultural activity across Scotland (and the wider EU) in response to a combination of challenging market conditions and the decoupling of Pillar I support.

24. As noted previously, it is possible that the extent of abandonment may be exaggerated in some cases where land has been effectively abandoned for some time, but is only now being reported as such because of the formal activity requirements. Equally, some parishes have gone against the trend and actually increased numbers of livestock and/or claimed areas. ^[5] Nonetheless, it is reasonably clear that some agricultural abandonment has occurred.

25. This abandonment has occurred despite on-going support, both via LFASS and the (larger) SFP - probably reflecting the switch to decoupled payments and poor market returns. The high share of poorer quality (Grade A) land within the total abandoned area indicates that more marginal land is being discarded first. This is as expected given the lower per ha market returns and lower per ha support payments associated with poorer quality land - businesses with a mix of land qualities will abandon the poorest first and businesses with only poorer quality land have the least scope for adjusting production systems. ^[6]

26. Where abandonment has occurred, questions arise as to the extent of any adverse impacts and the extent to which LFASS has reduced the degree of abandonment. Consideration of impacts is beyond the remit of this review, but it should be noted that environmental benefits derive from appropriate management, not necessarily simply continued current management. Equally, given the dominance of poorer quality land amongst the total abandoned area, the associated reductions in agricultural output and labour inputs may not be large at the national scale (but may still be significant at the local level).

Discussion

27. The pressure for abandonment arises from a combination of factors. First, when support payments are excluded, many cattle and sheep enterprises within the LFA are unprofitable (see later sections on Sustainable Farming Systems and on Additional Costs/Income Foregone). As such, continued land management is dependent on support payments off-setting market losses sufficiently to provide an acceptable return.

28. Second, the level of return deemed acceptable is influenced by a range of considerations. For example, the level of market returns achieved (which in turn depends on scale and efficiency) but also the importance of that enterprise to overall business or household income (which may draw on, inter alia, other enterprises, off-farm employment, pensions and investments), average incomes (in the local or wider economy), willingness to tolerate lower incomes to retain a familiar way of life, capital (land) value appreciation, and whether a family successor for the farm is in place. As such, reflecting variation in individual circumstances, the level of support needed to induce continued land management varies considerably across farms. This partly explains ^[7] continued agricultural production when average profitability is clearly poor: the question is perhaps not so much why some land has been abandoned but why more has not been.

29. Third, the decoupled nature of both LFASS and the SFP means that they exert no direct influence over intensity of production above some minimum level required to satisfy "active farmer" criteria. As such, support payments do not necessarily preclude partial abandonment (i.e. a general reduction in average production intensity and withdrawal of active management from individual parcels of land) nor total abandonment if market losses are high relative to total support.

30. Higher support payment rates per ha might reduce formal abandonment of poorer quality land, but would not necessarily induce significant land management changes: if the act of production incurs losses, decoupled support payments can really only induce minimum production since greater production imposes effectively an income loss on land managers. In this context, an understanding of the distribution of support payments across the LFA is of interest.