Scottish Government

Chapter 6 Economic Assessment of LFASS Scenarios

6.1 Introduction and objectives of the economic analysis

With £61 million per year the LFASS is the most important single support measure in the Scottish Rural Development Plan and any changes in this scheme can be expected to have major direct economic impacts on LFA agriculture. However, agriculture is known to have a range of important linkages to other economic sectors, such as upstream and downstream industries. This suggests that, in addition to the direct impact, changes in agricultural policies will also affect other sectors in the Scottish economy. These spill-over effects are at the centre of the analysis in this chapter.

The scope of the required economic analysis requires a quantitative economic modelling technique which explicitly takes into account linkages of agriculture with other sectors and assesses economy-wide effects, but, at the same time, allows for sufficient details in covering the agricultural sector. The basis of such multisectoral analysis is a consistent and complete data set on all transactions among sectors and institutions requiring for every income a corresponding expenditure. A Social Accounting Matrix provides such a framework: it is a simple and efficient framework to organize economic data and has been used widely to analyse multiplier effects resulting from external policy changes (Sadoulet and de Janvry, 1995).

Agriculture is acknowledged to have relatively high multiplier effects. For example, a review of agriculture's contribution to the Scottish society, economy and environment (Slee et al., 2001) mentions an output multiplier for agriculture of 1.83. Specifically for a range of different farm enterprises, Doyle (2000) describes high income multipliers (around 5) and employment multipliers (around 3) for pig and poultry farms. These multiplier effects provide indications on how changes in agricultural output, income and employment affect the whole economy and can be used as a first assessment and comparison of the economic impacts of different policy scenarios. However, Doyle also explains that the employment effects in primary sectors such as agriculture may be exaggerated due to the exclusion of self-employment occurring in these sectors. Moreover, multiplier effects depend on the well known instabilities in the agricultural sector and the availability of reliable input and output data and can change widely from year to year.

The main objective of chapter 6 is to evaluate the contribution of LFASS to the Scottish economy. To achieve this objective, two main tasks have been carried out:

• Development of a Social Accounting Matrix, which captures both sufficient details of LFA agriculture and the linkages to other sectors and the overall economy
• Based on the accounting framework ( SAM), to calculate multipliers to assess economic linkages of LFA agriculture with the rest of the economy

The developed disaggregated Social Accounting Matrix for Scotland (for the year 2001) is used to review the contribution of LFA agriculture to the Scottish economy. As outlined in chapter 2, in a next step economic modelling techniques are used to transform the Social Accounting Matrix into a simplified multiplier model to analyse the economic linkages of LFA agriculture with other sectors in the Scottish economy. The multiplier model is then applied to determine production multipliers of LFA agriculture and to review and validate these multipliers through the outcome of the farmer and business surveys. Finally, the multiplier model is used to provide indications of the economic impact of LFASS scenarios on the Scottish economy. However, some of the defined scenarios simulate changes at a scale which are difficult to capture in the applied models. Hence, the economic scenario analysis concentrates on the impacts of the abolishment of LFASS. The other LFASS scenarios will be discussed at the end of this chapter taking into account the outcome of the multiplier analysis and the survey presented in chapter 5.

6.2 LFA agriculture and the wider Scottish economy (base Scenario 1)

LFA agriculture contributes to the Scottish economy through direct effects such as generating gross output and value added, but also through indirect effects by buying goods and services from a range of upstream sectors and, on the other hand, selling goods and services as intermediate inputs to downstream sectors. In the following section the report reviews the direct contributions and economic linkages with other sectors of LFA agriculture by comparing the outlined coefficients and multipliers. We follow the assumption that the base Scenario 1 is described by the contribution and linkages of LFA agriculture to, and with, the wider economy in 2001 ²⁵.

Contributions of LFA agriculture to the Scottish economy

The Social Accounting Matrix describes the picture of an economy for a given year. The dataset includes a wide range of information, for example the sectoral contributions to gross output, intermediate consumption, value added as well as household income and consumption pattern, government earning and spending. Moreover, one can draw information from the SAM on the origin of supply and demand in the different sectors. Table 6.1 summarises the gross output of LFA agriculture in 2001 differentiated by farm type and commodity.

Table 6.1 shows that LFA farm types are specialised on livestock production, i.e. mainly sheep and cattle. The climatic and natural conditions in LFA regions do not allow a more diversified agricultural production on LFA farms. The importance of the LFA livestock sector is particularly evident from the contribution of LFA farming to the total output of the Scottish livestock sectors. About 58% of the beef output and nearly 77% of the output in the sheep sector originate from LFA farms in 2001. On the other hand, cereal and other crop output is very small due to often unfavourable soil conditions in LFAs. LFA cattle farms have the highest share (17.86%) in the total output of Scottish agriculture followed by LFA cattle and sheep farms (13.39%). The contribution of LFA sheep farms to the total Scottish agricultural output is with just under 4%, significantly lower. Looking at LFA agriculture as a whole, the table depicts that roughly one third of the agricultural output in Scotland is generated by LFA farms, which amounts to £887 million in 2001. Overall, however, the direct contribution of LFA agriculture to the total output of the Scottish economy as a whole is very limited (0.6%).

Table 6.1 Gross output of LFA agriculture per farm type and commodity (in £million)

Source: Calculated from the constructed Scottish SAM for 2001

The agricultural sector purchased intermediate inputs from other economic sectors of around £1450 million and provided intermediate inputs to other sectors of £912 million, which equals 2.8% of total intermediate consumption in Scotland. About one third of the purchases can be allocated to LFA farm types. In terms of employment, a total of 14,112 people were employed in LFA agriculture, of which 7,853 were full-time employed. The number of full-time employees, however, decreased between 1998 and 2001 by 5% and between 1998 and 2005 by 18% ( SEERAD, 2002 and 2006b). From the constructed SAM it can be calculated that factor income for agricultural labour (including family labour) accounted for less than one percent of total salaries in the Scottish economy in the chosen base year.

An important indicator for the economic contribution of LFA farming is value added, which measures the contribution to the economy of each individual industry or sector and is used in the estimation of Gross Domestic Product ( GDP). Table 6.2 summarises the value added generated by LFA agriculture and its distribution to factor income, again differentiated by farm types.

Table 6.2 Value added of LFA agriculture (in £ million)

Source: Calculated from the constructed Scottish SAM for 2001

LFA agriculture generates a total value added of £290 million in 2001 that is distributed to land rents (35.4%), family labour (30.4%), capital (25.4%) and hired labour (8.8%). With respect to the different LFA farm types, Table 6.2 shows that LFA cattle and LFA cattle and sheep farms contribute nearly four times as much as LFA sheep farms. Overall, Scottish agriculture generates 1.26% of total value added in Scotland, with one third originating from farming activities in LFAs and two thirds (or 0.82 %) from non- LFA agriculture. However, while at national level the overall economic contribution of agriculture is only minor, there are important regional differences which need to be considered. For example, Slee et al. (2001) point out in their comprehensive review of the contribution of agriculture to the Scottish society, environment and economy that in the Orkney Islands, Borders and Dumfries and Galloway regions, agriculture still accounts for more than 10% of the regional GDP.

Moreover, the SAM framework allows for conclusions regarding the origin of supply and demand in the economy and the shares of export and import values. For example, the SAM differentiates between domestically produced supply in the commodity accounts and imports from the United Kingdom and the Rest of the World. Figure 6.1 depicts the value shares of domestic supply of the main LFA farm products beef and sheep as well as pigs, divided into supply from LFA and non- LFA farm types, and the supply imported from the United Kingdom and the Rest of the World. According to Figure 6.1 imports from the UK and the Rest of the World account for about 30% of the value of beef, sheep and pig supply in Scotland, while imports from outside the UK account for more than 10% on all three livestock markets. LFA farming generates about 41% (or £422 million) of the value of the beef supply and 53% (£229 million) of the value of sheep supply in Scotland.

Figure 6.1 Origin of supply of livestock products in 2001 (in percentage of total supply value)

Source: Generated from constructed SAM

As the previous economic indicators, such as output and value added, have shown the overall contributions of LFA agriculture to the national Scottish economy are rather minor. However, LFA agriculture produce a large share of output in the cattle and sheep sectors and any significant decline in cattle and sheep production will thus have knock-on effects for lowland and non- LFA farms. Moreover, LFA agriculture generates indirect contributions to the Scottish economy through existing linkages with upstream and downstream sectors increasing the importance of agriculture for the wider economy, in particular at regional level. These linkages potentially lead to important spill-over effects of changes in LFA farming, mainly the upland livestock sector, to other economic sectors and the wider economy. The following section will review these linkages and quantify backward and forward multipliers for LFA farm types.

Backward and forward multipliers of LFA agriculture

Direct linkages (backward and forward) between agriculture and the rest of the economy are captured by the matrices of Leontief coefficients and output or supply coefficients. To explain the type of information these coefficients contain, Table 6.3 provides the Leontief coefficients for LFA farm types. From Table 6.3 we can see that out of the total production value of LFA sheep farms 13.5% are spent on animal feeds, nearly 10% goes to the manufacturing sector, more than 4% to health vet services and more than 3% each to fertilisers and pesticides. In total, intermediate inputs used in LFA sheep farm production amounts to 70.6%. About 33.5% are allocated to primary inputs (value added) with land rents taking the biggest share of 16.5%. LFA cattle and cattle & sheep farms have similar input pattern, whereby the cattle & sheep farms have the highest share allocated to value added. Overall, direct effects of backward linkages of LFA farm types can be identified for animal feeds, other manufacturing, other services, health and vet services, fertilisers and pesticides. In addition to these disaggregated sectors, a relatively large share is allocated to the aggregated SAM account of other aggregated sectors. Noticeably, between 8.9% and 13.5% of the production value is allocated to intra-sectoral inputs from agriculture.

Table 6.3 Input coefficients of LFA farm types (in percent of total gross output)

Source: Calculated from the constructed Scottish SAM

In addition to linkages of farm types with sectors providing inputs for farming, interesting direct linkages can be identified with the meat processing sector by using both the proportions of total output of the LFA farms that is sold to the meat processing sector and the shares of total input costs of the meat processing sector taken up by LFA agricultural commodities, based on the commodity output structure of LFA farm types.

Table 6.4 Shares of output from the LFA farm types supplied to the meat processing sector (in %)

Source: Calculated from the constructed Scottish SAM

The meat processing sector buys between 15.1% and 23.3% of the output of the three disaggregated LFA farm types. While the figure for the LFA sheep farms is relatively low compared to earlier years, the output shares of LFA cattle and cattle and sheep farms are similar to 1999 (Swales et al., 2003). From these figures one can expect relatively strong forward linkages from LFA agriculture to the meat processing sector.

Table 6.5 Shares of total input costs of the meat processing sector taken up by LFA agricultural commodity sector (in %)

Source: Calculated from the constructed Scottish SAM

Table 6.5 shows that more than 13% of the input costs of the meat processing industry can be allocated to the beef commodity sector followed by 5.2% to the sheep sector and 1.1% to the pig sector. Overall, LFA farming commodities take up nearly 20% of the input costs of the meat processing sector in 2001, which, again, is comparable to figures identified in previous studies. For example, Swales et al. (2003) showed that 21% of the input cost of the meat processing industry could be allocated to LFA farms in 1999. Similarly to the forward linkages of LFA agriculture (only vice versa), there are some noticeable backward linkages from the meat processing sector to LFA agriculture.

In a next step, backward and forward multiplier can be generated by using the inverse matrices of the coefficients matrices. These multipliers include direct and indirect effects of the economic linkages. The complete matrices of backward and forward multipliers are provided in the appendices 6.2.

Backward multipliers

Table 6.6 summarises the backward multipliers of the different LFA farm types on other economic sectors captured in the SAM. Backward linkages are reported with respect to the gross output of other industries (module activity-activity) and demand (supply) of the different commodities (module activity-commodity).

Table 6.6 Backward linkages of LFA farm types²⁷

Source: Calculated from the constructed Scottish SAM

Table 6.6 indicates that, for example, an increase in final demand for the output of LFA sheep farms of £1000 will generate an additional output of £80 in the animal feeds and other manufacturing sectors, £60 in the health and veterinary sector, as well as £150 in the other services sector and £220 in the aggregated sectors. Overall, LFA sheep farms have a cumulative industry output multiplier of 1.7, which means that the above increase in final demand for the output of sheep farms generates an additional output of £700 in all other industries. LFA cattle and sheep farms have a slightly higher multiplier in this module. It is evident from the above table that LFA farms have the strongest linkage with the industries other services, other manufacturing, animal feeds, and health and vet services. This picture is similar to findings in other country case studies, such as Hajnovicova (1999) and Hajnovicova and Lapisakova J. (2002) who also identified close linkages of farming with manufacturing and service sectors. Generally, the second (lower) module of the LFA farm type multiplier on the different commodities follows a similar pattern. The other point to note is that the single elements of the multipliers on the commodity demand are in most cases higher than the multipliers on the domestic gross output of the industries (activities). For example, the described increase in final demand for LFA sheep farm output generates an £80 increase in the gross output of the animal feeds industry, but the demand on the commodity market for animal feeds increases by £140. This can be explained by the fact that part of the additional supply needed to satisfy the increased demand for animal feeds will come from imports.

Comparing the backward linkages of the different industries, Table 6.7 depicts the cumulative multipliers for the partial backward linkages between the different industries (activities) and between the industries and commodities.

Table 6.7 Cumulative backward linkages of the different industries (activities)

Source: Calculated from the constructed Scottish SAM

Since different farm types have different input requirements it is not surprising that the multipliers vary between the different farm types. However, the range of the output multipliers is between 1.65 to 1.79 with cattle and dairy farms having relatively high multipliers. The same applies for the cumulative commodity multipliers, although these are generally lower. In comparison to other (non-farming) industries, Table 6.7 shows that the output multipliers (as well as the commodity multipliers) of LFA farm types are in the middle range ranking from 9 ^th to 11 ^th of the 24 sectors. Upstream and downstream industries such as dairy products, meat processing, animal feeds, fertilisers and also forestry harvesting have higher output (as well as commodity) multipliers.

The above calculations provide national multipliers. Multipliers analysed by Doyle (2000) and Midmore (1987) indicate significant differences between Scotland and Wales, suggesting that multipliers are spatially specific. In this context, Slee et al. (2001) argued that the magnitude of multipliers depends on the structure of the economy under investigation and extent of inter-industry dependencies. Such interdependencies, however, do not only vary between different national economies, but also between different regions within a national economy. The economic importance of agriculture differs between different regions in Scotland which consequently leads to variations in the interdependencies of the regional economy on agricultural activities. Regional differences are also confirmed by the surveys conducted in chapter 5. Moreover, the survey results indicate regional differences in how farmers would try to reduce cost as a response to LFASS changes, which would lead to different spill-over effects varying between sectors and regions. For example, reducing spending on machinery was mainly mentioned in Aberdeenshire, while reducing spending for animal feeds is a more likely response on the Western Isles.

Forward multipliers

Following the Rasmussen method (Rasmussen, 1958) and its modification introduced by Jones (1976) forward multipliers can be generated based on the output coefficients of a supply-driven model. Table 6.8 reports the forward multiplier of the LFA farm types with respect to other sectors.

Table 6.8 Forward linkages of LFA farm types

Source: Calculated from the constructed Scottish SAM

Similarly to backward linkages, LFA farm types have similar cumulative forward multipliers between 1.50 and 1.54 with cattle farms at the top end of the range. The forward linkages are, however, smaller than the backward linkages. It is evident from the above table that LFA farms have the strongest forward linkage with the meat processing industries indicating the importance of LFA livestock production for the wider economy. The forward linkages with the meat processing industry are even stronger than the linkages with the other aggregated sectors.

Table 6.9 Cumulative forward linkages of the different industries²⁸

Source: Calculated from the constructed Scottish SAM

The cumulative forward multipliers of the different farm types range from 1.48 (dairy farms) to 1.62 (cereals farms). In comparison to other (non-farming) industries, Table 6.9 shows that the forward multipliers of LFA farm types are in the lower middle range ranking from 13 ^th to 15 ^th of the 24 sectors. In contrast to the backward multipliers, downstream industries (expectantly) have smaller forward linkages, while the main input (upstream) industries for agriculture have rather high forward multipliers, since these are placed at the beginning of the production process.

However, it is important to note that the multiplier estimates presented in this study are based on a dataset of transactions in a given single year. The inherent instability of the agricultural sectors, and the difficulties involved in collecting reliable input-output data, means that that the magnitude of these national multiplier effects may vary widely from year to year (Slee et al., 2001). Multipliers are based on strong assumptions and the simplicity and transparency of multiplier models does not come without cost. If there are any kind of capacity constraints in any of the sectors, multiplier models will overestimate the total effects. Moreover, as prices are fixed, no substitution can take place, which also leads to an overestimation of the total response (Round, 2003). These limitations need to be taken into account when using multiplier models for policy analysis.

6.3 Economic implications of alternative LFASS scenarios

In this section, the report continues by discussing the economic impact of the alternative LFASS scenarios defined in chapter 2. However, due to the nature of the defined scenarios and the requested methodology for the economic analysis, the multiplier analysis will only be applied to quantify some indications of the economic impact of Scenario 5. For the other scenarios a review of the potential implications is provided taking into account the multipliers and linkages of LFA agriculture and the results from the survey in chapter 5.

6.3.1 Scenario 2

As outlined in chapter 2, Scenario 2 follows the proposed interim proposals for LFASS in the new Scottish Rural Development Programme ( SRDP) plus some specific modification such as a minimum eligibility area of 10 hectares, minimum payments of £750 and a maximum payment per recipient of £30,000. An indicative calculation conducted in chapter 5 (see Table 5.14) summarises changes in regional payment receipts under this scenario.

While the LFASS changes assumed in this scenario do change the distribution of funds within the LFA farming sector, the overall amount "injected" into LFA agriculture remains unchanged. This suggests that no major spill-over or multiplier effects on the wider economy are to be expected at national level. In terms of the national multiplier models developed in the previous section no changes in the exogenous government spending for LFA agriculture would be implemented. At regional level, however, Scenario 2 leads to important redistribution effects of the LFASS funds. Regions with larger and more productive farms would receive less LFASS funding, while other regions, characterised by extensive livestock production mainly on rough pasture, would receive higher LFASS payments.

The results of the survey suggest that, for example, farmers in the Western Isles would benefit from such a scenario. Total payment receipts in the Western Isles would increase by nearly 24%. But farmers in this region do obtain a large proportion of their income from alternative sources outside agriculture (the survey suggests an 84% income share from outside agriculture), which implies a relatively low economic dependency of household incomes on agricultural subsidy payments. This suggests that the response to such a LFASS scenario would only include minor adjustments in farm management with little changes in agricultural production and input use. On the other hand, depending on how a future LFASS is designed, mandatory scheme requirements and methods of payment calculation (for example with respect to livestock numbers) could imply an automatic increase in livestock numbers and agricultural production to qualify for the higher payments. But even if higher available LFASS funds do not lead to a noticeable increase in livestock numbers, the higher payments could buffer further extensification pressure resulting from, for example, other policy changes such as reductions in the SFP and thus help to maintain livestock numbers. This would avoid potential problems of undergrazing and habitat deterioration and sustain environmentally important extensive grazing habitats in these regions.

Table 5.5 ( chapter 5) shows that there a number of regions potentially affected by lower LFASS receipts. The affected regions include Dumfries & Galloway and the Scottish borders, where agriculture provides larger than average contributions to the regional economy. Given the relatively high importance of agricultural activities in these regions, potential farmers' responses such as reducing their spending on local businesses as well as labour input on the farms could lead to spill-over effects on other sectors and significant reductions in the regional economy. The multiplier analysis as well as the results from the survey suggest that a range of upstream and downstream sectors, such as the animal feeds, other manufacturing and meat processing sectors, are at highest risk of being affect by spill-over effects because of their strong linkages with the agricultural sector. Fewer inputs bought by the farming sector and less output provided reduces the income of the upstream and downstream sectors and, as the business survey confirmed, could potentially lead to further negative employment effects in these sectors.

Multiplier models are static and can not take into account substitution effects and potential restructuring of the farming sector in some regions triggered by lower subsidy payments. Lower subsidy payments reduce the dependency on such payments and increase the need, but also the opportunity, to restructure towards new businesses and income sources. While extensification is a widely anticipated response to subsidy decreases, some farmers might respond differently and restructure the business and diversify their income sources, both within and without agriculture (as indicated in Table 5.8, chapter 5). Extensification and reduction of labour could weaken the local economy, but diversification might imply positive effects for the local economy through employment creation, increase in economic activities and stronger collaboration within the community. While restructuring and diversification might be an option for a certain proportion of the farms, not all farms will have the opportunity to take up these options. Diversification and restructuring options, in particular on-farm, are generally limited in LFA regions.

Overall, the economic implications of Scenario S2 are of importance at regional level but less significant for the national economy. The key economic aspects of Scenario 2 are:

• Limited impact at national level, but important regional redistribution effects of the LFASS funds
• Higher economic incentives to maintain livestock numbers in very extensive systems ( e.g. Western Isles)
• Negative spill-over effects to upstream and downstream businesses in more agricultural areas ( e.g. Dumfries & Galloway and the Scottish borders)
• Incentive for restructuring and diversification of some farm businesses in regions with lower LFASS payments due to reduced dependency on subsidies.

6.3.2 Scenario 3

Scenario 3 assumes that LFASS no longer exists and upland support are focused primarily on achieving environmental goals via Tier 2 and Tier 3 of LMCs with an emphasis on the agri-environment measures (including access measures) within these tiers. It is important to note that this scenario does not assume that the increase in agri-environment funding are new funds injected into the agricultural sector, instead it is money shifted from one policy measure to another measure within the agricultural sector. Hence, no new money is injected into this sector and the potential implications of this scenario are a product of the abolishment of LFASS and an increase in agri-environment funding.

On the one hand, the loss of the LFASS payment will in many areas reduce the incentives to maintain livestock numbers and, as the results of the survey confirm (Table 5.8, chapter 5), lead to a decrease in cattle and sheep numbers and other extensification measures in LFA livestock systems, resulting in a loss of farm income (compare with the discussion in Scenario 5). On the other hand, the funds will then be redistributed to LFA farms through agri-environment measures in LMCs. While in total the same amount of payments goes back to LFA agriculture, the payments for the single farms might be lower or higher than their LFASS payment, depending on the corresponding contracts of each farm and the specific measures and associated payments included.

Agri-environment measures have specific requirements and conditions with respect to land management and in many cases, management requirements in agri-environmental schemes are likely to increase labour and capital requirements on the farm. But their economic impact on LFA farms depends on the type and design of the specific measures in Tier 2 and Tier 3 as well as the amount of associated financial support. The rules for the payment calculation are defined at European level setting the wider framework for the payment calculations. The new Rural Development Regulation determines that payments need to be based on income foregone or additional cost. Instead of the previous incentive element, farmers can claim transaction cost for agri-environment and animal welfare payments. In this context, a recent EU report ( EU-Commission 2005a) confirms the importance of the methods of payment calculations in agri-environment measures for environmentally friendly land management. Payment calculations normally take only into account variable costs or income forgone, which result from the participation in agri-environmental programmes. But in cases where land abandonment is a high risk, the continuation of farming and land management is in question. The report concludes that this implies that payment calculations would have to take into account the full costs of environmentally desirable land management, reflecting both fixed costs and variable costs, which would lead to considerably higher payments than existing ones that cover only marginal costs or marginal income forgone. Thus, payments need to be based on a harmonized methodology, but at the same time incorporating specific local circumstances to provide appropriate economic incentives.

Overall, however transferring funds from LFASS to agri-environment schemes is likely to lead to reductions in gross output of LFA agriculture. The reduction in livestock could be reduced by providing further incentives (in addition to the Scottish Beef Calf Scheme) through environmental payments to maintain livestock, e.g. cattle on LFA farms.

The reduction in agricultural output would lead to further indirect negative effects on the Scottish economy. Following the farm type multipliers generated in the section 6.3, a hypothetical £1m reduction in output of LFA cattle farms would lead to a further reduction in output in other sectors of £700,000 resulting in an overall reduction in gross output of £1.7m in the Scottish economy. But the potential reduction in agricultural activities and their spill-over effects to the wider economy cover only one side of the implications of Scenario 3. Agri-environment support entail widely recognised economic benefits both for the farms and the local economy such as employment creation, wildlife management and on-farm diversification and public good provision. These benefits can be divided into market benefits and non-market benefits.

A number of studies evaluated the economic benefits of agri-environment schemes (for example Hanley et al., 1996, CRER and CJC Consulting, 2002, Crabtree et al., 2002, Scottish Executive, 2004, Agra CEAS Consulting, 2005, Moran and Hussain, 2006), whereby the majority of studies focussed on the non-market benefits. Non-market benefits are analysed by comparing the magnitude of public expenditure for the schemes and the Willingness to Pay ( WTP) of the public for conserving the study sites. Moran and Hussain (2006) conclude in their recent review of economic benefits of agri-environment schemes that these studies show the cost-effectiveness of public expenditure for the covered study sites even in appraising non-market benefits in isolation.

Employment effects of nature heritage protection and management are well cited in the literature ( e.g. Environmental Resources Management, 2004, Shiel et al. 2002, SNH, 1997), but the specific linkages of market benefits to agri-environment measures are less evident in the literature. However, an often cited example for an agri-environment scheme, which successfully creates market benefits, is the Tir Gofal scheme in Wales. Agra CEAS Consulting (2005) concluded in their evaluation report that Tir Gofal is an important source of business for land-based contractors in Wales and makes an important contribution to rural employment opportunities and rural development. In particular, positive employment effects were identified in this evaluation, but also positive spill-over effects for the local economy through employment creation and local procurement of materials. While the assessment of Tir Gofal provides an interesting example for the economic potential of agri-environment support, it is difficult to draw conclusions from this example for Scottish agri-environment measures. Agri-environment schemes and measures vary to a great extent in their design with different specific requirements and payments. Hence, different agri-environment schemes are likely to have different (economic) impacts. It can be summarised that in general agri-environment support has, in general, the potential to provide significant economic benefits, in addition to environmental improvements, but further and more detailed economic assessments of the different agri-environment measures (to be) included in Tier 2 and Tier 3 of the LMCs are required to be able to compare the specific economic benefits of these measures and benefits generated by other schemes such as LFASS.

The key economic aspects of Scenario 3 are:

• Loss of LFASS is likely to lead to lower gross output and lower farm income with negative spill-over effects for the local/regional economy
• Economic impact of increased agri-environment support on farms depends on the type and design of the measures and associated payments
• Evidence for positive employment effects from agri-environment measures with positive spill-over effects for the local economy and provision of non-market benefits (public goods)
• More detailed economic assessment of agri-environment measures in Scotland requires further research

6.3.3 Scenario 4

Scenario 4 assumes the shift of LFA support from axis 2 to axes 1 and 3, which implies that no per hectare payments are required by EU guidelines. It is further assumed that financial support will mainly be provided for projects in relation to diversification, marketing and processing. Overall, LFA support in Scenario 4 can be seen as a structural payment aiming to improve the infrastructure and business environment for rural communities in remote areas assuming that farming activities in remote areas can only be maintained within viable rural communities. Similarly to Scenario 3, it is again important to note that this scenario does not assume that the increase in funding for socio-economic measures constitutes of new funds. Instead, in this case, it is money redistributed from one policy measure to other (socio-economic) measures in the Rural Development Plan. In contrast to Scenario 3, not the full amount of (former) LFASS funds is going back to farming; 50% of the funding is distributed to processing activities through projects conducted in collaboration with land managers and linked to support measures in LMCs. Hence, the potential implications of this scenario are a product of the abolishment of LFASS and an increase in project-based socio-economic funding.

Similar to the previous scenario, the removal of LFASS is, in the first instance, likely to lead to lower agricultural output in LFAs due to the reduction in incentives of maintaining production levels, such as livestock numbers. Lower agricultural output would again lead to the previously discussed negative spill-over effects to upstream and downstream businesses (see Scenario 5). On the other hand, the potential economic implications of an increase in funding for socio-economic measures are more complex since the socio-economic measures comprise of a range of different measures (diversification, marketing processing, etc.) and part of the funding is assumed to be used for processing activities, not necessarily conducted on-farm. Hence, part of the funding is transferred from primary agricultural production to processing activities.

Using the multiplier model for a first comparison of the impact of government spending for agriculture or other related sectors reveals that processing activities have a higher output multiplier than agriculture (compare with Table 6.7). Due to backward linkages from processing to primary agricultural production, an increase in final demand for output of the processing sector £1m would increase the output of the LFA farm types by £140,000, with the strongest increase for LFA cattle farms (compare with Appendix 6.3). However, given the implied reduction of incentives for maintaining livestock numbers in this scenario (due to the assumed abolishment of LFASS), care needs to be taken to avoid a situation, where livestock numbers have declined to an extent that the upland cattle and sheep sectors can not anymore provide sufficient livestock for lowland farms and the processing industry. It is important in this context to point out again that this scenario assumes that any support for processing activities will be used either for direct on-farm processing activities or collaborative projects involving the farming and wider rural community, hence benefiting the farming industry more directly than captured by the multiplier model. But evidence from an economic evaluation of the Processing and Marketing Grant Scheme in England reveals that the scheme has little impact on markets for agricultural production due to the limited budget and is not very effective in generating direct benefits for farmers and growers. On the other hand, positive employment effects have been acknowledged ( ADAS Consulting and University of Reading, 2003a).

In contrast to Scenario 3, the increase in socio-economic measures can be expected to have some positive effects on LFA agriculture, e.g. through marketing measures, improved infrastructure and on-farm diversification. The removal of LFASS would increase the need to restructure farm businesses. These measures potentially improve the competitiveness of LFA agriculture and diversify the farm income and create employment opportunities for the rural communities. On-farm diversification could be supported through measures similar to Agricultural and Farm Business Development Schemes. Such schemes provide the opportunity to promote the uptake of alternative production systems such as energy crops and biomass production. The outcome of an evaluation of the Agricultural Development Scheme in England indicated that, in the cases examined in the study, the Scheme promoted sustainable employment creation, expansion of trade associations which provide information to primary producers entering new markets and improved volumes or values for produce from primary producers ( ADAS and University of Reading, 2003b). In this context, the importance of existing and effective marketing and processing infrastructure, which need to be established and supported, has been pointed out in the literature (Towers et al., 2006). But the potential of diversification into alternative production systems strongly depends on the natural and climatic conditions and will be limited in many areas in Scottish LFAs.

The outcome of the survey suggests strong regional differences in farmers' responses to take up diversification options outside agriculture which would have a significant influence on the economic impact of such a scenario. Increased funding for diversification outside agriculture is unlikely to have a significant economic benefit in the Western Isles where already the bulk of the income of farm households is generated through non-farming sources. On the other hand, the survey revealed scope for further diversification outside agriculture in Aberdeenshire and Lochaber suggesting potential economic benefits from enhanced diversification measures in these areas.

Generally, non-farming diversification offers potential economic benefits for the farm households through income from activities outside agriculture and through collaborative projects with other rural interest groups. Diversification into tourism is often mentioned as a potential option for farm diversification including the provision of accommodation, farm shops and tea shops as well as direct marketing of quality farm products. Beside the direct economic benefits for the farm household, diversifying into tourism also extend the farmers' range of public goods provision from a traditional landscape manager to a landscape educator and interpreter and provider of the cultural experience of uplands to the public. Thus, there are economic gains for farmers and benefits for the public through increased public goods provision. On the other hand, tourism opportunities vary between different regions and some regions in Scottish LFAs have only a limited potential for tourism-related businesses. Although the "tourism market" is limited and diversification into tourism accommodation can clearly not be an option for all farmers, this supports the case for policy to promote local and farm-based (or farm-linked) tourism opportunities as much as possible depending on the local circumstances. Interestingly, Midmore et al. (2001) concludes in this context that in hill farming, environment-oriented activities and tourism opportunities need to be more integrated in the policy approach. However, he also points out that these opportunities only exist as long as enough environmental and cultural assets are present.

Overall, the effectiveness and success of the processing grants and diversification schemes has been hampered by limited available budgets. Moreover, these schemes have in most cases been implemented isolated from other support measures lacking the required integration with wider socio-economic and agri-environment measures. An increase in available funding and integration in a wider support framework such as LMCs is likely to improve the success and efficiency of these measures ( ADAS and University of Reading, 2003a and b, Burton et al., 2005, Wright et al., 2005c).

The key economic aspects of Scenario 4 are:

• Negative impact of LFASS removal on agricultural output and farm income needs to be weighed against the positive impact of increased funding for socio-economic measures to enhance competitiveness and income diversification
• Experience from English processing schemes reveals little impact on markets for agricultural production and farmers due to the limited budget, but positive employment effects have been acknowledged
• English case studies indicate that such socio-economic measures create employment, provide information on new markets and increase primary production
• Evidence suggests economic benefits from regional targeting of diversification measures (both within agriculture and outside)
• Increased budget and integration with other support are crucial to improve efficiency of diversification, marketing and processing measures

6.3.4 Scenario 5

Scenario 5 assumes the removal of all LFASS funds without reallocation in any aspect of rural development. The removal of LFASS without any provision of alternative funding mechanisms, as assumed in the two previous scenarios, would have the strongest impact on agricultural output of all scenarios and is likely to cause significant adjustments in the LFA farming sector. Using the demand driven multiplier model to simulate the removal of LFASS, the potential spill-over effects to other sectors and the wider economy are discussed. The absolute changes in gross output of the different sectors are reported in Table 6.10.

The removal of LFASS reduces the output of LFA cattle farms by nearly £23m and of sheep and mixed livestock farms in LFAs by around £20m each. In relative terms, however, the biggest reduction is depicted for LFA sheep farms with a reduction of slightly more than 20%. This confirms the farm level calculations conducted in chapter 3, which identified that the LFASS removal would represent a drop for the sample farms studied of roughly £10k (compared with the average LFASS receipts in Scotland as a whole of about £4,700) and would have a particular impact of LFA sheep farmers. The removal of LFASS has also a minor impact on the non- LFA farm types and the total output of these farm types declines by nearly £5m. In addition to the impact on the agricultural sector, Table 6.10 shows reductions in output in a number of other economic sectors due to the backward linkages of LFA agriculture. Output of animal feeds is declining by more than £5m, while other manufacturing and health and vet services experience a decline of about £4m and £3.5m, respectively. The reduction in the other services sector is roughly twice as high and the aggregated other sectors in the SAM suffer a decline of more than £13m. These figures are in line with the outcome of the business survey in chapter 5 predicting animal feeds, fertilisers and veterinary services as strongly affected sectors. In total, the £61m reduction in government spending for LFA agriculture leads to a decrease in the gross output of the Scottish economy of £104m, which reflects an average multiplier of the LFA farm types of 1.71. While the total output effect on the Scottish economy is rather small (-0.7%), the figures suggest that a range of non-farming sectors are affected by the LFASS removal.

Table 6.10 Percentage and absolute changes in gross output (in £ million)

Sources: Generated based on the Scottish SAM 2001

As outlined in section 6.2 the rather strong assumptions behind multipliers need to be considered suggesting that the above figures reflect an overestimation of the effects, e.g. the 20% decline of output of LFA sheep farms. Moreover, it is important to remember that such models do not allow for any substitution effects. For example, a reduction of subsidies for agricultural land in LFAs reduces the opportunity cost of using the land for other purposes such as afforestation and woodland creation. A recent scoping study on the opportunities for Scottish forestry (Towers et al., 2006) outlines that with a changing balance between agricultural subsidies and forestry payments, opportunity costs of forestry uptake decline and the need for rural diversification increases, potentially leading to higher forestry uptake in particular on lower quality land. An Irish case study simulating the consequences of the impact of the 2003 Mid-Term Review of the CAP reports an increase in forestry output (Dixon and Matthews, 2006). But as Towers et al. (2006) point out such land use change scenarios strongly depend on the perception and response of land managers to decoupling of agricultural subsidies in the longer term.

Moreover, substitution effects and occurring prices changes would lower the effects of the LFASS removal. Other studies which have applied sector models or general equilibrium models, which allow for substitution effects and price changes, show that the responses to policy changes are comparably smaller (for example Dixon and Matthews, 2006, Gelan and Schwarz, 2005). These studies also show that substitution effects, hence the shift of resources from agriculture to other sectors results also in positive spill-over effects for sectors not linked to agriculture and the overall economy. These positive effects are not captured in such multiplier model. Given the regional variations in the importance of agriculture for the Scottish economy as well as the differences in farmers' responses in the survey, any impact of LFASS changes or removal on other upstream and downstream sectors will significantly differ across the Scottish regions.

The key economic aspects of Scenario 5 are:

• Removal of LFASS leads to significant reductions of gross output of LFA farm types, with the biggest relative decline for LFA sheep farms
• Reductions in output in a range of other economic sectors due to the inter-sectoral linkages of LFA agriculture. The business survey indicates animal feeds, fertilisers and veterinary services as strongly affected sectors
• The removal of LFASS leads to a changing balance between agricultural subsidies and subsidies for other land uses. Lower opportunity costs of forestry and woodland creation and the need for rural diversification could lead to higher forestry uptake in particular on lower quality land
• Potential substitution effects of resources could lead to positive spill-over effects for the wider economy
• Evidence suggests large regional variations of the implications of LFASS removal, but further regional economic analysis would be required to derive more detailed conclusions